text
string |
|---|
# Introduction
The Relationship of Urinary Metabolites of [Carbaryl](https://www.ncbi.nlm.nih.gov/mesh/D012721)/[Naphthalene](https://www.ncbi.nlm.nih.gov/mesh/C031721) and [Chlorpyrifos](https://www.ncbi.nlm.nih.gov/mesh/D004390) with Human Semen Quality
# Abstract
*In the **Abstract** section:*
Most of the general population is exposed to carbaryl and other contemporary-use insecticides at low levels. Studies of laboratory animals, in addition[ to limi](https://www.ncbi.nlm.nih.gov/mesh/D012721)ted human data, show an association between carbaryl exposure and decreased semen quality. In the present study we explored whether environmental expo[sures to](https://www.ncbi.nlm.nih.gov/mesh/D012721) 1-naphthol (1N), a metabolite of carbaryl and naphthalene, and 3,5,6-trichloro-2-pyridinol (TCPY), a metab[olite of c](https://www.ncbi.nlm.nih.gov/mesh/C029350)hl[or](https://www.ncbi.nlm.nih.gov/mesh/C029350)pyrifos and chlorpy[rifos-me](https://www.ncbi.nlm.nih.gov/mesh/D012721)thyl,[ are associ](https://www.ncbi.nlm.nih.gov/mesh/C031721)ated w[ith decreased semen quality](https://www.ncbi.nlm.nih.gov/mesh/C012587) i[n hu](https://www.ncbi.nlm.nih.gov/mesh/C012587)mans. Subjects (n =[ 272) were r](https://www.ncbi.nlm.nih.gov/mesh/D004390)ecrui[ted through a Massa](https://www.ncbi.nlm.nih.gov/mesh/C007031)chusetts infertility clinic. Individual exposures were measured as spot urinary concentrations of 1N and TCPY adjusted using specific gravity. Semen quality was assessed as sperm concentration, percent mo[ti](https://www.ncbi.nlm.nih.gov/mesh/C029350)le sp[erm,](https://www.ncbi.nlm.nih.gov/mesh/C012587) and percent sperm with normal morphology, along with sperm motion parameters (straight-line velocity, curvilinear velocity, and linearity). Median TCPY and 1N concentrations were 3.22 and 3.19 μg/L, respectively. For increasing 1N tertiles, adjusted odd[s ra](https://www.ncbi.nlm.nih.gov/mesh/C012587)tios [(O](https://www.ncbi.nlm.nih.gov/mesh/C029350)Rs) were significantly elevated for below-reference sperm concentratio[n ](https://www.ncbi.nlm.nih.gov/mesh/C029350)(OR for low, medium, and high tertiles = 1.0, 4.2, 4.2, respectively; p-value for trend = 0.01) and percent motile sperm (1.0, 2.5, 2.4; p-value for trend = 0.01). The sperm motion parameter most strongly associated with 1N was straight-line velocity. There were suggestive, borderline-significant associations for TCPY with sp[er](https://www.ncbi.nlm.nih.gov/mesh/C029350)m concentration and motility, whereas sperm morphology was weakly and nonsignificantly assoc[iate](https://www.ncbi.nlm.nih.gov/mesh/C012587)d with both TCPY and 1N. The observed associations between altered semen quality and 1N are consistent with previous s[tudi](https://www.ncbi.nlm.nih.gov/mesh/C012587)es of[ c](https://www.ncbi.nlm.nih.gov/mesh/C029350)arbaryl exposure, although suggestive associations with TCPY a[re](https://www.ncbi.nlm.nih.gov/mesh/C029350) difficult to interpret because human and[ animal ](https://www.ncbi.nlm.nih.gov/mesh/D012721)data are currently limited.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
Despite the ubiquitous use of insecticides and subsequent exposure among the general population [Centers for Disease Control and Prevention (CDC) 2003; Hill et al. 1995; MacIntosh et al. 1999], there are limited human studies investigating associations between exposure to contemporary-use insecticides at environmental levels and male reproductive health. Human and animal data suggest a potential association between exposures to some commonly used insecticides and decreased semen quality. A study of workers that packaged carbaryl found an increased proportion of oligozoospermic (< 20 million sperm/mL) and teratospermic (> 60% abnormal sperm morphology) men compared with a reference group of chemical workers (Whorton et al. 1979; Wyrobek et al. 1981). Further support for carbaryl’s testicular toxicity comes from studies in laboratory rats that showed associations between carbaryl exposure and sperm shape abnormalities and chromosomal aberrations (Luca and Balan 1987), as well as dose–response relationships between carbaryl exposure and a decline in epididymal sperm count and motility and increased abnormal sperm morphology (Pant et al. 1995, 1996; Rybakova 1966; Shtenberg and Rybakova 1968). Carbaryl was also found to disrupt endocrine regulation of gonadal function in fish (Ghosh and Bhattacharya 1990).
Chlorpyrifos, a frequently used insecticide until being banned for residential use (Lewis 2000), is less studied than is carbaryl for its testicular toxicity but has been found to disrupt endocrine regulation in ewes (Rawlings et al. 1998). Recently, the CDC reported measurable levels of urinary 3,5,6-trichloro-2-pyridinol (TCPY), a metabolite of chlorpyrifos and chlorpyrifos-methyl, and 1-naphthol (1N), a metabolite of carbaryl and naphthalene, in > 90% and 75% of males in the United States, respectively (CDC 2003).[](https://www.ncbi.nlm.nih.gov/mesh/D012721)
The present study was designed to investigate the association between environmental exposure to the nonpersistent insecticides chlorpyrifos and carbaryl and altered semen quality among adult men. Insecticide metabolite levels in urine were used as biomarkers of chlorpyrifos and carbaryl exposure.[](https://www.ncbi.nlm.nih.gov/mesh/D004390)
## Materials and Methods
*In the **Materials and Methods** section:*
Study subjects were men who were partners in subfertile couples seeking infertility diagnosis from the Vincent Burnham Andrology lab at Massachusetts General Hospital (Boston, MA) between January 2000 and April 2003. The study was approved by the human studies institutional review boards of the Massachusetts General Hospital and the Harvard School of Public Health. After the study procedures were explained and all questions answered, subjects signed informed consent forms. Details of subject recruitment have been previously described (Hauser et al. 2003). Briefly, consecutive eligible men were recruited to participate.
Of those approached, 65% consented. Most men who declined to participate in the study cited lack of time on the day of their clinic visit as the reason for not participating. Men with a medical history of risk factors for infertility (e.g., varicocele or orchidopexy) were a priori excluded from the study analyses. None of the men reported occupational exposure to pesticides or other agents suspected to be associated with semen quality. A single spot urine sample was collected from each subject on the same day as the semen sample.
Urine samples were frozen at −20°C and mailed on dry ice to the CDC, where TCPY and 1N were measured as previously described by Hill et al. (1995). Briefly, samples were fortified with stable isotope analogs of the target analytes, and glucuronide or sulfate-bound metabolites were liberated using an enzyme hydrolysis. TCPY and 1N were isolated using liquid–liquid extraction, chemically derivatized, and measured using gas chromatography–chemical ionization–tandem mass spectrometry.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
Although creatinine concentrations are commonly used to adjust for variable urine dilution in spot samples when measuring pesticide metabolites, creatinine adjustment may not be appropriate for compounds that undergo active tubular secretion, which includes organic compounds such as TCPY and 1N that can be conjugated by the liver in the form of glucuronides or sulfates (Boeniger et al. 1993). Creatinine levels also vary by sex, age, muscle mass, race, diet, activity, and time of day. Therefore, adjusting urine insecticide metabolite concentrations using specific gravity (SG) may be more appropriate; thus, SG was used as the primary method for dilution adjustment in the present study. However, in addition to SG-adjusted results, volume-based (unadjusted) and creatinine-adjusted TCPY and 1N concentrations were also determined to allow for comparisons with exposure distributions from other studies. Samples with creatinine concentrations > 300 mg/dL or < 30 mg/dL, or with SG > 1.03 or < 1.01, were considered too concentrated or too dilute, respectively, to provide valid results (Teass et al. 1998) and were excluded.
Creatinine was measured photo-metrically using kinetic colorimetric assay technology with a Hitachi 911 automated chemistry analyzer (Roche Diagnostics, Indianapolis, IN). SG was measured using a handheld refractometer (National Instrument Company Inc., Baltimore, MD).[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
Measurement of the semen parameters (sperm concentration, motility, and morphology) has been described previously (Hauser et al. 2003). Briefly, we measured sperm count and motility by computer-aided semen analysis (CASA) using the Hamilton Thorne IVOS 10 Analyzer (Hamilton-Thorne Research, Beverly, MA). To assess sperm morphology, we evaluated 200 sperm using the Tygerberg Strict Criteria (Kruger et al. 1988). In addition, seven CASA motion parameters were measured. Measurement of these parameters has been previously described (Duty et al. 2004).
Briefly, CASA outcomes included a mathematically smoothed velocity (designated VAP), straight-line velocity (VSL), curvilinear velocity (VCL), amplitude of lateral head displacement (ALH) that corresponds to the mean width of the head oscillation as the cell swims, and beat cross frequency (BCF), which measures the frequency with which the cell track crosses the cell path in either direction. VAP, VSL, straightness (STR = VSL/VAP × 100), and linearity (LIN = VSL/VCL × 100) are indicators of sperm progression, whereas VCL, ALH, and BCF are indicators of sperm vigor. We also used STR and LIN to describe sperm swimming pattern. Some of the CASA parameters were strongly correlated with each other because they describe different aspects of the same movement. Measures of progression, VAP and VSL, were highly correlated, which indicated they were likely measuring a similar characteristic of sperm movement.
We chose VSL over VAP as a measure of progression because it is a direct measurement as opposed to a mathematically smoothed value. VCL was chosen as a measure of vigor and was strongly and positively correlated with ALH but not correlated with BCF. The two measures of swimming pattern (LIN and STR) were strongly correlated, indicating they were likely measuring a similar characteristic of sperm movement. We chose LIN as a measure of swimming pattern because the other parameters chosen for this study (VSL and VCL) are components of LIN and not of STR. Therefore, we chose measure of progression (VSL), vigor (VCL), and swimming pattern (LIN) for statistical analyses.
These three measures are also not as heavily dependent on the type of CASA instrument used, allowing for some comparison with results from other studies.
## Statistical analysis.
*In the **Statistical analysis.** section:*
Statistical analyses were performed using semen parameters both as a continuous measure and dichotomized using World Health Organization (WHO) reference values for sperm concentration (< 20 million sperm/mL) and motility (< 50% motile sperm; WHO 1999). We used the Tygerberg Strict Criteria for morphology to determine below-reference morphology (< 4% normal morphology) (Kruger et al. 1988). Men with values above reference values for all three semen parameters were used as comparison subjects in the logistic regression models. For the CASA motion parameters (VSL, VCL, and LIN), we used multiple linear regression models to assess associations with insecticide metabolites. Nine azoospermic men were excluded from the CASA analyses because motion parameters were not measurable.
Insecticide metabolite concentrations were used both as a continuous measure and categorized into tertiles. For metabolite values below the limit of detection (LOD), corresponding to 0.25 μg/L for TCPY and 0.40 μg/L for 1N, an imputed value equal to one-half the LOD was used. Normality of the metabolite concentrations and semen parameters was assessed, and appropriate transformations were performed before linear regression. Distributions of TCPY, 1N, and sperm concentration were log-transformed in the models. The remaining semen parameters and CASA parameters were normally distributed and not transformed.
Semen parameters were stratified by demographic categories to investigate the potential for confounding. Associations between demographic variables and insecticide metabolite levels were also explored. We considered smoking status, race, age, body mass index, a previous exam for infertility, and abstinence time as potential covariates. Inclusion of covariates in the models was based on statistical and biological considerations (Hosmer and Lemeshow 1989). Covariates were entered into the models individually in a forward stepwise manner.
Covariates that changed the exposure parameter estimate by greater than 10% were retained in the multivariate model and were considered confounders. There was evidence of confounding by both age and abstinence time in many, but not all, of the models for the various outcome measures. However, because there is evidence that age and abstinence time are associated with semen quality, we included them in all multivariate models (Blackwell and Zaneveld 1992; Kidd et al. 2001). Age was modeled as a continuous independent variable. Abstinence time was modeled as an ordinal variable with five categories: ≤ 2, 3, 4, 5, and ≥ 6 days.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
## Results
*In the **Results** section:*
A total of 330 eligible men provided a single semen and urine sample. The distributions of urinary levels of TCPY and 1N for the 330 men are presented in Table 1, as are adjusted metabolite distributions after excluding men with highly concentrated or dilute samples according to creatinine (23 of 330 men; n = 307) or SG (58 of 330 men; n = 272). SG-adjusted TCPY and 1N levels were moderately correlated (Spearman correlation coefficient = 0.3; p < 0.001). Demographic characteristics and semen parameters are described in Table 2. Subjects were primarily white (82%), with a mean (± SD) age of 36.2 ± 5.5 years, and 72% had never smoked.
The proportion of men with a previous exam for infertility was higher among all three of the below-reference semen parameter groups (48%, 36%, and 40% for sperm concentration, motility, and morphology groups, respectively) than among the comparison group (25%). The semen parameter categories were not mutually exclusive. A man could contribute data to one, two, or all three of the below-reference groups.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
Odds ratios (ORs) for the relationship between dichotomized semen parameters and SG-adjusted metabolite tertiles are presented in Table 3. Compared with men in the lowest 1N tertile, men in both the medium and high SG-adjusted 1N tertiles were more likely to have below-reference sperm concentration {ORs for increasing exposure tertiles = 1.0, 4.2 [95% confidence interval (CI), 1.4–13.0], 4.2 [95% CI, 1.4–12.6]; p-value for trend = 0.01} and sperm motility [1.0, 2.5 (95% CI, 1.3–4.7), 2.4 (95% CI, 1.2–4.5); p-value for trend = 0.01]. Although the ORs for the second and third tertiles were both significantly different from 1.0, the exposure–response trends were not monotonic. There were suggestive associations between SG-adjusted TCPY with sperm concentration (1.0, 2.1, 2.4; p-value for trend = 0.09) and sperm motility (1.0, 1.6, 1.7; p-value for trend = 0.09).
However, the estimates for the second and third tertiles suggest that the dose–response relationship was not monotonic. Sperm morphology was weakly associated with both TCPY and 1N.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
To further explore potential dose–response relationships, subjects were divided into quintiles based on SG-adjusted 1N and TCPY concentrations (Figures 1 and 2). Although not monotonic, there were relationships between increased 1N and sperm concentration (OR estimates for increasing exposure quintiles were 1.0, 0.7, 2.3, 3.6, 2.4; p-value for trend = 0.02) and decreased sperm motility (1.0, 0.8, 2.8, 2.0, 2.8; p-value for trend = 0.002). A suggestive relationship was found between 1N and abnormal sperm morphology (1.0, 1.1, 1.5, 1.4, 2.3; p-value for trend = 0.09). Point estimates for the associations between TCPY quintiles and below-reference sperm concentration, motility, and morphology were > 1.0, but none of them approached statistical significance.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
We conducted sensitivity analyses to test the robustness of the results. Associations between SG-adjusted exposure tertiles and below-reference semen parameters were recalculated after excluding nine azoospermic men. For 1N, ORs were moderately attenuated for sperm concentration (1.0, 3.0, 3.1; p-value for trend = 0.05) but were unchanged for sperm motility. ORs for the highest TCPY tertile with both sperm concentration and motility were slightly larger but remained of borderline statistical significance.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
We also reanalyzed the data after retaining the 58 men with SG < 1.01 or > 1.03 (n = 330). Estimates of relationships with 1N tertiles became moderately lower for sperm concentration (1.0, 3.0, 2.6; p-value for trend = 0.05) and motility (1.0, 2.2, 1.9; p-value for trend = 0.03). The suggestive relationship between TCPY tertiles and sperm concentration became slightly stronger (1.0, 1.8, 2.2; p-value for trend = 0.08), whereas relationships of 1N with sperm morphology and TCPY with sperm motility and morphology remained weak.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
Results of multivariate linear regression models for continuous semen parameters and continuous urinary metabolites are shown in Table 4. A suggestive association between SG-adjusted 1N concentration and decreased sperm concentration was found (p = 0.06). As in the logistic regression analysis, there was an association between 1N levels and a decreased percentage of motile sperm (p = 0.03). SG-adjusted TCPY did not show associations with decreased concentration or morphology, but there was a suggestive association with motility. Similar results were found in sensitivity analyses that excluded nine azoospermic men (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
Multivariate linear regression analyses for CASA motion parameters (Table 4) showed significant inverse associations for VSL and LIN with increased SG-adjusted TCPY (p-values < 0.05). SG-adjusted 1N levels were inversely associated with VSL (p = 0.02). CASA motion parameters were also modeled against tertiles of SG-adjusted TCPY and 1N. The association of TCPY with LIN became nonsignificant (linear regression coefficients for increasing exposure tertiles were 0, −1.16, −1.05; p-value for trend = 0.3). An inverse relationship remained for TCPY and VSL (0, −0.13, −2.79; p-value for trend = 0.05) and between 1N and VSL (0, −2.17, −3.50; p-value for trend = 0.01).
There was a suggestive inverse relationship between 1N and VCL (0, −0.49, −4.16; p-value for trend = 0.09).[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
In addition to SG-adjusted values, all statistical analyses were performed with unadjusted and CRE-adjusted TCPY and 1N concentrations (results available from the authors upon request). Results using unadjusted values were similar to those from SG-adjusted values. Creatinine-adjusted results differed from SG-adjusted results. The only relationship in the multivariate logistic models that approached statistical significance was between sperm motility and creatinine-adjusted 1N tertiles (1.0, 1.3, 1.7; p-value for trend = 0.08) and quintiles (1.0, 1.3, 1.6, 1.9, 1.8; p-value for trend = 0.07). No statistically significant associations were found between creatinine-adjusted metabolite levels and outcome measures in the multivariate linear regression analysis.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
## Discussion
*In the **Discussion** section:*
In the present study we found associations between urinary metabolites of contemporary-use insecticides and decreased sperm concentration and motility in humans. Specifically, we found statistically significant inverse dose–response relationships between 1N and sperm concentration and motility, as well as between 1N and VSL. Suggestive associations were found between 1N and sperm morphology, VCL, and LIN and between TCPY and sperm concentration, motility, and VSL.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
The present data were generally consistent with laboratory animal studies that have shown an association between exposure to carbaryl and decreased semen quality. A 90-day study of rats found statistically significant dose-related declines in epididymal sperm count and percent motile sperm, as well as increased sperm with abnormal morphology (Pant et al. 1995, 1996). In an earlier study, subacute and chronic reproductive effects of carbaryl were found in male rats (Rybakova 1966; Shtenberg and Rybakova 1968). Subacute exposure induced a decrease in motile sperm by an average of 40% after 50 days, whereas chronic exposure led to a significant decrease in motile sperm among even the lowest of the three exposed groups after 12 months.[](https://www.ncbi.nlm.nih.gov/mesh/D012721)
Limited animal studies have explored relationships between chlorpyrifos exposure and semen quality. Decreased sperm production and motility was observed in Holstein bulls 6 months after dermal lice treatment with an unknown amount of chlorpyrifos [Agency for Substances and Disease Registry (ATSDR) 1997; Everett 1982]. Other animal studies found no associations between chlorpyrifos exposure and altered male reproductive health (ATSDR 1997; Breslin et al. 1996).
However, semen quality was not assessed in these studies, and conclusions were reached in part based on the lack of observed changes in testicular weight. In the carbaryl studies, no change in rat testicular weight was reported for lower doses for which decreased semen quality was observed (Pant et al. 1995, 1996; Rybakova 1966).[](https://www.ncbi.nlm.nih.gov/mesh/D004390)
Human studies investigating exposure to carbaryl and chlorpyrifos and associations with male reproductive health are limited. Until recently, there were no known human male reproductive health studies that used biological measures of exposure to carbaryl and chlorpyrifos (ATSDR 1997). Swan et al. (2003) found elevated but nonsignificant ORs for low semen quality (sperm concentration, motility, and morphology below the population median) among 24 Missouri men with detectable 1N (OR = 2.7; 95% CI, 0.2–34.2) and TCPY levels (6.4; 95% CI, 0.5–86.3). The numbers of subjects were small, limiting statistical power. In a study among Chinese workers exposed to other organophosphate pesticides (ethylparathion and methamidophos), Padungtod et al. (2000) found significantly lower sperm concentration and sperm motility compared with nonexposed workers but no difference in sperm morphology.[](https://www.ncbi.nlm.nih.gov/mesh/D012721)
In the present study, the relationship between 1N and sperm concentration below the WHO reference value (WHO 1999) is consistent with two published reports on a cohort of carbaryl production workers (Whorton et al. 1979; Wyrobek et al. 1981). Whorton and co-workers found a higher percentage of exposed workers (15%) had sperm concentrations below the reference value of 20 million sperm/mL compared with non-exposed controls (5.5%, p = 0.07). In contrast to the present study, Wyrobek et al. (1981) reported an association between carbaryl exposure and sperm morphology. The distribution of abnormal sperm morphology was significantly higher for exposed workers (p < 0.005), and the proportion of teratospermic men was larger in the exposed group (29%) compared with controls (12%, p = 0.06). Because of logistical constraints, sperm motility was not measured in the published reports of the carbaryl production worker study.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
Functional defects of sperm may be an important factor in male infertility. The role of reactive oxygen species (ROS) in male infertility has been suggested in studies that found higher seminal ROS levels in infertile men compared with fertile controls (Agarwal et al. 1994; Pasqualotto et al. 2000). Sperm cells do not have cytoplasmic defense enzymes (e.g., catalase) that serve as ROS scavengers. Consequently, sperm, which have a high content of polyunsaturated fatty acids, are more susceptible to the oxidative deterioration of polyunsaturated fatty acids known as lipid peroxidation (Sharma and Agarwal 1996). Lipid peroxidation causes the plasma membrane to lose its fluidity and integrity, ultimately leading to loss of sperm function (Aitken 1995).
Loss of membrane fluidity also impairs the cell membrane ion exchange that controls sperm movement (Rao et al. 1989). Carbaryl causes lipid peroxidation at low concentrations by either efficiently lowering the intracellular level of glutathione, which is associated with an increase in ROS, or through the inhibition of excision esterases (Soderpalm-Berndes and Onfelt 1988). Thus, it is biologically plausible that exposure to carbaryl may be associated with altered semen quality, particularly sperm motility and sperm motion.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
Biomonitoring for insecticide metabolite concentrations in urine is a commonly used indicator of internal dose integrating the various routes through which the contaminant enters the body (Barr et al. 1999). However, nonpersistent insecticides are metabolized and excreted rapidly. For example, TCPY has an estimated half-life of 27 hr in humans (Nolan et al. 1984), and levels of both TCPY and 1N measured in urine reflect insecticide exposure in the previous 24–48 hr (Maroni et al. 2000). Spermatogenesis is a cyclical process that takes approximately 3 months. Although insecticide metabolite levels in urine can vary considerably over time, suggesting that a single urine sample may not be a reliable surrogate for longer-term exposure (MacIntosh et al. 1999), we recently showed that a single urine sample was predictive of the 3-month average urinary insecticide metabolite levels (Meeker et al., in press).
A single urine sample correctly classified men in the highest 3-month exposure tertile with a sensitivity (specificity) of 0.6 (0.9) for SG-adjusted 1N and 0.5 (0.8) for SG-adjusted TCPY.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
Distributions of unadjusted and creatinine-adjusted TCPY and 1N levels in the present study were compared with those recently reported for males in the National Health and Nutrition Examination Survey (NHANES) 1999–2000 (CDC 2003). Unadjusted TCPY concentrations were slightly higher in the present study, with median and 95th percentile values of 2.69 and 10.6 μg/L, respectively, compared with 1.90 and 9.9 μg/L from NHANES 1999–2000. Median and 95th percentiles for unadjusted 1N concentrations were also higher in the present study (2.86 and 13.3 μg/L, respectively, vs. 1.40 and 11.0 μg/L from NHANES 1999–2000). SG-adjusted TCPY and 1N distributions were not reported by NHANES 1999–2000 (CDC 2003).[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
In the present study, we obtained similar results using SG-adjusted or unadjusted urine metabolite levels, but our results were different for creatinine-adjusted levels. The inability to detect associations using creatinine-adjusted values may reflect tubular secretion of 1N and thus excretion rates of 1N that are independent of urine flow through the glomerulus and not directly related to the amount of creatinine that is filtered (Boeniger et al. 1993). Adjustment of 1N concentration by urinary dilution using creatinine may introduce additional nondifferential exposure measurement error, further limiting the ability to find associations between exposure and outcome.[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
Strengths of the present study include its size and high participation rate and the use of biological markers of exposure. To test the robustness of the data analysis, we used several modeling approaches in which exposures and outcomes were used as both continuous and categorical measures. The results were consistent across modeling approaches, suggesting that the data were not sensitive to the statistical analysis methods used. Study weaknesses included collecting only a single urine sample as an estimate of 3-month exposure and collecting only a single semen sample to assess semen quality. However, our earlier work supported the utility of a single urine specimen as predictive of 3-month average exposure (Meeker et al., in press).
In conclusion, associations between 1N and sperm concentration and motility were found that are consistent with animal studies of carbaryl exposure. The sperm motion parameter most strongly associated with urinary 1N was VSL, although suggestive associations of 1N with VCL and LIN were also found. There were also suggestive associations between TCPY and sperm concentration and motility, but they are difficult to interpret because there are currently limited human and animal data.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
Because most of the U.S. population is exposed to these insecticides (CDC 2003), the public health significance of an association with semen quality is potentially large. For instance, our results suggest that an interquartile range increase in carbaryl metabolite levels in urine is associated with a 4% decrease in sperm motility. Although this may not alter an individual man’s fertility, a 4% decrease in the mean of the distribution of sperm motility among U.S. men may result in a significant increase in the number of men in the lower tail of the sperm motility distribution, increasing the number of subfertile men. Further studies are needed to confirm these preliminary findings and assess the potential public health significance.[](https://www.ncbi.nlm.nih.gov/mesh/D012721)
# References
*In the **References** section:*
Adjusted ORs and 95% CIs for below-reference semen parameters by increasing quintiles of 1N for (A) sperm concentration (p-value for trend = 0.02), (B) motility (p-value for trend = 0.002), and (C) morphology (p-value for trend = 0.09). The quintiles of SG-adjusted 1N (μg/L) are as follows: Q1 (low), < LOD to 1.50; Q2, 1.50–2.67; Q3, 2.67–3.73; Q4, 3.73–5.86; Q5 (high), 5.86–159.7.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
Adjusted ORs and 95% CIs for below-reference semen parameters by increasing quintiles of TCPY for (A) sperm concentration (p-value for trend = 0.21), (B) motility (p-value for trend = 0.15), and (C) morphology (p-value for trend = 0.26). The quintiles of SG-adjusted TCPY (μg/L) are as follows: Q1 (low), < LOD to 1.45; Q2, 1.45–2.72; Q3, 2.72–3.85; Q4, 3.85–5.59; Q5 (high), 5.59–40.69.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
Distribution of insecticide (carbaryl and chlorpyrifos) metabolite levels in urine.[](https://www.ncbi.nlm.nih.gov/mesh/D012721)
Number of subjects.
LOD for 1N = 0.40 μg/L; 99.7% of samples > LOD. LOD for TCPY = 0.25 μg/L; 93.9% of samples > LOD.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
Excluded 58 samples with SG > 1.03 or < 1.01.
Excluded 23 samples with creatinine > 300 or < 30 mg/dL.[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
Demographic categories by semen parametersa (n = 330).
Information on race missing for one man and on smoking for three men.
Adjusted ORsa (95% CIs) for SG-adjusted metabolite tertiles (n = 272).b
ORs adjusted for age and abstinence time.
Excluded 58 subjects with SG > 1.03 or < 1.01.
Number of subjects in each exposure tertile with below-reference semen parameters. The semen parameter categories were not mutually exclusive; a man could contribute data to one, two, or all three of the below-reference groups.
SG-adjusted 1N tertiles: low, < LOD to 2.36 μg/L; medium, 2.36–4.02 μg/L; high, 4.02–159.7 μg/L.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
SG-adjusted TCPY tertiles: low, < LOD to 2.30 μg/L; medium, 2.30–4.42 μg/L; high, 4.42–40.7 μg/L.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
p < 0.05.
Adjusted regression coefficientsa,b for a change in semen parameters and sperm motion parameters associated with an interquartile range (IQR)c increase in SG-adjusted insecticide metabolite levels (n = 272).
Regression coefficients were adjusted for age and abstinence time.
Regression coefficients for motility, morphology, and motion parameters represent the change in semen parameter for an IQR change in insecticide metabolite concentration (0, no change in semen parameter for an IQR change in insecticide metabolite concentration; < 0, a decrease in semen parameter for an IQR change in insecticide metabolite concentration; > 0, an increase in semen parameter for an IQR change in insecticide metabolite concentration).
1N IQR = 1.80–5.02 μg/L; TCPY IQR = 1.76–5.01 μg/L.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
1N and TCPY were log transformed for regression analysis.[](https://www.ncbi.nlm.nih.gov/mesh/C029350)
Sperm concentration was log transformed. The coefficient represents a multiplicative change in sperm concentration per IQR change in TCPY or 1N (1.0, no change in sperm concentration for an IQR change in insecticide metabolite concentration; < 1.0, a multiplicative decrease in sperm concentration for an IQR change in insecticide metabolite concentration; > 1.0, a multiplicative increase in sperm concentration for an IQR change in insecticide metabolite concentration).[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
VSL, VCL, and LIN analyses not performed on 9 azoospermic men; n = 263. TCPY IQR = 1.76–5.08 μg/L; 1N IQR = 1.77–5.02 μg/L.[](https://www.ncbi.nlm.nih.gov/mesh/C012587)
p < 0.05.
|
# Introduction
Effects of apoE genotype on macrophage inflammation and heme oxygenase-1 expression
# Abstract
*In the **Abstract** section:*
In order to gain a more comprehensive understanding of the aetiology of apolipoprotein E4 genotype-cardiovascular disease (CVD) associations, the impact of the apoE genotype on the macrophage inflammatory response was examined. The murine monocyte–macrophage cell line (RAW 264.7) stably transfected to produce equal amounts of human apoE3 or apoE4 was used. Following LPS stimulation, apoE4-macrophages showed higher and lower concentrations of tumour [nec](https://www.ncbi.nlm.nih.gov/mesh/D008070)rosis factor alpha (pro-inflammatory) and interleukin 10 (anti-inflammatory), respectively, both at mRNA and protein levels. In addition, increased expression of heme oxygenase-1 (a stress-induced anti-inflammatory protein) was observed in the apoE4-cells. Furthermore, in apoE4-macrophages, an enhanced transactivation of the key redox sensitive transcription factor NF-κB was shown. Current data indicate that apoE4 macrophages have an altered inflammatory response, which may contribute to the higher CVD risk observed in apoE4 carriers.
Apolipoprotein E (apoE) is a polymorphic multifunctional protein with three common isoforms in humans (E2, E3, and E4). ApoE3 is the wild-type and most common isoform, while apoE4 carriers account for about 25% of the Caucasian population [1]. Presence of the apoE4 allele is associated with a 40–50% higher risk of cardiovascular disease (CVD) [2] and apoE4 is the major known genetic risk factor for maturity-onset Alzheimer’s disease (AD)
[3].
Although apoE4 is strongly linked to both diseases, the molecular basis of these associations remains uncertain. Traditionally, the differential risk has been attributed to the increased low density lipoprotein cholesterol (∼8%) observed in apoE4 carriers, but it is becoming increasingly evident that this alone cannot explain the disease differential [4].
ApoE is not only synthesised by the liver, but also in the brain and by resident macrophages [5] in the atherosclerotic wall, where it exerts atheroprotective actions, independent of its role in lipid metabolism [6]. These localised functions include regulation of smooth muscle cell migration and proliferation [7], inhibition of adhesion molecule expression in endothelial cells [8] and inhibition of platelet aggregation [9]. In addition to its paracrine effect on surrounding cells, apoE has also been shown to impact on macrophage function by promoting cholesterol efflux [10] and modulating NO production [11].
Inflammation and oxidative stress are key features of the pathology of atherosclerosis and AD. A limited number of studies, which have largely focussed on brain biology and neurodegeneration [12,13] have reported on the immuno-modulatory properties of apoE and its impact on inflammatory mediators [14]. However, little is known about the possible role of apoE genotype as a mediator of the macrophage innate immune and inflammatory responses in relation to CVD.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
Using a murine macrophage cell line, which has been stably transfected with human apoE3 or apoE4, we have recently showed that apoE isoform affects macrophage oxidative status [15], and now we hypothesise that this may be accompanied by an altered inflammatory response. Furthermore, the impact of genotype on the activity of the transcription factor nuclear factor κB (NF-κB), which is known to play a major role in modulating the inflammatory response, will be presented.
## Methods
*In the **Methods** section:*
Cell culture. RAW 264.7 murine macrophage cell lines, stably transfected with either human apoE3 or apoE4 were kindly given by Dr. B. Pitas (Gladstone Institute, UCSF, USA). Cells were genotyped for human apoE3 and apoE4 by the method of Hixson and Vernier [16] and apoE concentrations were measured in supernatants to ensure that secreted levels in 24 h were physiological and comparable among the two clones. Mean (SEM) levels of 1.38 (0.38) and 1.36 (0.38) μg/mg cell protein were secreted in 24 h by apoE3 and apoE4 cells, respectively, as has been previously reported [15]. Cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% foetal bovine serum, 4 mM l-glutamine, 100 U/ml penicillin and 100 μg/ml streptomycin and 100 μg/ml G-418.
Macrophages were grown in a humidified incubator at 37 °C and 5% CO2. Cells were incubated with lipopolysacharide (LPS) (Salmonella enteriditis, Sigma–Aldrich, Taufkirchen, Germany) for different time-periods depending on the outcome of interest.[](https://www.ncbi.nlm.nih.gov/mesh/D005973)
Cytokine levels. Cells were stimulated with increasing concentrations of LPS (0–10 μg/ml) for 4 h. Supernatants were collected for analysis 20 h later. Cytokine concentrations were measured using commercial ELISA kits according to the manufacturer’s instructions. Tumour necrosis factor alpha (TNFα) was determined by the Quantikine® mouse TNFα kit (R&D Systems, Wiesbaden, Germany), Interleukin (IL) 6 and 10 were measured using the Mouse Biotrak ELISA systems (Amersham Biosciences, Freiburg, Germany), and IL1β and macrophage inflammatory protein-1alpha (MIP1α) were determined with the mouse RayBio® ELISA kits from Ray Biotech (Norcross, USA).
Values were normalised for total cell protein which was determined using the BCA assay (Pierce Biotechnology, Rockford, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
Cytokine and heme oxygenase-1 (HO-1) mRNA levels. Cells were stimulated with LPS (1 μg/ml) for 1 h to determine TNFα mRNA levels and for 6 h to determine mRNA levels for IL1β, IL6, IL10, and MIP1α. The time-points were chosen on the basis of maximum mRNA expression for each cytokine in a time-course experiment (data not shown). Total RNA was isolated with the RNeasy Mini Kit (Qiagen, Hilden, Germany). One step RT-PCR was carried out using the QuantiTect®SYBR®Green RT-PCR kit (Qiagen) according to supplier instructions.
For HO-1, cells were stimulated with LPS (1 μg/ml) for 24 h and RNA was isolated by acid guanidinium thiocyanate–phenol–chloroform extraction and reverse transcription was carried out with oligo-dT primers for 1 h at 42 °C using MMLV reverse transcriptase, according to the manufacturer’s instructions (Promega, Madison, WI, USA). Real-time RT-PCR was performed with the SYBR® Green qPCR Kit (Finnzymes, Espoo, Finland). For all reactions, the Rotor Gene RG-3000 (Corbett Research) cycler was used and relative quantification of gene expression was calculated based on the 2−ΔΔCt method (β-actin or elongation factor 2 were used as housekeeping genes). Primers and cycling conditions are shown in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
Transcription factor activity. The NF-κB-secretory alkaline phosphatase (NF-κB-SEAP) (Clontech, BD Biosciences, Palo Alto, USA) reporter construct was used to measure the binding of transcription factors to the κ enhancer, and the activation of this pathway. Cells growing in 24 well plates were transiently transfected with 0.5 μg of the vector by SuperFect® transfection Reagent (Qiagen) according to the manufacturer’s protocol. Twenty-four hours later, cells were stimulated with varying concentrations of LPS (0–1 μg/ml). At 6, 12, and 24 h, the cell culture media was removed and stored for analysis.
The chemiluminescent SEAP assay (Clontech) was carried out as has been previously described [17]. Values were normalised for total cell protein determined by the BCA assay (Sigma).[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
Western blot analysis for heme oxygenase-1 (HO-1). Total cellular protein was isolated using ice-cold lysis buffer (1× PBS, 10 mM phenylmethylsulfonyl fluoride (PMSF), 10 mM leupeptine, 10 mM aprotinin and 1% Triton X-100). Samples were centrifuged (10 min, 8000g, 4 °C) and clear supernatants were collected. Twenty-five micrograms of protein was loaded on 12% SDS–PAGE gel followed by transfer to nitrocellulose membrane PROTRAN (Perkin-Elmer Life Sciences, Boston, USA). Following overnight blocking (4 °C in 5% non-fat milk), membranes were probed with polyclonal antibodies against HO-1 (Stressgen Biotech, Canada) and monoclonal antibodies against α-tubulin (both diluted 1:1000 in TTBS with 3% albumin) at room temperature for 1.5 h followed by anti-rabbit HRP-linked secondary antibodies (Cell Signalling Technology, USA) (1:10,000 in TTBS with 3% albumin) for 40 min at room temperature.
Blots were developed with SuperSignal West Pico Chemiluminescent Substrate (Pierce Biotechnology) according to the manufacturer’s instructions.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
Statistical analysis. Statistical calculations were conducted with SPSS Version 13.0. T-Tests (for independent samples) were performed to compare the outcomes between apoE3 and apoE4 cells. In the absence of normal distributed data, Mann–Whitney U-test was used. Results are expressed as means ± SEM and significance was accepted at P < 0.05.
## Results
*In the **Results** section:*
## Cytokine protein and mRNA levels
*In the **Cytokine protein and mRNA levels** section:*
Stimulation of cells with increasing concentrations of LPS (0–10 μg/ml) resulted in a dose–response accumulation of cytokines in the cell culture media of RAW 264.7-apoE3 and -apoE4 cells (Fig. 1). In the non stimulated cells, levels of cytokines were under the limit of detection. In the cases of IL1β and MIP1α (Fig. 1A and B), there was a tendency for apoE4 expressing cells to secrete higher levels of cytokines at the majority of LPS concentrations tested, although the inter-group differences did not reach statistical significance (Fig. 1A and B). In addition, no genotype mediated differences in the IL6 concentrations detected in the media were observed (Fig. 1C).
In contrast apoE4-macrophages produced 99%, 62%, 54%, and 83% more TNFα than E3 cells when stimulated at 0.1, 0.4, 0.8, and 10 μg/ml LPS (Fig. 1D). Furthermore, when IL10, an anti-inflammatory cytokine, was measured in the culture media, it was observed that apoE4-cells secreted lower concentrations of the cytokine, at all three LPS concentrations tested, with the differences reaching significance at the higher LPS concentrations (0.8 and 10 μg/ml) (Fig. 1E).[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
To examine whether the differences in cytokine accumulation in the cell culture media were associated with differences in cytokine gene expression, the mRNA levels were measured by quantitative reverse transcription PCR analysis. Gene expression profiles revealed to be comparable to the apoE genotype mediated differences in the cytokine levels in the media, with 68%, 60% and 32% higher levels of mRNA for IL1β, MIP1α, and TNFα, respectively observed in the apoE4 transfected cells (Fig. 2). No differences could be observed in the mRNA levels of IL6 between genotypes and similar to the cytokine accumulation, IL10 mRNA levels, although not statistically different were ∼15% lower in the apoE4 cells.
## NF-κB promoter activity
*In the **NF-κB promoter activity** section:*
Our results demonstrate stronger NF-κB pathway activation in the apoE4- versus the apoE3-macrophages (Fig. 3A). In non-stimulated cells, NF-κB activity was 80% higher in apoE4 cells. Following stimulation for 6 h at different LPS concentrations, NF-κB activity augmented in both cell lines. In apoE3 expressing macrophages, NF-κB activity increased ∼2.5-fold (P < 0.005) at a concentration of 0.01 μg/ml, with no further increase evident with increasing concentrations of LPS.
In E4 cells, a higher activation of NF-κB (∼4.2-fold change, P < 0.005) was evident in cells stimulated with 0.01 μg/ml LPS, with a maximum ∼5-fold increase relative to apoE3 controls evident following stimulation with 0.1 μg/ml LPS (Fig. 3A). Comparable differences between genotypes on NF-κB activity were observed at 12 and 24 h, with a time-dependent accumulation of alkaline phosphatase in the cell culture media evident after LPS (0.1 μg/ml) stimulation, with 100%, 64%, and 48% differences between genotypes evident at 6, 12, and 24 h, respectively (Fig. 3B).[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
## Heme oxygenase-1 (HO-1) protein and mRNA expression
*In the **Heme oxygenase-1 (HO-1) protein and mRNA expression** section:*
Western blotting analysis showed higher HO-1 protein concentrations in apoE4 macrophages in both untreated (controls) and LPS (1 μg/ml) treated cells (Fig. 4B). The gene expression profiles were consistent with protein levels, with control and LPS stimulated cells expressing apoE4, producing 235% and 180% higher HO-1 mRNA relative to apoE3-macrophages (Fig. 4A).[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
## Discussion
*In the **Discussion** section:*
We have previously reported a higher circulating [18] and macrophage [15] oxidative stress status associated with the E4 allele. Here, we extend these findings by reporting on an impact of apoE genotype on the inflammatory component of the innate immune response.
A murine monocyte-macrophage cell line (RAW 264.7) stably transfected to express either human apoE3 or apoE4 at similar concentrations and within the physiological range was used, so that the isoform-effects observed were apoE concentration-independent. In addition, LPS, a Toll-like receptor 4 (TLR 4) ligand that triggers cytokine expression by activation of a signalling cascade, was applied to investigate innate immune response, given that it is a commonly used approach, and that TLR4 activation is regarded to be relevant in the pathogenesis of atherosclerosis [19].[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
Following LPS stimulation, higher mRNA levels of the pro-inflammatory cytokines TNFα and, IL1β, and the chemokine, MIP1α, with a trend towards lower levels of the anti-inflammatory cytokine IL10 were evident in the apoE4 cell line. No differences were observed for IL6, which is known to act as both a pro- and anti-inflammatory mediator. Furthermore, these changes were reflected by the protein levels in the medium with the greatest differences evident for TNFα and IL10.[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
TNFα is pleiotropic and one of the most important pro-inflammatory and immuno-modulatory cytokines involved in the process of atherogenesis. For instance, TNFα participates in the recruitment and activation of inflammatory cells into the vessel wall by promoting matrix degradation [20] and by enhancing expression of adhesion molecules on endothelial cells [21]. In contrast, IL10 is a potent anti-inflammatory cytokine, which affects several signalling pathways and destabilizes the mRNA of pro-inflammatory genes, to contribute in the resolution of the inflammatory process [22].
Whether the higher levels of TNFα and IL1β observed in apoE4 macrophages were partly due to decreased levels of IL10 or due to other mechanisms cannot be concluded. The increased pro-inflammatory cytokine response observed in our apoE4 monocyte-macrophage cell model is in agreement with the limited amount of data available from other authors, which have examined associations between apoE genotype and inflammation in the brain. By using transgenic mice expressing human apoE3 or apoE4, Lynch et al. [14] determined that apoE4 mice showed elevated systemic and brain pro-inflammatory cytokines following intravenous administration of LPS, and Ophir et al. [12] demonstrated that the expression of inflammation genes was higher and more prolongated in the brains of apoE4 following intracerebroventricular injection of LPS.
Furthermore, Maezawa et al. demonstrated that the isoform-specific patterns in cytokine production (apoE4 > apoE3 > apoE2) was specific of microglia and could not be observed in astroglia cultures [13,23]. This is particularly relevant to our studies given that microglia share many functional characteristics with macrophages. In addition to the increased pro-inflammatory response, our findings reveal that apoE4 macrophages produce decreased amounts of IL10. A recent study by Tziakas et al. [24] shows, in accordance with our results, that apoE4 carriers with acute coronary syndrome and chronic stable angina patients have lower circulating levels of IL10 than the non-apoE4 patients.
Therefore taken together with the previous findings, our data support the hypothesis, that apoE4 carriers may show an “inflammatory imbalance” between pro- and anti-inflammatory mediators [24].[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
The regulation of cytokine production is highly complex, but the NF-κB signalling pathway is considered a key element, and is the most prominent and best characterised signal transduction pathway in TLR-mediated inflammation. In addition, the promoter regions of the cytokines presented in the current paper have putative κB binding sites (see http://people.bu.edu/Gilmore/nf-kb/). NF-κB is a redox sensitive transcription factor, and therefore it was hypothesised that the increased oxidative stress observed in the apoE4 macrophages could contribute to a higher cytokine production by enhancing the activation of NF-κB. Here, we demonstrate through a reporter gene assay, that the activity of NF-κB under basal conditions is higher in apoE4 than -E3 macrophages, and that upon stimulation with LPS, NF-κB activity increases in both cell lines, but the response is more augmented in the apoE4 macrophages. Again, this is in accordance with the results of Ophir et al. [12], who by means of microarray analysis, determined that the genes that were more differentially expressed between genotypes in the brain, where NF-κB-regulated and showed that NF-κB activation was more pronounced in the microglia of apoE4 versus apoE3 mice.[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
Under stress situations, the integrity of the vascular wall is maintained by several protective mechanisms, which in addition to anti-inflammatory cytokines include other proteins such as inducible heme oxygenase-1 (HO-1). HO-1 catalyses the rate limiting reaction in the degradation of heme, to yield biliverdin (with radical scavenging properties), carbon monoxide (a vascular modulator), and iron [25]. HO-1 is expressed in response to oxidative stress, and has been shown to be upregulated in atherosclerotic plaques [26] and in AD lesions
[27], and attenuate inflammation and the growth of atherosclerotic plaques in transgenic mice models [28,29]. However, so far the production and role of HO-1 has not been investigated in relation to the human apoE polymorphism. Here, for the first time, we show that HO-1 levels are significantly increased in apoE4-macrophages. ApoE4-macrophages demonstrated increased levels of HO-1 under baseline conditions, and a stronger up-regulation of HO-1 at the mRNA and protein levels following LPS application.[](https://www.ncbi.nlm.nih.gov/mesh/D001664)
Whether HO-1 exerts anti-inflammatory effects in apoE4 macrophages cannot be concluded from the current study. Although the inflammatory response was apparently aggravated in apoE4 cells, we cannot discard the possibility that this could be even more exacerbated without the up-regulation of HO-1 observed. Hence, one may consider, that induction of HO-1 by apoE4 represents a stress-induced protective response.
The current study is suggestive that an impact of apoE genotype on the monocyte inflammatory response may contribute to the higher CVD and AD risk observed in humans with an apoE4 genotype. However, further clarification of the molecular mechanisms, the complexity of apoE4–HO-1 interactions, as well as the impact of apoE genotype on inflammation using in vivo animal models and human trials is needed.
# References
*In the **References** section:*
Cytokine production in RAW 264.7-apoE3 and -apoE4 following stimulation with increasing concentrations of LPS (0–10 μg/ml) for 4 h. Supernatants were collected for ELISA analysis 20 h later. (A) IL1β, (B) MIP1α, (C) IL6, (D) TNFα, (E) IL10. Data are expressed as means ± SEM of three independent experiments performed in duplicate. ∗P < 0.05, ∗∗P < 0.01, comparing E3- vs. E4-cells at each LPS concentration.[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
Cytokine mRNA levels measured using reverse transcription real-time PCR in RAW 264.7-apoE3 and -apoE4 following stimulation with LPS (1 μg/ml) for 6 h or 1 h (TNFα). Results are calculated with the 2−ΔΔCt method and data are expressed as means ± SEM of three independent experiments performed in duplicate. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, comparing E3- vs. E4-cells at each LPS concentration.[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
(A) NF-κB activity detected with alkaline phosphatase reporter gene assay in RAW 264.7-apoE3 and -apoE4 following stimulation with increasing concentrations of LPS (0–1 μg/ml) for 6 h. (B) NF-κB activity following stimulation with LPS (0.1 μg/ml) for 6, 12, and 24 h. Results are calculated as chemiluminescence units corrected for total protein, and as fold change of apoE3 controls. Data are expressed as means ± SEM of three independent experiments performed in duplicate. ∗P < 0.05, ∗∗P < 0. 01, ∗∗∗P < 0.001 comparing E3- vs. E4-cells at each LPS concentration; (C) control non-stimulated cells.[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
(A) HO-1 mRNA levels detected with reverse transcription real-time PCR following stimulation with LPS (1 μg/ml) for 24 h. Results are calculated with the 2−ΔΔCt method and data is expressed as mean ± SEM of four independent experiments performed in duplicate. ∗P < 0.05, ∗∗∗P < 0.001, comparing E3- vs. E4-cells at each LPS concentration. (B) HO-1 levels as determined by Western blotting in relation to α-tubulin in RAW 264.7-apoE3 and -apoE4 under baseline conditions (control) and following stimulation with LPS (1 μg/ml) for 24 h.[](https://www.ncbi.nlm.nih.gov/mesh/D008070)
PCR primers and conditions
Note. F, forward primer; R, reverse primer; IL, interleukin; TNFα, tumour necrosis factor α; MIP1α, macrophage inflammatory protein 1 α; EF2, elongation factor 2; HO-1, heme oxygenase-1.
|
# Introduction
Reduction of [quaternary ammonium](https://www.ncbi.nlm.nih.gov/mesh/D000644)-induced ocular surface toxicity by emulsions: an in vivo study in rabbits
# Abstract
*In the **Abstract** section:*
Purpose
To evaluate and compare the toxicological profiles of two quaternary ammonium compounds (QAC), benzalkonium chloride (BAK), and cetalkonium chloride (CKC), in standard solut[ion or cationic emulsion form](https://www.ncbi.nlm.nih.gov/mesh/D000644)ul[ati](https://www.ncbi.nlm.nih.gov/mesh/D000644)ons[ in rabbit eyes using](https://www.ncbi.nlm.nih.gov/mesh/D001548) n[ewl](https://www.ncbi.nlm.nih.gov/mesh/D001548)y devel[oped in vivo and ex ](https://www.ncbi.nlm.nih.gov/mesh/C012817)vi[vo ](https://www.ncbi.nlm.nih.gov/mesh/C012817)experimental approaches.
Methods
Seventy eyes of 35 adult male New Zealand albino rabbits were used in this study. They were randomly divided into five groups: 50 µl of phosphate-buffered saline (PBS), PBS containing 0.02% BAK or 0.002% CKC (BAK Sol and CKC Sol, respectively), and emulsion containing 0.02% BAK or 0.002% CKC (BAK Em and CKC Em, respectively) were applied to rabbit eyes 15 times at 5-min intervals. The ocular surface changes induced by these eye drops were investigated using slit-lamp examination, flow cytometry (FCM), impression cytology (IC) on conjunctiva, and [corneal in vivo confocal ](https://www.ncbi.nlm.nih.gov/mesh/D010710)mi[cro](https://www.ncbi.nlm.nih.gov/mesh/D010710)sco[py ](https://www.ncbi.nlm.nih.gov/mesh/D010710)(IVCM). Standard i[mmu](https://www.ncbi.nlm.nih.gov/mesh/D001548)nohistology[ in](https://www.ncbi.nlm.nih.gov/mesh/C012817) c[ryo](https://www.ncbi.nlm.nih.gov/mesh/D001548)sections [was](https://www.ncbi.nlm.nih.gov/mesh/C012817) also examined for cluster of differentiation (CD) [45+](https://www.ncbi.nlm.nih.gov/mesh/D001548) infiltrati[ng ](https://www.ncbi.nlm.nih.gov/mesh/C012817)an[d t](https://www.ncbi.nlm.nih.gov/mesh/D001548)erminal [deo](https://www.ncbi.nlm.nih.gov/mesh/C012817)xynucleotidyl transferase-mediated dUTP-nick end labeling (TUNEL)+ apoptotic cells.[](https://www.ncbi.nlm.nih.gov/mesh/C027078)
Results
Clinical observations and IVCM showed that the highest toxicity was induced by BAK Sol, characterized by damaged corneal epithelium and a high level of inflammatory infiltration. BAK Em and CKC Sol presented moderate effects, and CKC Em showed the lowest toxicity with results similar to those of PBS. Conjunctival imprints analyzed by FCM showed a higher expression of RLA-DR and TNFR1 markers in BAK Sol-instilled eyes than in all other groups, especially at 4 h. Immunohistology was correlated with in vivo and ex vivo findings and confirmed this toxicity profile. A high level of infiltration of CD45+ inflammatory cells and TUNEL+ apoptotic cells was observed in limbus and conjunctiva, especially in QAC solution-receiving eyes compared to QAC emulsion-instilled eyes.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
Conclusions
The acute administration of 15 instillations at 5 min intervals was a rapid and efficient model to assess quaternary ammonium toxicity profiles. This model showed the highest toxicity, induced by the BAK solution, and the lowest level of toxicity, induced by the CKC emulsion. These in vivo and ex vivo experimental approaches demonstrated that ocular surface toxicity was reduced by using an emulsion instead of a traditional solution and that a CKC emulsion was safe for future ocular administration.[](https://www.ncbi.nlm.nih.gov/mesh/D000644)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
During the past few years, several oil-in-water emulsions have been introduced on the market for the treatment of dry eye syndrome and are used as tear substitutes or vehicles for active compounds. These oil-in-water emulsions can be visualized as tiny oil droplets suspended within an aqueous phase, and they are particularly suitable for dry eye syndromes because of the oil supplementation to the tear film. Oil-in-water emulsions can be charged positively (for example, the surface of the tiny droplets becomes cationic) by adding quaternary ammonium compounds (QACs) as cationic agents. This cationic emulsion technology presents the advantage of the electrostatic attraction between the positively charged emulsion with the negatively charged ocular surface (cornea and conjunctiva), which allows longer residence time and hence, an improved ocular bioavailability of the active compound.[](https://www.ncbi.nlm.nih.gov/mesh/D009821)
However, the use of QACs as cationic agents at relatively high concentrations in emulsion raises concerns regarding their toxicological profile. As the most commonly used preservative, benzalkonium chloride (BAK), has shown its high level of toxicity in vitro and ex vivo by stimulating epithelial cell death, acting as pro-inflammatory or pro-apoptotic mediators, inducing oxidative stress, and significantly altering the precorneal mucins. In vivo, these iatrogenic effects were most particularly found with the eye drops used for treating long-term pathologies such as glaucoma. The analysis of conjunctival epithelium using flow cytometry (FCM) showed increased human leukocyte antigen (HLA) DR class II antigens, interleukin (IL) synthesis such as IL-6, IL-8, IL-10, and the involvement of both T helper (Th)1 and TH2 systems through the overexpression of CC chemokine receptor (CCR) 5 and CCR4 in the conjunctiva of long term-treated glaucomatous patients. Moreover, BAK-induced conjunctival fibrosis is considered a relevant risk factor for glaucoma surgery failure.[](https://www.ncbi.nlm.nih.gov/mesh/D000644)
Indeed, several papers have reported that the efficacy of different preservatives including BAK is attenuated when incorporated within an oil-in-water emulsion. In emulsions, high concentrations of antimicrobial agents were needed to achieve effective preservative activity. In emulsion, BAK partitioned preferentially in the oil phase, resulting in only approximately 1.2% free BAK in solution in the aqueous phase. The antimicrobial activity and the correlated toxicity are driven by the free preservative in solution whereas the emulsion-bound preservative seems to lose its efficacy.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
BAK is composed of a mixture of alkylbenzyldimethyl ammonium chlorides bearing various alkyl chain lengths, each one with a different water solubility and water-octanol partition coefficient. For the development of cationic emulsions in ophthalmology, the use of QAC for their cationic property rather than their preservative effect is being considered. We suggested the use of lipophilic cetalkonium chloride (CKC), one of the longest alkyl-chain BAK components, as a cationic agent in ophthalmic emulsions. With a highly lipophilic QAC, the distribution between the oil and aqueous phases of the emulsion is modified because of the affinity toward the oil phase, further favoring the cationic agent role over the preservative role.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
It is important for further clinical developments to study the toxicological profiles of these new BAK or CKC-cationic emulsions and to compare them with solutions. Over the past few years, our group has developed new in vivo tools to explore the ocular surface of animal models. In vivo confocal microscopy (IVCM) offers a high definition of histological-like images that correspond very well with standard immunohistology of healthy and pathological conjunctivae or corneas. It can be used repeatedly in vivo to follow a disease course or a healing process. Moreover, the evaluation of impression cytology (IC) specimens with FCM has been widely used to detect inflammation, apoptosis, or TH1/TH2 profiles in patients, in rabbit conjunctivitis, and in rat ocular toxicity models.
Class II HLA-DR antigens and tumor necrosis factor (TNF)-related markers were thus found to be involved in toxicological or inflammatory pathways of the ocular surface.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
In this study, we combined two new investigative methods (IVCM for in vivo tissues images and IC for ex vivo epithelium inflammatory marker expression) in correlation with standard immunohistology for deep infiltration and apoptosis to assess the toxicological effects of BAK/CKC emulsion/solution formulations on the ocular surface of rabbits. We chose an experimental model described by Ichijima, consisting of 15 successive instillations in rabbit eyes at 5-min intervals. This model presents the advantages of inducing a toxic injury in a relatively short time and of emphasizing the effects of standard concentrations of compounds in which its toxicity could only be assessed over the long-term in standard instillation conditions. Our objectives were to evaluate the interest of QAC-containing emulsions compared to QAC-containing solutions, to compare BAK and CKC toxicity, and to assess the ocular safety of the newly developed CKC cationic emulsion.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
## Methods
*In the **Methods** section:*
## Animals and eye drop treatments
*In the **Animals and eye drop treatments** section:*
All experiments were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Male albino rabbits (New Zealand; two to three kilogram) were used. Before all experiments, the ocular surface integrity was examined by slip-lamp microscopy. A mixture of ketamine (35 mg/kg; Imalgène 500; Merial, Lyon, France) and xylazine (5 mg/kg; Bayer, Puteaux, France) was used to anesthetize the animals. Each group was composed of seven rabbits: five rabbits were used for clinical and IVCM observation, conjunctival imprints collection at hour (H) 4, day (D) 1, D4, and D7; two rabbits from each treatment were sacrificed for immunohistological procedures at D1, a time point chosen for the maximal inflammatory infiltration according to a preliminary seven-day study (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D007649)
In vivo confocal microscopy scale for the evaluation of ocular toxicity in the cornea, the limbus, and the conjunctiva (maximum score: 40)
We instilled 50 µl eye drops of sterile phosphate-buffered saline (PBS), 0.02% BAK solution (BAK Sol), 0.02% BAK in emulsion (BAK Em), 0.002% CKC solution (CKC Sol), or 0.002% CKC in emulsion (CKC Em) in rabbit eyes 15 times at 5 min intervals according to Ichijima et al.. All the eye drops were supplied by Novagali Pharma (Evry, France) and were sterile with physiologic pH and osmolality. We compared 0.02% BAK to 0.002% CKC since these two QAC concentrations confer equivalent positive charge to the emulsion surface (zeta potential around 20 mV), and similarly enhanced ocular delivery could be obtained with cyclosporine A (CsA)-emulsions containing 0.002% CKC or 0.02% BAK (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Clinical findings and Draize test
*In the **Clinical findings and Draize test** section:*
The first instillation was chosen as time zero (T0). During the instillations, the time when conjunctival redness appeared was recorded. At H4, D1, and D4, the eyes were examined using slit lamp microscopy for ocular irritation and scored according to a weighted scale for grading the severity of ocular lesions modified from the Draize Test. We especially evaluated the degree of redness, swelling (chemosis), and tearing of the conjunctiva; the degree and area of cornea opacity; and the increased prominence of folds and congestion of the iris. The possible maximum total score was 110 (conjunctiva=20, cornea=80, iris=10).
## In vivo confocal microscopy observation and scale
*In the **In vivo confocal microscopy observation and scale** section:*
The laser scanning IVCM Heidelberg Retina Tomograph (HRT) II/Rostock Cornea Module (RCM; Heidelberg Engineering GmbH, Heidelberg, Germany) was used to examine the entire ocular surface. The x-y position and the depth of the optical section were controlled manually; the focus position (µm) was automatically calculated by the HRT II/RCM. For all eyes, at least 10 confocal microscopic images of each layer in the conjunctiva/limbus/cornea were recorded and analyzed. The final scores were the averages of the 10 eyes of five animals.
An IVCM scale was established to quantify the ocular surface damage as presented in Table 1. Scores were obtained for five zones: the superficial epithelium, basal epithelium, and anterior stroma of the cornea, limbus, and conjunctival blood vessels. Cell morphology and nuclear aspects were evaluated, and the number of infiltrating inflammatory cells (lymphocytes, polymorphonuclear cells, or dendritic-like cells) was assessed by using the Cell Count® program (Heidelberg Engineering GmbH) associated with the HRT II/RCM. The maximal score was 40.
Microphotographs of typical clinical features. Microphotographs of typical clinical features of PBS- (A), BAK Sol- (B), BAK Em- (C), CKC Sol- (D), and CKC Em- (E) instilled rabbit eyes 4 h after repeated instillations are shown. BAK Sol induced diffuse hyperemia, chemosis, and purulent secretions on the conjunctiva. BAK Em and CKC Sol also induced mild conjunctival inflammation. CKC Em-receiving eyes presented no obvious abnormality on the conjunctiva and showed nearly the same aspect as the PBS-instilled eyes.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Conjunctival impression cytology collection
*In the **Conjunctival impression cytology collection** section:*
IC specimens were collected by techniques previously described. Two types of IC techniques were used for this study. Two nitrocellulose membranes (Millipore, Bedford, MA) were applied to the superior bulbar conjunctiva and then dipped into tubes containing 1.5 ml of cold PBS with 4% paraformaldehyde (PFA) for future cresyl violet cytology, and two Supor®-membrane (Gelman Sciences, Ann Arbor, MI) were dipped immediately after application into tubes containing 1.5 ml of cold PBS with 0.05% PFA and kept at 4 °C until FCM procedures.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Cresyl violet staining of conjunctival impression cytology and morphological evaluation
*In the **Cresyl violet staining of conjunctival impression cytology and morphological evaluation** section:*
The membranes dipped in 4% PFA were washed in distilled water, dehydrated into ethanol, and stained by cresyl violet solution (1%, number 5235, Merck, Fontenay-sous-Bois, France) for 30 min. The samples were then air-dried and mounted in a Eukitt medium (CML, Nemours, France).[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
We evaluated the morphology of the conjunctival ocular surface according to a modified Nelson’s classification, assessing the appearance of the epithelial cells (morphological changes in the cytoplasm and in the nucleus, the nucleocytoplasmic (N/C) ratio, and the metachromatic changes in the cytoplasm), inflammatory infiltration, and the density of goblet cells and subsequently assigning the grades to the ocular surface.
Draize test evaluation after PBS, BAK Sol, BAK Em, CKC Sol, and CKC Em instillations in rabbit eyes at H4 and D1. The asterisk indicates that p<0.01 compared to the PBS-instilled and CKC Em-instilled groups; the sharp (hash mark) denotes that p<0.05 compared to the BAK Sol-instilled group.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Flow cytometry analysis of rabbit impression cytology specimens
*In the **Flow cytometry analysis of rabbit impression cytology specimens** section:*
Conjunctival cells were extracted as previously described. Cells were extracted by gentle agitation and were analyzed on a flow cytometer (FC500; Beckman Coulter, Miami, FL). A direct immunofluorescence procedure was used to study the expressions of the class II antigen RLA (rabbit leukocyte antigen) DR (1:40; DakoCytomation, Clostrup, Denmark) and TNF-receptor 1 (mTNFR1, 1:40 dilution, R&D Systems, Minneapolis, MN). Mouse FITC-conjugated IgG1 (BD Biosciences PharMingen, San Diego, CA) was used as a negative control. For each antibody, a minimum of 1,000 conjunctival cells were analyzed, and the results were expressed as percentages of positive cells.
Soon after the FCM analysis, we stained the cell suspension with propidium iodide (PI 0.5 µg/ml; Sigma Chemical Company, St. Louis, MO). Immunoreactive cells were then spun down on a glass slide using a cytospin centrifuge (Shandon Cytospin 4; Thermo, Electron Corporation, Waltham, MA) and later observed and photographed under a confocal microscope (E800; PCM 2000; Nikon, Tokyo, Japan).[](https://www.ncbi.nlm.nih.gov/mesh/D011419)
## Cryosections and immunohistology
*In the **Cryosections and immunohistology** section:*
Two rabbits in each group were euthanized with a lethal dose of pentobarbital at D1. Enucleated eyes were fixed in 4% PFA and embedded. The 10 µm cryosections were incubated with antibodies directed against rabbit CD45 (1:50; CBL1412; Cymbus Biotechnology, Chandlers Ford, UK) to detect inflammatory cell infiltration. Sections were stained with secondary antibody and later with PI. To detect apoptotic cells, a terminal deoxynucleotidyl transferase-mediated dUTP-nick end labeling (TUNEL) assay (Roche Diagnostics, Meylan, France) was used.
Cryosections were first permeabilized and then incubated with an apoptosis detection kit including the 10-μl TUNEL enzyme and 90-μlL TUNEL label at 37 °C for 1 h. After three washes in PBS, the slides were stained with PI.[](https://www.ncbi.nlm.nih.gov/mesh/D010424)
Images were digitized using an Olympus BX-UCB fluorescent microscope (Olympus, Melville, NY) equipped with a DP70 Olympus digital camera and image analysis software. Positive cells to the different markers were counted in a masked manner in four different rabbit eyes in at least five areas.
## Statistical analysis
*In the **Statistical analysis** section:*
Results were expressed as means±standard error (SE). Draize and IVCM scores were compared using nonparametric comparisons (Mann–Whitney). The groups for analysis in IC expression with FCM and immunopositive cells counts were compared using factorial analysis of variance (ANOVA) followed by the Fisher’s method (Statview V; SAS Institute Inc., Cary, NC).
HRT II IVCM images of rabbit ocular surface. HRT II IVCM images of rabbit ocular surface after PBS (A), BAK Sol (B), BAK Em (C), CKC Sol (D), and CKC Em (E) instillations at D1 are displayed. Results are shown in the superficial epithelium (line 1), basal epithelium (line 2: 10–15 μm from the superficial epithelium layer), anterior stroma (line 3: 50–65 μm from the superficial epithelium layer), and conjunctival substantia propria (line 4: 60–90 μm from the superficial epithelium layer). BAK Sol-receiving eyes showed the greatest damage in the epithelium and the greatest inflammatory infiltration in the basal epithelium and anterior corneal stroma. BAK Em and CKC Sol induced intermediate toxicity. These three groups induced inflammatory cells rolling in conjunctival blood vessels. CKC Em presented almost the same aspects in all ocular surface structures as the PBS-instilled group. The scale bar indicates 100 μm.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Results
*In the **Results** section:*
## Clinical findings
*In the **Clinical findings** section:*
Four hours after the first instillation (i.e., 2.75 h after the previous one), BAK Sol induced diffuse hyperemia, chemosis, and purulent secretions on the conjunctiva when compared with the PBS-instilled eye. BAK Em and CKC Sol also induced mild conjunctival hyperemia but less than what was induced by BAK Sol with no obvious chemosis or purulent secretion (Figure 1). CKC Em-receiving eyes (Figure 1E) presented no redness, chemosis, or secretions on the conjunctiva and showed nearly the same aspect as the PBS-instilled eyes.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
PBS did not induce any redness during the instillation period. BAK Sol induced conjunctival redness very quickly 13±1.07 min after the first instillation (p<0.0001 compared to BAK Em and CKC Em). BAK Em and CKC Sol groups started to show visible redness at 34±2.08 min and 23±3.50 min, respectively, with no significant difference between the two groups. CKC Em induced a slight redness close to the end of the experiment 60±4.47 min after the first instillation (p<0.0001 compared to all other groups, except PBS).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Draize test
*In the **Draize test** section:*
At H4, the BAK Sol-, BAK Em-, and CKC Sol-instilled groups presented higher Draize Test scores than the PBS-instilled group (p<0.01 for the three groups; Figure 2). CKC Em presented no difference with the PBS group (p>0.05). The ocular toxicity score was the highest in the BAK Sol group (5±0.4), which had higher scores than the BAK Em (2±0.4), CKC Sol (3±0.6), and CKC Em groups (0.4±0.3; p<0.05 for the three groups). The BAK Em- and CKC Sol-instilled groups also showed higher ocular toxicity than the CKC Em-instilled group (p<0.01 for the two groups).[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
At D1, the PBS, BAK Em, CKC Sol, and CKC Em eyes all returned to normal aspects without significant differences among them. BAK Sol still induced substantial ocular abnormalities (p<0.01 compared to the PBS and CKC Em groups, p<0.05 compared to the BAK Em and CKC Sol groups). Only at D4 did the BAK Sol-instilled eyes return to a normal ocular aspect (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
In vivo confocal microscopy scores in the five tested groups. The toxicity of CKC Em was less than that of BAK Sol, BAK Em, and CKC Sol with no significant differences with the PBS-instilled groups at all time points (H4, D1, and D4). BAK Sol presented the highest IVCM toxic score at H4 on D1 with intermediate results for BAK Em and CKC Sol. The asterisk indicates that p<0.01 compared to PBS and CKC Em; the sharp (hash mark) denotes that p<0.05 compared to BAK Sol; and the filled diamond symbol indicates that p<0.05 compared to CKC Sol.[](https://www.ncbi.nlm.nih.gov/mesh/C012817)
## In vivo images of rabbit ocular surface after instillations
*In the **In vivo images of rabbit ocular surface after instillations** section:*
Figure 3 shows the IVCM images of the rabbit corneal epithelium (line 1), basal epithelium (line 2), anterior stroma (line 3), and conjunctival stroma (line 4) after application of PBS (Figure 3A), BAK Sol (Figure 3B), BAK Em (Figure 3C), CKC Sol (Figure 3D), and CKC Em (Figure 3E) at D1.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Surface epithelium
*In the **Surface epithelium** section:*
PBS-instilled rabbits presented almost a normal corneal epithelium (Figure 3A) with a regular polygonal mosaic appearance and brightly reflective nuclei. No obvious desquamation, swelling of epithelium, or inflammation was detected. BAK Sol (Figure 3B) induced partial desquamation of epithelial cells. The cells presented an irregular shape with abnormal reflectivity patterns and swelling cells, observed as a loss of cell borders. Inflammatory infiltrates were also found.
Compared to the BAK Sol instillation, fewer abnormalities were observed for BAK Em (Figure 3C) and CKC Sol (Figure 3D) instillation with partial desquamation of epithelial cells and irregular cell shapes. The CKC Em group (Figure 3E) showed almost the same epithelial aspects as did PBS-instilled rabbits without obvious epithelium abnormality or inflammatory infiltration.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Basal epithelium
*In the **Basal epithelium** section:*
PBS (Figure 3A) and CKC Em (Figure 3E) induced no obvious inflammation in this layer whereas BAK Sol (Figure 3B) induced the greatest infiltration (129±13.29 inflammatory cells/mm2, p<0.001 compared to all other groups). These bright hyperreflective inflammatory infiltrates were also found at a moderate level in the BAK Em-instilled eyes (Figure 3C; 55±6.00 inflammatory cells/mm2, p<0.001 compared to CKC Em) and in the CKC Sol-instilled eyes (Figure 3D; 55±11.12 inflammatory cells/mm2, p<0.001 compared to CKC Em).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Anterior stroma
*In the **Anterior stroma** section:*
One day after instillation of BAK Sol (Figure 3B), slight inflammatory infiltration and slight disorganization of the anterior stroma was recorded by IVCM. No abnormality was observed in the anterior stroma in the other groups (Figure 3A,C–E).[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
Conjunctival impression cytology stained by cresyl violet at D1. PBS (A) presented normal aspects of the conjunctival epithelium with no inflammatory infiltration. BAK Sol (B) induced numerous polymorphonuclear inflammatory cells with almost no normally shaped epithelial cell visible. BAK Em (C) and CKC Sol (D) both showed epithelial damage with inflammatory infiltration. CKC Em-instilled (E) rabbit eyes presented normal epithelial patterns without inflammatory infiltration. (original size 40×).[](https://www.ncbi.nlm.nih.gov/mesh/C028911)
## Posterior stroma and endothelium
*In the **Posterior stroma and endothelium** section:*
At all the observation times, no abnormality was observed in any group (data not shown).
## Limbus
*In the **Limbus** section:*
Minimal inflammatory cells were observed after PBS and CKC Em instillations. In the BAK Sol-instilled group, we observed that the inflammatory infiltrations in the peripheral cornea and limbus area were more abundant than in all the other groups (data not shown). We also observed the presence of capillary buds from limbal vessels at this time. Moderate inflammatory infiltration was also observed in the BAK Em and CKC Sol groups.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Conjunctiva
*In the **Conjunctiva** section:*
Blood vessels in hyperreflective conjunctiva were observed by IVCM. The PBS-instilled rabbit presented normal conjunctival blood vessel aspects with no rolling inflammatory cells (Figure 3A). After the applications of BAK Sol (Figure 3B), BAK Em (Figure 3C), and CKC Sol (Figure 3D), inflammatory cells rolling along vascular walls were consistently recorded in blood vessels. In contrast, CKC Em-instilled eyes (Figure 3E) presented almost normal blood vessel aspects as did PBS-instilled rabbits with no obvious rolling cells. At H4 (images not shown), IVCM showed the same tendency of toxic ranking with BAK Sol inducing the worst aspect of epithelium and the greatest inflammatory infiltration; BAK Em and CKC Sol induced moderate abnormalities in cornea, limbus, and conjunctiva; and the CKC Em group showed almost the same images as did the PBS-instilled group.
At D4 (images not shown), the abnormalities found in the ocular surface decreased in all groups. BAK Sol, BAK Em, and CKC Sol groups still presented abnormal aspects in limbus and conjunctival blood vessels. These slight abnormalities disappeared at D7 after instillations (data not shown). According to the IVCM observations, the CKC Em-instilled eyes presented no difference in the ocular surface compared to PBS-instilled eyes from H4 to the end of experiment.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## In vivo confocal microscopy scale evaluation
*In the **In vivo confocal microscopy scale evaluation** section:*
An IVCM scoring system was used to quantify toxic patterns. As shown in Figure 4, at H4 and D1, BAK Sol induced the highest IVCM score compared to PBS, CKC Em (p<0.01 compared to the two groups), BAK Em, and CKC Sol (p<0.05 compared to the two groups). At these time points, the BAK Em- and CKC Sol-instilled groups showed higher IVCM scores than did the PBS- and CKC Em-instilled groups (p<0.01 for the two groups). BAK Em eyes presented lower IVCM scores than did CKC Sol eyes (p<0.05) at H4 and D1. At H4, D1, and D4, the CKC Em-instilled group always presented similar IVCM scores to the PBS-instilled group with no statistical differences.
At D4, scores decreased for every treatment except BAK Sol; BAK Em and CKC Sol still presented higher IVCM scores than the PBS (p<0.01 for the three groups) and CKC Em (p<0.01 for the three groups) groups. At D7, the IVCM scores of BAK Sol, BAK Em, and CKC Sol eyes returned to normal aspects (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
## Impression cytology staining
*In the **Impression cytology staining** section:*
Impression cytology of conjunctival surface after PBS instillation at D1 showed normal, polyedric conjunctival epithelial cells (Figure 5A) with a prominent nucleus. The nuclear cytoplasmic ratio ranged from 1/2 to 1/3. There were no inflammatory infiltrating cells. The goblet cells were clearly visible among or beside the epithelial cells. In the BAK Sol-instilled group (Figure 5B), the conjunctival epithelium was barely recognized because of the very intense infiltration of polymorphonuclear cells.
Goblet cells completely disappeared after BAK Sol treatment at this time. The lesions observed in conjunctival IC after cresyl violet staining are summarized in Table 2. BAK Em (Figure 5C) and CKC Sol (Figure 5D) induced moderate toxicity in conjunctival epithelium, which showed aspects of anisocytosis and anisonucleosis. Inflammatory infiltration was observed, and the density of goblet cells decreased after these two treatments. The CKC Em eyes (Figure 5E) presented a nearly normal conjunctival epithelium aspect with no obvious inflammatory infiltration.
Goblet cells were clearly present with no morphological abnormalities.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
Description of cresyl violet conjunctival impression cytology results at D1 following repeated applications of quaternary ammonium compound emulsions and solutions[](https://www.ncbi.nlm.nih.gov/mesh/C028911)
## Expression of rabbit leukocyte antigen (RLA) DR and tumor necrosis factor receptor type 1 (TNFR1) evaluated by flow cytometry
*In the **Expression of rabbit leukocyte antigen (RLA) DR and tumor necrosis factor receptor type 1 (TNFR1) evaluated by flow cytometry** section:*
The baseline range of RLA DR- and TNFR1-positive cells in IC specimens from normal rabbits was approximately 3% - 6%. At H4, BAK Sol induced 65.4%±4.7% of cells positive for RLA DR (Figure 6A). These were mostly inflammatory cells (Figure 6B) as viewed after cytospin centrifugation. The RLA DR expression in the other groups was much lower than that of BAK Sol group with statistical significance (p<0.0001 compared to PBS, BAK Em, CKC Sol, and CKC Em groups). This high level of expression decreased one day after instillation; BAK Sol still induced 29.6%±3.7% of RLA DR-positive cells with significant differences with the other groups that showed normal levels (p<0.05 compared to the four other groups).
At D4, RLA DR-positive cells returned to the normal level (approximately 5%) after BAK Sol instillation.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
Impression cytology evaluated by flow cytometry and viewed after cytospin centrifugation. Percentages of RLA-DR- (A) and TNFR1- (C) positive cells after multiple instillations of PBS, BAK Sol, BAK Em, CKC Sol, CKC Em is displayed. The astersisk indicates that p<0.0001 compared with PBS, BAK Em, CKC Sol, and CKC Em-instilled groups, and the doube asterisk means that p<0.05 compared with PBS, BAK Em, CKC Sol, and CKC Em-instilled groups. Positive cells for RLA-DR (green, B) and TNFR1 (green, D) were viewed after propidium iodide staining (red) and cytospin centrifugation. The scale bar indicates 100 μm.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
TNFR1 was also found at high levels at H4 after the instillation of BAK Sol (Figure 6C; 44.3%±5.9% of positive cells; p<0.0001 compared to all other groups). The cells positive to TNFR1 were mostly inflammatory cells as well as some typical conjunctival epithelial cells (Figure 6D). At D1, this strong expression of TNFR1-positive cells decreased to 18.2%±7.1%, which was still much higher than in the other groups (p<0.05 compared to all other groups). At D4, this expression returned to about 4% with no difference with all other groups.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
## Immunostaining of CD45 and terminal deoxynucleotidyl transferase-mediated dUTP-nick end labeling markers in cryosections
*In the **Immunostaining of CD45 and terminal deoxynucleotidyl transferase-mediated dUTP-nick end labeling markers in cryosections** section:*
Immunostaining of CD45+ inflammatory cells (line 1 for limbus, line 2 for conjunctiva) and TUNEL+ apoptotic cells (line 3 for limbus, line 4 for conjunctiva) at D1 are shown in Figure 7. The immunopositive cell counts are presented in Figure 8. PBS-instilled rabbits only showed a few CD45+ inflammatory cells in the limbus (Figure 7A; 180±43 cells/mm2) and conjunctiva (Figure 7A; 431±63 cells/mm2) zones. Immunohistology clearly showed that BAK Sol instillation induced numerous CD45+ inflammatory cells infiltrating the limbus (Figure 7B) and conjunctiva (Figure 7B) zones, 1160±134 cells/mm2 and 1290±139 cells/mm2, respectively (Figure 8A, p<0.005 compared to the PBS and CKC Em groups). These inflammatory cells were especially located in the limbal or conjunctival stroma but also beneath the epithelial layers.
BAK Em presented moderate inflammatory infiltration, 900±121 cells/mm2 in the limbus (Figure 7C) and 860±34 cells/mm2 in the conjunctiva (Figure 7C; p<0.005 compared to PBS; and p<0.05 compared to CKC Em). BAK Em induced significantly fewer CD45+ cells than did BAK Sol (p<0.005). CKC Sol also presented moderate inflammatory infiltration (Figure 7D for limbus with 790±59 cells/mm2; Figure 7D for conjunctiva with 890±60 cells/mm2; p<0.005 compared to PBS and p<0.05 compared to CKC Em) with no difference with BAK Em treatment. After CKC Em instillation, occasional inflammatory cells were found in the limbal zone (Figure 7E, 170±40 cells/mm2) and beneath the conjunctival epithelium (Figure 7E, 460±34 cells/mm2) with no difference with the PBS group. In corneal tissue, very slight CD45+ expression was found in the BAK Sol-instilled group, and no other treatments induced inflammatory cells in the cornea (data not shown).
Few apoptotic cells were observed in the limbal zone (Figure 7A, 110±38 cells/mm2) and in the conjunctiva (Figure 7A, 280±53 cells/mm2) after instillation of PBS. When BAK Sol was instilled in the ocular surface of rabbits, numerous apoptotic cells were found in the limbal zone (Figure 7B and Figure 8B, 710±82 cells/mm2) and conjunctiva (Figure 7B, 820±80 cells/mm2; p<0.005 compared to the PBS and CKC Em groups). Compared to BAK Sol, the BAK Em treatment induced fewer apoptotic cells in the limbal zone (Figure 7C, 370±47 cells/mm2, p<0.005 compared to the PBS and BAK Sol groups, p<0.05 compared to the CKC Em group) and in the conjunctiva (Figure 7C, 390±43 cells/mm2; p<0.005 compared to BAK Sol with no differences with PBS or CKC Em-instilled groups). Apoptosis was found in the limbal zone after CKC Sol application (Figure 7D, 440±69 cells/mm2); the conjunctival stroma also contained numerous apoptotic cells in subepithelial and deeper zones (Figure 7D, 590±41 cells/mm2; p<0.005 compared to the PBS and BAK Sol groups; p<0.05 compared to the CKC Em group). CKC
Em induced fewer apoptotic cells in the limbal (Figure 7E, 190±35 cells/mm2) and conjunctival zones (Figure 7E, 250±48 cells/mm2) with no difference compared to PBS. In corneal tissue, few apoptotic cells were found in the superior epithelium of the cornea after BAK Sol and CKC Sol instillations. In other treatments, no apoptotic cells were found in the corneal layers (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
Immunostaining of inflammatory (CD45) and apoptotic (TUNEL) markers in cryosections. Immunostaining of inflammatory (CD45 in green: line 1 limbus, line 2 conjunctiva) and apoptotic (TUNEL in green: line 3 limbus, line 4 conjunctiva) markers in cryosections of rabbit eyes instilled with PBS (A), BAK Sol (B), BAK Em (C), CKC Sol (D), and CKC Em (E) at D1 is shown. Nuclei were stained in red with propidium iodide. The scale bars indicate 100 μm. Immunohistology clearly showed that BAK Sol instillation induced numerous CD45+ inflammatory cells and TUNEL+ apoptotic cells infiltrating the limbus and conjunctiva zones. BAK Em and CKC Sol also induced moderate inflammatory/apoptotic cells. After CKC Em or PBS instillation, occasional inflammatory/apoptotic cells were found.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Discussion
*In the **Discussion** section:*
Lipid emulsions are known to improve the tolerance of topically applied ophthalmic drugs. Amphotericin B emulsion was found to be better tolerated than the commercial solution of Fungizone® in rabbits. Several clinical studies of emulsion-based eye drops of artificial tears or cyclosporine showed good overall safety, efficacy, and comfort in normal subjects and dry eye patients. Compared to other ophthalmic vectors, cationic vehicles have a better spreading capacity and improve ocular bioavailability. Formulation of cyclosporine A in a cationic emulsion results in significantly improved (11 fold) corneal and conjunctival delivery in rabbits compared to olive oil.
Compared with the anionic emulsion, indomethacin in cationic emulsion provided significantly higher drug levels in the aqueous humor and sclera or retina, and submicron cationic lipid emulsion enhanced the ocular bioavailability of cyclosporine A. The main mechanism involved is that cationic emulsions interact with the negatively charged ocular surface such as the cell membranes of the conjunctival and corneal epithelia. In an effort to optimize cationic emulsions, we proposed the use of QAC to provide the cationic charge. BAK cationic emulsion of cyclosporine significantly improved the penetration of cyclosporine over a negatively charged emulsion. The use of lipophilic QAC, such as CKC rather than BAK, allows obtaining a positive charge with a lower amount of QAC because of its optimal oil/aqueous interface distribution.
In the present study, we studied the toxicity of BAK and CKC cationic emulsions compared with their respective BAK and CKC solutions, and we observed a reduction of QAC toxicity when it was incorporated into the emulsion.[](https://www.ncbi.nlm.nih.gov/mesh/D000666)
The model consisting of repeated applications in rabbit eyes (15 times at 5 min intervals) does not reflect the real ocular surface reactions in patients, but it may emphasize the action of low toxic compounds and mimic repeated administrations. As the repetition of instillations in a short period of time causes drastic stimulation of the ocular surface, this model is useful for between-drug comparisons and testing a specific compound’s absence of toxic effects, which is a good indicator of further absence of ocular toxicity in a more conventional use over the long-term. This model combined rapidity and efficiency in comparing the toxicity of several products without modifying their concentration and/or composition. It was therefore used in the past to distinguish the toxicity ranking after applications of 0.02%, 0.01%, and 0.005% BAK and clearly showed the different levels of epithelial deterioration. In our study, this model also clearly demonstrated the differences between emulsion and solution formulations.
Pertinent and reliable animal models are in great need for testing new formulations and new preservatives. To simulate long-term toxicity of QAC at reasonable time intervals and over a short period of time, previous studies have used high concentrations (50 fold to 500 fold the commercial concentrations) or repeated instillations over a long period of time to detect toxic effects, which may be complex in animal models. A short duration of testing, such as 14 or 28 days as used in standard toxicological evaluations, at commercial concentrations in young healthy animals may not reflect the clinical use when the drug is administered over the long-term sometimes in association with other drugs or preservative-containing eye drops or in patients with preexisting or concomitant ocular surface impairment. This point raises the problem of validating reliable tests for toxicological purposes and may explain why clinical trials often fail to show mild or subclinical toxic effects that are observed in patients treated for long periods of time, such as in glaucomatous patients in whom the ocular surface has widely shown inflammatory changes and clinical impairment.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
Compared to the previous study by Ichijima [29], mainly describing the effects on the superficial corneal epithelium, we developed new in vivo tools and observed the entire ocular surface, the cornea, the limbus, and the conjunctiva, at various depths. Moreover, we studied inflammatory markers in conjunctival imprints and confirmed this analysis by using standard immunohistology after animal sacrifice. This approach provides much broader and more reliable information than clinical assessment alone as in the standard Draize test or even in investigations at the corneal level alone as conjunctival inflammation may not be assessed by corneal-based examinations. Moreover, this complementary set of investigations decreases the need for large animal series as IVCM or conjunctival IC may be used repeatedly for monitoring drug toxicity or efficacy.
Counts of positive cells for CD45 and TUNEL markers. Counts of positive cells for CD45 (A) and TUNEL (B) after applications of PBS, BAK Sol, BAK Em, CKC Sol, or CKC Em at D1 are displayed in this graph. The asterisk means that p<0.005 compared to PBS; the sharp (hash mark) indicates that p<0.005 compared to BAK Sol; and the double “S” symbol (§) denotes that p<0.05 compared to CKC Em. A high level infiltration of CD45+ inflammatory cells and TUNEL+ apoptotic cells was found in the limbus and conjunctiva, especially in quaternary ammonium compounds solution-receiving eyes compared to quaternary ammonium compounds emulsion-instilled eyes.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
The toxicity ranking of the four eye drops tested in our study was distributed as follows BAK Sol > BAK Em ≈CKC Sol > CKC Em, this latter formulation being almost nontoxic in our experimental conditions. Interestingly, when formulated in an emulsion, both BAK and CKC presented reduced toxicity compared to the same concentration in solutions. This reduction of toxicity of BAK/CKC in emulsion is attributed to the distribution of the QAC within the oil phase, leading to a low concentration of QAC in the aqueous phase.
One of the proposed mechanisms of QAC for antimicrobial activity is their intercalation within the bacterial membrane; in emulsion, they behave in a similar way by distributing at the emulsion droplet surface. We therefore suggest that the QAC that is bound to the emulsion surface is not available for binding with bacterial membranes and that only the remaining free QAC is responsible for the preservative effect and the toxicity. In the formulation tested in the present study, eventual toxicity was reduced not only by the use of cationic emulsion but also by selecting a highly lipophilic QAC such as CKC to improve the compound partition within the emulsion. The lipophilicity of CKC partition is almost 100-fold more important in octanol than in water, which is four to five times higher than BAK. Compared to BAK, CKC distributed even more preferentially into the oil phase (allowing reduced cationic agent content) with a lesser amount dissolved in the water phase and free for possible toxic damage.
Moreover, QAC could enhance membrane fluidity, and CKC was found more efficient than BAK. In our experiments, CKC Em exhibited almost the same aspect as the PBS-instilled group and did not induce any obvious ocular toxicity during all the observation times. Consequently, CKC-associated emulsions present the advantages of cationic emulsions and reduced free QAC and demonstrated no obvious toxicity. Pharmaceutical companies require new in vivo and in vitro tools to test the possible toxicity of their newly manufactured drugs. In vivo tests present the advantages of mimicking the real ocular environment, especially regarding the composition of the lachrymal film and drug metabolism in ocular tissues.
At the same time, animal experimentation guidelines require refining the tests to reduce the number of animals used. In addition to demonstrating the tolerance of emulsion-containing eye drops, this study has also proved that it is possible to refine the classical scoring elaborated by Draize test, based on clinical evaluations in 1944. However, as the gold standard in ocular toxicology, it lacks precise and objective criteria, especially at cellular levels. The Draize test is therefore more of a good standard for eliminating truly toxic drugs in a screening approach than in a predictive evaluation of the real use of eye drops in further clinical development. We developed both the IVCM and flow cytometry on IC for our toxicological models.
Used in combination, each of these techniques was able to detect and analyze the microstructures of the animal’s ocular surface (cornea, limbus, and conjunctiva) as well as the surface markers expressed by conjunctival epithelium in toxic conditions. Based upon histological precise patterns of the ocular surface for three-dimensional visualization, IVCM offers the advantage of examining the same animal in vivo during experimental procedures and could not be replaced with any other standardized method except histology after sacrifice, which would have required a much higher number of animals to provide successive time points. Moreover, specifically in the rabbit model, IVCM also allows us to explore the limbus and conjunctiva blood vessels that are difficult to see in small eyes such as rat eyes. The IVCM scale could also provide a classification of global toxicity in the cornea, limbus, and conjunctiva. This scale system could be used as an effective tool for evaluating ocular surface toxicity in a standardized way in many laboratories worldwide.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
Conjunctival IC also usefully completed the information provided by IVCM. IC specimens were evaluated using a standard cytology method to identify the different cell populations in the conjunctival epithelium and then using FCM to assess the levels of inflammation or apoptosis. Thus, IC specimens were also efficient ex vivo tools for evaluating and comparing different toxic agents. As for human eyes, rabbit conjunctival imprints collect enough cells to ensure reliable results obtained from FCM. In our study, RLA-DR and TNFR1 were two inflammatory markers that were found to be significantly increased after toxic injuries induced by BAK Sol, confirming local inflammatory infiltration even one day after instillation.[](https://www.ncbi.nlm.nih.gov/mesh/D001548)
In this study, we proposed a set of tools for exploring a drug’s toxicity in all components of the ocular surface, and this set is more pertinent, complete, and reliable than if those tools were used individually as in former studies. We also showed that the use of long-chain CKC-cationic emulsions should be further developed in eye drops because of their reduced toxicity, the improvement in drug ocular delivery, and finally for the comfort brought to patients.[](https://www.ncbi.nlm.nih.gov/mesh/C012817)
# References
*In the **References** section:*
|
# Introduction
β1-6 branching of cell surface glycoproteins may contribute to uveal melanoma progression by up-regulating cell motility
# Abstract
*In the **Abstract** section:*
Purpose
This study investigated the influence of integrin expression as well as the oligosaccharide structure of surface N-glycoproteins on cell behavior of two primary uveal (92–1 and Mel202) and two primary cuta[neous (FM55P an](https://www.ncbi.nlm.nih.gov/mesh/D009844)d IGR-39) melanoma cell lines.
Methods
Cell adhesion to fibronectin and cell migration on fibronectin (wound healing) were selected as the studied cell behavior parameters. The percentage of cells positive for expression of selected integrins was estimated by flow cytometric analysis. The influence of β1–6 branched complex-type N-oligosaccharides on wound healing on fibronectin was investigated. Cell surface β1–6 branched N-oligosaccharides were measured by their specific binding to PHA-L followed by flow cytometry, and the fibronectin receptors bearing β1–6 GlcNAc branched N-linked glycans were identified. In addition, the transcript of GnT-V (the enzyme that catalyzes the addition[ of N-acetylglucosamine to the core mannose o](https://www.ncbi.nlm.nih.gov/mesh/D009844)f di- and tri-antennary N-glycans through a β1–6 linkage) was an[alyzed by semiquantitative RT–PC](https://www.ncbi.nlm.nih.gov/mesh/D009844)R.[](https://www.ncbi.nlm.nih.gov/mesh/D000117)
Results
Unlike the two examined cutaneous melanoma cell lines, neither of the uveal melanoma cells adhered to fibronectin. The adhesion efficiency of IGR-39 cells was twice that of FM55P cells. In contrast, uveal melanoma cells repaired scratch wounds on fibronectin-coated surfaces twice as fast as cutaneous melanoma cells did. The expression of α3β1, α4β1, α5β1, and αvβ3 integrins, acting as fibronectin receptors, differed between the tested cell lines, and no distinct pattern distinguished uveal melanoma from cutaneous melanoma except for high expression of α4β1 integrin on both FM55P and IGR-39 cells. The results also demonstrated that the high levels of α3β1, α4β1, and α5β1 integrin expression on IGR-39 cells promoted their strong attachment to fibronectin-coated surfaces. In addition, 92–1, Mel202, and FM55P cells showed no or low adhesion to fibronectin, perhaps the result of low expression of fibronectin receptors excluding high expression of α4β1 integrin in FM55P cells. Cell migration was significantly decreased in three out of four PHA-L-treated cell lines, suggesting that β1–6 branched complex type N-oligosaccharides are critical for 92–1, Mel202, and FM55P cell motility. Semiquantitative RT–PCR analysis showed that the tested cells did not differ in mRNA levels of β1–6 –N-acetylglucosaminyltransferase V. However, FACS analysis showed that 92–1, Mel202 and IGR-39 cells expressed significantly higher amounts of β1–6 branched N-oligosaccharides on the cell surface than FM55P cells did. All examined α3, α5, αv, and β1 integrin subunits were shown to bear β1–6 branched N-linked glycans.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
Conclusions
The role of integrins and their N-glycosylation in the regulation of uveal melanoma growth and progression is largely unknown. These results reveal that cell surface complex-type N-glycans with GlcNAc β1–6 branches are important factors determining the migration of primary uveal melanoma cells on fibronectin.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Uveal melanoma (UM) is the most common primary intraocular tumor in the adult population. Although significant advances have been made in the ability to diagnose and treat primary tumors of UM, the mortality rates for UM have changed little during the past few decades. Because metastatic disease can appear as late as 12 to 15 years after enucleation of the ocular melanoma, the disease is likely to have already disseminated at the time of diagnosis, either as circulating malignant cells or as occult micrometastatic lesions. UMs metastasize preferentially to the liver, and the average survival period for UM is less than one year after clinical diagnosis of the lesions.
Little is known about the molecular mechanisms underlying the metastatic potential of UM. Although UMs and cutaneous melanomas (CM) are of similar embryological origin and as such share several common features including morphology and the properties of the melanogenesis pathway, they differ significantly in their epidemiological, cytogenetic, and immunological characteristics as well as biologic behavior. UMs almost exclusively metastasizes through the blood and preferentially to the liver, whereas CMs are capable of both lymphatic and hematogenous (often less organ-specific) spread to almost any organ in the body. Early events in the formation of metastases include the escape of tumor cells from the primary tumor site followed by invasion into the surrounding stroma; these events depend on the interaction of tumor cells with the extracellular matrix (ECM). Recently we showed that primary UMs and primary CMs also differ in their adhesion to selected components of the ECM, including fibronectin (FN), laminin, and type IV collagen.
Successful metastasis requires an ordered series of steps, including detachment of tumor cells from the primary neoplasm, invasion into and migration through ECM, entry into blood as well as lymph vessels, transport along the circulatory system, adhesion to the endothelium, extravasation, and outgrowth in a distant organ. Receptors that mediate cell-cell and cell-ECM adhesion have been shown to be key components in the metastatic cascade. It is reported that during the progression of malignant CM, the expression profiles of several adhesion molecules undergo changes which are directly or inversely correlated with its metastatic potential. Loss, overexpression or malfunctioning of adhesion molecules may contribute to the detachment of tumor cells from the primary tumor, local invasion and metastasis.
The aims of this study were (1) to compare two human primary UM cell lines (92–1, Mel202) and two human primary CM cell lines (FM55P, IGR-39) in terms of their adhesion and migration (wound healing) to FN; (2) to determine the repertoire of integrins acting as FN receptors on these melanoma cells; (3) to test whether the oligosaccharides of surface N-glycoproteins influence melanoma cell behavior; (4) to measure the surface expression of β1–6 branched N-oligosaccharides on melanoma cells; and (5) to identify the FN receptors bearing β1–6 branched N-oligosaccharides.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
## Methods
*In the **Methods** section:*
## Chemicals
*In the **Chemicals** section:*
Phaseolus vulgaris agglutinin (PHA-L), PHA-L immobilized on cross-linked 4% agarose, proteinase inhibitor cocktail, protamine sulfate, penicillin-streptomycin solution, Tween 20, bovine serum albumin (BSA), and poly-L-lysine were obtained from Sigma-Aldrich (St. Louis, MO). Fetal bovine serum (FBS) was from GibcoBRLTM (Paisley, UK). The polyvinylidene difluoride (PVDF) membranes were products of Millipore (Bedford, MA, USA). We purchased 4-nitroblue-tetrazolium salt and 5-bromo-4-chloro-3-indolylphosphate solution from Roche Diagnostics GmbH (Mannheim, Germany) and obtained 96-well and 24-well plates coated with FN from BD Biosciences (San Diego, CA). Antibodies used in flow cytometric analysis as well as for immunodetection are listed in Table 1 and Table 2, respectively.
All remaining chemicals were of analytical grade, commercially available.[](https://www.ncbi.nlm.nih.gov/mesh/D012685)
Range, specificity, and supplier of monoclonal antibodies used for flow cytometry experiments.
## Cell lines and culture condition
*In the **Cell lines and culture condition** section:*
Included in the study were four cell lines received from the ESTDAB Melanoma Cell Bank (Tűbingen, Germany). Two cell lines were derived from a primary UM (92–1 and Mel202) and two others were derived from a primary CM (FM55P and IGR-39).
All cell lines were cultured in RPMI-1640 medium (GibcoBRLTM), supplemented with 10% heat-inactivated FBS, and penicillin-streptomycin solution. Cells were fed biweekly and grown to confluence as a monolayer in 5% CO2-enriched atmosphere at 37 °C in a humidified incubator, and passaged by treatment with 0.05% trypsin-EDTA solution (Sigma-Aldrich). Experiments were initiated when cells had reached subconfluence.[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Adhesion assay
*In the **Adhesion assay** section:*
Melanoma cell adhesion to FN was assessed according to the protocol previously described. The reference value for 100% attachment was estimated from cells in wells coated with 500 μg/ml poly-L-lysine. All data are given as relative percentages of adhesion compared to adhesion on poly-L-lysine (taken as 100%). Values are expressed as mean ± standard deviation of three separate experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D011107)
Adhesion of uveal (92–1, Mel202) and cutaneous melanoma (FM55P, IGR-39) cells to fibronectin. Each result is the average of three independent experiments done in triplicate. All data are given as percentage of adhesion relative to adhesion on poly-L-lysine (taken as 100%). Error bars indicate standard deviations. Asterisk (*) indicates p<0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011107)
## Wound healing assay
*In the **Wound healing assay** section:*
Cells were grown on FN to confluence. Then the medium was aspirated, and the cell-coated surface was scraped with a 200 μl pipette tip in a single stripe. The scrape-wounded surface was washed once with PBS and twice with RPMI-1640 supplemented with FBS, and then the wounds in the cultures were allowed to heal for 24 h at 37 °C. In some experiments, wound healing in culture medium containing 25 μg/ml PHA-L was examined. The applied dose of PHA-L had no effect on the viability or growth rate of the tested cells as demonstrated by trypan blue exclusion and 3[4,5-dimethyldiazol-2-yl]-2,5diphenyltetrazolium bromide (MTT) tests (data not shown).
Migration of cells into wounded areas was observed with an inverted microscope and photographed. The average extent of wound closure was quantified by multiple measurements of the width of the wound space for each of these cases. Twenty measurements of two separate trials were made and averaged for all these conditions. Values are expressed as mean ± standard deviation of three separate experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Flow cytometric analysis
*In the **Flow cytometric analysis** section:*
Expression of human integrin subunits was assessed by flow cytometry as previously described. Briefly, cells (1 × 105) were incubated for 45 min on ice with antibodies against 50 μl/ml α3, 50 μl/ml α4, 75 μl/ml α5, integrin subunits as well as against 10 μl/ml αvβ3, integrin, or 50 μl/ml normal mouse IgG as negative control, Cells were then washed in phosphate-buffered saline (PBS, pH 7.2) and then incubated with 50 μl/ml fluorescein isothiocyanate (FITC)-conjugated antimouse IgG (Fab’)2 fragments for 45 min on ice. PHA-L binding to cells was performed according to the method of Chakraborty et al.. Briefly, cells (1 × 105) were incubated with 10 μg/ml FITC-conjugated PHA-L (Vector, Burlingame, CA) in PBS containing 2% BSA, for 45 min on ice. The cells were then washed in PBS, and assessed for fluorescence in a FACSCalibur flow cytometer (BD Biosciences, San Diego, CA). A total of 104 cells were analyzed for each immunofluorescence profile.
Antibodies used in flow cytometric analysis are listed in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
Effect of phaseolus vulgaris agglutinin on repair of wounds in monolayers of 92–1, Mel202, FM55P, and IGR-39 cells. A line was scratched with a plastic pipette tip through the confluent monolayer of cells maintained in serum-containing RPMI 1640 on a fibronectin-coated surface. The wounded cultures were allowed to heal for 24 h at 37 °C in the presence or absence of 25 μg/ml phaseolus vulgaris agglutinin (PHA-L) in serum-containing RPMI 1640. A: Panels show migration of cells in the presence or absence of PHAL after 24 h. B: The extent of wound closure was quantified by measurements of the width of the wound space for each case. For this value, the width was measured at twenty different locations in the wound and the mean value was compared to the width of the original closure (0 h). Values are means ± standard deviation of three separate experiments. Asterisk (*) indicates p<0.05.
## Expression of mRNA for β1,6-N-acetylglucosaminyltransferase V
*In the **Expression of mRNA for β1,6-N-acetylglucosaminyltransferase V** section:*
RNA was extracted using RNeasy Mini Kit (Qiagen, Hilden, Germany). The concentration and quality of RNA samples were measured with a Spectrophotometer UV/VIS (Beckman) Then 1 μg of total cellular RNA was reverse transcribed by reverse transcriptase Omniscript (Qiagen) with oligo dT23 according to manufacture protocol. PCR amplification of the sample was performed with both specific primer pairs for each of the studied glycosyltransferases: β1–6-N-acetylglucosaminyltransferase V – MGAT-5, and human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) genes. The PCR reaction comprised 30 cycles and consisted of denaturing at 94 °C (1 min), annealing at 60.5 °C (1 min), and extension at 72 °C (2 min).
The PCR mixture contained: 2.5 μl 10x PCR buffer, 5 μl Q-solution, 1 μl MgCl2, 2 μl sample cDNA, 1.6 μl 10 mM dNTPs, 0.25 μl Taq DNA polymerase (Qiagen), 1 μl of each glycosyltransferase specific primer (concentration 10 μM), and water in a final volume of 22 μl. The negative control reaction was performed simultaneously. Reaction products obtained after 30 cycles were electrophoresed on 2% agarose containing ethidium bromide. Glycosyltransferase mRNA expression of each sample was determined in at least two independent experiments (separate RNA isolation) and was normalized relative to GAPDH values. Sequences of forward (F) and reverse (R) oligonucleotide primers for MGAT-5 gene and length of the amplification products were as follows: F: 5′-GTG GAT AGC TTC TGG AAG AA-3′ R: 5′-CAG TCT TTG CAG AGA GCC-3′
(856 bp).[](https://www.ncbi.nlm.nih.gov/mesh/D015636)
Range, specificity, and supplier of monoclonal and polyclonal antibodies used for immunodetection of integrin chains in material recovered after precipitation with phaseolus vulgaris agglutinin bound to agarose.[](https://www.ncbi.nlm.nih.gov/mesh/D012685)
## Precipitation with Phaseolus vulgaris agglutinin lectin
*In the **Precipitation with Phaseolus vulgaris agglutinin lectin** section:*
After reaching early confluency, cells were washed twice with PBS, harvested with a rubber policeman and pelleted by centrifugation. Then the cells were homogenized on ice in 10 mM Tris/HCl, pH 7.5, containing 1 mM EDTA and a proteinase inhibitor cocktail, followed by incubation with the same buffer containing additionally 1% Triton X-100 and 0.3% protamine sulfate for 1 h on ice. Finally, cell extracts were clarified by centrifugation at 16000 g for one hour.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
Precipitation with the use of immobilized PHA-L lectin was performed according to the modified method of Seales et al.. Briefly, 1 mg cell extracts were incubated overnight at 4 °C with 50 μl of PHA-L immobilized on cross-linked 4% beaded agarose, 3 mg lectin/ml packed gel. PHA-L/glycoprotein complexes collected by brief centrifugation were then washed three times with 10 mM HEPES, pH 7.5, containing 150 mM NaCl, followed by one wash with PBS. Glycoproteins were released from the complexes by boiling in electrophoresis sample buffer before being subjected to SDS–PAGE. Plain agarose was used as negative control (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D012685)
Expression of integrins on human uveal (92–1, Mel202) and cutaneous melanoma (FM55P, IGR-39) cells. Melanoma cells were examined by flow cytometry for the expression of α3β1, α4β1, α5β1, and αvβ3 integrins, and data were compared to cells incubated with normal mouse IgG. Fluorescein isothiocyanate (FITC)-conjugated rabbit anti-mouse IgG (Fab’)2 fragments were used for detection. Fluorescence signals of 10,000 cells were counted for each integrin subunit tested. Histograms of cells versus log fluorescence were generated. A: Panels show FACS profile for integrin-positive cell lines. Colored areas indicate the fluorescence profile of cells after indirect fluorescence staining with anti-integrin monoclonal antibodies. Open histograms represent background fluorescence. Relative fluorescence is shown as a logarithmic scale of 4 log cycles on the x-axis, and cell number as a linear scale on the y-axis. Data from one of three similar experiments are presented. The negative control for each line is different in some experiments because the experiments were not run on the same occasion. B: Diagram shows percentage of melanoma cells expressing α3β1, α4β1, α5β1, and αvβ3 integrins. C: Diagram shows quantitation of data from flow cytometric analyses. Values are means ± standard deviation of three separate experiments. Asterisk (*) indicates p<0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D016650)
## SDS–PAGE and western blotting
*In the **SDS–PAGE and western blotting** section:*
For electrophoresis, equal volumes of proteins (1/20 of precipitated materials) mixed with sample buffer and heated, were separated on 10% SDS-polyacrylamide gels under nonreducing condition according to the method of Laemmli. Following separation, the proteins were transferred onto PVDF membranes in buffer consisting of 25 mM Tris, 0.192 M glycine, and 20% methanol, pH 8.4, overnight at constant amperage 0.1 A with cooling (Bio-Rad). Polyacrylamide gels were calibrated for molecular weight determination using the Sigma Standard Kit for electrophoresis in SDS (205–29 kDa).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## Identification of the proteins bearing β1–6 GlcNAc branched N-linked oligosaccharides
*In the **Identification of the proteins bearing β1–6 GlcNAc branched N-linked oligosaccharides** section:*
Materials precipitated with PHA-L lectin were separated by SDS–PAGE and blotted onto PVDF membrane. Next the blots were blocked with 1% BSA in 20 mM Tris/HCl, pH 7.6, containing 0.15 M NaCl and 0.1% Tween 20 (TBS/Tween). Afterwards, the membranes were incubated for 2 h in 1% BSA in TBS/Tween containing a 1:1000 dilution of one of the following antibodies specific for different integrin subunits: α3, α5, αv, and β1. Following a triple wash with TBS/Tween, the membranes were incubated for 1 h with the secondary antibodies either alkaline phosphatase conjugated goat anti-rabbit IgG (for α3, α5, αv, integrin subunits; 1:250 dilution in TBS/Tween with 1% BSA) or alkaline phosphatase coupled goat anti-mouse IgG (for β1 integrin subunit; 1:500 dilution in TBS/Tween with 1% BSA). Visualization of immunoreactive proteins was achieved with the use of 4-nitroblue-tetrazolium salt/5-bromo-4-chloro-3-indolylophosphate solution.
Antibodies used for immunodetection are listed in Table 2.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## Statistics
*In the **Statistics** section:*
Flow cytometric analysis of phaseolus vulgaris agglutinin binding on the surface of human uveal (92–1, Mel202) and cutaneous melanoma (FM55P, IGR-39) cells. A: Histogram of fluorescence intensity with or without fluorescein isothiocyanate (FITC)-conjugated phaseolus vulgaris agglutinin. B: Quantification of data from flow cytometric analyses. Fluorescence intensity relative to negative control, representing means from three pooled experiments. Asterisk (*) indicates p<0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D016650)
Statistical analysis was performed with the use of Dunkan’s new multiple range test. A p value less than 0.05 was considered statistically significant.
## Other methods
*In the **Other methods** section:*
Protein concentrations were measured according to Peterson with BSA as standard.
## Results
*In the **Results** section:*
Cell-FN interaction, mediated through several different receptors, has been implicated in a wide variety of cell activities, including important roles at several stages of tumor development. In the first part of this study we compared primary UM and primary CM in terms of their adhesion and migration to FN. UM cells did not adhere to FN. As shown in Figure 1, the two examined CM cell lines attached to FN, but the adhesion efficiency of IGR-39 cells (80%) was twice that of FM55P cells (40%). Interestingly, UM cells repaired scratch wounds twice as fast as CM cells did (36% and 18% wound closure after 24 h, respectively; Figure 2).
Among the various classes of cell adhesion molecules, integrins are particularly associated with cell adhesion to extracellular matrices, and altered levels of integrin expression are related to tissue invasion and metastasis in many types of cancer. In the second part of this study we used flow cytometry to characterize UM and CM cells with respect to their cell surface integrins acting as receptors for FN (α3β1, α4β1, α5β1, αvβ3), applying specific monoclonal antibodies that recognize different integrin heterodimers or integrin subunits (Table 1). The flow cytometry data are summarized in Figure 3. FN receptor expression differed between the tested cell lines, but no distinct pattern distinguished UM from CM except for high expression of α4β1 integrin on both FM55P and IGR-39 cells. The results also showed that the high levels of α3β1, α4β1 and α5β1 integrin expression on IGR-39 cells correlate with strong attachment to FN-coated surfaces, and the high expression of α4β1 integrin on FM55P probably was enough to make them adhere weakly to FN.
Interestingly, the expression of α5β1 integrin, which is known to be a major FN receptor, was low on 92–1 (16%), Mel202 (29%), and FM55P (22%) cells.
Most integrins are able to bind different ligands with different affinities. The affinity of integrins may vary depending on the cell type in which they are expressed or as the result of conformational changes. Although the molecular basis of adhesion molecule-ligand interaction is not fully understood, integrin glycosylation represents a kind of regulation by which a wide variety of these receptors have their specificity and affinity modulated in several cell lines. Because one of the common structural alterations in cell surface glycans observed in various human and rodent tumors is highly elevated expression of β1–6-N-acetylglucosamine (β1–6 GlcNAc) branched tri- and tetraantennary complex type N-glycans, we also tested their influence on melanoma cell migration. Addition of PHA-L, whose preferred ligands are β1–6 branched N-glycans, reduced the rate of 92–1, Mel202, and FM55P cell migration into scratch wounds on FN-coated wells by 79%, 93%, and 63%, respectively, indicating the participation of β1–6 branched N-oligosaccharides in this process, but it had no effect on the migration rate of IGR-39 cells (Figure 2).[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
The formation of β1–6 branches on the trimannosyl terminus of N-linked oligosaccharides is controlled via the activity of GnT-V, the enzyme that catalyzes the addition of N-acetylglucosamine to the core mannose of di- and triantennary N-glycans through a β1–6 linkage. Analysis of the transcript of GnT-V by semiquantitative RT–PCR showed that the tested cells did not differ in their mRNA levels of GnT-V (data not shown). To assess GnT-V activity in vivo, we measured cell surface β1–6 branched N-oligosaccharides via their specific binding to PHA-L and detection by flow cytometry. It showed that 92–1, Mel202, and IGR-39 cells expressed significantly higher amounts of β1–6 branched N-oligosaccharides on the cell surface than FM55P cells did, as reflected in the mean fluorescence intensity of the cells (Figure 4): the former cells showed mean fluorescence intensity twice that of FM55P cells. Presumably the enhanced PHA-L binding was restricted to the cell surface, because the binding and wash procedures were performed on ice with no previous lysis of cells.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
To identify the glycoproteins bearing β1–6 GlcNAc branched N-glycans from four melanoma cell lines, we precipitated clarified lysates of 92–1, Mel202, FM55P, and IGR-39 with PHA-L lectin. The glycoproteins recovered after precipitation were electrophoresed under nonreducing conditions, blotted onto a PVDF membrane, and probed with antibodies specific for different integrin subunits: α3, α5, αv, and β1. Immunodetection clearly indicated the presence of β1–6 GlcNAc branched N-glycans on these integrin chains (Figure 5).[](https://www.ncbi.nlm.nih.gov/mesh/D000117)
Immunodetection of α3, α5, αv, and β1 in materials obtained after precipitation of 92–1, Mel202, FM55P, and IGR-39 cell extracts with phaseolus vulgaris agglutinin bound to agarose. One mg of the cell extracts were incubated overnight with phaseolus vulgaris agglutinin (PHA-L) immobilized on cross-linked 4% beaded agarose. Glycoproteins were released from the complexes by boiling in electrophoresis sample buffer before being subjected to 10% SDS–PAGE. Following separation, the proteins were blotted onto PVDF membrane. After being blocked the blots were incubated with one of the following antibodies specific for different integrin subunits: α3, α5, αv, and β1. Next, the membranes were incubated with the secondary antibodies either alkaline phosphatase conjugated goat anti-rabbit IgG (for α3, α5, αv, integrin subunits) or alkaline phosphatase coupled goat anti-mouse IgG (for β1 integrin subunit). Visualization of immunoreactive proteins was achieved with the use of 4-nitroblue-tetrazolium salt/5-bromo-4-chloro-3-indolylophosphate solution. Lane S shows position of molecular weight markers.[](https://www.ncbi.nlm.nih.gov/mesh/D012685)
## Discussion
*In the **Discussion** section:*
Many studies have shown tumor cells to be generally less adhesive and to deposit less ECM than their normal counterparts. The loosened matrix adhesion of tumor cells may permit them to leave their original site in the tissue. Previously we demonstrated that although both CM and UM are similarly derived from neuroectodermal tissues, they differ in their adhesion to type IV collagen, laminin and FN. To study the biologic mechanisms that underlie this distinctive biologic behavior, we investigated the influence of the surface expression of the integrins that act as FN receptors, as well as the expression of β1–6 branched N-linked oligosaccharides on surface proteins, on UM (92–1, Mel202) and CM (FM55P, IGR-39) cell behavior. In this study we showed that the UM cells did not adhere to FN and that they repaired wounds twice as fast as CM cells did.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
It is well documented in the literature that the expression pattern of adhesive molecules differs widely between normal and malignant tissues. The gain or loss of adhesive molecule expression on cancer cells appears to be a natural consequence of their adaptation and survival in a different environment. Integrins, which are the major class of adhesion molecules responsible for mediating cellular interaction with the ECM, seem to have an important role in various aspects of cancer development. Several studies have focused on the integrin expression of melanoma cells, but only the expression of α2β1, α3β1, α6β1 and αvβ3 integrins was found to be associated with tumor progression in CMs. No correlation between integrin expression and cell type or aggressive behavior has been has been confirmed in UM.
The present study found no distinct pattern of FN receptor expression among primary UMs and CMs, except for high expression of α4β1 integrin on both FM55P and IGR-39 cells. Contrasting with our results showing a high level of α4β1 integrin on CM cells are previous findings that α4β1 integrin expression was rare in CM. The divergent outcomes may be attributable to differences in the techniques and antibodies used. The high levels of α3β1, α4β1, and α5β1 integrin expression on IGR-39 cells seemed to be associated with their strong adhesion to FN. Interestingly, 92–1, Mel202, and FM55P cells showed no or weak adhesion to FN, perhaps the result of low expression of FN receptors except for that of α4β1 integrin on FM55P cells.
The high expression of α4β1 integrin on FM55P cells probably was enough to make them adhere weakly to FN. These results were further confirmed by assays of adhesion inhibition in the presence of specific anti-integrin monoclonal antibodies (data not shown).
There is increasing evidence that progression of cancer from a tumorigenic to metastatic phenotype is directly associated with an increased level of β1–6 branched N-oligosaccharides as the result of hyperactivity of GnT-V. Perhaps the most interesting findings of this study are those related to the involvement of β1–6 branched N-linked oligosaccharides on surface glycoproteins in FM55P, 92–1 and Mel202 cell migration. Although literature data, including results obtained in our laboratory, describe the involvement of β1–6 branched N-glycans in CM cell adhesion and migration, the present study is the first to present data with respect to UM. The functional significance of increased β1–6 branching in N-glycoproteins has not been well established, but it has been associated with several hallmarks of tumor progression: decreased substrate adhesion, loss of contact inhibition, increased migration in vitro, and increased metastasis in vivo. We recently showed that changes in the number of proteins acting as substrates for GnT-V were associated more with melanoma development and progression than with expression of cell adhesion molecules. It is believed that expression of β1–6 branched N-oligosaccharides on integrins and other adhesion receptors may facilitate the turnover of cell-cell and cell-ECM contacts to enhance cell motility.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
Here we provided evidence that β1–6 branched N-oligosaccharides on the cell surface of 92–1, Mel202, and FM55P cells facilitated their migration on FN. How is the effect of the sugar moieties achieved? Cell migration is influenced by the strength of transient cell-substratum attachment and depends on ligand and integrin levels, as well as integrin-ligand binding affinities. Cell migration is most rapid at an intermediate ratio of cell-substratum adhesion to contractile force; then the cell can still extend lamellae and form new attachments at the cell front but break the attachment at the rear. In 92–1, Mel202, and FM55P cells, altered surface glycosylation might induce functional changes in adhesion proteins and in this way decrease the binding capacity of integrins to FN, possibly by holding the conformation in the low-affinity form to the ligand.
Indeed, studies with tri- and tetraantennary minimal energy conformers indicate that the β1–6 branch is folded back to the protein structure, and this in turn could likely modulate the integrin conformation, and thus also integrin function (affinity). It has also been shown that β1–6 branched N-glycans may reduce the stability of the integrin receptor aggregates that maintain firm cell-substratum attachment, and thereby facilitate cell motility. From flow cytometric studies it is known that 92–1 and Mel202 cells possess more glycoproteins bearing β1–6 branched N-oligosaccharides than do FM55P cells, so it is not surprising that 92–1 and Mel202 cells seemed more sensitive to PHA-L treatment as judged by their 79% and 93% versus 63% decrease in wound healing assays. Interestingly, IGR-39 cells, which possess an amount of β1–6 branched N-oligosaccharides on the cell surface similar to the level on UM cells, were not sensitive to PHA-L treatment in the wound healing assay. Possibly the high level of FN receptors on these cells had a stronger effect on the degree of transient cell-substratum attachment than did modulation of integrin-FN binding by β1–6 branched N-oligosaccharides.
This suggestion needs to be confirmed.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
Malignant cells acquire invasive potential by accumulating features including increased cell motility, secretion of proteolytic enzymes, and alteration of cell-substratum and cell-cell adhesion. Tumor cell adhesion is a fundamental event in the formation of distant tumor metastases; during the formation of metastases, malignant cells often show decreased cell-cell and cell-ECM interaction at the primary tumor site and must establish new adhesive interactions at secondary sites. Elevated expression of PHA-L reactive oligosaccharides in carcinomas is usually associated with tumor progression and metastasis, as has been shown in breast and colon cancers or melanoma. Inhibition of the expression of β1–6 branched N-oligosaccharides through different strategies always results in the loss of metastatic ability. As 92–1 and Mel202 cells did not adhere to FN and were twice as mobile as CM cells, and since the presence of β1–6 branched N-oligosaccharides on their surface enhanced their motility, it is tempting to speculate that these cells may have also been more metastatic, but this requires confirmation.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
One aspect of metastasis that has intrigued scientists for over a century is organ-specific metastasis. As mentioned, the metastatic behavior of UMs and CMs in the body differs greatly. It is believed that molecules on the surface of tumor cells are the principal regulators of adhesion to organ components. Indeed, most of the cell lines expressing β1–6 branched N-oligosaccharides have been shown to metastasize to either the liver or the lungs. β1–6 branched N-oligosaccharides possibly influence adhesion by providing specific ligands to the lectin receptors on the target site, because the terminal substitution on these glycans influences the choice of metastatic site.
It has been demonstrated that β1–6 branched N-oligosaccharides substituted with polylacNAc possibly metastasize to the lungs, while cells expressing unsubstituted multiantennary N-oligosaccharides home in the liver which express galectin-1. UM cells metastasize specifically to the liver, so it would be useful to find clear evidence for the role of galectin-1 in liver-specific metastasis.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
In 92–1, Mel202, and IGR-39 cells the levels of cell surface β1–6 branched N-oligosaccharides were significantly higher than in FM55P cells, but it did not appear to have resulted from differences in the mRNA expression of GnT-V in these cells as shown by semiquantitative RT–PCR. Nevertheless, it should be stressed that GnT-V activity may undergo regulation not only at expression level. Although little is known about the cellular regulators of GlcNAc transferases, there are few reports describing decrease in GnT-V activity in response to the metastasis gene nm23-H1 and the tumor supressor gene p16, and loss of β1–6 GlcNAc branching of β1 integrins and concurrent dramatic reduction in migration through ECM after overexpression of 16-kDa membrane subunit of vacuolar H+-ATP-ase. Moreover, it has been shown that the basal activity of GnT-V is also regulated by Ras/Raf-1/MEK/MAPK cascade and phosphatidylinositol-3-kinase/protein kinase B signaling pathway and changes also during the cell cycle. In addition, β1–6 branching is also dependent upon GnT-V having access to suitable oligosaccharide acceptors.
GnT-V competes for the same substrate as N-acetylglucosaminyltransferase III (GnT-III). Substitution by GnT-III effectively reduces β1–6 branching because GnT-V cannot act on such bisected precursor, resulting in lowering tumor cell metastasis. Although it is well documented in the literature that β1–6 branched N-glycans contribute to cancer progression, the role of integrins with a bisecting GlcNAc cannot be neglected. It has been shown that the modification of α5β1 integrin by bisecting GlcNAc inhibited cell spreading and migration on FN, subsequently leading to the down-regulation of integrin-mediated signaling.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
The present studies on UM cells showed remarkable homogeneity of their adhesion and migration properties, expression of FN receptors, level of β1–6 branched N-oligosaccharides on the cell surface, and the influence of these glycans on cell migration. The roles of integrins and their N-glycosylation in the regulation of UM growth and progression are largely unknown. To our knowledge, this study is the first to demonstrate the role of β1–6 branched N-oligosaccharides on surface glycoproteins in the migration of UM cells. Further studies of other melanomas are needed to confirm these interesting findings.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
# References
*In the **References** section:*
|
# Introduction
Functional identification of NR2 subunits contributing to NMDA receptors on substance P receptor-expressing dorsal horn neurons
# Abstract
*In the **Abstract** section:*
NMDA receptors are important elements in pain signaling in the spinal cord dorsal horn. They are heterotetramers typically composed of two NR1 and two of four NR2 subunits: NR2A-2D. Mice lacking specific NR2 subunits show deficits in pain transmission yet subunit location in the spinal cord remains unclear. We have combined electrophysiological and pharmacological approaches to investigate the composition of functional NMDA receptors expressed by lamina I, substance P receptor-expressing (NK1R+) neurons, as well as NK1R- neurons. Under low Mg2+ conditions (100 μM), the conductance of NMDA receptors at -90 mV (g(-90 mV)) with NR2A or NR2B subunits (NR2A/B) is low com[pare](https://www.ncbi.nlm.nih.gov/mesh/D008274)d to conductance measured at the membrane potential where the inward current is maximal or maximal inward current (MIC) (ratio of ~0.07 calculated from Kuner and Schoepfer, 1996). For NR2C or NR2D subunits (NR2C/D), the ratio is higher (ratio ~0.4). NK1R+ and NK1R- neurons express NMDA receptors that give ratios ~0.28 and 0.16, respectively, suggesting both types of subunits are present in both populations of neurons, with NK1R+ neurons expressing a higher percentage of NR2C/D type NMDA receptors. This was confirmed using EAB318, an NR2A/B preferring antagonist, and UBP141, a mildly selective NR2C/D antagonist to increase and decrease the g(-90 mV)[/g(MIC](https://www.ncbi.nlm.nih.gov/mesh/C498147)) ratios in both subpopulations of neur[ons.](https://www.ncbi.nlm.nih.gov/mesh/D002264)
## Background
*In the **Background** section:*
NMDA receptors in the spinal cord dorsal horn are key elements in the initiation of changes in synaptic strength [1] and pain hypersensitivity [2,3]. These receptors consist of two obligatory NR1 subunits and two NR2 subunits, of which there are four types encoded by distinct genes: NR2A, NR2B, NR2C and NR2D [4].
The incorporation of different NR2 subunits has a major impact on the functional properties of the NMDA receptor, critically influencing agonist and antagonist affinity, receptor deactivation kinetics, channel conductance and interactions with intracellular proteins [3]. Additionally, NMDA receptors with NR2A or NR2B show higher Mg2+ sensitivity at negative membrane potentials than those with NR2C or NR2D [5,6].[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
Involvement of NMDA receptors in dorsal horn function has been demonstrated through experiments interfering with expression of different NMDA receptor subunits. Knockdown of the NR1 subunit of NMDA receptors to eliminate functional NMDA receptors in the spinal cord reduces hyperalgesia and allodynia in a number of animal models but does not alter acute pain responses [7-9]. NR2A knockout mice show some reduced forms of hypersensitivities [10-12]. However, these NR2A knockouts display normal acute pain responses [12], formalin-induced hyperalgesia
[13] and nerve ligation or injury-induced allodynia [14,15]. NR2B knockout mice do not survive postnatally [16,17], therefore NR2B specific antagonists have been used to study the role of this protein in pain hypersensitivity. Intrathecal administration of NR2B antagonists blocks or decreases PGE2 or NMDA induced allodynia [11] as well as capsaicin-induced hyperalgesia
[18]. NR2D knockout mice fail to develop nerve ligation [12], PGE2 [19] or PGF2alpha-induced allodynia [11,20].
Overall, these data suggest that different NR2 subunits are involved in dorsal horn circuits important for the development of hyperalgesia or allodynia but their specific functions remain unresolved.[](https://www.ncbi.nlm.nih.gov/mesh/D005557)
Lamina I of the spinal cord is a critical site for nociceptive processing, receiving abundant monosynaptic input from nociceptors. The main output neurons of lamina I, the substance P receptor-expressing (NK1R+) projection neurons, are essential in mediating pain hypersensitivity [21,22]. NK1R+ neurons express NMDA receptors [23,24] but little is known about the subtypes of NMDA receptors they express.
In this paper, we have taken advantage of the different magnesium sensitivities and pharmacology of NMDA receptors with different NR2 subunit composition to identify functionally expressed NMDA receptors on NK1R+ and NK1R- dorsal horn neurons in lamina I.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
## Methods
*In the **Methods** section:*
## Transverse slice preparation
*In the **Transverse slice preparation** section:*
Lumbar spinal cords were obtained from rats of postnatal day 14 (P14) to P19. The animals were first anesthetized with isoflurane and then decapitated. All experiments were conducted with the approval of the Columbia University Institutional Animal Care and Use Committee and in accord with the Guide for the Care and Use of Laboratory Animals. The spinal cords were excised and placed in ice-cold oxygenated high Mg2+ Krebs solution (95% O2/5% CO2 saturated Krebs solution, in mM: NaCl 125 or sucrose 250, KCl 2.5, NaHCO3 26, NaH2PO4 1.25, glucose 25, MgCl2 6, CaCl2 1.5, pH 7.4) plus 1 mM kynurenic acid. After removal of the dura mater and arachnoid membranes, all ventral roots were cut close to the cord and the spinal cord embedded in low melting agarose (Invitrogen Life Technologies) for slicing.
Transverse slices (350–450 μm) with attached dorsal roots were obtained using a Leica VT1000S vibrating blade microtome. Slices were then incubated in oxygenated high Mg2+Krebs solution (no sucrose included) at 36°C for 1 hour before recording.[](https://www.ncbi.nlm.nih.gov/mesh/D007530)
## Recording from pre-identified NK1R+ and NK1R- lamina I neurons
*In the **Recording from pre-identified NK1R+ and NK1R- lamina I neurons** section:*
The labeling of NK1R+ dorsal horn neurons with fluorescent dye has been described elsewhere [25,26]. In brief, spinal cord slices were incubated in high Mg2+ Krebs solution containing 20 – 40 nM tetramethylrhodamine-conjugated substance P (TMR-substance P) for 20 – 30 minutes at room temperature following 1 hour of recovery at 36°C. After unbound substance P was washed away for at least 20 minutes in an incubation chamber containing oxygenated high Mg2+ Krebs solution, slices were transferred to a submersion style chamber for recording. NK1R+ neurons were identified as expressing NK1R by clear, intense labeling with TMR-substance P. NK1R- neurons were chosen as lamina I neurons showing no evidence of TMR-substance P staining.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
## Recording solutions
*In the **Recording solutions** section:*
Intracellular solution used for most of these experiments had the following composition (in mM): Cs-methylsulfonate 130, Na-methylsulfonate 10, EGTA 10, CaCl2 1, HEPES 10, QX-314·Cl 5, Mg2+-ATP 2, pH adjusted to 7.2 with CsOH, osmolarity adjusted to 290 with sucrose. For some experiments in which intracellular Ca2+ needed to be strongly chelated, BAPTA intracellular solution was used. It was composed of (in mM): Cs-methylsulfonate 50, Na-methylsulfonate 10, BAPTA·Cs 40, CaCl2 4, HEPES 10, QX-314·Cl or QX-222·Cl 5, Mg2+-ATP 2, TEA·Cl 10, pH adjusted to 7.2 with CsOH, osmolarity about 310. Junction potentials were measured empirically and corrected in the bath before GOhm seal formation for each cell.[](https://www.ncbi.nlm.nih.gov/mesh/C045880)
Modified Krebs solutions were used for the extracellular bath. To prevent possible neurotoxicity associated with Ca2+ influx through activated NMDA receptors, we replaced 95–98% of the extracellular Ca2+ with 3 mM Ba2+. The barium Krebs comprised: NaCl 125, KCl 2.5, NaH2PO41.25, NaHCO3 26, glucose 25, MgCl2 0.1, CaCl2 0.04–0.1, BaCl2 3 and pH 7.4. TTX (0.5 μM), SR95531 (5–10 μM) and strychnine (1 μM) were included in the extracellular solutions to eliminate action potential generation and involvement of inhibitory circuits.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Analysis of NMDA induced membrane currents
*In the **Analysis of NMDA induced membrane currents** section:*
To obtain the current-voltage relationships of NMDA receptor-mediated currents, NMDA (15 μM) was superfused onto pre-identified, lamina I neurons for 2–3 minutes following several minutes of baseline, whole-cell recording. Triangle voltage ramp commands (the ramp up and ramp down were 0.9 sec duration each) were applied continuously at low frequency (0.05 Hz). Digital sampling frequency was 10 KHz. NMDA applications were repeated 2–3 times before NMDA co-application with antagonists. The data for the first NMDA application were not included for analysis due to changing baseline conditions.
Current responses to triangle voltage ramps before and after recovery from NMDA application were averaged as a control current then subtracted from each triangle ramp made during NMDA induced currents. The resulting NMDA current ramps were plotted as a function of membrane potential and further analyzed. To minimize noise for measuring the following parameters, NMDA current ramps were subjected to a rolling average procedure over a 100 msec time frame.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
For each voltage ramp during NMDA applications, the membrane current at -90 mV (I(-90 mV)), the maximal inward current (MIC), and the membrane potential for the MIC (VMIC) were measured (Figure 1E). The current measured at -90 mV holding potential was determined as I(-90 mV). The MIC was initially determined as the most negative current value in the rolling average. The VMIC was then determined as the voltage corresponding to the MIC.
Because each ramp had an up and a down phase, each parameter from a ramp current had a pair of values and they were averaged for following analysis. The conductance at -90 mV and MIC (g(-90 mV) and g(MIC) respectively) as well as conductance ratio were then calculated based on the formulae:[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
To compare NMDA receptor g(-90 mV)/g(MIC) ratios under different pharmacological conditions, we averaged three ratio values calculated for each NMDA application near the peak NMDA response at -70 mV. The ratios under different pharmacological conditions or represented by different neuron populations were then compared.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
Only cells starting with reversible NMDA induced membrane currents, in which the difference between g(-90 mV)/g(MIC) ratios during wash-in and wash-out of NMDA was less than 0.15, were included for analysis. Cells with high membrane holding current (> -100 pA) were discarded.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
## Materials
*In the **Materials** section:*
SR 95531 hydrobromide and QX-222·Cl were purchased from Tocris Cookson (Bristol, UK). QX-314·Cl was purchased from Sigma-Aldrich or Alomone labs (Jerusalem, Israel). Low melting point agarose and TMR-substance P were purchased from Invitrogen Corp. Some TMR-substance P was synthesized and purchased from AnaSpec, Inc. Strychnine was obtained from Sigma-Aldrich.
EAB318 was provided by Wyeth Research. EAB318 has IC50 of 20, 80 and 3500 nM for NMDA receptors with NR2A, NR2B and NR2C respectively [27]. UBP141 was synthesized as described [28]. The Ki of UBP141 for NMDA receptors with NR2A – NR2D are 14, 19, 4 and 2.7 μM respectively [28].[](https://www.ncbi.nlm.nih.gov/mesh/C049853)
## Results
*In the **Results** section:*
## The I-V relationship of NMDA currents induced by superfusion of NMDA onto dorsal horn neurons
*In the **The I-V relationship of NMDA currents induced by superfusion of NMDA onto dorsal horn neurons** section:*
We investigated the total population of functional NMDA receptors expressed by different classes of lamina I neurons. NMDA was bath-applied onto spinal cord slices to activate all functional NMDA receptors. To identify the type of NMDA receptors expressed by pre-identified lamina I neurons in the spinal cord dorsal horn, we took advantage of the differential sensitivity to Mg2+ inherent in NMDA receptors composed of different NR2 subunits. NMDA receptors containing NR2A or NR2B subunits show more negative slope conductance at negative membrane potentials than those containing NR2C or NR2D [6].
Because the measurable difference in Mg2+ sensitivity is enhanced when extracellular Mg2+ concentration is low, 100 μM extracellular Mg2+ was used throughout these experiments. SR95532 (10 μM), strychnine (1 μM) and TTX (0.5–1 μ;M) were always included to eliminate the inhibitory currents and action potential triggered responses. Most of the Ca2+ in the Krebs was replaced with Ba2+ to diminish evoked neurotransmitter release and Ca2+ dependent currents in the cells.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
NK1R+ and NK1R- dorsal horn neurons were visually identified for whole-cell recording following incubation of spinal cord slices in TMR-substance P [25,26] as shown in Figure 1A and 1B, in which fluorescence and IR-DIC images are shown respectively. Bath application of NMDA (15 μM) to these neurons generated inward currents at a holding potential of -70 mV as shown in Figure 1C. To determine the voltage-dependent Mg2+ sensitivity of these receptors, triangle voltage ramp commands from -100 mV to + 50 mV and back to -100 mV were applied before, during and after NMDA application at 0.05 Hz (Figure 1C and 1D). After subtraction of control current, the resulting ramps of NMDA receptor-mediated current were plotted as a function of command voltage as shown in Figure 1E. The voltage sensitivity of the NMDA current generated by the ascending ramp command is similar to that generated by the descending command (Figure 1E). In addition, the current-voltage relationships of these NMDA induced currents had an average reversal potential of -2.0 ± 1.0 mV (n = 15), close to the predicted NMDA receptor reversal potential.
The pair of NMDA current responses obtained from each triangle voltage command were used to determine the current at -90 mV (I(-90 mV)), the maximal inward current (MIC) and the membrane potential at which MIC occurs (VMIC) as illustrated in Figure 1E (see Methods). The membrane conductance at -90 mV (g(-90 mV)) and MIC (g(MIC)) were then calculated as shown in Figure 1F.[](https://www.ncbi.nlm.nih.gov/mesh/C005358)
Parameters representing the voltage sensitivity of current flow through NMDA receptors were determined following NMDA application onto pre-identified dorsal horn neurons. (A) NK1R+ neuron in the superficial dorsal horn was selectively labeled by TMR-substance P in a transverse slice (40 × objective). (B) IR-DIC image showing the NK1R+ neuron patched with a pipette. (C) A representative trace shows an NMDA-evoked inward current with 100 μM Mg2+ in the bath. (D) Current responses to voltage ramps are shown at expanded time base. (E) The voltage dependence of NMDA receptor-mediated currents derived from NMDA evoked currents. (F) The same as in Figure E. The conductance of NMDA receptors at -90 mV (g(-90 mV)) and MIC (g(MIC)) are illustrated.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
The voltage sensitivity of NMDA currents depends predominantly on the voltage dependence of Mg2+ block of the receptors [29,30]. The voltage sensitivity of the agonist activated NMDA receptors in our experiments was quantified by dividing g(-90 mV) with g(MIC). The ratio was then compared to the value derived from heterologous expression data using specific NR1 and NR2 subunit combinations. From such data we calculated that NMDA receptors containing NR1/NR2A or NR1/NR2B show g(-90 mV)/g(MIC) ratios of about 0.07 and that their VMIC is between -37 and -40 mV. Conversely, NMDA receptors containing NR1/NR2C or NR1/NR2D have ratios around 0.39 and VMICs around -52 to -57 mV (extracted from Kuner et al.
[6]). Thus, for example, lower g(-90 mV)/g(MIC) ratios near 0.07 and less negative VMICs indicate expression of NR2A/B-containing NMDA receptors with high Mg2+ sensitivity.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
In Figure 2A, the control response of an NK1R+ neuron to 15 μM NMDA is shown. The ramp response recorded when the agonist response to NMDA was at its greatest is plotted at the bottom. The value of the g(-90 mV)/g(MIC) ratio is 0.2. This is intermediate between the values for receptors that include NR2A/B and NR2C/D subunits suggesting that NMDA receptors with both types of NR2 subunits are present on this NK1R+ lamina I neuron. The VMIC value (-48 mV in this example) is also intermediate between the VMIC values of NR2A/B and NR2C/D subunits, supporting the interpretation that the NMDA receptors expressed by this NK1R+ neuron are heterogeneous in NR2 subtype expression.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
The voltage-dependence of NMDA receptor-mediated currents is a good indicator of NMDA receptor subtype expression.A-C show current responses to NMDA bath applied to the same NK1R+ neuron as in Figure 1. (A) NMDA-evoked current response is the same as in Figure 1. (B) Co-application of NMDA and UBP141, an NR2C/D preferring antagonist, induced a smaller inward current. The I-V relationship shows a more pronounced negative slope conductance. (C) Co-application of NMDA and EAB318, an NR2A/B preferring antagonist, induced NMDA receptor-mediated current with less negative slope conductance.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
Although most of the extracellular Ca2+ was replaced with Ba2+ in these experiments, it was still possible that the remaining bath Ca2+ or Ca2+ released from endoplasmic reticulum (ER) was sufficient to trigger activation of other currents, altering g(-90 mV)/g(MIC) ratio and VMIC values. In 9 of 15 cells recorded, intracellular solution containing 40 mM BAPTA was used to fully suppress accumulation of intracellular Ca2+ associated with NMDA receptor activation. There was no significant difference between the g(-90 mV)/g(MIC) values when recording with BAPTA or EGTA intracellular solutions. Thus the data from these two groups were pooled.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Pharmacological test of NR2 subunit confirms the expression of NR2A/B and NR2C/D type NMDA receptors
*In the **Pharmacological test of NR2 subunit confirms the expression of NR2A/B and NR2C/D type NMDA receptors** section:*
Next we used pharmacology to confirm that g(-90 mV)/g(MIC) ratio and VMIC are good indicators of NMDA receptor subtypes expressed by dorsal horn neurons. EAB318 (100 – 200 nM) and UBP141 (20–30 μM) block different NMDA receptor subtypes: UBP141 [28] is a mildly selective NR2C/D preferring antagonist while EAB318 is a NR2A/B selective blocker [27]. If both categories of NR2 subtypes are present, EAB318 should make the NMDA evoked current more NR2C/D like and UBP141 should make the current more NR2A/B like.
As predicted, when NMDA was co-applied with UBP141, the current-voltage relationship had a smaller g(-90 mV)/g(MIC) ratio and less negative VMIC (Figure 2A and 2B). This suggests that in the presence of UBP141, a higher proportion of NR2A/B type NMDA receptors dominate the current. When NMDA was co-applied with EAB318, the current voltage relationship shifted to a larger g(-90 mV)/g(MIC) ratio and a more negative VMIC, suggesting that a higher proportion of NR2C/D type NMDA receptors were revealed (Figure 2C).[](https://www.ncbi.nlm.nih.gov/mesh/C498147)
To ensure that the shift of g(-90 mV)/g(MIC) ratio and VMIC were genuinely associated with selective block of a subpopulation of NMDA receptors and not simply caused by errors associated with measuring smaller amplitude NMDA evoked currents, the relationship between g(-90 mV)/g(MIC) and NMDA evoked current amplitude was plotted as in Figure 3A. As the NMDA plus antagonists washed onto the dorsal horn neuron under study, the impact of the two antagonists on g(-90 mV)/g(MIC) were different. UBP141 depressed the amplitude of NMDA evoked currents and g(-90 mV)/g(MIC) values. EAB318 also depressed the amplitude of NMDA evoked current but caused a large shift to higher g(-90 mV)/g(MIC) values. In the two situations, UBP141 and EAB318 depressed the peak amplitudes of NMDA induced currents from -306 ± 48 to -130 ± 222 pA (n = 15, p < 0.01 for paired t-test) and -108 ± 16 pA (n = 15, p < 0.01 for paired t-test) respectively.
Figure 3B shows the individual and mean g(-90 mV)/g(MIC) ratios determined when NMDA evoked currents were maximal under different drug conditions. UBP141 significantly decreased the g(-90 mV)/g(MIC) ratio from 0.23 ± 0.03 to 0.15 ± 0.01 (n = 15, p < 0.01 for paired t-test) while EAB318 significantly increased the ratio from 0.23 ± 0.03 to 0.36 ± 0.05 (n = 15, p < 0.01 for paired t-test) (see Methods). The upper and lower broken horizontal lines represent the g(-90 mV)/g(MIC) for pure NR2A/B and NR2C/D-containing NMDA receptors respectively as calculated using the data of Kuner and Schoepfer [6]. Comparing our data to these benchmarks shows that the two antagonists are pushing the g(-90 mV)/g(MIC) values in the directions predicted from heterologous expression data, assuming both types of subunits are present in the neuron tested.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
The NMDA g(-90 mV)/g(MIC) ratio is not a function of the amplitude of NMDA-evoked membrane current. (A) The g(-90 mV)/g(MIC) ratio was calculated for each test ramp throughout NMDA applications in the absence and presence of antagonists and plotted as a function of NMDA-evoked current amplitude. (B) Summary of the antagonist effects on the g(-90 mV)/g(MIC) ratio. (C) The VMIC was calculated for each test ramp throughout NMDA applications in the absence and presence of antagonists and plotted as a function of NMDA-evoked current amplitude for the same data as (A). (D) Summary of the antagonist effects on the VMIC (n = 15). (E) The same data as in (B) but grouped according to the sequences of drug application.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
We also analyzed VMICs under different drug conditions. Figure 3C, calculated from the same data as Figure 3A, shows the action of the two antagonists on VMIC. As expected, UBP141 caused the VMIC to become somewhat more positive or more NR2A/B like, while EAB318 caused the VMIC to become more negative or more NR2C/D like. Figure 3D is the summary showing that UBP141 significantly shifted the mean VMIC, measured when the currents evoked by NMDA were maximal (see Methods), from -50 ± 2 mV to -45 ± 2 mV (n = 15, p < 0.01 for paired t-test) while EAB318 significantly changed the mean VMIC to -53 ± 2 mV (n = 15, p < 0.05 for paired t-test).[](https://www.ncbi.nlm.nih.gov/mesh/D002264)
To rule out the possibility that the sequence of antagonist co-application with NMDA may have some effects on the g(-90 mV)/g(MIC) ratio, we grouped the experiments according to the order of drug application. In 9 of 15 cells tested, UBP141 was co-applied with NMDA before EAB318 co-application with NMDA. In 6 of 15 cells tested, UBP141 was co-applied after EAB318. UBP141 significantly decreased the g(-90 mV)/g(MIC) from 0.23 ± 0.04 to 0.13 ± 0.02 (n = 9, p < 0.01 for paired t-test) and EAB318 significantly increased the g(-90 mV)/g(MIC) from 0.23 ± 0.04 to 0.37 ± 0.07 (n = 9, p < 0.01 for paired t-test) when UBP141 was applied first (Figure 3E left side). Similarly, EAB318 significantly increased the g(-90 mV)/g(MIC) from 0.23 ± 0.04 to 0.33 ± 0.06 (n = 6, p < 0.05 for paired t-test) and UBP141 significantly decreased the g(-90 mV)/g(MIC) from 0.23 ± 0.04 to 0.17 ± 0.02 (n = 6, p < 0.05 for paired t-test) when EAB318 was applied earlier (Figure 3E right side).[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
## Comparison of NMDA receptor types between NK1R+ and NK1R- neurons
*In the **Comparison of NMDA receptor types between NK1R+ and NK1R- neurons** section:*
Having verified that these two approaches, Mg2+ sensitivity and pharmacological blockade, allowed us to distinguish different NMDA receptor subtypes, we asked if NK1+ and NK1- neurons differed in the proportions of NR2A/B and NR2C/D containing receptors that they expressed. Indeed, although co-application of NMDA with either UBP141 or EAB318 significantly changed the g(-90 mV)/g(MIC) ratios, the ratio changes were not the same for all neurons tested. Figure 4A1 shows the UBP141-induced change in the g(-90 mV)/g(MIC) ratio in individual lamina I neurons pre-identified as either NK1R+ or NK1R- neurons. The g(-90 mV)/g(MIC) ratios of most neurons showed high sensitivity to UBP141, suggesting that most dorsal horn neurons express some NMDA receptors that include NR2C/D subunits. On average, NK1R+ neurons had a significantly higher g(-90 mV)/g(MIC) ratio than NK1R- neurons (0.26 ± 0.03, n = 11 v.s. 0.16 ± 0.02, n = 4, p < 0.05 for unpaired t-test), indicating that in these neurons, a higher percentage of NMDA receptors include NR2C/D subunits than in NK1R- neurons (Figure 4A2).[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
NK1R+ neurons express higher proportion of NMDA receptors with NR2C/D subunits than do NK1R- neurons. (A1) Individual NK1R+ and NK1R- neurons showed different degree of g(-90 mV)/g(MIC) ratio decrease following exposure to UBP141. (A2) Overall, NK1R+ neurons show significantly higher g(-90 mV)/g(MIC) ratio than NK1R- neurons (p < 0.05). (B1) Individual NK1R+ and NK1R- neuron showed increased g(-90 mV)/g(MIC) ratio when NMDA was applied in the presence of EAB318. (B2) In summary, EAB318 significantly increased g(-90 mV)/g(MIC) ratios in both NK1R+ and NK1R- lamina I neurons.[](https://www.ncbi.nlm.nih.gov/mesh/D002264)
To confirm this observation, we also observed the effect of EAB318 on the g(-90 mV)/g(MIC) ratio of individual neurons in Figure 4B1. EAB318 increased the g(-90 mV)/g(MIC) ratio in most, but not all dorsal horn neurons tested, again suggesting that most of them expressed NMDA receptors with both high and low Mg2+ sensitivity, consistent with the data of UBP141. On average, EAB318 caused the g(-90 mV)/g(MIC) ratio measured from NK1R+ neurons to become more NR2C/D like than from NK1R- neurons (Figure 4B2), consistent with the interpretation that NK1R+ neurons express a higher percentage of NMDA receptors with NR2C/D subunits.[](https://www.ncbi.nlm.nih.gov/mesh/C498147)
## Discussion
*In the **Discussion** section:*
We have identified NMDA receptor subtypes expressed by two populations of dorsal horn neurons; NK1R+ and NK1R- lamina I neurons. Based on our experiments, both highly Mg2+ sensitive (NR2A/B) and poorly Mg2+ sensitive (NR2C/D) NMDA receptors are expressed by NK1R+ neurons. NR2C/D subunits are less strongly expressed by NK1R- lamina I neurons and therefore the NR2A/B receptor subtypes dominate more strongly there.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
## Ratio assay confirmed by pharmacology
*In the **Ratio assay confirmed by pharmacology** section:*
The main approach for identification of NR2 subunit expression by different neurons in our study was to establish and apply a ratio assay for NMDA receptor-mediated currents recorded in 100 μM Mg2+. After identifying conditions to minimize the change of the g(-90 mV)/g(MIC) ratio during activation of NMDA receptors, including recording in the presence of Ba2+ and using low concentrations of agonist, it was possible to repeatedly measure the g(-90 mV)/g(MIC) throughout the duration of NMDA application with minimal variation in the ratio in many of the neurons tested. The ratio values observed, particularly in NK1R+ neurons, indicated that functional receptors composed of NR2A/B and NR2C/D subunits are present. The reversible shift of the ratios to larger values in the presence of the NR2A/B antagonist, EAB318, and to smaller values in the presence of the NR2C/D preferring antagonist, UBP141, confirm this interpretation. In addition, the action of these drugs in shifting the measured g(-90 mV)/g(MIC) ratio in the predicted direction strongly supports the use of this ratio assay to identify natively expressed NR2 subunits.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
## NR2A/B and NR2C/D subunits expressed by subpopulations of lamina I neurons
*In the **NR2A/B and NR2C/D subunits expressed by subpopulations of lamina I neurons** section:*
While published evidence suggests expression of both NR2A/B and NR2C/D subunit types in the dorsal horn generally, our data, collected on subpopulations of lamina I neurons, show cell specific differences. Previous reports indicate that NMDA receptors with NR2A and NR2B subunits are expressed in superficial dorsal horn based on in situ hybridization [31-33], single cell PCR [34] and immunostaining [14,35,36] studies. Our observations show strong evidence of NR2A and/or NR2B expression in both NK1R+ and NK1R- lamina I neurons.
Earlier studies suggest that NR2D mRNA is expressed by many and NR2C mRNA by few dorsal horn neurons [34]. In addition, more NR2D mRNA is expressed in adult dorsal horn and embryonic spinal cord than NR2C mRNA [5,37,38]. Further support for the presence of NR2D is that NMDA receptors with NR2D-like single channel conductance have been reported for lamina II neurons in rat dorsal horn [39,40]. Based on our experiments, we have found that NK1R+ neurons express NR2C/D subunits more strongly than the NK1R- neurons.
While it remains uncertain which NMDA receptors with low Mg2+ sensitivity are expressed by these lamina I neurons, NR2D is the best candidate.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
NK1R+ lamina I neurons represent a comparatively uniform population of neurons that are predominantly projection neurons [21]. The NK1R- neuron population is heterogeneous, including inhibitory and excitatory interneurons as well as a small population of NK1R- projection neurons [41]. Within the NK1R- population of neurons, some of the variability of NR2 subunit identity may represent different receptor configurations on different subpopulations of dorsal horn neurons.
At the whole cell level, particularly for NK1R+ neurons, we have evidence that NMDA receptors with NR2C/D subunits are present. NMDA receptors with these less Mg2+ sensitive NR2 subunits could be expressed at synapses, extrasynaptically or both. Momiyama (2000) has suggested an extrasynaptic localization of NR2D containing NMDA receptors by lamina II neurons in the dorsal horn. Because of their higher binding affinity with glutamate, these extrasynaptic receptors may be more sensitive to ambient glutamate levels in the extracellular space that could accumulate due to glial release [42,43], spill over associated with high amounts of activity, and to injury [44]. Activation of these receptors would be expected to have a potent impact on neuronal cell function due to their lowered Mg2+ sensitivity, prolonged time over which they open following glutamate binding, and lack of desensitization [5,6,45].[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
## Other factors that could influence NMDA receptor conductance ratio
*In the **Other factors that could influence NMDA receptor conductance ratio** section:*
One concern with our approach to NR2 subunit identification is the possibility that changes in membrane currents secondary to NMDA receptor activation will alter the g(-90 mV)/g(MIC). It is because of this concern that we have recorded in low Ca2+ solution with added Ba2+ and limited our analysis to those neurons showing no change in g(-90 mV)/g(MIC) while NMDA washes on and off the spinal cord slices. Even more importantly, we have used pharmacological tools as an independent test of subunit composition under these carefully controlled drug application conditions. In some of the neurons excluded from these studies, NMDA-induced currents showed strongly increased g(-90 mV)/g(MIC) ratios during wash-out of NMDA (data not shown). The underlying mechanism for this is not clear.
For the data that met the criteria for our study, we have confirmed identification of subunit composition by the use of NMDA receptor specific compounds to alter conductance ratio in predictable ways. The opposing effects of EAB318 and UBP141 on g(-90 mV)/g(MIC) supports our interpretation of conductance ratio in terms of subunit composition.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Significance
*In the **Significance** section:*
We have taken advantage of the Mg2+ sensitivity of NMDA receptors to identify NMDA receptors of different NR2 subunits in identified subpopulations of lamina I neuorns and confirmed this with pharmacology. We show that individual neurons express NMDA receptors with different NR2 subunits at different ratios. When comparing identified populations of lamina I neurons, NK1R+ neurons express a higher mean ratio of NR2C/D type NMDA receptors compared with NK1R- neurons. NR2D has been suggested to have a role in the development of allodynia or hyperalgesia in several different pain models [12] and lamina I, NK1R+ neurons are importantly involved in the expression of allodynia [46]. In this context, it is possible that these receptors may contribute to development of NR2D-dependent allodynia.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interests.
## Authors' contributions
*In the **Authors' contributions** section:*
CKT, EJK and ABM conceived and designed the experiments. CKT and ABM analyzed the data and wrote the manuscript. CKT carried out the experiments. EJK provided the essential compounds UBP141 and EAB318. All authors read and approved the final manuscript.[](https://www.ncbi.nlm.nih.gov/mesh/D002264)
|
# Introduction
Association of the [Asn](https://www.ncbi.nlm.nih.gov/mesh/D001216)306[Ser](https://www.ncbi.nlm.nih.gov/mesh/D012694) variant of the SP4 transcription factor and an intronic variant in the β-subunit of transducin with digenic disease
# Abstract
*In the **Abstract** section:*
Purpose
SP4 is a transcription factor abundantly expressed in retina that binds to the GC promoter region of photoreceptor signal transduction genes. We have previously shown that SP4 may be involved in the transcriptional activation of these genes alone or together with other transcription factors such as SP1, neural retina leucine zipper protein (NRL), and cone-rod homeobox gene (CRX). Since mutations in NRL and CRX are involved in inherited retinal degenerations, SP4 was considered a good candidate for mutation screening in patients with this type of diseases. The purpose of this work, therefore, was to investigate possible mutations in SP4 in a cohort of patients affected with different forms of retinal degenerations.
Methods
270 unrelated probands with various forms of retinal degeneration including autosomal dominant and autosomal recessive retinitis pigmentosa (RP), autosomal dominant and autosomal recessive cone-rod dystrophy (CRD), and Leber's congenital amaurosis (LCA), were screened for mutations in the SP4 gene. Single strand conformation polymorphism (SSCP) analysis was performed on the six SP4 gene exons including flanking regions followed by direct sequencing of SSCP variants.
Results
Nine different sequence variants were found in 29 patients, four in introns and five in exons. Many of the probands were previously screened for mutations in the genes encoding the α-, β- and γ-subunits of rod-specific cGMP phosphodiesterase (PDE6A, PDE6B, PDE6G), the β-subunit of rod-specific transducin (GNB1), and peripherin/rds (RDS). One group of seven probands of Hispanic background that included five with arRP, one with RP of unknown inheritance (isolate) and 1 with arCRD carried an Asn306Ser mutation in SP4. Of the seven, the isolate case was homozygous and the other 6 heterozygous for the variant. Two arRP and the arCRD probands carried an additional intronic GNB1 variant. DNA from the family members of the arCRD proband could not be obtained, but for the other two families, all affected members and none of the unaffected carried both the SP4 Asn306Ser allele and the GNB1 intronic variant.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
Conclusions
If mutations in SP4 do cause retinal degenerative disease, their frequency would be low. While digenic disease with the SP4 Asn306Ser and the GNB1 intronic variant alleles has not been established, neither has it been ruled out. This leaves open the possibility of a cooperative involvement of SP4 and GNB1 in the normal function of the retina.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Transcription factors have been shown to play an important role not only in the biology of photoreceptors and other retinal cells, but also as sites of mutations causing degenerative disease. For example, mutations in CRX and NRL cause various forms of progressive retinal dystrophies [1-4]. The SP family of transcription factors (SP1-5) is formed by a group of proteins that selectively bind to the "GC box" in the promoter region of many genes [5-7] via three putative zinc finger domains of the C2H2 type [8]. Based on sequence similarities, SP1, SP3, and SP4 are more closely related to each other than to SP2 and SP5 [6,9] and the three have similar affinity for the GC box
[10,11]. While SP1 and SP3 are ubiquitously expressed, SP4 is most abundant in the developing nervous system, particularly in the hippocampus [12] and retina [13], although it is also expressed in other tissues [10,14]. Sp1 and Sp3 knockout mice all die after E10 or at birth, respectively.
Sp4 knockout mice appear to develop normally to birth, but after birth, many pups die by P28 and those that survive are small and have other abnormalities [12]. In the last few years, we have studied extensively the SP4 transcription factor and have found that it is present in all retinal layers, interacts with CRX and NRL and activates transcription of several rod specific genes including PDE6B and RHO, encoding the β-subunit of cGMP-phosphodiesterase (β-PDE) and rod opsin, respectively [13,15,16]. Because of its specific involvement in transcription of rod genes and because of the history of transcription factor mutations causing retinal degeneration, we considered SP4 a good candidate gene to screen for missense mutations in patients with various forms of retinal degeneration. To this end, we screened the 6 exons of the SP4 gene in a group of 270 patients with various forms of retinal degeneration that had been screened previously for a number of photoreceptor genes [17-21].
Although we could not establish that mutations in the SP4 gene cause retinal degenerative disease, neither could we rule this possibility out because the families of two patients in which an SP4 missense mutation and a GNB1 intron 2 variant were present, segregated with disease. Interestingly, the inherited retinal degeneration of the Rd4 mouse is caused by an inversion of mouse chromosome 4, and the site of the telomeric breakpoint is precisely on intron 2 of the Gnb1 gene [22].
## Methods
*In the **Methods** section:*
## Patients
*In the **Patients** section:*
270 patient probands of mixed ethnicities (56% European, 17% Asian, 13% Black, 14% Hispanic) were screened for variants in the six exons of the SP4 gene, including 49 with autosomal dominant retinitis pigmentosa (adRP), 103 with autosomal recessive retinitis pigmentosa (arRP), 26 with autosomal dominant cone-rod dystrophy (adCRD), 52 with autosomal recessive cone-rod dystrophy (arCRD), and 40 with Leber's congenital amaurosis (LCA). Many of the above patients had been previously screened for mutations in the genes encoding rod- αPDE, βPDE, γPDE, rod β-transducin and RDS-peripherin. 95 controls with a similar ethnic distribution (58% European, 16% Asian, 13% Black, 13% Hispanic) were screened for each of the above genes and SP4 as well. Written informed consent was obtained in compliance with the tenets of the declaration of Helsinki and with the approval of the office of Human Research Protection of the School of Medicine, University of California, Los Angeles.
## Polymerase chain reaction
*In the **Polymerase chain reaction** section:*
Blood was drawn in 10-20 ml aliquots and DNA was extracted from the leukocytes by standard methods. Initial screening was done by SSCP as described previously [17-21]. The exons of SP4 were amplified by polymerase chain reaction (PCR) directly from genomic DNAs with appropriate primers pairs. Each PCR amplicon included 50-150 nt of intronic flanking sequence on each side of the exon. The PCR protocol was 94 °C for 3 min followed by 30 cycles of 94 °C for 45 s, 55-60 °C for 45 s and 72 °C for 45 s, followed by 5 min at 72 °C.
The sequences of primer pairs are presented in Table 1.
Primer sequences used for SP4 and GNB1 screening.
## Single strand conformation polymorphism
*In the **Single strand conformation polymorphism** section:*
Amplicons were separated by electrophoresis in 7% acrylamide gels and analyzed by standard P32 autoradiography or silver staining methods to reveal polymorphisms as described previously [17-21].[](https://www.ncbi.nlm.nih.gov/mesh/D020106)
## Sequencing
*In the **Sequencing** section:*
Amplicons carrying polymorphisms were purified using the QIA QUICK PCR purification kit (Qiagen, Valencia, CA) and sequenced using the Dyenamic ET Terminator cycle sequencing kit (Amersham, Piscataway, NJ).
## Results
*In the **Results** section:*
SSCP screening of the SP4 gene showed 9 sequence variants in 29 patients, five present in exons (Table 2). The heterozygous Leu241Val and Pro286Ala missense variants, both present in exon 3 of arRP probands did not segregate with disease in the corresponding families. An Asn306Ser missense variant in exon 3 was present in both alleles of one isolate RP proband, in one allele of five arRP probands and in one arCRD proband. This missense variant was also present in 1/95 controls. Interestingly, although only 14% of the 270 patients and 13% of the 95 controls were Hispanic, all seven patients and the 1 control that carried Asn306Ser were Hispanic.
Thus, 18.4% (7/38) of the Hispanic patients carried Asn306Ser while none of the other patients did, including 0/151 patients of European origin. The other 2 coding region variants were both silent. Neither was present in 95 controls. The remaining 4 intronic sequence variants were present in patients and absent from controls with the exception of -121 A to C which was present in one control (Table 2).[](https://www.ncbi.nlm.nih.gov/mesh/D007930)
Sequence variants detected in the screening of the SP4 gene.
Table 3 shows the results of previous screenings of several photoreceptor genes in the six probands with the Asn306Ser mutation in one allele. Patient 856 has arCRD while the other 5 patients have arRP. Three of the six probands also carried an A-G variant in intron 2 of the GNB1 gene. DNAs of the family members of one of these probands could not be obtained (family 2177). However, in the families of the other two probands, the SP4 missense and the GNB1 intronic variant segregated with disease (Figure 1A,B).
None of the other variants in the screened genes segregated with disease. We found no additional variants in the genes encoding α-, β- and γ-cGMP-phosphodiesterase, RDS/peripherin or the β-subunit of transducin in the RP isolate patient homozygous for Asn306Ser.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
Sequence variants in five genes of six patients with the heterozygous SP4 Asn306Ser mutation.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
Pedigrees number 485 and number 1526 segregate the SP4/GNB1 alleles. Filled symbols indicate individuals with retinitis pigmentosa. Probands are indicated by arrows. The allele designation SP4 denotes the Asn306Ser allele, GNB1 denotes the A-G substitution in intron 2 of the GNB1 gene, + denotes normal alleles.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
## Discussion
*In the **Discussion** section:*
Even though 2/3 of mice born with the Sp4 gene deleted die within the first four weeks of life [14], the surviving mice have many abnormalities including severe retinal degeneration (our data, not shown). Therefore, we considered the human SP4 gene a good candidate for the site of missense mutations causing retinal degeneration, given its involvement in the transcription of several photoreceptor genes including PDE6B [13,15,16] and the phenotype of the Sp4 knockout mouse.
We found nine unique sequence variants in the SP4 gene of 29 patients affected with different types of inherited retinal degenerative disease. Of these, five variants were in one of the six exons of SP4. Two of the variants coded for the same amino acid (Ala276Ala and Gln451Gln). Two more variants predicted amino acid changes in the SP4 protein, Leu241Val and Pro286Ala, but neither segregated with disease in the corresponding families. The fifth missense Asn306Ser mutation was present in seven probands.
One of the probands that had the homozygous Asn306Ser mutation was an isolate with RP, so we could not tell if the two Asn306Ser alleles were causing disease. In family 449, an affected sibling of the proband did not carry the Asn306Ser allele; for the probands of families 856 (arCRD) and 1824, neither a second variant SP4 allele nor a variant in any of the other genes previously screened could be identified. The three remaining probands all carried Asn306Ser and an intronic A-G substitution 103 bp upstream of the 3' splice site of intron 2 of the GNB1 gene. We could not obtain DNAs from the family of one of the probands (2177), but in the families of the other two probands (485 and 1526) only the two affecteds in each pedigree carried both alleles (Figure 1A,B). Both of these families had arRP.
All seven families carrying Asn306Ser were of Hispanic background and so was the 1 control carrying the same mutation. Thus, 18.4% of the Hispanic patients carried this allele while none of the patients from other backgrounds did (0/151 Europeans, 0/35 Blacks and 0/46 Asians). The higher frequency of Asn306Ser in Hispanic patients (p<0.001 applying the Fisher's exact test) compared to a relatively low frequency in Hispanic controls (1/12=8.3%) suggests that this variant may be pathogenic.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
There are several additional reasons that implicate SP4 Asn306Ser as a mutant allele that may contribute to autosomal recessive disease. (1) We have no family history for the proband that carried two alleles of Asn306Ser. Therefore, the two Asn306Ser alleles together could be the cause of that isolate patient's disease. (2) In each of the three families where the Asn306Ser allele did not segregate with disease, two mutant alleles in other genes may have rendered the presence of the heterozygous Asn306Ser coincidental and unrelated to disease. However, this does not rule out the possibility that two Asn306Ser alleles could cause disease.
(3) For the two families where only the affecteds carried both the Asn306Ser allele and the GNB1 variant intronic allele, pathogenesis is possible at least genetically. Furthermore, GNB1 has three GC boxes in its promoter and SP4 interacts with GC boxes. Thus, digenic disease may be plausible. To answer the question of pathogenicity of Asn306Ser, functional assays of the protein carrying this variant would have to be conducted. Nevertheless, the possibility that Asn306Ser may be pathogenic is supported by the fact that Asn306Ser is in the transactivation domain of the SP4 protein and in one of six glycosylation sites (N-X-S/T; Figure 2).
Changing asparagine to serine eliminates this site of posttranslational modification and this may affect the function of the protein. Interestingly, at position 306 there is an asparagine only in the human SP4 sequence while in the mouse, rat, dog and cow there is a threonine. Therefore, asparagine is not a conserved residue. With regard to the GNB1 intronic variant, it is not in a splice site or a consensus branch point, but it may be in a heretofore-unknown regulatory region.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
Partial nucleotide sequence of the cDNA encoding human SP4 and the corresponding predicted amino acid sequence. Boxes indicate potential N-glycosylation sites (N-X-S/T). Asterisk (*) indicates the location of the A to G mutation (AAC to AGC causes the change of Asn306 to Ser).[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
For the intronic GNB1 variant, a DNA fragment including exon 2, intron 2 (carrying the variant), and exon 3 would have to be expressed to determine if this variant caused a splicing problem. However, the sequences adjacent to the A-G substitution do not correspond to consensus branch point sequences. Another possibility is that the A-G substitution would disrupt an enhancer or a repressor sequence causing altered expression of the GNB1 gene. Although there is no direct evidence that the Asn306Ser mutation in the SP4 gene and the intron 2 A-G variant of the GNB1gene together are responsible for disease in the affected individuals, digenic disease cannot be ruled out without further testing the pathogenicity of the alleles. It is certainly plausible that the protein products of a phototransduction gene like GNB1 and a transcription factor that may influence its expression like SP4 can together cause digenic disease when one allele of each carries a mutation.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
# References
*In the **References** section:*
|
# Introduction
Efficacy of [artesunate-amodiaquine](https://www.ncbi.nlm.nih.gov/mesh/C515299) for treating uncomplicated falciparum malaria in sub-Saharan Africa: a multi-centre analysis
# Abstract
*In the **Abstract** section:*
Background
Artesunate and amodiaquine (AS&AQ) is at present the world's second most widely used artemisinin-based combination therapy (ACT). It was ne[cessary to evaluate the ef](https://www.ncbi.nlm.nih.gov/mesh/C515299)fi[cacy ](https://www.ncbi.nlm.nih.gov/mesh/C515299)of ACT, recently adopted by the World Health Organiz[ation (WHO)](https://www.ncbi.nlm.nih.gov/mesh/D037621) and deployed over 80 countries, in order to make an evidence-based drug policy.
Methods
An individual patient data (IPD) analysis was conducted on efficacy outcomes in 26 clinical studies in sub-Saharan Africa using the WHO protocol with similar primary and secondary endpoints.
Results
A total of 11,700 patients (75% under 5 years old), from 33 different sites in 16 countries were followed for 28 days. Loss to follow-up was 4.9% (575/11,700). AS&AQ was given to 5,897 patients. Of these, 82% (4,826/5,897) were included in randomized comparative trials with polymerase chain reaction (PCR) genotyping results and compared to 5,413 patients (half receiving an ACT).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
AS&AQ and other ACT comparators resulted in rapid clearance of fever and parasitaemia, superior to non-ACT. Using survival analysis on a modified intent-to-treat population, the Day 28 PCR-adjusted efficacy of AS&AQ was greater than 90% (the WHO cut-off) in 11/16 countries. In randomized comparative trials (n = 22), the crude efficacy of AS&AQ was 75.9% (95% CI 74.6–77.1) and the PCR-adjusted efficacy was 93.9% (95% CI 93.2–94.5). The risk (weighted by site) of failure PCR-adjusted of AS&AQ was significantly inferior to non-ACT, superior to dihydroartemisinin-piperaquine (DP, in one Ugandan site), and not different from AS+SP or AL (artemether-lumefantrine). The risk of gametocyte appearance and the carriage rate of AS&AQ was only greater in one Ugandan site compared to AL and DP, and lower compared to non-ACT (p = 0.001, for all comparisons). Anaemia recovery was not different than comparator groups, except in one site in Rwanda where the patients in the DP group had a slower recovery.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Conclusion
AS&AQ compares well to other treatments and meets the WHO efficacy criteria for use against falciparum malaria in many, but not all, the sub-Saharan African countries where it was studied. Efficacy varies between and within countries. An IPD analysis can inform general and local treatment policies. Ongoing monitoring evaluation is required.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Background
*In the **Background** section:*
Artemisinin-based combination therapy (ACT) is now the treatment of choice for uncomplicated Plasmodium falciparum malaria. As of February 2009, more than 80 countries worldwide have adopted ACT as first-line therapy. Currently, four forms of ACT are recommended by the World Health Organization (WHO): artemether and lumefantrine (AL), artesunate and amodiaquine (AS&AQ), artesunate and mefloquine (AS+MQ) and artesunate and sulphadoxine-pyrimethamine (AS+SP). The choice of ACT for a country or a region depends on a number of considerations. A critical element is the level of underlying resistance to the longer-acting partner drug in the combination.
This is particularly important for amodiaquine (AQ) and sulphadoxine-pyrimethamine (SP) in Africa, where both drugs have been widely used as monotherapies.[](https://www.ncbi.nlm.nih.gov/mesh/D037621)
The WHO recommends that countries use ACT, which is at least 90% effective and introduce new forms of ACT that are at least 95% effective after discounting reinfections (PCR-adjusted) and that the Day 28 efficacy of respective partner drugs alone should exceed 80%. Concerns have been raised over ACT including amodiaquine (AQ) meeting such criteria in areas where AQ has been widely used as monotherapy. The second most widely used ACT, AS&AQ has been adopted as first-line treatment in 18 countries in Africa (Burundi, Cameroon, Chad, Congo, Côte d'Ivoire, Democratic Republic of Congo, Equatorial Guinea, Gabon, Ghana, Guinea, Liberia, Madagascar, Mauritania, Sénégal, Sao Tomé & Principe, Sierra Leone, Sudan [South], Zanzibar) and Indonesia. The AS&AQ combinations have been available in a non-fixed formulation (AS+AQ) as either a loose combination or as blister-packed tablets from several pharmaceutical companies, and more recently, as a fixed-dose drug combination (ASAQ). Therefore, the widespread use of these combinations calls for a comprehensive synthesis of published and unpublished available results to properly inform policy decisions.[](https://www.ncbi.nlm.nih.gov/mesh/D000655)
The efficacy and tolerability of AS&AQ has been tested formally in several clinical trials in different epidemiological settings in Africa. Following a systematic review of published and unpublished comparative and non-comparative trials (Olliaro et al, personal communication), investigators were contacted for individual patient data. This resulted in a pooled multi-centred analysis (; Bonnet et al, unpublished data, 2004; van den Broek, unpublished data, 2005; Cohuet et al, unpublished data, 2004; Grandesso, unpublished data, 2004), which cannot be strictly defined as a meta-analysis since it does not include exhaustively all the trials with AS&AQ, an element that may introduce a selection bias. However, to date, this analysis with 26 trials enrolling 11,700 patients mostly in randomized comparative trials with genotyping, is the largest analysis at the individual patient level ever compiled for an ACT.
In addition to examining the primary parasitological outcomes, this study also analysed the resolution of parasitaemia, fever, gametocytaemia and anaemia, and modelled risk factors for treatment failure.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Methods
*In the **Methods** section:*
## Study sites, design and patients
*In the **Study sites, design and patients** section:*
The studies were identified through a systematic review of comparative and non-comparative clinical trials conducted in sub-Saharan Africa, using any formulation of AS&AQ for treating uncomplicated falciparum malaria with follow-up of at least 28 days, regardless of language or publication status (published, unpublished, in press, technical reports) (Olliaro et al, personal communication).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Published studies were identified through electronic searches up to April 2007 of MEDLINE, EMBASE, LILACS, the Cochrane Infectious Diseases Group's trials register and the Cochrane Central Register of Controlled Trials (CENTRAL) using the following search terms: malaria, amodiaquine, artesunate and artemisinins. Unpublished studies were identified through personal contacts and by manually searching the reference lists of studies identified by the above-mentioned methods, contacting individual researchers working in the field, and examining WHO records. For all studies identified, the corresponding author was contacted and asked to provide individual patient data.[](https://www.ncbi.nlm.nih.gov/mesh/D000655)
The following aspects of methodological quality of the received data sets/publications were assessed: generation of allocation sequence, adequacy of concealment of the allocation of treatment, degree of blinding, and completion of follow-up. Generation of the allocation sequence and allocation concealment were classified as adequate, inadequate, or unclear. Blinding was classified as open, single, or double. Losses to follow-up (regardless of reasons) were computed and considered adequate if less than 10%. Finally, based on the presented power calculation, sample size estimation was assessed, as was whether an intention-to-treat analysis could be computed.
The last patient included in this analysis was enrolled in December 2006. Studies involving pregnant women or severe malaria, studies performed outside sub-Saharan Africa, as well as economic and pharmacokinetic analyses, were excluded.
A total of 46 trials were identified, of which 42 compared AS&AQ with other anti-malarial drugs; four were non-comparative. Seven studies were excluded, four because follow-up was limited to 14 days, and three because they were not in Africa (Afghanistan, Colombia, Indonesia). Of the remaining, 39 identified studies fulfilling the inclusion criteria, 14 had to be excluded because individual patient data were not made available. The remaining 25 studies, and a randomized comparative trial conducted in a common site in Uganda-Apac comprised a total of 11,700 patients treated with either a non-fixed AS+AQ (N = 4,914) or fixed ASAQ (N = 1,073). In randomized comparative trials the AS&AQ groups (82% N = 4,826/5,897) were compared to anti-malarial drugs: AQ (N = 648), AS (N = 279), AS+SP (N = 1,005), AQ+SP (N = 1,257), AL (N = 1,319), dihydroartemisinin-piperaquine (DP, N = 463), chloroquine + SP (CQ+SP, N = 699), or quinine + chloroquine (Q+CQ, N = 43)
(Figure 1).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Flow chart of comparative and non comparative studies by anti-malarial drug.
## Treatments
*In the **Treatments** section:*
AS&AQ treatment regimens. AS&AQ products were either loose or co-blister-packed combinations, individually formulated products (AS+AQ), or fixed-dose co-formulations (ASAQ). In general, the loose AS+AQ were dosed based on body weight, while in a few studies the co-blister-packed AS+AQ and the co-formulated ASAQ were based on age and weight range.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The majority (82%, 4,914/5,987) of the patients were treated with individually formulated AS and AQ. The target dose was AS 12 mg/kg over 3 days and AQ 30 mg/kg over 3 days, except in Uganda where AQ was given at 25 mg/kg (10 mg/kg on Days 0 and 1 then 5 mg/kg on Day 2). The co-blister-packed AS+AQ (AS 50 mg + AQ 153 mg base for each of the days of treatment, dose ratio = 3.1) was used in two studies in Senegal containing for each of the days of treatment 1 tablet of AS 50 mg and 1 of AQ 153 mg (base), dosed either by age or by weight. When used by age, the dosing categories were: (i) <1 y: 1/2 tablet; (ii) ≥ 1 to 6 y: 1 tablet; (iii) ≥ 7 to <13 y: 2 tablets; (iv) ≥ 13 y: 4 tablets.
The fixed-dose combinations of ASAQ were available as two-and three-tablet strength products given by age. The fixed-dose combination was also given either once or twice a day. For the two-tablet strength fixed-dose combination ASAQ (paediatric AS 25 mg + AQ 67.5; adult AS 100 mg + AQ 270 mg, dose ratio = 2.7), the dosing categories were: (i) 0–1 months: 1/2 paediatric; (ii) 2–11 months: 1 paediatric; (iii) 1–6 years: 2 paediatric; (iv) 7–13 years: 1 adult; and (v) ≥ 14 years: 2 adult. For the three-tablet strength ASAQ, age- and weight-based doses were administered once-a-day for three days: one tablet/day for children up to 13 years of age (≤ 35 kg) or two tablets/day for adolescents aged 14 years and above and adults (≥ 36 kg). Doses available were: infants (2 to 11 months) received AQ 25 mg/AS 67.5 mg; young children (1 to 4 years) received AQ 50 mg/AS 135 mg; children (6 to 13 years) received 1 tablet/day of AQ 100 mg/AS 270 mg, and adults (14 years or more) received two tablets (AQ 100 mg/AS 270 mg) per day.[](https://www.ncbi.nlm.nih.gov/mesh/D000077332)
## Comparator treatment regimens
*In the **Comparator treatment regimens** section:*
(i) for the ACT groups: AL (20 mg artemether/120 mg lumefantrine given according to weight as 1 [5–14 kg], 2 [15–24 kg], 3 [25–34 kg], and 4 [≥ 35 kg] tablets given twice daily for 3 days); DP (around 2.3 mg/kg/day dihydroartemisinin and 18.4 mg/kg for 3 days); AS+SP (AS 4 mg/kg/day; SP 25 mg/kg of sulphadoxine and 1.25 mg/kg/of pyrimethamine administered in a co-formulated tablet [SP] as a single dose)[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
(ii) for the non-ACT groups: AQ+SP (AQ 10 mg/kg/day for 3 days and SP 25 mg/kg of sulphadoxine and 1.25 mg/kg/of pyrimethamine administered in a co-formulated tablet [SP] as a single dose); CQ (25 mg/kg chloroquine over 3 days) and SP; AQ only (10 mg/kg/day for 3 days); AS5 only (AS 12 mg/kg over 5 days).[](https://www.ncbi.nlm.nih.gov/mesh/C001205)
## Study endpoints and statistical analysis
*In the **Study endpoints and statistical analysis** section:*
The primary endpoint was treatment efficacy by Day 28 defined prospectively in all studies as the treated population free of failure (PCR not adjusted: recurrence, or PCR confirmed: recrudescence). Data were standardized in order to be pooled and to allow for a modified intention-to-treat analysis (mITT). Patients lost to follow-up (or missing a weekly visit) or with any P. falciparum infection during the follow-up were censored for the primary outcome at the time they were last seen. Efficacy was measured using Kaplan-Meier survival analysis.
Patients were censored as non-failures when last seen if: (a) lost to follow-up with no evidence of failure, (b) having withdrawn consent, (c) taking a drug with anti-malarial activity, (c) having another illness or being a protocol violation, or (e) having never started the assigned treatment.
A treatment failure was defined as any of the following: (i) the development of danger signs or severe malaria or death or drug-induced vomiting requiring rescue treatment (ii) Day 2 parasitaemia > Day 0 parasitaemia, (iii) Day 3 parasitaemia >25% of the baseline, (iv) Day 7 parasitaemia, and (v) a recurrent parasitaemia within 28 days i.e. a conversion from a positive to a negative smear result sustained to Day 28. Patients who had a treatment failure with missing PCR samples or indeterminant PCR results were classed as recrudescent failures to prevent from overestimating efficacy levels. Treatment failure was considered the sum of early and late treatment failures, as defined by the WHO.
The secondary outcomes and analytical method by Day 28:
(i) Parasitaemia clearance times (negative slide)
(ii) Elevated temperature (≥ 37.5°C)
(iii) The risks of recurrence or recrudescence (using PCR) in AS&AQ groups compared to the comparator groups were assessed by Cox regression stratified by site in an attempt to account for potential statistical heterogeneity (assessments made consistently within each study but at different times across studies), and presented as Adjusted Hazard Ratio (AHR).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
(iv) The predictors of recurrent and recrudescent infections in AS&AQ groups were similarly assessed by Cox regression stratified by site.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
(v) Gametocytaemia was defined as any positive slide for gametocytes, and was analysed as a binary variable. The predictors of patent gametocytaemia on admission were assessed by logistic regression controlled by sites. The overall cumulative incidence probability of gametocyte appearance on Day 28 was defined as the first positive slide during the follow-up being the censoring event and a measured by Kaplan-Meier survival analysis. The carriage rates were calculated in person-week-gametocyte (PGW), compared using the Mantel-Haenszel method to estimate a combined odds ratio between treatments and presented as Rate Ratio (RRMH). The risks of gametocyte appearance post-admission were assessed by Cox regression stratified by site.
(vi) Fractional change in haemoglobin or haematocrit between Days 0 and 7 and Days 0 and 28, and anaemia (the cut off was set at Hct <30% or a Hgb <10 mg/L) at baseline and recovery (when HCT became ≥ 30% or Hgb ≥ 10 mg/L) were compared using a student-t paired analysis. Student-t test was used for comparisons between paired mean results. Anaemia defined as <30% Hct or <10 g/dL was categorized into 4 grades:
a. grade 1 was ≥ 10 g/dL or ≥ 30% Hct
b. grade 2 was mild (8–9.9 g/dL, or 25%–29.9% Hct)
c. grade 3 was moderate (5–7.9 g/dL, or 15%–24.9% Hct)
d. grade 4 was severe anaemia (<5 g/dL, or <15% Hct)
Parasitaemia, fever, and gametocyte clearance times for each treatment group were compared using the Mann-Whitney test. The proportions of patients remaining febrile on Day 2 or with a positive slide on Day 3 in each treatment group were compared by logistic regression controlled by sites.
For patients who cleared parasites, fever, or anaemia (i, ii, vi respectively), the time of the first negative result (followed by negative counts) was taken as time of clearance. Patients were seen each day for as long as they stayed parasitaemic and thus parasite clearance could be assessed as those who cleared 24, 48, and 72 hours after treatment administration. However given the potential absence of data after treatment (72 hours) for fever and parasite clearance, multivariate analysis was focused on the proportion of patients remaining parasitaemic on Day 3 in comparative trials.
Additionally, any positive gametocyte counts (v) detected any time after treatment start defined gametocytes carriage. Gametocyte carriage rate was expressed in PGW per 1,000 person-weeks followed-up calculated as the total length of gametocyte carriage divided by the total number of persons exposed.
Spearman bivariate correlation was noted rs. Confidence intervals were calculated at 95%, and comparisons considered significant when P < 0.05. Data were analysed using Stata v10 (Stata Corp.).
## Heterogeneity
*In the **Heterogeneity** section:*
Differences in settings, age group, use of PCR, trial design and study procedures were included in the assessment of study heterogeneity using Cochran's Q test and the I2 test.
## PCR methods
*In the **PCR methods** section:*
PCR was performed on paired samples to compare the parasites' genotypes and thus distinguish between new and recrudescent infections. Allelic variation within the merozoite surface protein 1 and 2 (MSP1 and MSP2, respectively), and glutamate reach protein (GLURP) was used in Angola, Congo, DRC, Kenya, Senegal, Sudan, and Uganda. MSP1 and MSP2 were analysed in Angola, Burkina Faso, Gabon, Guinea, Madagascar, Mali, Democratic Republic of Congo (DRC), Rwanda, Sudan, Sierra Leone, and Uganda. MSP2 alone was used in Zanzibar and Uganda.
In Uganda, selected regions of the MSP1, MSP2 and 6 microsatellite markers were amplified using PCR and characterized based on sequence and size polymorphisms identified by gel electrophoresis.
A recrudescent infection was defined as one that matched in size at least one allele of both the MSP1 and MSP2 loci present in the first sample. Thus, if any clone of a polyclonal primary infection was detected during a second episode, this was considered to be a recrudescence and thus a treatment failure.
## Ethical issues
*In the **Ethical issues** section:*
All the studies had been approved by the relevant ethics and institution review committees.
## Results
*In the **Results** section:*
## Characteristics of included studies
*In the **Characteristics of included studies** section:*
This review extends from February 1999 in Kenya to December 2006 in Senegal and included only children aged between four and 59 months in half (13/26) of the studies. The proportion of patients lost to follow-up was <10% in all the included trials. A total of 11,700 patients were enrolled in 16 countries at 33 different sites. Individually, the trials enrolled between 27 and 890 ASAQ-treated patients. The total number of patients treated with ASAQ was 5,987, of which 82% (N = 4,896) were enrolled in randomized comparative trials.
In comparator arms (N = 5,713), 49% (N = 2,787) were treated with other ACTs and 51% (N = 2,926) treated with non-ACTs (Figure 1). The overall loss to follow-up by Day 28 was 4.9% (575/11,700).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Study design
*In the **Study design** section:*
Eighteen trials were randomized, comparative, and open label; Grandesso S, unpublished data, 2004; Cohuet S, unpublished data, 2004; Bonnet M, unpublished data, 2004; van den Broek I, unpublished data, 2005), four were single blinded, 1 was placebo-controlled, and 3 were non-comparative.
Twenty-two (22) of the 25 studies included applied the Consort guidelines. In the comparative trials, methods of assigning patients to treatments varied from not specified (N = 6), to allocation by bloc (N = 7), age group (N = 3), computer generation (N = 2), sequential alternation (N = 1), slip of paper (N = 1), stratified (N = 1), and variable blocks (N = 1). Treatment allocation was double blinded (N = 1), not reported (N = 3), not specified (N = 8), not done (N = 1), staff blinded (N = 7), and concealed in sealed envelopes (N = 2). In all cases, allocation was not disclosed to investigators at study site until a patient had given written consent to participate in the study. All treatments were supervised over the three-day course.
All studies considered followed patients for 28 days or more, but in order to standardize the analysis, only results up to 28 days were included.
In all the included studies, the primary treatment outcome was treatment efficacy. In 22 trials, parasites were genotyped to distinguish new from recrudescent infections (10,077/11,700; 86% of the patients), one of which was not comparative (126/11,700 or 1% of the patients). One study compared non-fixed and fixed combinations of AS&AQ. Comparative studies included 4,896 patients on AS&AQ and 5,713 on other anti-malarials: AL in 11 sites (24%), DP in four sites (8%), AS+SP in nine sites (18%) for ACT groups, and AQ+SP in eight sites (22%), CQ+SP in four sites (12%), AQ in seven sites (11%), AS in two sites (5%), and Q+CQ in one site (1%) (Figure 1).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Day 0 Hgb and Hct levels were available in 87.1% of all the patients (10,944/11,700), 46.9% of whom (N = 4,779) on AS&AQ. The mean change in Hgb or Hct (± SD) could be calculated using a student paired analysis in 2,405 patients (1,361 on AS&AQ), and in 5,388 patients on Day 28 (2,753 on AS&AQ) respectively. All trials recorded gametocyte carriage at study enrolment and during follow-up.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Two studies were multi-centre: (i) a double-blinded comparison of Placebo+AQ to AS+AQ in Senegal, Gabon, and Kenya; and (ii) an open-label comparison of a fixed-dose combination ASAQ vs. AL in Senegal (2 sites), and Cameroon, Madagascar, and Mali (1 site each). Three (12%, 3/25) unpublished reports from Epicentre (n = 3) were included, and represented 5% of the total patients (639/11,700). One study from Uganda was conducted in four geographical areas with different transmission intensities.[](https://www.ncbi.nlm.nih.gov/mesh/D000655)
## Demography
*In the **Demography** section:*
The majority of the patients treated with AS&AQ were between six months to five years of age (N = 4,153, 69%, of whom 20%, N = 1,177 were one year old or younger). The five to 14 years olds were 22% (N = 1,307), and adults (15 years or older), 8% (N = 449). The median age was three years, and the range was six months to 89 years. Most of the adults (81%, N = 364) were from Senegal. Uganda contributed the largest percentage of children by country (24%, N = 1,002).
In randomized comparative trials, the proportions of children less than five years of age were 81% in both the ASAQ groups (3,918/4,826) and the comparator groups (4,618/5,711).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The minimum patient weight was 5 kg and mean (SD) weight was 10.7 kg (2.8) for young children, 25 kg (8.4) in 5–14 year olds, and 54 kg (10.6) in adults. The proportion of male participants was 53%, ranging from 43% in Sierra Leone to 57% in Senegal.
## Heterogeneity (cont.)
*In the **Heterogeneity (cont.)** section:*
While all trials had similar endpoints, there were differences in trial design, age group, and PCR genotyping, and substantial heterogeneity was detected due to the inclusion of non-comparative studies and large differences between field sites (I2 test = 83%, p = 0.001, Cochran Q test for heterogeneity). Therefore in an attempt to account for statistical heterogeneity non-comparative studies were not included in most of the analysis and all analysis were stratified by site.
## Primary efficacy outcomes
*In the **Primary efficacy outcomes** section:*
Of the 5,987 patients treated with AS&AQ, 1% (45/5,987) were censored on Day 0 from the analysis for protocol violation, consent withdrawal, or self-medication as predefined by the mITT analysis. Of the patients treated with AS&AQ, 1,235 were censored by Day 28 for the efficacy analysis (Table 1).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Censoring events by drug category for the efficacy analysis (both comparative and non-comparative studies included)
## Crude Day 28 outcomes (not PCR adjusted)
*In the **Crude Day 28 outcomes (not PCR adjusted)** section:*
All studies (comparative and non-comparative). The details of the Kaplan-Meier analysis for a hypothetical cohort of 1,000 patients free of failure are shown in Table 2 for AS&AQ and comparators (ACT and non-ACT).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Crude efficacy (not PCR-adjusted), number and quotients of failures by Day and drug categories within 28 days*
The overall Day 28 efficacy of AS&AQ was 78.3% (95% CI 77.2–79.4) and 75.9% (95% CI 74.6–77.1) for all trials and comparative trials only, respectively. Efficacies of other forms of ACT were 83.2% (95% CI 81.8–84.6), and for non-ACT 55.8% (95% CI 54.0–57.7)(Table 2). Crude AS&AQ efficacy varied widely across study sites (p = 0.001, log rank test), ranging from 30% (95% CI 25.4–34.5) in Tororo, Uganda, to 100% (95% CI 97.2–100) in Cameroon. The median (range) time to failure in AS&AQ groups was 21 days (0–28). Efficacy by site was positively correlated to median time to failure (rs = 0.624, p = 0.001).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## PCR-adjusted Day 28 outcomes
*In the **PCR-adjusted Day 28 outcomes** section:*
Among AS&AQ recipients, genotyping was available in 21 randomized comparative studies at 23 sites (78% or 4,577/5,987). Of these patients, 257 had a recrudescent infection (Table 3) of which 90% (232/257) were PCR confirmed, 24 were danger signs or severe malaria, and one was resulting from the clinician decision. Overall, 93.9% (95% CI 93.2–94.5) of the patients treated with AS&AQ had cleared their primary infection and were free of recrudescence within 28 days. The corresponding figures were for other forms of ACT 94.8% (95% CI 93.8–95.6) and for non-ACT 80.6% (95% CI 78.8–82.0).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
PCR-adjusted efficacy, number of failures by Day, and drug categories within 28 days (only in studies using PCR results)*
The WHO criterion of >90% efficacy after genotyping was not met in 10 of 23 sites from 16 countries with PCR results: Uganda-Amudat, Kenya-Migori, Zanzibar-Micheweni, Uganda-Arua, Sierra Leone-Kailahun, Uganda-Apac, South Sudan-Nuba, Congo-Kinbanda, Rwanda-Rukara, and DRC-Boende (Table 4). Where AS&AQ efficacy PCR-adjusted per site was <90%, the comparator arm was >90% in 3 sites: in Congo-Kinbanda (AL), Uganda-Amudat (AS+SP), and Zanzibar-Micheweni (AL), and not significantly superior to AS&AQ (P > 0.05 for all comparisons, logrank test).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
AS&AQ groups: crude and PCR-adjusted Day 28 efficacy results presented by site and country[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The mean efficacy after PCR adjustment by country (Figure 2b) was below 90% in five countries: Kenya (1 site) 89.5% (95% CI 83.4–93.5), South Sudan (2 sites) 88.4% (95% CI 79.8–93.4), Sierra Leone (one site without comparator group) 88.3% (95% CI 76.6–93.1), DRC (two sites) 87.5% (95% CI 78.5–92.9), and Congo (one site) 84.2% (95% CI 74.8–90.4). In all these countries the upper limit of the 95% CI was >90%. Conversely, the lower limit of the 95%CI was above 90% in seven countries (Angola, Burkina-Faso, Cameroon, Guinea, Madagascar, and Senegal).
Crude (A) and PCR-adjusted (B) Day 28 efficacy with ASAQ by country stratified by site (mean and 95% CI). Note: The dotted horizontal line in panel B shows the WHO-recommended threshold of efficacy.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Out of the five countries where the mean efficacy PCR-adjusted of AS&AQ was <90%, two countries had two studies conducted in a different site: Sudan (Malakal and Nuba) where the mean PCR-adjusted efficacy of AS+SP of both sites was >90% (91.7, 95%CI 85.3–96.4); and DRC (Boende and Kamalo) where the mean PCR-adjusted efficacy of AS+SP was <90% [84.7, 95%CI 76.5–89.1]).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Secondary efficacy outcomes
*In the **Secondary efficacy outcomes** section:*
## Parasite clearance
*In the **Parasite clearance** section:*
The overall geometric mean of parasite counts on admission (Day 0) was 21,541/μL with wide site/country variations; the geometric means ranged from 3,303/μL in Burkina Faso to 40,492/μL in Dabola, Guinea (p = 0.001). Using multivariate analysis based on randomized comparisons, and controlling by sites, higher parasitaemia was found in younger patients (age as continuous variable) (p = 0.001), and in patients without gametocytes on admission compared to patients with gametocytes (p = 0.001). In the randomized trials, the median parasite clearance time for AS&AQ was Day 2 ranging from Day 1 to Day 7 in 5,853 patients (2.2% did not complete treatment or were regarded as treatment failures). Time to parasite clearance varied from site to site (p = 0.001), and was longer in patients with higher Day 0 parasitaemia (p = 0.001).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The proportion of patients remaining parasitaemic was 66.4% (1634/2462) on Day 1, 8.5% (440/5170) on Day 2, 1.8% (104/5460) on Day 3, and 0.6% (35/5507) on Day 7, including recurrences (Figure 3).
Prevalence rate of patients remaining with parasitaemic or febrile in the first 7 days of follow-up.
The risk of being parasitaemic on Day 3 in randomized controlled trials was not different between AS&AQ (2.9%, 90/3125) and other ACTs or AS alone (2.6%, 92/2778, p = 0.476, weighted by site). However, patients treated with non-ACT were at a higher risk of remaining parasitaemic on Day 3 (16%, 415/2609, p = 0.001, weighted by site): in AQ groups (19%, 114/611, p = 0.001, weighted by site), in AQ+SP (8%, 103/1256, p = 0.001, weighted by site), and in CQ+SP groups (26%, 184/699, p = 0.001, weighted by site).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The multivariate analysis stratified by site based on randomized comparisons confirmed these results: no difference between AS&AQ and comparators were found using a three-day ACT (p > 0.268 for all comparisons). A significant better parasite clearance was seen with AS&AQ than with non-ACT (the OR for being parasitaemic with AQ, CQ+SP, and AQ+SP was 16.43, 73.11 and 16.71, respectively, p = 0.001 for all comparisons).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Fever clearance
*In the **Fever clearance** section:*
All patients included in the trials had fever (axillary temperature ≥ 37.5°C) or a recent history of fever. For the 77.7% (4,614/5,940) AS&AQ patients who were actually febrile on admission, the median fever clearance time was Day 1 (Figure 3). The proportion of patients with fever decreased to 7.4% (372/5,040, 95% CI 6.7%–8.1%) on Day 1, 2.4% (119/4,998, 2.0%–2.8%) on Day 2, and 2.4% (102/4,308, 1.9%–2.8%) on Day 3. Three patients were febrile on Day 21: in one patient from Guinea fever decreased quickly on Day 2 (37.9°C) but remained between 37.5°C and 37.9°C until Day 21, and cleared on Day 28; two other patients who had previously cleared fever became febrile on Day 21 and cleared on Day 28.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Based on randomized comparative studies, the proportion of patients remaining febrile on Day 2 was lower in the AS&AQ groups compared to AS5 (0.8%, 2/251 vs. 4.0%, 10/252, p = 0.020), AS+SP (1.4%, 7/487 vs. 4.1%, 20/488, p = 0.011), CQ+SP (1.5%, 7/745 vs. 4.7%, 33/697, p = 0.001), AL (2.8%, 33/1,178 vs. 6.3%, 63/1,000, p = 0.001). No difference was detected between AS+AQ and AQ alone, DP, or AQ+SP and between the non-fixed and fixed ASAQ products (p > 0.147 for all comparisons).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Risks of failure in randomized comparisons
*In the **Risks of failure in randomized comparisons** section:*
Recurrence. The randomized comparative clinical trial conducted in Burkina Faso did not detect any difference in crude efficacy (PCR not adjusted) between the loose (AS+AQ) and the fixed-dose (ASAQ) combinations (p = 0.510). Based on randomized comparative studies with AS&AQ (N = 4,896) compared to other anti-malarial treatments (N = 5,713) and using multivariate analysis stratified by site (Figure 4a), patients treated with DP, AL, and AS+SP were at lower risk of failure (p = 0.001, for all comparisons) compared to AS&AQ, while patients treated with AQ alone, AS5 and CQ+SP were at a higher risk (p = 0.001, for all comparisons). The risk of failure was not different between AS+AQ and AQ+SP (p = 0.812).[](https://www.ncbi.nlm.nih.gov/mesh/D000077332)
Overall risks of failure of artesunate-amodiaquine by comparator: (A) crude, (B) PCR-adjusted Day 28 outcome. Note: The forest plot represents the risk of failure of artesunate amodiaquine versus comparators in randomized comparative studies. Results were stratified by site. The size of boxes is proportional to the number of patients included. The square represents the adjusted hazard ratio and 95% CI.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Recrudescence. In the non-AS&AQ comparator arms, genotyping results were available for all studies analysed (N = 21). In Burkina Faso, no difference was detected between the fixed-dose 95.0% (95% CI 92.7–97.3) and the non-fixed combination 95.7% (95% CI 93.7–97.7) (p = 0.645). Based on comparative randomized trials, and using multivariate analysis stratified by site, the risk of failure compared with AS&AQ was (i) lower with DP (p = 0.001); (ii) higher with AQ+SP, AQ alone, AS5, and CQ+SP (p = 0.001 for all comparisons); (iii) not different as compared with AS+SP and AL (p = 0.346; p = 0.158, respectively) (Figure 4b).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Predictors of failure
*In the **Predictors of failure** section:*
Recurrence. In AS&AQ groups using multivariate analysis stratified by site and controlling for potential independent factors (age, parasitaemia, and gametocyte on admission), younger children (age in continuous in terms of per 1 year increase of age AHR = 0.93, 95% CI 0.90–0.97, p = 0.001), and anaemic compared to non-anaemic patients were at a higher risk of failure (AHR = 1.17, 95% CI 1.02–1.35, p = 0.022). Likewise, when reinfections (PCR confirmed) were included for analysis, younger (AHR = 0.96, 95% CI 0.92–0.99, p = 0.023) and anaemic patients (AHR = 1.21, 95% CI 1.04–1.42, p = 0.014) were at higher risks.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Recrudescence. The median time to recrudescence (PCR confirmed) with AS&AQ was Day 21. Using similar analysis as previously, younger patients were also at higher risk for recrudescence (AHR = 0.88, 95% CI 0.81–0.95, p = 0.001), and no other independent factor was detected.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Gametocytaemia
*In the **Gametocytaemia** section:*
On admission. In AS&AQ groups, the prevalence rate of gametocytaemia on admission was 12.9% (95% CI 5.4%–20.5%), ranging from 0% in Zanzibar-Micheweni and Cameroon-Mendong, to 51.7% in Uganda-Apac. Using multivariate analysis and controlling for site, younger patients (p = 0.001) and patients with lower parasitaemia (p = 0.001) were at a higher risk for gametocytaemia on admission. The overall cumulative incidence probability of gametocyte presence on Day 28 was 31.4% (95% CI 22.7–39.7). The peak of gametocyte prevalence was on Day 2 (19.3%, 677/3,516).
The gametocyte carriage rate was 71 PGW per 1,000 weeks of follow-up, and the mean duration per patient was 14.5 (SD ± 11.7) days. In patients who did not have gametocytes on admission, the cumulative probability was 20.4% (95% CI 18.3–22.5), and the gametocyte carriage was 36/1,000 PGW. The mean duration was 6.2 (SD ± 2.5) days, and the maximum incidence rate was reached on Day 2: 10.4% (95% CI 9.3–11.5) when 48% of the cases occurred (Figure 5).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Incidence rate of gametocyte appearance in AS&AQ groups by Day in patients without gametocyte on admission.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Gametocyte clearance. The overall median clearance time was Day 14 in patients treated with AS&AQ, and varied widely by site (ranging from Day 1 to 28). Clearance time could not be calculated for 7.7% (46/594) of the patients with gametocytes on admission who had been lost to follow-up or censored due to failure, leaving 548 patients for the analyses on Day 28 (all having cleared their gametocytes by then). There was no difference in clearance time between patients who had gametocytes on admission and those who developed gametocytaemia post-admission (p = 0.378). However, while the peak time to clearance was Day 14 for the patients who did not have gametocyte on admission (32%), gametocyte clearance for those who had gametocytes on admission was almost evenly distributed throughout treatment and follow-up (20% on Day 2 and Day 21)
(Figure 6).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Gametocyte clearance time distribution in AS&AQ groups in patients with and without gametocyte on admission.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Results by drug treatment. Using survival analysis to examine the cumulative probability of gametocyte appearance in patients who did not have gametocytes on admission, and measuring levels of carriage expressed in PGW by site, different profiles were obtained depending on the anti-malarial used. No difference in gametocyte appearance was detected between the loose and fixed AS&AQ combinations (AHR = 1.07, p = 0.587).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Overall, using multivariate analysis based on randomized comparative studies and stratified by site (Figure 7), the risk of gametocyte appearance post-admission compared to AS&AQ groups was higher with AQ (AHR = 2.59, p = 0.001), CQ+SP (AHR = 2.29, p = 0.001), and AQ+SP (AHR = 1.77, p = 0.001); lower with AL (AHR = 0.57, p = 0.001) and DP (AHR = 0.39, p = 0.001); and not different with AS+SP (AHR = 0.88, p = 0.288).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Overall risks of gametocyte appearance in artesunate-amodiaquine groups by drug comparator. Note: The forest plot represents the risk of failure of artesunate amodiaquine versus comparators in randomized comparative studies. Results were stratified by site. The size of boxes is proportional to the number of patients included. The square represents the adjusted hazard ratio and 95% CI.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Gametocyte carriage rate in patients without gametocyte on admission in randomized comparative trials.
No difference was detected between the loose and the fixed AS&AQ combinations in Burkina Faso (p = 0.824). The overall carriage rate was 57% shorter with AL (13/1000 PGW) compared with AS&AQ (27/1000 PGW, RRMH [Mantel-Haenszel rate ratio] = 0.48, 95% CI 0.31–0.63, p = 0.001)(Figure 8). However, results by site showed that AL was only superior to AS+AQ (RR = 0.05, 95% CI 0.01–0.14, p = 0.001) in one site in Uganda (out of 3 Ugandan sites, and not in the other sites). Compared with DP groups in Rwanda, no difference in gametocyte carriage was detected compared to AS+AQ (p = 0.817). As a result, the overall carriage rate was 70% shorter in DP groups (RRMH = 0.25, 95% CI 0.11–0.41, p = 0.001, weighted by site).
However, the carriage rate was only significantly lower in DP compared to AS&AQ groups in one Ugandan site (RR = 0.12, 95% CI 0.06–0.26, p = 0.001). Conversely, AS+SP (57/1000 PGW) increased the overall gametocyte carriage rate by 8% versus AS+AQ, but not significantly so (53/1000 PGW, RRMH = 1.15, 95% CI 0.78–1.69, p = 0.514, weighted by site).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Relative difference in gametocyte carriage rate (person-gametocyte-week, PGW) in artesunate amodiaquine groups and comparators in randomized comparative studies.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Treating patients with AQ+SP significantly increased the overall carriage rate by 101% from 36/1000 to 72/1000 PGW vs. AS+AQ (RRMH = 1.95, 95% CI 1.56–2.40, p = 0.001, weighted by site). Using CQ+SP (157/1000 PGW) also increased the overall carriage rate by 194% compared to AS+AQ (53/1000 PGW, RRMH = 3.00, 95% CI 3.36–3.71, p = 0.001, weighted by site). Overall gametocyte carriage with AQ alone (45/1000 PGW) was also superior by 209% compared to AS+AQ combination (15/1000 PGW, RRMH = 3.34, 95% CI 1.97–5.71, p = 0.001, weighted by site).[](https://www.ncbi.nlm.nih.gov/mesh/C001205)
## Anaemia
*In the **Anaemia** section:*
The prevalence of anaemia at baseline was 49.4% (5031/10194), ranging from low levels (2.9%, 4/140) in Senegal-Mlomp, to high levels (86.7%, 91/105) in Zanzibar-Micheweni.
Predictors of anaemia. Using multivariate analysis and controlling for sites, patients in the AS&AQ group who had gametocyte on admission were at a higher risk for anaemia (AOR = 1.56, 95% CI 1.23–1.98, p = 0.001), as were younger children (AOR = 0.66, 95% CI 0.62–0.70, p = 0.001).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Effects of AS&AQ treatment on anaemia in randomized trials. In patients treated with AS&AQ and followed until Day 28, 62% (1,764/2,863) were anaemic at enrolment. Of these, 25% (709/2,863) had mild (8–9.9 g/dL, grade 2), 26% (745/2,863) moderate (5–7.9 g/dL, grade 3) and 11% (310/2,863) severe anaemia (<5 g/dL, grade 4). By Day 28, 38% (678/1,764) of the patients had recovered, 62% (1,086/1,764) remained anaemic, and 9% (104/1,099) who were not anaemic on admission became anaemic (of which 9% [9/104] had severe anaemia). For 88% (274/310) of the patients who had severe anaemia on admission, the severity of anaemia was reduced.
By Day 28, anaemia in these 274 patients became moderate in 273 patients (99%) and mild in 1 patient (<1, 1/274). Severe anaemia remained unchanged in 12% (36/310). Overall, less than 1% (9/2553) of the patients developed severe anaemia post treatment. Paired analysis of Day 0 and Day 7 showed a significant transient decline in Hgb count (-28 g/dL, SD 1.17, -3%, 95% CI -5 to -1, p = 0.001) followed by a significant increase on Day 28 (+1.16 g/dL, SD 1.63, +13%, 95% CI 12 – 15, p = 0.001) compared to Day 0.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Results by drug treatment (Table 5).
Haemoglobin and haematocrit values and changes between Day 0–14, and Day 0–28 (AS&AQ and comparators)[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
In randomized comparative trials, in patients followed up until Day 28, 46% (353/762) in the AS&AQ groups, and 32% (246/471) of the patients in the AL groups were anaemic on admission (p = 0.044). On Day 28, in AS&AQ groups, 54% (190/353) of the patients recovered, 46% (163/353) remained anaemic, and 10% (39/409) who were not anaemic on admission became anaemic (none had severe anaemia <5 g/dL). On Day 28, in the AL groups, 56% (138/246) recovered from anaemia, 44% (108/246) remained anaemic, and 8% (19/225) became anaemic (none had severe anaemia). There was no difference in proportions of patients recovering, remaining, or becoming anaemic between these groups. A paired analysis of Days 0 and 7 showed a significant transient decline in both groups in Hgb count (-5%, 95% CI -3 to -7, p = 0.001; -2%, 95% CI -4 to -1, p = 0.004, respectively).
The mean paired difference decrease was greater in AS&AQ compared to AL groups (p = 0.001). By Day 28, the relative mean paired difference increased significantly in both groups with no difference between the two groups (+9%, 95% CI +7 to +11, p = 0.001; +10%, 95% CI +7 to +12, p = 0.001, respectively; p = 0.917).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Similar paired analysis results of Day 0 and Day 14 (Table 5) were observed in randomized comparative trials with AQ+SP and DP, as well between Days 0 and 28 in patients treated with AS&AQ (p = 0.001 for all comparisons). No significant differences were observed in variations of Hgb between treatment arms, except in Rwanda where the recovery in relative mean paired Hct difference in the AS+AQ group (+10%, 95% CI 8 to 11) was significantly higher than in the DP group (6%, 95% CI 4 to 8) (p = 0.021).[](https://www.ncbi.nlm.nih.gov/mesh/C001205)
In trials comparing AS+SP and AS+AQ (Table 5), a similar transient decrease was observed in various settings on Day 7 until recovery on Day 28 (+13% for both comparisons) without any variation difference comparing the drugs (Day 7: p = 0.372, Day 28: p = 0.772).[](https://www.ncbi.nlm.nih.gov/mesh/C001205)
In Uganda, there was no significant difference in haemoglobin levels in the AS+AQ group on Day 14 compared to Day 0 (-0.4%, p = 0.551) whereas in the comparative AQ+SP group, a significant relative mean paired increase was detected (+1.3%, p = 0.016). The mean paired difference between the two groups being significant (p = 0.039). On Day 28 the variation was no longer different between the two groups (+15%, +16%, respectively, p = 0.406).[](https://www.ncbi.nlm.nih.gov/mesh/D000077332)
In Burkina Faso no difference was detected between the loose and the fixed AS&AQ combinations between Days 0 and 28 (+17%, +18%, respectively, p = 0.946).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Discussion
*In the **Discussion** section:*
This individual patient analysis has pooled data from 26 drug trials in a majority of paediatric malaria cases in sub-Saharan Africa identified through a systematic search and has focused on efficacy; safety will be reported separately. The trials reported herein were conducted between July 1999 and December 2006; thus, this analysis provide recent information on the current situation. Both absolute and comparative efficacy results varied between crude and PCR-adjusted results (i.e. whether reinfections are counted or discounted in the analysis).
The WHO recommends using treatments that are at least 90% effective after discounting reinfections. Overall, AS&AQ had an efficacy of ~94% after excluding reinfections by PCR. However, 10 sites in eight countries (out of 28 sites in 16 countries) failed to meet the WHO, Day 28, PCR-adjusted cut-off of >90% efficacy. These sites were in Congo, DRC, Kenya, Sierra Leone, South Sudan, Rwanda, Uganda, and Zanzibar. However, at other sites in some of these countries (DRC, Rwanda, South Sudan, Uganda, and Zanzibar), the PCR-adjusted efficacy exceeded 90%.
Moreover, the PCR-adjusted efficacy in the comparator arm was >90% only in three sites (Kindamba-Congo, Amudat-Uganda, Micheweni-Zanzibar), and was not significantly superior to AS&AQ.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The definition of recrudescent failure was strict since recurrent parasitaemia that could not be successfully genotyped by PCR (indeterminate case) was considered conservatively as a recrudescent failure. This was done in order to prevent from introducing overestimation bias in assessing AS&AQ efficacy levels, in comparison to other attrition methods by modified intent-to-treat analysis that would increase the level of efficacy by excluding the PCR indeterminate cases from the analysis. Compared to other treatments, AS&AQ was either superior to non-ACTs or not different from AS+SP and AL but inferior to DP.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
As expected, the AS&AQ crude efficacy, which counts reinfections as failures, was much lower, ~78% (with wide inter-country variability), than was the PCR-adjusted efficacy (~94%). During the 28 days of follow-up, the quotients of failure in the AS&AQ groups were the greatest on Day 21 and Day 28, in contrast with the other forms of ACT, for which the peak was reached on Day 28. When the risk of a reinfection is high in areas of intense transmission, treatment with longer post-treatment protection (AL, AS+SP, DP) fared better than AS&AQ. This probably reflects the relatively shorter residence time of AQ in the human body such that concentrations of the active metabolite, monodesethyl-amodiaquine, might be lower or absent when a reinfection occurs compared to other partner drugs combined to artemisinin derivatives. In the crude analysis of efficacy, AS&AQ was inferior to DP, AL and AS+SP.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Whether a short or a longer residence time for a drug is preferable is a matter of debate. Operationally, post-treatment protection against reinfection is a positive feature as it minimizes the number of treatments needed by the individual, the frequency of contacts with health providers, the risk of cumulative toxicity, and the costs (direct and indirect) incurred by households and health systems. Conversely, persisting concentrations of low drug levels may be insufficient to inhibit the replication of parasites arising from a new infection and potentially select for the parasites that can tolerate those levels. Furthermore, results depend on the study design and the duration of follow-up. It might be difficult to judge the operational implications of reinfections and re-treatment based on studies of treatment of single episodes of malaria; prospective cohort studies are best suited to assess the consequences of repeat treatments.
Based on the Day 28 efficacy results of these studies, AS&AQ would be suitable according to WHO standards as a potential alternative treatment for P. falciparum malaria in Angola, Burkina Faso, and Mali, where the current first line is AL. AS&AQ satisfied the criteria for continued use in some of the countries where is the current first-line treatment (Cameroon, Guinea, Madagascar, Gabon, Senegal, and Zanzibar), but AS&AQ would not qualify in some sites in Sierra Leone, Congo, DRC, North and South Sudan, Rwanda, Kenya, Uganda. However, where the AS&AQ efficacy PCR-adjusted was <90%, and the one of the comparators was >90%, the comparator groups were never significantly superior to AS&AQ whether in Congo (AL), Rwanda (DP), or Uganda (AS+SP).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
This multi-centre analysis provides also interesting information on malaria and response to treatment. It confirms that children under 5 years of age are particularly vulnerable, as they are more likely to have on presentation higher baseline parasitaemia, be anaemic and carry gametocytes, and have a higher risk of failure compared to older children for all treatments evaluated, consistent with a lack of malaria-acquired immunity.
Being young and anaemic increases the risk for antimalarial treatment to fail to clear parasites and to be reinfected after clearing the current infection, suggesting a relationship between anaemia and transmission intensity, and between anaemia and susceptibility to infection. Conversely, young age alone predicts recrudescence after initial clearance.
Young children are a major reservoir of gametocytes and hence the engine of malaria transmission. Gametocyte carriage is highest when asexual parasitaemia is low.
Fever clearance was fast with AS&AQ, similar to other forms of ACT and AQ+SP, but faster than AL, and other non-ACT. Parasite clearance was fast with AS&AQ, generally faster than non-ACT and similar to other forms of ACT.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The presence of gametocytes on admission ranged from 0 to ~50% across the studies, and was related to young age and low asexual parasitaemia. The cumulative risk of gametocytes appearing post-treatment was ~20% with 36 PGW carriage per 1000 weeks of follow-up. The gametocyte clearance time in AS&AQ groups was the same (median 14 days) whether patients presented with gametocytes or developed gametocytaemia thereafter, but the peak distributions of time to clearance were Day 2 and Day 14, respectively. Compared to AS&AQ, the risk of appearance of gametocytes was higher and the carriage duration was longer with the non-ACTs and AS+SP, but lower with DP and AL in one Ugandan site, consistent with their better efficacy against asexual parasites.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Endemic countries are faced with the challenge of identifying the treatment(s) best adapted to their needs. To inform decisions, both locally generated data and more general information are needed. Systematic reviews and meta-analyses are useful to summarize evidence and assist policy makers. Pooling individual patient data offers advantages over aggregate patient data meta-analysis because it allows standardizing patient attrition and analyses. Each study can then be re-analysed based on common criteria for efficacy and safety and different drug regimens can be combined and compared.
Data can also be combined and analysed together while stratifying by site. Efficacy analyses can be done on modified intent-to-treat basis of all randomized patients and use Kaplan-Meier product-limit estimates of time to event. This is now the preferred analytical method for anti-malarial drug efficacy trials.
Individual studies are not usually designed and, therefore, not powered to detect differences in a variety of secondary outcomes (e.g. gametocyte carriage, parasite or fever clearance time). Results of this analysis of individual patient data were presented using similar methods to that used for a conventional meta-analysis of trials (for instance in Cochrane's review) with graphical representation of risks, recommended for communicating in medical research. Compared to a meta-analysis from published studies, combining and standardizing these data at patient level increases statistical power by facilitating analytical practice (sub-group and multivariate analyses stratified by site) despite significant heterogeneity between trials. It also enables standardized estimates of drug efficacy across different studies, and the identification of at-risk groups to help target public health strategies.
However, this individual patient multi-centre analysis is not without limitations. First, the analysis included only half, 25 of the 46 trials that met criteria of quality for inclusion. It also has excluded additional trials published past August 2008, due to the time needed to adequately harmonize published data, obtain additional reported data and conduct the analyses. This might be a source of bias. The Worldwide Antimalarial Resistance Network (WARN) intends to create a living database, which might become the basis for updated assessments of drug efficacy.
Second, these results apply primarily to children under five years of age (75% of the patients enrolled) and less to older children or adults. However, young children are indeed those at higher risk and are the primary target of malaria interventions. Finally, this analysis showed heterogeneity of study results both across and within countries, a finding that illustrates the challenges faced when making drug policy decisions. Differences in efficacy between sites might have resulted from the variability in the composition of the study drug, as well as PCR methods that have been used according to sites facilities.
A Cochrane systematic review and meta-analysis which includes AS&AQ has just been published with consistent results.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
At a bare minimum, malaria control programmes need up-to-date, dynamic, and comparative data on anti-malarial drug efficacy and safety in order to recommend optimal drug treatments for their countries. Prospective multi-centre analysis could be a key element for deciding drug policy at national and regional levels.
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interests.
## Authors' contributions
*In the **Authors' contributions** section:*
JZ and PO designed the analyses and were the primary writers of the manuscript. JZ pooled the data and conducted the analysis. All the principal investigators from the AS&AQ Individual Patient Data (IPD) study group contributed data and participated in the writing or approved the manuscript. We thank the collaborating centres for sharing their data.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
|
# Introduction
[ABL-N](https://www.ncbi.nlm.nih.gov/mesh/C551742)-induced apoptosis in human breast cancer cells is partially mediated by c-Jun NH2-terminal kinase activation
# Abstract
*In the **Abstract** section:*
Introduction
The present study was designed to determine the possibility of acetylbritannilactone (ABL) derivative 5-(5-(ethylperoxy)pentan-2-yl)-6-methyl-3-methylene-2-oxo-2,3,3a,4,7,7a-hexahydrobenzofur[an-4-yl 2-(6-methoxyn](https://www.ncbi.nlm.nih.gov/mesh/C487680)ap[hth](https://www.ncbi.nlm.nih.gov/mesh/C487680)alen-2-yl)pro[panoate (ABL-N) as a novel therapeutic agent in human breast cancers.](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Methods
We investigated the effects of ABL-N on the induction of apoptosis in human breast cancer cells and further examined the underlying mechanisms. Moreover, tumor growth inhibition of ABL-N was done in xenograft models.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Results
ABL-N induced the activation of caspase-3 in estrogen receptor (ER)-negative cell lines MDA-MB-231 and MDA-MB-468, as evidenced by the cleavage of endogenous substrate Poly (ADP-ribose) polymerase (PARP). Pretreatment of cells with pan-caspase inhibitor z-VAD-fmk or caspase-3-specific inhibitor z-DEVD-fmk inhibited ABL-N-induced apoptosis. ABL-N treatment also resulted in an increase in the expression of pro-apoptotic members (Bax and Bad) with a concomitant decrease in Bcl-2. Furthermore, c-Jun-NH2-terminal kinase (JNK) and p38 mitogen-activated protein (MAP) kinase (p38) were activated in the apoptosis induced by ABL-N and JNK-specific inhibitor SP600[125 a](https://www.ncbi.nlm.nih.gov/mesh/C551742)nd JNK small interfering RNA (siRNA) antagonized ABL-N-mediated apoptosis. However, the p38-specific inhibitor SB203580 had no effect upon these processes. Moreover, neither of the caspase inhibitors prevented ABL-N-induced JNK activation, indicatin[g that JN](https://www.ncbi.nlm.nih.gov/mesh/C096713)K is upstream of caspases in ABL-[N-initiate](https://www.ncbi.nlm.nih.gov/mesh/C110772)d apoptosis[. Add](https://www.ncbi.nlm.nih.gov/mesh/C551742)itionally, in a nude[ mice](https://www.ncbi.nlm.nih.gov/mesh/C551742) xenograft experiment, ABL-N significantly inhibited the tumor growth of MDA-MB-231 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Conclusions
ABL-N induces apoptosis in breast cancer cells through the activation of caspases and JNK signaling pathways. Moreover, ABL-N treatment causes a significant inhibition of tumor growth in vivo. Therefore, it is thought that ABL-N might be a potential drug for use in breast cancer prevention and intervention.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Breast cancer is one of the most common cancers among women in both developed and underdeveloped countries. It is the malignancy with the highest incidence and death rate for women [1,2]. However, the efficacy of the present drugs is very limited, and it is urgent to find the anticancer compounds that can target multiple points in the apoptotic cascade to achieve synergistic actions. Chinese herbs have obtained considerable attention for the prevention and treatment of certain cancer types in clinical studies [3-6]. In many cases, the extracts obtained from the plants are not highly effective and require chemical modification for improved potency and toxicity profile
[7-9]. Thus, studies of naturally plant-based agents could supply new strategies for the management of cancer and related diseases [7,10,11].
Recently, several phytochemicals that have been used in clinical cancer chemotherapy were derived from herbs and plants, such as paclitaxel [5,12], etoposide [13], camptothecin [4] and vinca alkaloids [14]. Acetylbritannilactone (ABL) is a sesquiterpene lactone abundant in Inula britannica L, which is used to treat bronchitis and inflammation. In the previous work, it is demonstrated that ABL inhibits the expression of inflammation-associated genes and it possesses anticancer properties [15-19].
In the course of our continuing search for cytotoxic ABL analogues, we synthesized the compound 5-(5-(ethylperoxy)pentan-2-yl)-6-methyl-3-methylene-2-oxo-2,3,3a,4,7,7a-hexahydrobenzofuran-4-yl 2-(6-methoxynaphthalen-2-yl)propanoate (ABL-N), which in preliminary studies showed exceptional anti-proliferative activity against several human cancer cell types. Here, we showed that ABL-N was more potent than ABL in the ability to induce apoptosis, at a low concentration, of human breast cancer cells and investigated the therapeutic potential of the ABL-N and its underlying mechanism of action.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
## Materials and methods
*In the **Materials and methods** section:*
## Preparation of ABL and ABL-N
*In the **Preparation of ABL and ABL-N** section:*
Silica gel column chromatography was used to isolate ABL from Inula britannica L grown in Shan-xi Province in China. ABL-N was synthesized to improve efficacy and pharmacologic characteristics by substitution at C-6 of ABL (Figure 1a). These compounds were characterized by nuclear magnetic resonance and mass spectroscopy. The purified ABL and ABL-N were dissolved in ethanol at 1,000-fold final concentration and added to cells in exponential growth. The effects of ABL and ABL-N on our experiments were compared with the same concentration of ethanol as vehicle.[](https://www.ncbi.nlm.nih.gov/mesh/C487680)
Effect of ABL and ABL-N on cancer cell lines. (a) The chemical structures of ABL and ABL-N. (b) The differences of growth inhibition activity between ABL and ABL-N in MDA-MB-231 cells. (c) Effects of ABL-N on the viability of various cancer cell lines. Cells were treated with ABL-N for 24 hours and cell viability was determined by the MTT assay. The IC50 is the concentration of ABL-N that reduced the cell viability by 50% under the experimental conditions (n = 6). (d) Time- and dose-dependent inhibition of cell viability by ABL-N treatment in MDA-MB-231, MCF-7, MDA-MB-468 and NBECs. Cells were treated with 5, 10, 20 or 40 μM ABL-N for 1, 3, 6, 12 and 24 hours and cell viability was assessed by MTT assay (n = 6). Results represent the means ± SE from three independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C487680)
## Reagents
*In the **Reagents** section:*
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT), DMSO, 4,6-diamidino-2-phenylindole (DAPI), small interfering RNA (siRNA) specific for human JNK mRNA and control siRNA were obtained from Sigma Chemicals (St. Louis, MO, USA). LipofectAMINE 2000, Dulbecco's modified Eagle's medium (DMEM), penicillin, and streptomycin were purchased from Invitrogen (Carlsbad, CA, USA). The antibodies specific for Poly (ADP-ribose) polymerase (PARP), c-Jun NH2-terminal kinase (JNK), phospho-JNK, p38 MAP kinase (p38) and phospho-p38 were obtained from Cell Signaling Technology (Beverly, MA, USA). Antibodies against extracellular signal-regulated kinase (ERK), phospho-ERK, caspase-3, caspase-9, c-Jun, phospho-c-Jun, Bcl-2, Bax, Bad and secondary antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Pan-caspase inhibitor (z-VAD-fmk) was from Promega (Madison, WI, USA) and caspase-3-specific inhibitor (z-DEVD-fmk) was obtained from CalBiochem (San Diego, CA, USA).
Unless otherwise indicated, all other reagents used in this study were obtained from Sigma Chemicals.[](https://www.ncbi.nlm.nih.gov/mesh/C022616)
## Cell lines and culture conditions
*In the **Cell lines and culture conditions** section:*
The human breast cancer cell lines (MCF-7, MDA-MB-468 and MDA-MB-231), the human prostate carcinoma cells (Du145 and PC-3), and the human colon carcinoma cells (LoVo and HT-29) were from the American Type Culture Collection. Normal human breast epithelial cells (NBECs) were cultured and characterized as described previously from reduction mammoplasty specimens [20,21]. Cells were grown in a 5% CO2 atmosphere at 37°C in DMEM supplemented with 100 units/ml penicillin, 100 μg/ml streptomycin, 1% nonessential amino acids, and 10% fetal bovine serum (v/v). All treatments were carried out on cells at 60 to 80% confluence.[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
## Cell viability assay
*In the **Cell viability assay** section:*
Loss of cell viability was measured by the MTT assay. Cells were seeded at 1 × 104 cells/well in 96-well plates and allowed to grow in the growth medium for 24 hours. Cells were then treated with indicated concentrations of ABL-N for various time periods. After drug treatment, cells were incubated with 5 mg/ml MTT for two hours, and subsequently solubilized in DMSO. The absorbency at 570 nm was then measured using an enzyme-linked immunosorbent assay (ELISA) reader.
The IC50 is the concentration agent that reduced the cell viability by 50% under experimental conditions. Experiments were repeated at least three times, and the data were expressed as the means ± SE.[](https://www.ncbi.nlm.nih.gov/mesh/C022616)
## Nuclear staining assay
*In the **Nuclear staining assay** section:*
After treatment, the cells were harvested, washed with phosphate-buffered saline (PBS), and fixed in 70% ethanol for 30 minutes. The fixed cells were placed on slides and stained with 1 mg/ml DAPI for 15 minutes. Excess dye was washed off with PBS. Nuclear morphology was observed via a fluorescence microscope.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## Cell cycle analysis
*In the **Cell cycle analysis** section:*
Cell cycle distribution was determined using flow cytometry analysis. Briefly, after treatment, cells were harvested with 0.25% trypsin, washed in PBS and centrifuged. The cells were fixed in ice-cold 75% ethanol for at least 30 minutes. Cells were washed and resuspended in PBS containing 25 μg/ml RNase and 0.5% Triton X-100. Samples were then incubated with 50 μg/ml propidium iodide (PI) at 37°C for 30 minutes and analyzed in a flow cytometer (Becton Dickinson, San Jose, CA, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## Cell apoptosis assay
*In the **Cell apoptosis assay** section:*
Cell apoptosis was measured by ELISA and flow cytometry, respectively. For ELISA, the cells seeded in 96-well plates (1 × 104 cells/well) were treated with ABL-N at 5, 10, 20 and 40 μM for 24 hours. Both floating and adherent cells were collected and lysed. Each concentration of ABL-N was repeated five times. The induction of apoptosis by the agent was evaluated with a Cell Death Detection ELISAPlus kit (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instruction.
Photometric enzyme immunoassay was used to quantitatively determine the formation of cytoplasmic histone-associated DNA fragments in the form of mononucleosomes and oligonucleosomes after apoptosis of the cells [22]. Measurements were made using an ELISA reader at 405 nm and the results were calculated as the ratio of the absorbance of the ABL-N-treated cells/absorbance of untreated cells.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
For flow cytometry, briefly, after treatment, cells were collected and stained with Annexin V and PI staining using annexin V-FITC apoptosis kit (BD Biosciences, San Diego, CA, USA) according to the manufacturer's instruction. Early apoptotic (annexin V-positive and PI-negative) cells were distinguished from late apoptotic (annexin V and PI double-positive) or necrotic (PI-positive) cells by a flow cytometric analysis.[](https://www.ncbi.nlm.nih.gov/mesh/D017304)
## Western blot analysis
*In the **Western blot analysis** section:*
The cellular lysates were subjected to Western blot analysis as described previously [16,23]. Scanned images were quantified using TotalLab TL120 software (Nonlinear Dynamics Ltd., Newscastle, United Kingdom).
## JNK activity assay
*In the **JNK activity assay** section:*
Kinase activity of JNK was assayed with a nonradioactive assay kit according to enclosed manufacturer's procedures of Cell Signaling Technology. Briefly, after the cells were treated, the lysates were prepared using a lysis buffer (20 mM Tris (pH 7.5) containing 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium pyrophosphate, 1 mM β-glycerolphosphate, 1 mM phenylmethylsulfonyl fluoride, 1 mM Na3VO4, and 1 μg/ml leupeptin). JNK in approximately 250 μg of proteins in each sample lysate was pulled down selectively by an N-terminal c-Jun (residues 1 to 89) fusion protein that were bound to glutathione sepharose beads at 4°C overnight with gentle rocking. Thereafter beads were washed twice with the lysis buffer and twice with kinase buffer (25 mM Tris (pH 7.5) containing 5 mM β-glycerolphosphate, 2 mM dithiothreitol, 0.1 mM sodium orthovanadate, and 10 mM MgCl2). After the washings, pellets were suspended in 50 μl of kinase buffer supplemented with 200 μM ATP and incubated for 30 minutes at 30°C, during which c-Jun fusion proteins were phosphorylated by the activated JNK.
JNK activity was analyzed by Western blotting using a specific phospho-c-Jun (Ser63) antibody. To determine the direct effect of ABL-N on JNK activity, in vitro cell-free kinase assays were also performed using purified recombinant GST-JNK1 fusion proteins (SignalChem, Richmond, British Columbia, Canada). ABL-N and purified GST-JNK1 fusion proteins were incubated for 12 hours and JNK activity assay was also performed in a similar manner.[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
## Measurement of caspase activities
*In the **Measurement of caspase activities** section:*
Cells were treated with ABL-N, and the caspase-2, caspase-3/7, caspase-6, caspase-8, and caspase-9 activities in the cleared lysates were measured by using Caspase-Glo 2, Caspase-Glo 3/7, Caspase-Glo 6, Caspase-Glo 8, and Caspase-Glo 9 assays (Promega) according to the manufacturer's protocols. Luminescence was quantified using an ELISA reader. Blank values were subtracted, and increases in caspase activities were expressed as fold increase and calculated based on activities measured from untreated cells. Each sample was measured in triplicates.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## Transfection of siRNA
*In the **Transfection of siRNA** section:*
siRNAs specific for JNK (JNK siRNA) and control siRNA were transiently transfected into cells using transfection reagent (Lipofectamine 2000; Invitrogen) according to the manufacturer's protocol. JNK protein levels were analyzed by Western blotting to confirm adequate silencing of the genes at 48 hours.[](https://www.ncbi.nlm.nih.gov/mesh/C086724)
## Human breast tumor xenograft experiments
*In the **Human breast tumor xenograft experiments** section:*
The animal study was performed via a protocol approved by the governmental committee for animal research. Female BALB/c nude mice (four- to five-weeks old) were injected s.c. with MDA-MB-231 cells (6 × 106) per mouse at both flanks. After six days, when tumors reached a size of about 100 mm3, mice were randomly grouped and treated by daily i.p. injection with ABL-N at concentrations of 15 mg/kg body weight, or vehicle (10% DMSO, 10% ethanol in water). Mice were weighed at least twice a week to assess toxicity of the treatment, and the tumor size was measured every other day using calipers and their volumes were calculated according to a standard formula: width2 × length/2. Mice were sacrificed after 34 days of treatment when control tumors reached about 1,600 mm3.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## Statistical analysis
*In the **Statistical analysis** section:*
Normal distribution of data was verified using the Shapiro-Wilk test. Statistical analysis was performed with the two-tailed, unpaired Student's t-test using SPSS 11.0 software. Results are given as means ± SE. A value of P <0.05 was considered statistically significant.
## Results
*In the **Results** section:*
## ABL-N reduces the viability of various carcinoma cell lines
*In the **ABL-N reduces the viability of various carcinoma cell lines** section:*
The inhibitory effects of ABL and ABL-N on various carcinoma cell lines were estimated using the MTT cellular survival assay. MDA-MB-231 cells were treated for 24 hours with various concentrations of ABL and ABL-N, and cell viability was expressed as percentage of untreated cells. Figure 1b showed that ABL did not affect cell viability at up to 40 μM; whereas this same concentration of ABL-N inhibited cell growth by 85%. It indicated that ABL-N was dramatically more effective than ABL in inhibiting the growth of cells.[](https://www.ncbi.nlm.nih.gov/mesh/C487680)
Next, we examined the effects of ABL-N on the viability of other human carcinoma cell lines. The results showed that ABL-N treatment inhibited cell growth with similar IC50 (approximately 12 to 20 μM) after a 24-hour treatment (Figure 1c). It indicated that ABL-N was a broad-spectrum inhibitory agent of the human carcinoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Furthermore, the sensitivity of three human breast cancer cells to ABL-N was assessed. These cells were chosen because they represent estrogen receptor (ER)-negative (MDA-MB-231 and MDA-MB-468) and ER-positive (MCF-7) cell lines. The data showed that the inhibitory effects of ABL-N on cell viability occurred very fast. Treating MDA-MB-231 cells for 12 hours with 20 and 40 μM ABL-N reduced cell viability by 51% and 62%, respectively (Figure 1d). Similar results were observed in MDA-MB-468 and MCF-7 cells (Figure 1d).
These results showed that there was no relationship between estrogen receptor status and cytotoxic effects of ABL-N.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Moreover, we assessed whether ABL-N had any differential sensitivity to normal versus cancer cells. As shown in Figure 1d, we found that the sensitivity of the NBECs to ABL-N was much lower, with ABL-N only having a significant effect on the viability (28% reduction, P <0.05) of the NBECs following 24 hours of treatment at the 40 μM dose, suggesting that ABL-N may have potential selectivity toward tumor cells.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## ABL-N arrests cells in G2/M phase of the cell cycle
*In the **ABL-N arrests cells in G2/M phase of the cell cycle** section:*
Because ABL-N can effectively inhibit cell viability, we reasoned that this effect might be attributable to its ability to interfere with the cell cycle. MDA-MB-231 cells were incubated with 20 μM ABL-N for 6, 12 and 24 hours, and the cell cycle analysis was done by PI uptake. Figure 2a shows that the ratio of cells in G2/M phase and the cells with hypodiploid DNA contents (sub-G1) were significantly accumulated over the treatment periods. In contrast, the cell cycle profile did not change over a 24-hour period in cells treated with the vehicle only (data not shown). Moreover, we also analyzed cell cycle in MDA-MB-468 and MCF-7 cells and found the G2/M arrest induced by ABL-N as well (data not shown).
Thus, it suggested that the inhibitory effect of ABL-N on breast cancer cells might be, at least in part, due to a G2/M arrest of the cell cycle.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Effect of ABL-N on cell cycle progression and apoptosis in breast cancer cells. After treatment with 20 μM ABL-N for indicated times, MDA-MB-231 cells were harvested. (a) The percentage of cell cycle distribution for cells done in triplicate with similar results. (b) Nuclear condensation was shown by DAPI-staining assay (magnification, ×100). (c) DNA fragmentation was evaluated using a Cell Death Detection ELISAPlus kit. The data are expressed as means ± SE of three separate experiments.*P <0.05, **P <0.01, as compared with the group without ABL-N treatment. (d) The apoptotic status was determined by Annexin V/PI staining method. Percentages of negative (viable) cells, annexin V-positive (early apoptotic) cells, PI-positive (necrotic) cells, or annexin V and PI double-positive (late apoptotic) cells were shown (mean of three independent experiments) by a flow cytometric analysis.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## ABL-N induces apoptosis in breast cancer cells
*In the **ABL-N induces apoptosis in breast cancer cells** section:*
We next analyzed whether the ABL-N-induced cell viability reduction in human breast cancer cells involved apoptosis. MDA-MB-231 cells were treated with 20 μM ABL-N and apoptosis was assayed by two different methods. DAPI staining showed that the condensed and fragmented nuclei increased with ABL-N treatment (Figure 2b). Nucleosome fragmentation (an indicator of apoptosis) further determined by Cell Death Detection ELISAPLUS confirmed that cells treated for six hours with 20 μM ABL-N underwent apoptosis, with the highest percentage of apoptotic cells seen at 24 hours (Figure 2c). In addition, an annexin V-binding assay showed that ABL-N treatment induced apoptosis but not necrosis in MDA-MB-231 cells (Figure 2d).
We also obtained similar results when MDA-MB-468 and MCF-7 cells were treated with ABL-N (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## ABL-N induces the activities of caspase in breast cancer cells
*In the **ABL-N induces the activities of caspase in breast cancer cells** section:*
To investigate the activation of caspases in ABL-N-induced apoptosis, the proteolytic activation of caspase-3 and caspase-9 was examined. As shown in Figure 3a, ABL-N (20 μM) treatment resulted in a significant increase in the active form of cleaved caspase-3 and cleaved caspase-9 in MDA-MB-231 cells. The cleavage products were detectable as early as six hours after exposure of cells to ABL-N. Furthermore, caspase activities were measured with Caspase-Glo assays. As shown in Figure 3b, treatment with ABL-N induced the activation of caspases-3/7, -8 and -9 in MDA-MB-231 cells. Under similar conditions, ABL-N also stimulated caspase-2 and -6 activities in a lesser extent.
Similar results were found in MDA-MB-468 cells (data not shown). However, in ER-positive cell MCF-7, the activities of caspase-3/7 and -9 were not affected by ABL-N, although caspase-2 and -6 were activated to a level similar to that seen in MDA-MB-231 and MDA-MB-468 cells (data not shown). However, we found that caspase-8 activity is significantly higher in MCF-7 cells treated with ABL-N, with the activity being increased by 2.8-fold of the untreated cells at 24 hours. The activation of caspases in MDA-MB-231 cells was further confirmed by detecting the degradation of PARP, which is an endogenous substrate of activated caspase-3 and its cleavage is considered a hallmark of apoptosis [24-26].
Treatment of MDA-MB-231 cells with ABL-N resulted in cleavage of PARP to an 85 kDa fragment (Figure 3c).[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Induction of caspase activities by ABL-N. MDA-MB-231 cells were treated with 20 μM ABL-N for indicated times. (a) Whole cell protein lysates were prepared and subjected to Western blot analysis for detection of cleavage of caspase-3 and caspase-9. (b) Induction of caspase activities by ABL-N in MDA-MB-231 cells. (c) Cleavage of PARP was induced by ABL-N. (d) Cells were pretreated with 50 μM of either z-VAD-fmk or z-DEVD-fmk for one hour, followed by 20 μM ABL-N for 24 hours, and caspase-3 activity, DNA fragmentation and sub-G1 DNA contents were determined. The data are expressed as means ± SE of three separate experiments. (e) Abrogation of PARP cleavage by caspase inhibitors.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## Caspase inhibitors attenuate the induction of apoptosis by ABL-N
*In the **Caspase inhibitors attenuate the induction of apoptosis by ABL-N** section:*
To define the role of caspase activation in ABL-N-induced apoptosis, we treated MDA-MB-231 cells with pan-caspase inhibitor z-VAD-fmk (50 μM) and caspase-3-specific inhibitor (z-DEVD-fmk) (50 μM) before challenge with ABL-N (20 μM). z-VAD-fmk pretreatment for one hour abrogated ABL-N-induced apoptosis as measured by the nucleosome fragmentation and the appearance of sub-G1 cells (Figure 3d). Similar results were observed in both MDA-MB-468 and MCF-7 cells (data not shown). The caspase-3-specific inhibitor z-DEVD-fmk also reduced ABL-N-induced apoptosis in MDA-MB-231 cells (Figure 3d). Moreover, both of the caspase inhibitors abolished caspase-3 activity in MDA-MB-231 cells (Figure 3d).
In addition, the cleavage of PARP was attenuated by pretreatment of cells with z-VAD-fmk or with z-DEVD-fmk in MDA-MB-231 cells (Figure 3e). These results suggested that activation of a caspase cascade was essential for ABL-N-induced apoptosis in breast cancer cells.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## ABL-N modulates the expression of Bcl-2 family proteins in breast cancer cells
*In the **ABL-N modulates the expression of Bcl-2 family proteins in breast cancer cells** section:*
Bcl-2 family of proteins, including Bcl-2 and Bcl-2-related family members such as Bcl-xL, Bad and Bax, plays an important role in the regulation of apoptosis [27,28]. Thus, we evaluated the effects of ABL-N on the expression of anti-apoptotic protein Bcl-2 and the pro-apoptotic proteins Bax and Bad in breast cancer cells. Figure 4a showed a marked increase in the level of Bax and Bad, which started at six hours and peaked at 24 hours of treatment with ABL-N in MDA-MB-231 cells. In contrast, reduced Bcl-2 protein appeared later at 12 hours. Then, the ratio of Bax and Bcl-2, which is the determining factor for the induction of apoptosis [29], was measured by a densitometric analysis of the bands.
As shown in Figure 4b, ABL-N treatment resulted in a time-dependent increase in Bax/Bcl-2 ratio in MDA-MB-231 cells that favors apoptosis.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Effects of ABL-N on the expression of Bcl-2 family. (a) Treatment of ABL-N decreased the level of the anti-apoptotic protein Bcl-2, but increased the expression of the pro-apoptotic proteins Bax and Bad in MDA-MB-231 cells. (b) Effect of ABL-N on the Bax/Bcl-2 ratio in MDA-MB-231 cells. The data obtained from the Western blot analysis of Bax and Bcl-2 were used to evaluate the effect of ABL-N on the Bax/Bcl-2 ratio. The densitometric analysis of Bax and Bcl-2 bands was performed using TotalLab TL120 software, and the data (relative density normalized to β-actin) were plotted as Bax/Bcl-2 ratio. Data are expressed as means ± SE of three separate experiments with similar results. *P <0.05, **P <0.01, as compared with the group without ABL-N treatment.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## ABL-N induces the activation of JNK and p38 signaling in breast cancer cells
*In the **ABL-N induces the activation of JNK and p38 signaling in breast cancer cells** section:*
Activation of mitogen-activated protein (MAP) kinase is involved in many aspects of the control of cellular proliferation and apoptosis in response to a variety of extracellular stimulus [30-32]. We therefore examined the effects of ABL-N on the activation of several MAP kinase pathways. The cell lysates from ABL-N-treated MDA-MB-231 cells were subjected to Western blot analysis using antibodies that specifically recognize the phosphorylated forms of JNK, p38 and ERK. As shown in Figure 5a and 5b, ABL-N treatment induced activation of JNK and p38 in a time-dependent manner.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Time course of MAP kinase activation by ABL-N. MDA-MB-231 cells were treated with vehicle (0.5% ethanol) or 20 μM of ABL-N and harvested at the indicated times. (a) JNK; (b) p38; and (e) ERK activation was determined by Western blot analysis using antibodies that recognize the phosphorylated form of the respective active MAP kinase. (c) JNK activity was determined by an in vitro kinase assay as described in Materials and methods. Phosphorylation of c-Jun, which represents the intrinsic activity of JNK, was visualized by Western blotting. (d) ABL-N (20 μM) and GST-JNK1 fusion proteins (1 μg) were incubated for 12 hours or cells were treated with ABL-N (20 μM) for 12 hours and lysates were prepared, then JNK activity was measured as described above. Phosphorylation of c-Jun was visualized by Western blotting. The same blot was stripped and reprobed with a c-Jun antibody to monitor equal loading of proteins. Data are representative of three independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
The activation of JNK by ABL-N was further confirmed by the analysis of phosphorylation of its downstream substrate c-Jun (Figure 5a). It showed that c-Jun was phosphorylated following ABL-N treatment, which occurred over the same sustained period as JNK activation. However, no evident JNK and c-Jun phosphorylation could be observed within three hours after ABL-N treatment. Thus, JNK activation induced by ABL-N is a delayed and sustained, but not an early and transient, process. To gain further insight into the mechanism by which ABL-N treatment affects JNK, we assayed JNK activity in ABL-N-treated cells as well as the direct effect of ABL-N on GST-JNK1 fusion proteins activity.
As depicted in Figure 5c, JNK activity began to increase after treatment with 20 μM ABL-N for three hours and maximum activation was achieved 12 hours after treatment. In addition, using GST-JNK1 fusion proteins, we found that the JNK activity was unaffected by the presence of ABL-N (Figure 5d), indicating ABL-N activated JNK indirectly by activating signaling molecules located upstream in the JNK cascades. On the contrary, ERK was revealed to be constitutively activated in the cells and no significant change of ERK expression and phosphorylation was observed after ABL-N treatment (Figure 5e).[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## Suppression of JNK antagonized ABL-N-induced caspase-3 activity and PARP cleavage
*In the **Suppression of JNK antagonized ABL-N-induced caspase-3 activity and PARP cleavage** section:*
To determine the role of the activation of JNK and p38 in ABL-N-induced apoptosis, MDA-MB-231 cells were treated with the JNK inhibitor SP600125 or the p38 inhibitor SB203580 and their effects on cell apoptosis were examined. Additionally, we transfected MDA-MB-231 cells with JNK siRNA or with a control siRNA for 48 hours. Transfection of JNK siRNA markedly reduced the expression of JNK protein compared with that in cells transfected with the control siRNA (Figure 6a). As shown in Figure 6b, reduction of cell viability by ABL-N was effectively abolished by SP600125, but SB203580 only had a slight effect on the decreased viability by ABL-N. The cells transfected with the JNK siRNA also blocked ABL-N-induced loss of cell viability. These results suggested that the activation of JNK signaling is responsible for the ABL-N-induced apoptosis.
Similar results were further confirmed by flow cytometic analysis to determine the sub-G1 DNA contents (Figure 6c). Similarly, JNK inhibition by its specific inhibitor or siRNA could effectively antagonize ABL-N-induced caspase-3 activity and PARP cleavage (Figure 6d). Moreover, the stimulatory effect of ABL-N on c-Jun phosphorylation and JNK phosphorylation was not altered by SB203580, but significantly reduced by SP600125 (Figure 6e). These results suggested that activation of JNK, but not p38, was important for ABL-N-induced apoptosis.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Role of JNK in ABL-N-induced apoptosis. MDA-MB-231 cells were incubated with ABL-N (20 μM) and either SB203580 (20 μM) or SP600125 (30 μM), or a combination of both for 24 hours. (a) MDA-MB-231 cells were transfected with JNK siRNA (25 nM) or control siRNA (25 nM) for 48 hours, and cell lysates were subjected to Western blot with antibodies to JNK. (b) ABL-N-induced cell death was abrogated by inhibition of JNK using MTT assay. (c) Sub G1 DNA content was analyzed by flow cytometry. (d) Caspase-3 activity was determined by Caspase-Glo assay and the cleavage of PARP was analyzed via Western blotting. Data are expressed as the means ± SE of three independent experiments. (e) Effect of MAP kinase inhibitors on ABL-N-induced JNK activation. (f) Effect of caspase inhibitors on JNK activation. Cells pretreated with or without z-VAD-fmk (50 μM) or z-DEVD-fmk (50 μM, one hour) were further incubated with vehicle or 20 μM ABL-N for 24 hours. Data are representative of three independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Furthermore, to address the possible role of caspase cleavage in the activation of JNK pathway in ABL-N-induced apoptosis, we observed the effect of caspase inhibitors on JNK activation by ABL-N. The results indicated that z-VAD-fmk (50 μM) or z-DEVD-fmk (50 μM) pretreatment had no effect on c-Jun and JNK activation induced by ABL-N (Figure 6f), further suggesting that caspases were downstream targets of JNK signaling in response to ABL-N in breast cancer cells.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## ABL-N inhibits the growth of human breast cancer xenografts
*In the **ABL-N inhibits the growth of human breast cancer xenografts** section:*
Because ABL-N treatment showed the effective growth inhibition in cultured breast cancer cells, we subsequently carried out in vivo study using MDA-MB-231-derived cancer xenografts in nude mice. As shown in Figure 7, the i.p. treatment with ABL-N (15 mg/kg) caused a significant inhibition of tumor growth as early as 20 days after treatment and persisted after 34 days. Furthermore, animals showed no body weight loss, decreased activity, or anorexia (data not shown). This initial in vivo experiment suggested that ABL-N might be an effective anticancer agent at dosages that induced negligible toxic effects.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
The effect of ABL-N on the growth of MDA-MB-231 xenografts. Female nude mice bearing MDA-MB-231 tumor xenografts were treated with ABL-N (15 mg/kg), or vehicle (10% DMSO, 10% ethanol in water), and tumors were measured with calipers on alternate days. Points = mean tumor volume in each experimental group containing six mice; bars = SD; *, P < 0.05; **, P < 0.01, compared with vehicle treated group.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## Discussion
*In the **Discussion** section:*
Induction of apoptosis in malignant cells is a critical property of chemopreventive agents. ABL, which has been shown to be potently anti-tumorigenic, has pro-apoptotic features in many carcinoma cell types [17-19]. We modified ABL to improve potency and pharmacologic characteristics, and obtained a highly active derivative ABL-N, which showed exceptional anti-proliferative activity against several human cancer cell types. The differences in biological activity of ABL and ABL-N could be attributed to the difference in their structures (Figure 1a).
The increased activity of ABL-N could be, most likely, attributed to (S)-2-(6-methoxynaphthalen-2-yl)propionyl- side chain of ABL-N, which is substituted with hydroxyl group at C-6 in ABL. The purpose of this study was to investigate the possible role of caspase activation and JNK signaling in ABL-N-induced apoptosis in ER-positive and ER-negative breast cancer cells.[](https://www.ncbi.nlm.nih.gov/mesh/C487680)
Apoptosis is governed by a complex network of anti-apoptotic and pro-apoptotic effector molecules. Among proteins implicated in control of apoptosis, members of the Bcl-2 family have been demonstrated to be associated with the mitochondrial membrane integrity [33]. They can activate or inhibit the release of downstream factors such as cytochrome c which leads to the activation of caspase-3 and PARP in the execution of apoptosis [34].
Bax exerts pro-apoptotic activity by translocation from the cytosol to the mitochondria, where it induces cytochrome c release, while Bcl-2 exerts its anti-apoptotic activity, at least in part, by inhibiting the translocation of Bax to the mitochondria [35,36]. Notably, the ratio of pro- and anti-apoptotic protein expression, such as Bax/Bcl-2, is critical for the induction of apoptosis, and it decides the susceptibility of cells to undergo apoptosis [29]. In the present study, we showed that treatment of the MDA-MB-231 cells with ABL-N resulted in significant decrease in the Bcl-2 protein and increase in the Bax protein, thus shifting the Bax/Bcl-2 ratio in favor of apoptosis.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
Caspases are known to act as key mediators of apoptosis and to contribute to the apoptotic morphology through the cleavage of various cellular substrates. Caspase-3 is an executioner member that can cleave specific substrates such as PARP [37,38]. Here, the results that caspase-3 activation was accompanied by PARP cleavage in MDA-MB-231 and MDA-MB-468 ER-negative cells indicated that caspase-3 might play a key role as an important executioner in ABL-N-induced apoptosis in these cell lines. It was further substantiated by the data that caspase-3 inhibitor z-DEVD-fmk suppressed cell apoptosis induced by ABL-N. However, the absence of caspase-3 activation in MCF-7 ER-positive cells [39] raised the possibility that ABL-N might induce apoptosis in MCF-7 cells through caspase-3-independent pathway.
It is demonstrated that curcumin and tributyrin could induce apoptosis but independent of caspase-3 activation in MCF-7 cells [40,41]. Therefore, other caspases might take part in ABL-N-induced apoptosis in MCF-7 cells. This was supported by our results that the pan-caspase inhibitor z-VAD-fmk could abrogate ABL-N-induced nucleosome fragmentation and the appearance of sub-G1 cells. Our study further found that ABL-N induced significant activation of caspase-8 in the breast cancer cells, suggesting that ABL-N may activate the death receptor pathway and induce the expression of death receptors or death ligands such as tumor necrosis factor-α and Fas ligand [42].
Thus, death receptor- or mitochondria-mediated activation of the caspase may be a potential mechanism underlying ABL-N-induced apoptosis in breast cancer cells. Recently, it was reported that another factor contributing to the activation of caspase is related to the inhibitors of apoptosis proteins (IAPs), which have been revealed to inhibit apoptosis due to their function as direct inhibitors of caspases [43,44]. Many anticancer agents such as curcumin [45], epigallocatechin [46] and esculetin [47], have been shown to interfere with IAPs and increase apoptosis rate. Consequently, we can hypothesize that ABL-N possibly invokes similar pathways and further study is necessary to investigate the precise mechanisms responsible.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
It is reported that the JNK pathway is very important in cell apoptosis induced by various cytotoxic compounds [30,46,47]. Studies reported here showed that ABL-N induced activation of JNK beginning one hour after ABL-N treatment and produced a sustained elevation for at least 24 hours in MDA-MB-231 cells, which is much earlier than the activation of caspase and the apoptosis. Moreover, the JNK activation correlated with the increased phosphorylated c-Jun, which appeared to play a significant role in ABL-N-induced apoptosis and to correlate with the activation of caspase. ABL-N could also activate JNK in cells as monitored by JNK activity assay. Several JNK isoforms, including JNK2 and JNK3, had kinase activity in the absence of any activating upstream kinase, which showed autophosphorylation activity resulting in constitutive activation
[48]. Therefore, we use recombinant JNK1 proteins to determine the direct effects of ABL-N on JNK. However, ABL-N did not activate GST-JNK1 fusion proteins in vitro kinase activity assay, indicating ABL-N might act on JNK by the upstream activators. Many studies have reported that activation of JNK by its upstream kinases or molecules is critical for its function. A number of factors, such as MAP kinase kinase 4 (MKK4)
[49], epidermal growth factor receptor (EGFR) [50], insulin growth factor receptor (IGFR) [51] and spleen tyrosine kinase (SYK) [52], have been implicated in the activation of JNK. Thus, further studies will be required to determine how ABL-N acts upstream of JNK.
In addition, our data showed that neither of the caspase inhibitors prevented ABL-N-induced JNK activation, while JNK-specific inhibitor or JNK siRNA, at least partially inhibited ABL-N-induced apoptosis, indicating that JNK is upstream of caspases in ABL-N-initiated apoptosis. Therefore, the results showed that both the caspases and JNK pathway were necessary to ABL-N-induced apoptosis in breast cancer cells because interfering with either pathway could attenuate apoptosis. Specifically, it was important to note that interference with JNK using the specific inhibitor or siRNA did not result in total abolition of ABL-N-induced cell death, as shown by MTT and flow cytometry assays. This may contribute to partial inhibition of JNK by SP600125 or siRNA, since JNK inhibition led to significant but not total reduction in phosphorylation of c-Jun. Thus, it is likely that JNK-independent mechanisms may participate in ABL-N-induced apoptosis.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## Conclusions
*In the **Conclusions** section:*
In summary, our studies suggest for the first time that ABL-N significantly induces apoptosis in breast cancer cells. This induction is associated with the activation of caspases and JNK signaling pathways. Moreover, ABL-N treatment caused a significant inhibition of tumor growth in vivo. Therefore, it is thought that ABL-N might be a potential drug for use in breast cancer prevention and intervention.[](https://www.ncbi.nlm.nih.gov/mesh/C551742)
## Abbreviations
*In the **Abbreviations** section:*
ELISA: enzyme-linked immunosorbent assay; ER: estrogen receptor; ERK: extracellular signal-regulated kinase; IAP: inhibitors of apoptosis protein; JNK: c-Jun NH2-terminal kinase; MAP kinase: mitogen-activated protein kinase; MTT: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide; PARP: poly (ADP-ribose) polymerase; p38: p38 MAP kinase; PBS: phosphate-buffered saline; siRNA: small interfering RNA.[](https://www.ncbi.nlm.nih.gov/mesh/C022616)
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interests.
## Authors' contributions
*In the **Authors' contributions** section:*
BL designed the study, carried out many of the experiments, and drafted the manuscript. RHS, JJW and YPZ participated in the design of the study and data interpretation. MH, DQZ and JKW led the conception and design of the study and the revisions to the manuscript and supervised this research project. All authors read and approved the final manuscript.
## Acknowledgements
*In the **Acknowledgements** section:*
This research was supported by the National Natural Science Foundation of P.R China (No. 30971457 and 30973820) and Hebei Province Natural Science Foundation of P.R. China (C2008001049).
|
# Introduction
[(E)-4-(4-Hydroxy-3-methoxybenzylideneamino)-3-[1-(4-isobutylphenyl)ethyl]-1H-1,2,4-triazole-5(4H)-thione](https://www.ncbi.nlm.nih.gov/mesh/D013871)
# Abstract
*In the **Abstract** section:*
The asymmetric unit of the title compound, C22H26N4O2S, contains two crystallographically independent molecules (A and B). The isobutyl unit of molecule [B is disord](https://www.ncbi.nlm.nih.gov/mesh/D013871)ered over two orientations with refined occupancies of 0.785 (6) and 0.215 (6). In each molecule, intramolecular C—H⋯S hydrogen bonds generate S(6) ring motifs. The essentially planar 1,2,4-triazole rings [r.m.s. deviations of 0.00[4 (2) an](https://www.ncbi.nlm.nih.gov/mesh/D006859)d 0.011 (2) Å, in A and B respectively] form dihedral ang[les of 85.86 (](https://www.ncbi.nlm.nih.gov/mesh/C045575)12), 8.38 (10)°, respectively, with the isobutyl-substituted phenyl ring and the 2-methoxyphenol substituent in molecule A [89.26 (13) and 2.46 (10)°, respectively, in B]. In the crystal stru[cture, intermol](https://www.ncbi.nlm.nih.gov/mesh/D006139)ecular N—H⋯N and N—H⋯S hydrogen bonds link neighbouring molecules, generating R 2 2(7) ring motifs. Thes[e molecu](https://www.ncbi.nlm.nih.gov/mesh/D006859)les are further interconnected into extended chains along [20] by intermolecular O—H⋯O hydrogen bonds. The crystal structure is further stabilized by π–π [centroid-centroid distance = 3.6299 (13) Å] [and C—H⋯](https://www.ncbi.nlm.nih.gov/mesh/D006859)π interactions. A short O⋯O contact of 2.781 (2) Å is also observed.
## Related literature
*In the **Related literature** section:*
For general background to and applications of the title compound, see: Bekircan & Bektas (2006 ▶); Fun et al. (2009 ▶); Koparır et al. (2005 ▶). For hydrogen-bond motifs, see: Bernstein et al. (1995 ▶). For bond-length data, see: Allen et al. (1987 ▶). For a closely related 1,2,4-triazole structure, see: Fun et al. (2009 ▶).
For the stability of the temperature controller used for the data collection, see: Cosier & Glazer (1986 ▶).[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Experimental
*In the **Experimental** section:*
## Crystal data
*In the **Crystal data** section:*
C22H26N4O2S[](https://www.ncbi.nlm.nih.gov/mesh/D013871)
M r = 410.53
Triclinic,
a = 9.8646 (3) Å
b = 14.2026 (5) Å
c = 16.6758 (6) Å
α = 69.048 (2)°
β = 79.881 (2)°
γ = 85.946 (2)°
V = 2147.83 (13) Å3
Z = 4
Mo Kα radiation[](https://www.ncbi.nlm.nih.gov/mesh/D008982)
μ = 0.18 mm−1
T = 100 K
0.31 × 0.22 × 0.15 mm
## Data collection
*In the **Data collection** section:*
Bruker SMART APEXII CCD area-detector diffractometer
Absorption correction: multi-scan (SADABS; Bruker, 2005 ▶) T min = 0.948, T max = 0.974
39640 measured reflections
7435 independent reflections
5503 reflections with I > 2σ(I)
R int = 0.060
## Refinement
*In the **Refinement** section:*
R[F 2 > 2σ(F 2)] = 0.044
wR(F 2) = 0.111
S = 1.04
7435 reflections
571 parameters
H atoms treated by a mixture of independent and constrained refinement[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Δρmax = 0.25 e Å−3
Δρmin = −0.25 e Å−3
Data collection: APEX2 (Bruker, 2005 ▶); cell refinement: SAINT (Bruker, 2005 ▶); data reduction: SAINT; program(s) used to solve structure: SHELXTL (Sheldrick, 2008 ▶); program(s) used to refine structure: SHELXTL; molecular graphics: SHELXTL; software used to prepare material for publication: SHELXTL and PLATON (Spek, 2009 ▶).
## Supplementary Material
*In the **Supplementary Material** section:*
Supplementary data and figures for this paper are available from the IUCr electronic archives (Reference: SJ2704).
# supplementary crystallographic information
*In the **supplementary crystallographic information** section:*
## Comment
*In the **Comment** section:*
1,2,4-triazoles and their derivatives are found to be associated with various biological activities with for example anti-convulsant, anti-fungal, anti-cancer, anti-inflammatory, anti-bacterial properties (Bekircan & Bektas, 2006) and also act as effective pesticides (Koparır et al., 2005). Several compounds containing 1,2,4-triazole rings are well known as drugs. Furthermore, in recent years, some Schiff base derivatives of 1,2,4-triazoles have been found to possess pharmacological activities (Fun et al., 2009). As part of our ongoing work on Schiff base derivatives, we report here the crystal structure of this new Schiff base.[](https://www.ncbi.nlm.nih.gov/mesh/C045575)
In the asymmetric unit of the title 1,2,4-triazole compound, there are two crystallographically independent molecules, designated A and B (Fig. 1). In molecule B, the isobutyl unit is disordered over two positions with a refined site-occupancy ratio of 0.785 (6):0.215 (6). In each molecule, intramolecular C7A—H7AA···S1A and C7B—H7BA···S1B hydrogen bonds (Table 1) generate six-membered rings, producing S(6) ring motifs (Fig. 1, Bernstein et al., 1995). The 1,2,4-triazole rings (N2/C8/N3/N4/C9) are essentially planar, with maximum deviations of -0.004 (2) and -0.011 (2) Å, respectively, for atoms C8A and C8B.
In molecule A, the 1,2,4-triazole ring makes dihedral angles of 85.86 (12) and 8.38 (10)°, respectively, with isobutyl-substituted phenyl ring (C11-C16) and 2-methoxyphenol moiety (C1-C6/C21/O1/O2); the comparable angles for molecule B are 89.26 (13) and 2.46 (10)°, respectively. The bond lengths (Allen et al., 1987) and angles are within normal ranges and comparable to a closely related structure (Fun et al., 2009).[](https://www.ncbi.nlm.nih.gov/mesh/C045575)
In the crystal structure (Fig. 2), intermolecular N3A—H1N3···N4B and N3B—H2N3···S1A hydrogen bonds (Table 1) link neighbouring molecules into R22(7) ring motifs (Bernstein et al., 1995). Intermolecular O2A—H1O2···O2B hydrogen bonds (Table 1) interconnect these hydrogen bond ring motifs into one-dimensional extended chains along [201]. An interesting feature of the crystal structure is the short intermolecular O1A···O2B contacts [symmetry code: -1+x, y, 1+z] with a distance of 2.781 (2) Å, which is significantly shorter than the sum of the van der Waals radii of the oxygen atoms (3.04 Å). The crystal structure is further stabilized by intermolecular C5B—H5BA···Cg1 as well as Cg2···Cg3 interactions [centroid-centroid distance = 3.6299 (13) Åiv; Cg1, Cg2 and Cg3 are the centroids of C11A-C16A, C1A-C6A and C1B-C6B phenyl rings, respectively].[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Experimental (cont.)
*In the **Experimental (cont.)** section:*
The title Schiff base compound was obtained by refluxing 4-amino-5-[1-(4-isobutylphenyl)ethyl]-4H-1,2,4-triazole-3-thiol (0.01 mol) and 4-hydroxy-3-methoxybenzaldehyde (0.01 mol) in ethanol (20 ml) for 6 h, with the addition of three drops of concentrated sulphuric acid. The solid product obtained was collected by filtration, washed with ethanol and dried. It was then recrystallized using ethanol. Single crystals suitable for X-ray analysis were obtained from ethanol by slow evaporation.[](https://www.ncbi.nlm.nih.gov/mesh/D012545)
## Refinement (cont.)
*In the **Refinement (cont.)** section:*
Atoms H1N3, H2N3, H1O2 and H2O2 were located from difference Fourier map and allowed to refine freely. All other hydrogen atoms were placed in their calculated positions, with C—H = 0.93 – 0.98 Å, and refined using a riding model, with Uiso = 1.2 or 1.5 Ueq(C). A rotating group model was used for the methyl groups. The isobutyl unit of molecule B is disordered over two positions with refined occupancies of 0.785 (6) and 0.215 (6). The same Uij parameters were used for the atom pair C19B/C19C.
The reflection (010) was omitted as the intensity was affected by the beam backstop.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Figures
*In the **Figures** section:*
The asymmetric unit of the title compound, showing 50% probability displacement ellipsoids for non-H atoms and the atom-numbering scheme. Intramolecular hydrogen bonds are shown as dashed lines. Open bonds indicate the minor disorder component.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
The crystal structure of the title compound, showing R22(7) ring motifs being linked into one-dimensional extended chains. Only the major disorder component is shown. H atoms not involved in intermolecular interactions (dashed lines) have been omitted for clarity.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Crystal data (cont.)
*In the **Crystal data (cont.)** section:*
## Data collection (cont.)
*In the **Data collection (cont.)** section:*
## Refinement (cont.)
*In the **Refinement (cont.)** section:*
## Special details
*In the **Special details** section:*
## Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2)
*In the **Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2)** section:*
## Atomic displacement parameters (Å2)
*In the **Atomic displacement parameters (Å2)** section:*
## Geometric parameters (Å, °)
*In the **Geometric parameters (Å, °)** section:*
## Hydrogen-bond geometry (Å, °)
*In the **Hydrogen-bond geometry (Å, °)** section:*
Symmetry codes: (i) x−1, y, z; (ii) x+1, y, z; (iii) x+1, y, z−1; (iv) −x+1, −y+2, −z+1.
# References
*In the **References** section:*
Hydrogen-bond geometry (Å, °)[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
|
# Introduction
N-Carbamothioyl[amino-7-oxabicyclo[2.2.1]hept-5-ene-2,3-dicarboximide](https://www.ncbi.nlm.nih.gov/mesh/D007094)
# Abstract
*In the **Abstract** section:*
The title compound, C9H9N3O3S, comprises a racemic mixture of chiral molecules containing four stereogenic[ centres.](https://www.ncbi.nlm.nih.gov/mesh/D007094) The cyclohexane ring tends towards a boat conformation, while the tetrahydrofuran ring a[nd the dihyd](https://www.ncbi.nlm.nih.gov/mesh/C506365)rofuran ring adopt envelope conformations. The dih[edral angle betwe](https://www.ncbi.nlm.nih.gov/mesh/C018674)en the thiose[micarbazide f](https://www.ncbi.nlm.nih.gov/mesh/D005663)ragment and the fused-ring system is 77.20 (10)°. The crystal struc[ture is stabilized](https://www.ncbi.nlm.nih.gov/mesh/C005151) by two intermolecular N—H⋯O hydrogen bonds.
## Related literature
*In the **Related literature** section:*
For the use of 7-oxa-bicyclo[2,2,1]hept-5-ene-2,3-dicarboxylic anhydride in clinical practice, see: Deng & Hu (2007 ▶). For the pharmacological activity of its derivatives, see: Hart et al. (2004 ▶). For bond lengths and angles in related structures, see: Goh et al. (2008 ▶).[](https://www.ncbi.nlm.nih.gov/mesh/C035993)
## Experimental
*In the **Experimental** section:*
## Crystal data
*In the **Crystal data** section:*
C9H9N3O3S[](https://www.ncbi.nlm.nih.gov/mesh/D007094)
M r = 239.25
Orthorhombic,
a = 8.3978 (8) Å
b = 8.9032 (9) Å
c = 13.5930 (14) Å
V = 1016.31 (18) Å3
Z = 4
Mo Kα radiation[](https://www.ncbi.nlm.nih.gov/mesh/D008982)
μ = 0.31 mm−1
T = 298 K
0.45 × 0.43 × 0.40 mm
## Data collection
*In the **Data collection** section:*
Bruker SMART CCD area-detector diffractometer
Absorption correction: multi-scan (SADABS; Bruker, 1997 ▶) T min = 0.872, T max = 0.885
5015 measured reflections
1791 independent reflections
1632 reflections with I > 2σ(I)
R int = 0.023
## Refinement
*In the **Refinement** section:*
R[F 2 > 2σ(F 2)] = 0.029
wR(F 2) = 0.071
S = 1.07
1791 reflections
145 parameters
H-atom parameters constrained
Δρmax = 0.14 e Å−3
Δρmin = −0.16 e Å−3
Absolute structure: Flack (1983 ▶), 728 Friedel pairs
Flack parameter: 0.01 (9)
Data collection: SMART (Bruker, 1997 ▶); cell refinement: SAINT (Bruker, 1997 ▶); data reduction: SAINT; program(s) used to solve structure: SHELXS97 (Sheldrick, 2008 ▶); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008 ▶); molecular graphics: SHELXTL (Sheldrick, 2008 ▶); software used to prepare material for publication: SHELXTL.
## Supplementary Material
*In the **Supplementary Material** section:*
Supplementary data and figures for this paper are available from the IUCr electronic archives (Reference: BX2331).
# supplementary crystallographic information
*In the **supplementary crystallographic information** section:*
## Comment
*In the **Comment** section:*
7-Oxa-bicyclo[2,2,1]hept-5-ene-2,3-dicarboxylic anhydride has been widely employed in clinical practice, as it is less toxic and much easier to be synthesized [Deng et al., 2007]. Its derivatives are also pharmacologically active [Hart et al., 2004]. We report here the crystal structure of the title compound, (I) which comprises a racemic mixture of chiral molecules containing four stereogenic centres.
The cyclohexane ring tends towards a boat conformation, the tetrahydrofuran ring and the dihydrofuran ring adopt envelope conformations (Fig. 1). The bond lengths and bond angles are normal range and comparable to those in the similar compound [Goh, et al., 2008] as representative example. The dihedral angle between the thiosemicarbazide fragment and fused-ring system is 77.20 (10)°. The crystal structure is stabilized by two intermolecular N—H···O and one intramolecular N—H···N hydrogen bonds (Table 1, Fig. 2).[](https://www.ncbi.nlm.nih.gov/mesh/C035993)
## Experimental (cont.)
*In the **Experimental (cont.)** section:*
A mixture of exo-7-oxa-bicyclo[2,2,1]hept-5-ene-2,3-dicarboxylic anhydride (0.332 g, 2 mmol) and thiocarbanide (0.182 g, 2 mmol) in methanol (5 ml) was stirred for 5 h at room temperature, and then refluxed for 1 h. After cooling the precipitate was filtered and dried, the title compound was obtained. The crude product of 20 mg was dissolved in methanol of 10 ml. The solution was filtered to remove impurities, and then the filtrate was left for crystallization at room temperature. The single-crystal suitable for X-ray determination was obtained by evaporation from the methanol solution after 5 d.[](https://www.ncbi.nlm.nih.gov/mesh/C035993)
## Refinement (cont.)
*In the **Refinement (cont.)** section:*
H atoms were initially located from difference maps and then refined in a riding model with C—H = 0.93–0.96 Å and Uiso(H) = 1.2Ueq(C) or 1.5Ueq(methyl C).
## Figures
*In the **Figures** section:*
The molecular structure of (I), with the atom-numbering scheme. Displacement ellipsoide are drawn at 30% probability level.
The crystal packing of (I), viewed along b axis. Dashed lines indicate hydrogen bonds.
## Crystal data (cont.)
*In the **Crystal data (cont.)** section:*
## Data collection (cont.)
*In the **Data collection (cont.)** section:*
## Refinement (cont.)
*In the **Refinement (cont.)** section:*
## Special details
*In the **Special details** section:*
## Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2)
*In the **Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2)** section:*
## Atomic displacement parameters (Å2)
*In the **Atomic displacement parameters (Å2)** section:*
## Geometric parameters (Å, °)
*In the **Geometric parameters (Å, °)** section:*
## Hydrogen-bond geometry (Å, °)
*In the **Hydrogen-bond geometry (Å, °)** section:*
Symmetry codes: (i) x+1/2, −y+1/2, −z+1; (ii) −x+3/2, −y+1, z+1/2.
# References
*In the **References** section:*
Hydrogen-bond geometry (Å, °)
|
# Introduction
The evolution of farnesoid X, vitamin D, and pregnane X receptors: insights from the green-spotted pufferfish (Tetraodon nigriviridis) and other non-mammalian species
# Abstract
*In the **Abstract** section:*
Background
The farnesoid X receptor (FXR), pregnane X receptor (PXR), and vitamin D receptor (VDR) are three closely related nuclear hormone receptors in the NR1H and 1I subfamilies that share the property of being activated by bile salts. Bile salts vary significantly in structure across vertebrate species, suggesting that receptors binding these molecules may show adaptive evolutionary changes in resp[onse. We h](https://www.ncbi.nlm.nih.gov/mesh/D001647)av[e previous](https://www.ncbi.nlm.nih.gov/mesh/D001647)ly shown that FXRs from the sea lamprey (Petromyzon marinus) and zebrafish (Danio rerio) are activated by planar bile alcohols found in these two species. In this report, we characterize FXR, PXR, and VDR from the green-spotted pufferfish (Tetraodon nigriviridis), an actinopterygian fish th[at unlike the](https://www.ncbi.nlm.nih.gov/mesh/D002777) zebrafish has a bile salt profile similar to humans. We utilize homology modelling, docking, and pharmacophore studies to understand the structural features of the Tetraodon receptors.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
Results
Tetraodon FXR has a ligand selectivity profile very similar to human FXR, with strong activation by the synthetic ligand GW4064 and by the primary bile acid chenodeoxycholic acid. Homology modelling and docking studies suggest a ligand-binding pocket architecture more similar to human and rat FXRs than to lamprey or zebrafish FXRs. Tetraodon PXR was activated by a variety of bile acids and steroids, although not by the larger synthetic ligands that activate human PXR such as rifampicin. Homology modelling predicts a larger ligand-binding cavity than zebrafish PXR. We also demonstrate that VDRs from the pufferfish and Japanese medaka were activated by small secondary bile acids such as lithocholic acid, whereas the African clawed frog VDR was not.[](https://www.ncbi.nlm.nih.gov/mesh/C412815)
Conclusions
Our studies provide further evidence of the relationship between both FXR, PXR, and VDR ligand selectivity and cross-species variation in bile salt profiles. Zebrafish and green-spotted pufferfish provide a clear contrast in having markedly different primary bile salt profiles (planar bile alcohols for zebrafish and sterically bent bile acids for the pufferfish) and receptor selectivity that matches these differences in endogenous ligands. Our observations to date present an integrated picture of the co-evolution of bile salt structure and changes in the binding pockets of three nuclear hormone receptors across the species studied.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Background
*In the **Background** section:*
Nuclear hormone receptors (NHRs) are transcription factors that work in concert with co-activators and co-repressors to regulate gene expression [1,2]. Most of the NHRs in vertebrates are ligand-activated, although some NHRs function in a ligand-independent manner. Examples of ligands for NHRs include a range of endogenous compounds such as bile acids, retinoids, steroid hormones, thyroid hormone, and vitamin D. A few NHRs, such as the pair of xenobiotic sensors, pregnane X receptor (PXR; NR1I2; also known as steroid and xenobiotic receptor or SXR) and constitutive androstane receptor (CAR; NR1I3), are activated by structurally diverse exogenous ligands. NHRs share a conserved domain structure, which includes, from N-terminus to C-terminus, a modulatory A/B domain, the DNA-binding domain (DBD; C domain), the 'hinge' D domain, the ligand-binding domain (LBD; E domain), and a variable C-terminal F domain that is absent in some NHRs [2,3].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
NHRs show varying degrees of sequence conservation across vertebrate species in the LBD that likely reflects, at least in part, cross-species variation in the physiologically important ligands. The xenobiotic sensors PXR and CAR show extensive cross-species amino acid sequence divergence across vertebrate species [4,5] in addition to evidence of positive (Darwinian) selection in the LBD from phylogenetic analyses [6,7]. Cross-species differences in xenobiotic ligands are a possible driving force for changes in the LBD sequence and structure. We and others have also provided data consistent with the hypothesis that the structure of the LBD of NHRs in the NR1 H and NR1I subfamilies may have co-evolved with the endogenous ligands in some species in evolution [4,6-12].[](https://www.ncbi.nlm.nih.gov/mesh/D015262)
Bile salts are one class of NHR ligands that show substantial cross-species differences in chemical structure [13]. Bile salts are amphipathic, water-soluble end-metabolites of cholesterol that facilitate intestinal absorption of lipids, enhance proteolytic cleavage of dietary proteins, and exert antimicrobial activity in the small intestine [14,15]. Bile salts exhibit significant structural diversity across vertebrate species [13,16-18] and include bile alcohols (which have a hydroxyl group on the terminal carbon atom of the side-chain) and bile acids (which have a carboxylic acid group on the side-chain) (Figure 1)
[15]. Primary bile salts are defined as those synthesized by the liver, which is accomplished by a complicated biosynthetic pathway that starts with cholesterol [19,20].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
Representative bile salts and their structures. All bile salts are derived from cholesterol (topmost structure), illustrated with the carbon atoms numbered and the steroid rings labelled A, B, C, and D. Jawless fish, lobe-finned fish, and a limited number of actinopterygian fish use 5α bile alcohols such as 5α-cyprinol-27-sulfate that have an overall planar and extended structure of the steroid rings (see representation of A, B, and C rings on the right side). Most actinopterygian fish (including medaka and Tetraodon nigrivirdis) use 5β bile acids that have an overall bent structure of the steroid rings. One of the two most common primary salts in mammals is chenodeoxycholic acid (CDCA), the stem C24 bile acid that has the basic 3α,7α-dihydroxylation pattern. Lithocholic acid is one of the smallest naturally occurring bile acids and results from bacterial enzyme-mediated deconjugation and dehydroxylation of primary bile acids. The sodium and calcium salts of lithocholic acid have very low solubility at body temperature. Additionally, lithocholic acid is toxic in humans and other mammals.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
Secondary bile salts result from modification of primary bile salts by host bacteria in the intestine [15]. Common alterations in bile acids catalyzed by enzymes from anaerobic bacteria in the intestinal tract include deconjugation and dehydroxylation, with one possible end result being unconjugated and poorly water-soluble bile acids such as lithocholic acid (3α-hydroxy-5β-cholan-24-oic acid; LCA), a compound known to be toxic to mammalian species including humans [15,21]. There have been few studies of intestinal bile salts in animals that utilize bile alcohols, but one study of the Asiatic carp (Cyprinus carpio) showed that the bile alcohol sulfates synthesized in the liver of this species did not undergo hydrolysis in the intestinal tract and thus did not generate potentially toxic unconjugated secondary bile alcohols [22].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
Variation in the stereochemistry of the junction between rings A and B of the bile salt steroid nucleus determines whether bile salts have a flat (planar) or a 'bent' orientation. Based on surveys of the bile salt composition in 1153 vertebrate species, we have proposed that 5α (planar) bile alcohols are the likely ancestral ligands (the paleomorphic state) while 5β (bent) bile acids are the derived (apomorphic) phenotype [13,23-25]. Common 5β-bile acids include chenodeoxycholic acid (CDCA; 3α,7α-dihydroxy-5β-cholan-24-oic acid) and cholic acid (CA; 3α,7α,12α-trihydroxy-5β-cholan-24-oic acid), the two dominant primary bile acids in humans and other mammals (Figure 1) [13,24].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
PXRs are activated by a structurally diverse array of endogenous and exogenous molecules that includes bile salts, steroid hormones, prescription medications, herbal drugs, and endocrine disruptors [26,27]. PXR regulates the transcription of enzymes and transporters involved in the metabolism and elimination of potentially harmful compounds, including sulfation of toxic bile acids [28]. Previous studies have shown substantial cross-species differences in PXR ligand specificity, including in the selectivity for bile salts [4,7,8]. Mammalian PXRs are activated by a broad range of bile salt structures (both ancestral and evolutionarily derived), while chicken (Gallus gallus) and zebrafish (Danio rerio) PXRs are activated by a structurally narrow range of bile salts [4,6,7,9,29]. We and others have proposed that the evolution of PXRs has been driven by at least two factors: adaptation to evolutionary changes in bile salt (and perhaps other endogenous molecular) structure and function as a xenobiotic sensor [4,7,9,12].
Compared with published X-ray crystallographic structures of human PXR [30-35], a homology model of the zebrafish PXR is predicted to have a smaller ligand-binding pocket (LBP) than human PXR, with a flat ligand-binding surface well-suited to binding the planar 5α-bile alcohols that are the primary bile salts of zebrafish and other cypriniform fish [9].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
FXR serves as the major transcriptional regulator of bile salt synthesis, in part by controlling the expression of cytochrome P450 (CYP) 7A1, the rate-limiting enzyme in the synthetic pathway [36]. Mammalian FXRs are activated best by primary bile acids CDCA and CA [37-39]. FXR is typically expressed at high levels in the liver, intestine, kidney, and adrenal glands. A second FXR, termed FXRβ (NR1H5), is found in some animal species (although it is a pseudogene in the genome of some mammals such as humans and other primates) but does not appear to be involved with bile salt binding or regulation [40].
Throughout this manuscript, FXR refers to NR1H4, or what might be termed FXRα in species possessing two FXRs.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
The African clawed frog (Xenopus laevis) expresses an unusual FXR (also termed FOR, FXR-like orphan receptor) that has a 33 amino acid insert, not found in mammalian FXRs, in helix 7 of the ligand-binding domain (LBD) [41]. We have previously demonstrated that sea lamprey (Petromyzon marinus), zebrafish, and African clawed frog FXRs are activated by bile alcohols [9]. Sea lamprey and zebrafish FXRs are selective for planar 5α bile alcohols.
The Xenopus FXR isoform 1 was activated by 5α- and 5β-bile alcohols (paralleling the complex bile salt profile of this amphibian [23]) but was poorly activated by bile acids [9,41].[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
The vitamin D receptor (VDR; NR1I1) is known to mediate the action of 1,25-dihydroxyvitamin D (calcitriol), a hormone whose 'classic' function is to regulate calcium and phosphorus homeostasis. However, VDRs are now known to be involved in a wide range of physiological processes including immune system modulation, skin development, and regulation of the metabolism of toxic compounds [42-45]. Two VDR genes have been found in the genome of many actinopterygian fish (a likely by-product of whole genome duplication prior to actinopterygian fish radiation) [46-48], with VDRα and VDRβ both shown to be functional in the Japanese medaka (Oryzias latipes)
[49]. Mammalian VDRs are activated at low affinity by a narrow range of 5β-bile acids, particularly the toxic secondary bile acid LCA and its derivatives [7,50-52]. VDR activation in the intestine has been shown to upregulate the expression of enzymes (e.g., CYP3A) that can metabolize and reduce the toxicity of LCA [52-56]. We previously determined that the VDR from the sea lamprey (Petromyzon marinus) was insensitive to activation by a wide array of bile salts, including bile acids and bile alcohols with a range of substituents. We proposed the hypothesis that activation of VDRs by bile acids is a 'derived' trait, possibly as an adaptation to evolutionary changes in bile salt pathways and vertebrate physiology that allow for generation of toxic secondary bile acids [6,7,9,10].[](https://www.ncbi.nlm.nih.gov/mesh/D002117)
In this report, we characterize the ligand selectivity of FXR, VDR, and PXR from the green-spotted pufferfish (Tetraodon nigriviridis), an actinopterygian fish that synthesizes mainly the 5β-bile acids CA and CDCA [23], thereby having a bile salt profile similar to most mammals. We also cloned and characterized VDRs from three additional non-mammalian species: medaka, African clawed frog, and chicken (Gallus gallus). In terms of primary bile salts, the medaka synthesizes a mixture of 5β C24 and C27 bile acids [57,58]. The African clawed frog has a complicated mixture of C26 and C27 bile alcohols, C24 bile acids, and C27 bile acids in its bile [9,23].
The chicken synthesizes mainly CA and CDCA as primary bile salts [25].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
We generated homology models of Tetraodon FXR (tnFXR) and PXR (tnPXR) using rat FXR [59] and human PXR [32] crystallographic structures, respectively, as templates. We then performed computational ligand docking experiments to these receptors. This allowed for comparison to the structures of the corresponding mammalian receptors, helping to formulate hypotheses to rationalize the structural changes that have occurred during receptor evolution. Additionally, we build on our previous docking studies with bile acids and human VDR (hVDR) and expand structure-activity series of bile acids at this receptor. Combining information from docking experiments and cross-species sequence comparisons, we generated site-directed mutations and chimeras to better understand the molecular determinants of bile salt activation of VDRs.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Results
*In the **Results** section:*
## Bile salt profile of Tetraodon nigriviridis
*In the **Bile salt profile of Tetraodon nigriviridis** section:*
We previously reported a survey of the biliary bile salts of 213 fish species from 38 different orders [23]. Included in this survey were six species from Tetraodontiformes, an order of actinopterygian fish that includes Tetraodon nigriviridis. The primary bile salts from all six species of Tetraodontiformes examined were the common 5β-bile acids CA and CDCA. This is illustrated in Additional file 1 by high-performance liquid chromatography (HPLC) and electrospray ionization-tandem mass spectrometry (ESI/MS/MS) (Additional file 1: figures S1A-S1C). The primary bile salt pool of Tetraodon is quite similar to that of many mammals (including humans) and birds [13,16,17,24,25].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Ligand selectivity of the Tetraodon FXR
*In the **Ligand selectivity of the Tetraodon FXR** section:*
Analysis of the draft genome of Tetraodon nigriviridis revealed a single putative ortholog to human FXRα (hFXR) and another single putative ortholog to human PXR (hPXR). We did not find any evidence of an ortholog to FXRβ in the Tetraodon genome. The LBDs of tnFXR and tnPXR were cloned and functionally expressed as GAL4/LBD chimeras (concentration-response data in Figure 2 and 3). tnFXR was activated most strongly by GW4064 (Figure 2A), a synthetic ligand known to activate mammalian FXRs with high efficacy [60], and activated weakly by T-0901317 (Figure 2B), an agonist of mammalian liver X receptors (LXRs; NR1H2 and NR1H3), FXRs, and PXRs [34,61,62].
Conjugated and unconjugated 5β bile acids also activated tnFXR, including CDCA, tauro-CDCA, 3-oxo-LCA, and LCA (Figure 2C,D; Additional file 2). The EC50 of bile acids for activation of tnFXR ranged from 8.8 to 37.6 μM, with maximal effects varying from 15-61% of that produced by GW4064 (which served as the reference compound for comparisons of maximal effect or efficacy). The planar bile alcohol 5α-cyprinol sulfate (the main bile salt of cypriniform fish but only a very minor component of Tetraodon bile salts) weakly activated tnFXR. A 76-compound library of known NHR ligands was screened for additional activators of tnFXR. This screening identified three activators of tnFXR in addition to bile acids which included two farnesol derivatives (farnesol and S-farnesyl-L-cysteine methyl ester) and 6-formylindolo-[3,2-b]-carbazole.[](https://www.ncbi.nlm.nih.gov/mesh/C412815)
FXR concentration-response curves. Ligand activation of FXRs from different species. A) The synthetic mammalian FXR activator GW4064 activates human FXR (hFXR) with high efficacy and submicromolar potency. GW4064 also activates Tetraodon FXR (tnFXR) and zebrafish FXR (zfFXR) but not Xenopus laevis FXR(xlFXR). B) T-0901317 activates hFXR, zfFXR, and tnFXR (although with lower efficacy compared to GW4064) but not xlFXR. C) The primary bile acid chenodeoxycholic acid (CDCA) activates hFXR and tnFXR but not zfFXR or xlFXR. D) The secondary bile acid lithocholic acid (LCA) activates hFXR, tnFXR, and zfFXR, but not xlFXR. The ordinate indicates fold induction compared to vehicle control in luciferase-based assay. Note that the scale of the ordinate is different in A) through D).[](https://www.ncbi.nlm.nih.gov/mesh/C412815)
PXR concentration-response curves. Ligand activation of PXRs from different species. A) T-0901317 activates human PXR (hPXR) and chicken PXR (chPXR) but not Tetraodon PXR (tnPXR) or zebrafish PXR (zfPXR). B) The primary conjugated bile acid taurochenodeoxycholic acid (TCDCA) activates hPXR and tnPXR but not chPXR or zfPXR. C) The steroid 5β-pregnane-3,20-dione (5β-pregnanedione) activates all four PXRs shown. D) Nifedipine activates hPXR, chPXR, and zfPXR, but not tnPXR. The ordinate indicates fold induction compared to vehicle control in luciferase-based assay. Note that the scale of the ordinate is different in A) through D).[](https://www.ncbi.nlm.nih.gov/mesh/C423915)
## Ligand selectivity of the Tetraodon PXR
*In the **Ligand selectivity of the Tetraodon PXR** section:*
With respect to bile salts, tnPXR had similar structure-activity relationships to tnFXR, namely activation by 5β-bile acids with one or two hydroxyl groups on the nuclear rings (Figure 3; Additional file 2). Unlike tnFXR, tnPXR was not activated by GW4064 or farnesol compounds. tnPXR was also activated by a variety of androstane, estrane, and pregnane steroids, similar to PXRs from mammals, chicken, and zebrafish (Figure 3C) [4]. tnPXR was not activated by several xenobiotics known to be agonists of mammalian PXRs including hyperforin (pharmacologically active component of the herbal antidepressant St. John's wort), nifedipine (Figure 3D), rifampicin, and SR12813, but was activated by n-butyl 4-amino benzoate, an agonist of Xenopus laevis PXRs [4,63] (Additional file 2).[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Structure-activity relationship of bile acids for activation of human and mouse VDRs
*In the **Structure-activity relationship of bile acids for activation of human and mouse VDRs** section:*
Expanding on previous studies from our group [6,7,10] and others [50,52], we determined the structure-activity relationships of naturally occurring and synthetic bile acids for transactivation of GAL4/LBD constructs of hVDR and mouse VDR (mVDR). LCA and derivatives (e.g., 3-oxo-LCA, LCA acetate) activated both hVDR and mVDR (Figure 4; Additional file 3). Iso-LCA, with a single 3β-hydroxy group on the steroid rings (as opposed to 3α-hydroxy of LCA), weakly activated hVDR and mVDR. In contrast, naturally occurring 5β-bile acids with two or three hydroxyl groups on the nuclear rings were inactive including CDCA (3α,7α-dihydroxy-5β-cholan-24-oic acid), DCA (3α,12α-dihydroxy), and CA (3α,7α,12α-trihydroxy). We also tested three 5β-bile acids not known to occur naturally in bile that have a single hydroxyl substituent on the steroid rings on a carbon other than C-3 (7α-hydroxy, 7β-hydroxy, and 12α-hydroxy-5β-cholan-24-oic acids), as well as unsubstituted 5β-cholanic acid (no hydroxyl groups on any of the steroid rings).
All four of these bile acids were inactive with respect to activation of hVDR and mVDR. Thus, bile acids with hydroxyl groups at the C-7 or C-12 position are unfavourable for activation of hVDR (Figure 4). Unsubstituted 5α-cholanic acid, which would have an overall planar orientation of the steroid rings, weakly activated hVDR and mVDR. Two 5α-cholanic acid derivatives (3β-hydroxy and 3-oxo) were inactive (Additional file 3).[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
Structure-activity for human and mouse VDR. Structure-activity data for bile acid activation of human and mouse VDRs. The bile acids that activate human and mouse VDRs are small, hydrophobic bile acids that are related to lithocholic acid (LCA) and have either a hydroxyl or oxo group at C-3 (LCA, 3-oxo-LCA, LCA acetate, iso-LCA) or have unsubstituted steroid ring (5α-cholanic acid). Addition of a 7α-hydroxy group to LCA (resulting in the bile acid CDCA) or even single steroid ring substitutions at C-7 or C-12 result in bile acids that do not activate the VDRs. Nor-LCA, which has a shortened bile acid side-chain, is also inactive.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Activation of non-mammalian VDRs by bile salts
*In the **Activation of non-mammalian VDRs by bile salts** section:*
The African clawed frog VDR (Xenopus laevis VDR; xlVDR) was not activated by any bile salts tested, including bile alcohols. In contrast, chicken VDR (chVDR), medaka VDRα (olVDRα), Tetraodon VDRα (tnVDR), and zebrafish VDRα (zfVDRα) were each activated by LCA and/or its derivatives (3-keto-LCA and LCA acetate) but not by bile acids with two or more hydroxyl groups such as CDCA, DCA, or CA (Figure 5 and 6; Additional file 3). The efficacies of LCA, 3-oxo-LCA, and LCA acetate (in comparison to 1,25α-dihydroxyvitamin D3) for activation of chicken, medaka, Tetraodon, zebafish VDRs were lower than for hVDR and mVDR (Figure 6; Additional file 3).[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
Transactivation of full-length teleost VDRs. HepG2 cells were transiently transfected with pRL-CMV, XREM-Luc and either medaka VDRα-pSG5, zebrafish VDRα-pSG5, or Tetraodon VDRα-pSG5 as described in Methods. Cells were exposed to 100 μM of either lithocholic acid (LCA), 3-keto-LCA, or LCA acetate for 24 hours. VDR response was measured via dual-luciferase assays. Data is represented as the mean fold induction normalized to control (DMSO) ± SEM.[](https://www.ncbi.nlm.nih.gov/mesh/D008095)
VDR concentration-response curves. Ligand activation of VDRs from different species, chimeric human-frog VDRs, and mouse VDR site-directed mutants. A) The plot shows activation of human, chicken, Xenopus laevis, and Tetraodon VDRs by 1,25-dihydroxyvitamin D3. Also shown are activation data for a chimeric receptor that is mostly human VDR (hVDR) with the H1-H3 insert replaced by the corresponding sequence from Xenopus laevis VDR (xlVDR) and the converse chimeric receptor that is mostly xlVDR with the H1-H3 insert replaced by the corresponding sequence from hVDR. B) The plot shows activation of mouse VDR (mVDR) and six site-directed mutants of mVDR (R269A, R269E, H392A, H392Y, F417A, and F417D) by 1,25-dihydroxyvitamin D3. C) Lithocholic acid (LCA) activates human, chicken, and Tetraodon VDRs, as well as the chimeric receptor hVDR with xlVDR h1-H3 insert, but does not activate xlVDR and the chimeric receptor xlVDR with the hVDR H1-H3 insert. D) Lithocholic acid strongly activates mVDR, weakly activates the site-directed mutants F417A and F417D, and does not activate R269A, R269E, H392A, and H392Y. The ordinate indicates fold induction compared to vehicle control in luciferase-based assay. Note that the scale of the ordinate is different in A) through D).[](https://www.ncbi.nlm.nih.gov/mesh/D002117)
## Structure-directed mutagenesis experiments
*In the **Structure-directed mutagenesis experiments** section:*
We previously used molecular modelling computational docking studies to understand the structural basis of bile acid activation of hVDR and mVDR [9]. These studies predicted an electrostatic interaction between Arg-274 (hVDR numbering) and the bile acid side-chain, and a hydrogen bond between the 3α-hydroxyl group of LCA and His-397 in helix 11 (note corresponding residue numbers are 5 lower for mVDR; e.g., Arg-269 in mVDR is equivalent to Arg-274 in hVDR). This hydrogen bonding brings LCA close to the activation helix 12 where LCA forms hydrophobic contacts with Val-398 and Phe-422 that would stabilize the helix in the optimal orientation for coactivator binding. Site-directed mutagenesis by Adachi et al. supported this conclusion and indicated that alteration on this Arg residue of hVDR (e.g., Arg274Leu) significantly disrupted the receptor response to LCA [51].
Additional file 4 displays the surface around the ligand binding pocket of hVDR, showing that it is predominantly hydrophobic in the middle with more polar features on its ends.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
We next performed site-directed mutagenesis experiments to confirm the docking model of the bile acid to VDR, and to try to rationalize the cross-species differences in activation of VDR by bile salts. These mutations were performed in mVDR, which generally has higher maximal activation by bile acids but shows a similar selectivity for bile acids to hVDR. Three residues, previously identified by the hVDR docking model as key to bile acid activation - Arg-269 (R269; charge clamp to carboxylic acid group on bile acid side-chain), His-392 (H392; hydrogen bond to 3α-hydroxy group of LCA), Phe-417 (F417; stabilization of helix 12) - were mutated individually to two other amino acid resides. This produced a total of six-directed mutants: R269A, R269E, H392A, H392Y, F417A, and F417D.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
We also made site-directed mutations at amino acid positions that showed differences between VDRs that are bile acid-sensitive (human, mouse, chicken, green-spotted pufferfish) and bile acid-insensitive (African clawed frog, sea lamprey, zebrafish). These mVDR mutations changed the amino acid residue(s) in mVDR to the corresponding residue(s) found in the bile-salt insensitive VDRs: Q286E (to corresponding residue found in xlVDR), S293 D (xlVDR), S293G (sea lamprey VDR), N319 D (lamprey VDR), N319K (zfVDR), and RCR363-365LCK (xlVDR), RCR363-365RIQ (zfVDR), RCR363-365ACR (lamprey VDR) (see Additional file 5). Lastly, we also tested the role of the H1-H3 'insertion' domain in mediating bile acid activation. Thus, we generated hVDR
that lacked amino acid residues 165-218 (hVDR/Δins) and an hVDR construct where the insert domain was replaced with the insertion domain from the bile acid-insensitive xlVDR (hVDR/xlVDRins). The analogous constructs were also created for xlVDR: one lacking the insertion domain (xlVDR/Δins) and the other with the insertion domain from hVDR (xlVDR/hVDRins).[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
The two mutations at Arg269 (R269A and R269E) reduced the apparent affinity of 1α,25-dihydroxyvitamin D3 by over two orders of magnitude and abolished activation by LCA and 3-oxo-LCA. Similar effects were produced by the two mutations at His392 (H392A and H392Y) which reduced the apparent affinity of 1α,25-dihydroxyvitamin D3 by approximately 10-fold and 100-fold, respectively, and abolished activation by LCA and 3-oxo-LCA. The two mutations at Phe417 (F417A and F417D) both reduced the apparent affinity of 1α,25-dihydroxyvitamin D3 by approximately two orders of magnitude while retaining activation by LCA and 3-oxo-LCA but not LCA acetate. Interestingly, unsubstituted 5α-cholanic acid activated the R269A, R269E, H392A, H392Y, F417A, and F417D mutants weakly.[](https://www.ncbi.nlm.nih.gov/mesh/D001120)
In general, the mutations at Glu-286, Ser-293, Asn-319, and Arg/Cys/Arg (363-365) had little effect on activation by 1α,25-dihydroxyvitamin D3 or bile acids compared to wild-type mVDR. Similarly, the four constructs that involved either deletion (hVDR/Δins and xlVDR/Δins) or swapping of the H1-H3 insertion domain (hVDR/xlVDRins and xlVDR/hVDRins) had little effect on activation by 1α,25-dihydroxyvitamin D3 or bile acids compared to wild-type hVDR. Thus, the H1-H3 insertion domain does not appear to play a major role in selectivity for bile acids.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
## Pharmacophore models for Tetraodon VDR, PXR, and FXR
*In the **Pharmacophore models for Tetraodon VDR, PXR, and FXR** section:*
Using the structure-activity data for tnVDR, tnPXR, and tnFXR (summarized in Additional File 6), we developed common features pharmacophore models for these three receptors (Figure 7). The tnVDR pharmacophore, mapped onto 1,25-dihydroxyvitamin D3 in Figure 7A, has four hydrophobes and two hydrogen bond acceptors. The tnPXR pharmacophore, mapped onto taurochenodeoxycholic acid in Figure 7B, shows three hydrophobes and one hydrogen bond acceptor, a smaller pharmacophore than that determined for hPXR [64]. The tnFXR pharmacophore, mapped onto GW4064 in Figure 7C, shows four hydrophobes, one hydrogen bond acceptor, and one negative ionizable feature.[](https://www.ncbi.nlm.nih.gov/mesh/D002117)
Pharmacophore models of Tetraodon VDR, PXR, and FXR. Pharmacophore models were developed for A) Tetraodon VDR (tnVDR), B) Tetraodon PXR (tnPXR), and C) Tetraodon FXR (tnFXR) using the HipHop method in Catalyst™. The structure-activity data used to develop the pharmacophore models are summarized in Additional File 6. A) The tnVDR pharmacophore contains 4 hydrophobes (cyan) and two hydrogen bond acceptors (green). 1,25-Dihydroxyvitamin D3 (calcitriol) was mapped to the pharmacophore. B) The tnPXR pharmacophore contains three hydrophobes (cyan) and one hydrogen bond acceptor (green). Taurochenodeoxycholic acid was mapped to the pharmacophore. C) The tnFXR pharmacophore contains four hydrophobes (cyan), one hydrogen bond acceptor (green), and one negative ionizable feature (blue). GW4064 was mapped to the pharmacophore.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Homology modeling of Tetraodon PXR
*In the **Homology modeling of Tetraodon PXR** section:*
Structural studies of the LBD of hPXR reveal a large (~1,300 Å3), roughly spherical, hydrophobic LBP with the flexibility to accommodate large molecules such as hyperforin and rifampicin [30,32-34]. We can also see this by surface analysis of the co-crystallized ligands that cover a molecular weight range of 273-714 Da and a calculated ALogP (measure of hydrophobicity) range of 3.54-10.11 [65]. The homology model we generated of the LBD of tnPXR showed an LBP predicted to be slightly smaller (1,230 Å3) compared to hPXR (Figure 8), but larger than the estimated volumes of homology models of the LBPs of zebrafish PXR (~1,000 Å3) and Xenopus laevis PXRα (~860 Å3) that we reported previously [9,66]. The homology model results for tnPXR are consistent with the pharmacophore model for this receptor described above (Figure 7B).[](https://www.ncbi.nlm.nih.gov/mesh/C001654)
PXR-ligand interactions. Interaction map between the secondary bile acid lithocholic acid and PXRs. Key residues involved in binding are listed for human, Tetraodon (tn), and zebrafish (z) PXRs (the residue number is for human). Conserved resides throughout the three species are highlighted by black boxes. The hydrogen bond is shown by a green dashed line, the electrostatic charge interaction is presented as a blue dashed line, and the hydrophobic interaction is depicted as an orange dashed line. The ligand is shown as ball and stick presentation and colored according to atom types (gray = carbon, red-oxygen).[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
We performed molecular docking studies with two steroids (5β-pregnan-3,20-dione and 5α-androstan-3α-ol) and one bile acid (CDCA) that activated tnPXR in the luciferase assays (Figure 9). Docking studies show that the two steroids bind similar to the published crystallographic structure of 17β-estradiol bound to hPXR [35]. The carbonyl group of pregnanedione or hydroxyl group of androstanol at the C-3 position has a hydrogen bond with Thr-111, which corresponds to Ser-247 in hPXR. For pregnanedione, the carbonyl group at C-20 forms another hydrogen bond with Gln-305, as shown in Figure 6.
Molecular docking indicated that the electrostatic interactions between CDCA and the conserved Arg-151 residue on helix 5 are important for binding to tnPXR. The hydroxyl group at the C-3 position forms a hydrogen bond with Glu-149, which is different from tnFXR, for which a His residue is involved in hydrogen bonding. For tnPXR, Tyr is in the position of this His residue, and adopts a different side-chain rotamer orientation to provide van der Waals contacts with the hydrocarbon scaffold of CDCA.[](https://www.ncbi.nlm.nih.gov/mesh/D013256)
Homology model of Tetraodon PXR. Plot of interactions of ligands with tnPXR: A) the steroid 5β-pregnan-3,20-dione (pregnanedione) and B) the bile acid CDCA. The legend indicates the types of amino acid residues and amino acid-ligand interactions.[](https://www.ncbi.nlm.nih.gov/mesh/D013256)
Homology modeling and docking studies of zebrafish PXR [9] show an important amino acid sequence difference (Met-243 in human PXR versus Phe at the corresponding position in zebrafish PXR). This residue is located at the bottom of the PXR LBP and has direct van der Waals contacts with A-ring and B-ring of the hydrocarbon scaffold of bile salts, specifically the methyl group at the C-10 position. In zebrafish PXR, the bulky and more rigid benzyl side-chain of the phenylalanine significantly narrow this portion of the zebrafish PXR LBP compared with human PXR which has the flexible side-chain of methionine-243, which may lead to preference for planar bile alcohols by zebrafish PXR. tnPXR has valine at the position corresponding to methionine-243 in human PXR, which allows enough room for either bent or planar formations of the A/B ring of bile salts.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## Homology modeling of Tetraodon FXR
*In the **Homology modeling of Tetraodon FXR** section:*
For the homology model of tnFXR, molecular docking studies show that GW4064 adopts a similar orientation in the structural model as in the hFXR crystal structure [60]. The carboxyl group of GW4064 forms a strong electrostatic interaction with the conserved Arg residue in helix 5 of tnFXR. In addition, extensive hydrophobic contacts occur between the di-chloro-substituted benzyl ring of GW4064 and residues in helices 5, 7, 11 and 12 of tnFXR, which would stabilize the active conformation of receptor. CDCA also activated tnFXR in our experiments, but with weaker activity compared with its activity in hFXR.
Sequence alignment indicated that most of the ligand contact residues identified in the crystallographic structure of hFXR [60] are conserved between tnFXR and hFXR. However, one important hydrogen bond interaction between the C-3 hydroxyl group of CDCA and hFXR is not present for tnFXR because Tyr-358 (hFXR) in helix 7 is substituted for by Phe in tnFXR.[](https://www.ncbi.nlm.nih.gov/mesh/C412815)
There is an opening to solvent for the FXR LBP as seen in crystallographic structures of human or rat FXRs [59,60] and our homology models (Figure 10). The width of this opening is mostly controlled by two residues - Met-262 and Ile-332 in human FXR. For tnFXR, the residue corresponding to hFXR Ile-332 is alanine. Our model suggests that the LBP of tnFXR has a wider opening to solvent than hFXR, which may explain how tnFXR and not hFXR can accommodate cyprinol 27-sulfate, a C27 bile acid that has a longer side-chain and more hydroxyl groups than CDCA. The pharmacophore model for tnFXR summarized in Figure 7C is consistent with the homology modelling and docking results, in having multiple hydrophobic interactions, a hydrogen bond acceptor, and negative ionizable features for the bound ligand.[](https://www.ncbi.nlm.nih.gov/mesh/D008715)
FXR-ligand interactions. Interaction map between the primary bile acid chenodeoxycholic acid and FXRs. Key residues involved in binding are listed for human, mouse (m), Tetraodon (tn), and sea lamprey (sl) FXRs (the residue number is for human). Conserved resides throughout the four species are highlighted by black boxes. Hydrogen bonds are shown by green dashed lines, the electrostatic charge interaction is presented as a blue dashed line, and the important cation-π interaction for FXR is depicted as a brown dashed line. The ligand is shown as ball and stick presentation and colored according to atom types (gray = carbon, red-oxygen).[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Discussion
*In the **Discussion** section:*
In this study, we characterize the FXR, VDR, and PXR of the green-spotted pufferfish Tetraodon nigriviridis, an actinopterygian fish that synthesizes mainly 5β-bile acids. This bile salt profile is similar to humans and most mammals, as well as many other actinopterygian fish, some reptiles, and some birds [13,23]. We find that the FXR and PXR of Tetraodon (tnFXR and tnPXR) are activated by common bent-shaped 5β-bile acids such as CDCA and less well by planar 5α bile alcohols, matching the endogenous primary bile salt profile of this species.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
Our experimental and molecular modelling studies of tnFXR reveal a receptor that differs in multiple respects from sea lamprey and zebrafish FXRs [9] but is instead more similar to medaka FXRα [58] and mammalian FXRs. Homology modelling of tnFXR predicts an LBP with a wide opening to solvent (wider than crystallographic structure of hFXR [60]) and an overall architecture that differs markedly from homology models of sea lamprey and zebrafish FXRs, which are both predicted to have narrow LBPs that can accommodate planar bile alcohols but cannot accommodate bent bile acids such as CDCA. Docking studies of lamprey and zebrafish FXRs predict good binding to planar 5α-bile alcohols but not to bent 5β-bile acids, a prediction that is consistent with reporter assay experiments [9]. For tnFXR, docking studies show that the FXR-selective agonist GW4064 and bile acid CDCA adopt orientations in the LBP similar to those observed in the crystallographic structures of GW4064 bound to hFXR
[60] and rat FXR bound to an analog of CDCA, respectively [59], lending further support to the conclusion that tnFXR has an overall LBP structure more similar to human and rat FXRs than to sea lamprey or zebrafish FXRs.[](https://www.ncbi.nlm.nih.gov/mesh/D002777)
Tetraodon FXR has similar ligand selectivity to medaka FXRα (to which it shares high sequence identity in the LBD), including strong activation by GW4064 and the primary bile acid CDCA (both unconjugated and taurine-conjugated) [58]. Given that the activation of medaka FXRα upregulates the transcription of genes important in bile salt synthesis and transport, including the genes for CYP7A1 and bile salt export protein as well as the NHR repressor small heterodimer partner (SHP, NR0B2) [67], Tetraodon FXR seems likely to also be involved in regulation of bile salt biology. Tetraodon FXR and medaka FXRα have a substantially different ligand selectivity from sea lamprey FXR, which is not activated by GW4064, CDCA (or other 5β-bile acids), or T-0901317 [9].
This is not surprising because the primary bile salts of the sea lamprey (planar 5α-bile alcohols) are quite different from those of Tetraodon (CDCA, CA) and medaka (CDCA, CA, and C27 bile acids) [23,57,58]. The FXRs of Tetraodon and medaka also differ significantly from an FXR cloned and characterized from the little skate (Leucoraja erinacea, a cartilaginous fish), which was found to be insensitive to bile salts, even the 5β-bile alcohols (e.g., 5β-scymnol sulfate) produced by the little skate and many other cartilaginous fish [68]. However, the skate FXR showed significant differences in sequence from other vertebrate FXRs, including novel insertions, and there is the possibility that this receptor is actually orthologous to mammalian FXRβ (NR1H5), which are activated not by bile salts but instead by other steroidal compounds such as lanosterol [40].[](https://www.ncbi.nlm.nih.gov/mesh/C412815)
Homology modelling of tnPXR predicts a receptor with an LBP with a volume (1,230 Å3) significantly larger than the estimated volumes of the LBPs of zebrafish PXR (~1,000 Å3) and Xenopus laevis PXRα (~860 Å3) [9,66] but smaller than the LBP of hPXR in crystallographic structures [30,32-34]. The homology model and docking studies of tnPXR are consistent with our studies of recombinant tnPXR in luciferase reporter assays that show activation by a wider range of ligands than can activate either zebrafish or Xenopus PXRs. For example, tnPXR is activated by a broad range of 5β- and 5α-bile salts and steroids similar to hPXR. In contrast, studies of recombinant zebrafish PXR show activation by a narrow range of planar 5α-bile alcohols [4,6,7].
The pharmacophore model for tnPXR is similar to models of hPXR [64,69,70] which have four hydrophobic features along with a hydrogen bond acceptor, although the tnPXR pharmacophore had three hydrophobic features and a hydrogen bond donor, consistent with a somewhat more restricted LBP limited to smaller ligands than hPXR [29,64]. Like hPXR, tnPXR has a pharmacophore model markedly different from that of chicken and zebrafish PXRs [29].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
Overall, the ligand selectivity of tnFXR and tnPXR contrasts with that of sea lamprey FXR, zebrafish FXR, and zebrafish PXR, which are activated better by 5α bile alcohols, which are the main bile salts of jawless fish (Agnatha, which includes extant lampreys and hagfish) and Cypriniformes (which includes zebrafish) [9,23]. Unlike lampreys and zebrafish, Tetraodon nigriviridis has a bile salt profile of mainly bent 5β-bile acids such as CDCA, and would thus be predicted to have bile-salt-regulating NHRs that can recognize Tetraodon endogenous bile acids and not the 'ancestral' bile alcohols that comprise only a trace fraction of the Tetraodon bile salt pool. These results lend further significant support to the hypothesis that structural changes to the LBD of FXR and PXR throughout evolution have paralleled cross-species changes in bile salt profile [6,7,10,11,66].
In Additional file 7 bile salt variation and NHR bile salt sensitivity are overlaid on a fish phylogeny. So far, only a limited number of fish species have been characterized with respect to their FXRs, VDRs, and/or PXRs. As discussed in Additional file 7 a number of fish groups would be of particular interest for future studies of NHRs, including the Elasmobranchii (sharks, skates, rays), Chimaeriformes (chimaerae), Myxiniformes (hagfish), and lobe-finned fish (lungfish and coelacanths).[](https://www.ncbi.nlm.nih.gov/mesh/D002777)
Our present and previous studies [6,7,9,10] with VDRs from various non-mammalian species have revealed some VDRs that are insensitive to bile acids (sea lamprey, African clawed frog) and others that are activated by secondary bile acids such as LCA (human, mouse, chicken, medaka, Tetraodon). In humans and some other mammals, enzymes from anaerobic bacteria in the caecum deconjugate and dehydroxylate primary bile acids such as CDCA, leading to secondary bile acids such as LCA that can produce toxicity to the intestinal and hepatobiliary tracts [15,21]. We and others have speculated that the ability of VDRs to bind secondary bile acids was acquired during vertebrate evolution as a protective mechanism against the potential toxicity of poorly water-soluble secondary bile acids such as LCA [6,7,51,52]. One challenge to this hypothesis is that there has been little study to date of the secondary bile salts (and the anaerobic bacterial intestinal flora that could generate such bile salts) of non-mammalian species. For example, it is not known whether Tetraodon or other actinopterygian fish species have physiologically important amounts of secondary bile acids in the intestinal tract.
With these caveats in mind, we summarize our studies of VDRs across different species.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
The structure-activity and molecular modelling data are both consistent with a tight, hydrophobic binding pocket for bile acids in the human VDR LBP that can bind bile acids with an oxo or single hydroxyl group at the C-3 position but not bile acids with substituents at the C-7 or C-12 positions (including unnatural bile acids with a single hydroxyl group on C-7 or C-12 and no substituent on C-3). Molecular modeling predicts an hVDR bile acid binding pocket that is predominantly hydrophobic but with polar features that permit hydrogen bond interaction with the 3α-hydroxy or 3-oxo group on the A ring of the bile acid and electrostatic interactions with the bile acid side-chain as described above. Bile acids that have hydroxyl groups on steroid rings B and/or C (CDCA, DCA, CA, 7α-hydroxy-5β-cholanic acid, 7β-hydroxy-5β-cholanic acid, and 12α-hydroxy-5β-cholanic acid) do not interact favourably with the more hydrophobic and sterically constrained portion of the hVDR bile acid binding pocket, consistent with the lack of activity of these bile acids in transactivation assays. On the contrary, the unsubstituted and hydrophobic B, C, and D rings of LCA complement well the lipophilic portion of hVDR pocket and can form numerous van der Waals interactions. We suggested that the unique physicochemical arrangement of the hVDR ligand binding cavity provides the structural basis for selective activation by LCA and its derivatives [9].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
It is less clear, even with our mutagenesis studies, exactly what changes mediate the cross-species differences in VDR pharmacology. Our studies rule out the helix 1 - helix 3 'insertion' domain as playing a role in the cross-species differences in activation by secondary bile acids. This insert is disordered in human and mouse VDRs and shows highly variable sequences and lengths across species [71-74]. Our site-directed mutagenesis experiments do confirm the previous prediction that Arg-274 and His-397 play a key role in interacting with bile acids [50,51].
Interestingly, Arg-274 and His-397 are conserved in the bile salt-insensitive xlVDR, indicating that other determinants underlie the differences in bile salt pharmacology between xlVDR and bile-salt responsive VDRs.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Conclusions
*In the **Conclusions** section:*
Our studies provide further important evidence of the relationship between FXR, PXR, and VDR ligand selectivity and cross-species variation in bile salt profiles. Zebrafish and the green-spotted pufferfish Tetraodon nigriviridis provide a clear contrast in having markedly different primary bile salt profiles (planar bile alcohols for zebrafish and sterically bent bile acids for the pufferfish) and correspondingly contrasting receptor selectivity that matches the differences in endogenous ligands. In contrast, Tetraodon has a bile salt profile and receptor ligand selectivity similar to humans. Our observations present an integrated picture of co-evolution of bile salt structure and the binding pockets of three nuclear hormone receptors.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Methods
*In the **Methods** section:*
## Bile sample analysis
*In the **Bile sample analysis** section:*
Bile salts from the green-spotted pufferfish were analyzed by HPLC and ESI/MS/MS using methods previously described [23,75].[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Chemicals
*In the **Chemicals** section:*
The sources of chemicals were as follows: GW4064 (Sigma, St. Louis, MO, USA); T-0901317 (Axxora, San Diego, CA, USA); 3α,7α,12α,24-tetrahydroxy-5α-cholan-24-sulfate (5α-petromyzonol sulfate; Toronto Research Chemical, Inc., North York, ON, Canada); 1α,25-(OH)2-vitamin D3, Nuclear Receptor Ligand Library (76 compounds known as ligands of various NHRs; BIOMOL International, Plymouth Meeting, PA, USA). All other commercially purchased steroids and bile salts were obtained from Steraloids (Newport, RI, USA). Norlithocholic acid, 3β-hydroxy-5α-cholan-24-oic acid, 3β-hydroxy-5β-cholan-24-oic acid, 7α-hydroxy-5β-cholan-24-oic acid, 7β-hydroxy-5β-cholan-24-oic acid, and 12α-hydroxy-5β-cholan-24-oic acid were generously donated by the laboratory of A.F. Hofmann (University of California - San Diego, La Jolla, CA, USA).[](https://www.ncbi.nlm.nih.gov/mesh/C412815)
## Cloning and expression of full-length VDRs
*In the **Cloning and expression of full-length VDRs** section:*
Full-length VDRα sequences from medaka (Oryzias latipes), zebrafish, and green spotted pufferfish were identified within each species genome by conducting a generalized BLAST search using a query sequence representing the P box domain of human VDR located within the highly conserved DNA binding region for the NHR superfamily. Full-length transcripts were subsequently determined by identifying transcriptional start and stop codons for each gene. cDNAs containing a complete open reading frame for each gene were produced by PCR using primer sets that spanned the entire nucleic acid sequence for each gene including start and stop codons. cDNAs were produced from extracts of fish liver total RNA. Livers were homogenized with 1 mL RNA Bee (Tel-Test, Inc.
Friendswood, TX, USA) using a stainless steel Polytron homogenizer (Kinematica, Lucerne, Switzerland) followed by cleanup and on- column DNase treatment using an RNeasy Mini Kit (Qiagen, Valencia, CA, USA).
RNA was eluted with 30 μl RNase-free water. RNA quantity and quality were verified using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) and NanoDrop® ND-1000 spectrophotometer (ThermoScientific, Wilmington, DE, USA). First strand cDNA was made from total RNA (1-3 μg) and diluted with RNase-free water to a final volume of 10 μl, and 1 μl oligo(dT)15 (500 μg/ml; Promega, Madison, WI, USA) and 1 μl 10 mM dNTPs were mixed with diluted RNA to yield a final volume of 20 μl. The mix was heated to 65°C for 5 min and chilled on ice for 2 min. Following centrifugation, 4 μl 5X first-strand buffer (Invitrogen, Carlsbad, CA, USA), 2 μl of 0.1 M DDT, and 1 μl RNase OUT Inhibitor (40 U/μl; Invitrogen) were added to each reaction and heated to 37°C.
Following a 2 min incubation, 1 μl Superscript Reverse Transcriptase (200 U/μl; Invitrogen) was added to each reaction and mRNA reverse transcribed at 37°C for 1 h. All reverse transcription (RT) reactions were then inactivated by incubating at 70°C for 15 min. cDNAs were stored at -20°C until PCR. PCR primers for teleost VDRα were designed using PrimerQuest (Integrated DNA Technologies, Coralville, IA, USA). PCR primers were flanked by restriction sites for incorporation and transfer between appropriate cloning and expression vectors. For each 25-μl PCR reaction, first-strand cDNAs were amplified using 2 μl (100-300 ng) first-strand cDNA, 9 μl RNase-free water, 0.75 μl 10 μM forward primer (0.3 μM), 0.75 μl 10 μM reverse primer (0.3 μM), and 12.5 μl 2X Advantage Taq PCR Master Mix (Clontech, Mountain View, CA, USA). PCR reaction conditions were: 95°C for 1.5 min followed by 35 cycles of 94°C for 15 s, 55°C for 30 s, and 72°C for 1 min. PCR products for each teleost VDRα sequence were cloned into the TA cloning vector pCR2.1 (Invitrogen, Carlsbad, CA) as per manufacturer's suggestions.
VDR's were subsequently excised from pCR 2.1 using EcoRI and BamHI and inserted unidirectionally into the expression vector pSG5. Proper orientation of VDRα's within the vector was confirmed by PCR screening and sequencing in both directions.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
HepG2 cells were cultured in T75 flasks with vented caps (Corning, Corning, NY) using MEM containing head-inactivated fetal bovine serum (10%), 1X sodium pyruvate, 1X nonessential amino acids, and 1% penicillin/streptomycin. Cells were maintained at 37°C with 5% CO2 and split regularly at 70-80% confluency. For transient transfection, HepG2 cells were seeded at a density of 1 × 105 cells per well in 24 well plates in antibiotic-free MEM. Wells were transfected over night with either empty pSG5 vector (control), medaka VDRα, zebrafish VDRα, or Tetraodon VDRα. Each well contained 299 ng pSG5-VDRα for either (medaka, zebrafish, or Tetraodon), 64 ng hCYP3A4-Luc reporter construct (XREM), and 15 ng of the Renilla normalizing plasmid (pRL-CMV).
Luciferase reporter assay experiments were performed as previously described [58].[](https://www.ncbi.nlm.nih.gov/mesh/D011773)
Plasmids containing human organic anion transporting polypeptide (Oatp1a1; Slco1a1), as well as the reporter construct tk-UAS-Luc and the 'empty' vector PM2, were generously provided by SA Kliewer, JT Moore, and LB Moore (GlaxoSmithKline, Research Triangle Park, NC, USA). To permit comparison between species and to avoid mismatching of non-mammalian receptors with mammalian retinoid X receptor, co-factors, and chromatin remodeling factors, all receptors were studied as LBD/GAL4 chimeras. For the GAL4/LBD expression constructs, the reporter plasmid is tk-UAS-Luc, which contains GAL4 DNA binding elements driving luciferase expression. The cloning of LBDs from hFXR, hVDR, mouse VDR (mVDR), and zebrafish VDR has been previously reported [6,7,9].
The LBD of chicken VDR was cloned from RNA extracted from the LMH cell line. The LBD of Xenopus laevis VDR was cloned from RNA extracted from the A6 cell line. The LBDs of tnFXR, tnVDR, and tnPXR were cloned from total RNA extracted from Tetraodon liver using sequence information from the draft Tetraodon nigriviridis genome [76]. The LBDs of chVDR (amino acid residues 113-451), xlVDR (amino acid residues 91-422), tnFXR (amino acid residues 180-463), tnVDR (amino acid residues 91-426), and tnPXR (amino acid residues 202-483) were inserted into the pM2-GAL4 vector to create GAL4/LBD chimeras.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Site-directed mutations of mVDR were performed using the QuikChange II mutagenesis kit (Stratagene, La Jolla, CA, USA). All mutations were confirmed by sequencing of both DNA strands. To explore the importance of the H1-H3 insert in ligand selectivity, four constructs were created. Thus, we generated hVDR/Δins that lacked amino acid residues 165-218 (hVDR/Δins) and a hVDR construct where the insert domain was replaced with the insertion domain from bile acid-insensitive xlVDR (hVDR/xlVDRins; hVDR residues 158-223 replaced by xlVDR residues 160-218). The analogous constructs were also created for xlVDR:
one lacking the insertion domain (xlVDR/Δins; missing amino acid residues 166-211) and the other (xlVDR/hVDRins; xlVDR residues 160-218 replaced by hVDR residues 158-223). The chimeras were generated by synthesis of double-stranded DNA (Genscript, Piscataway, NJ) which was inserted into the pM2-GAL4 vector.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## Cell lines
*In the **Cell lines** section:*
The creation of a HepG2 (human liver) cell line stably expressing the human Na+-taurocholate cotransporter (NTCP; SLC10A1) has been previously reported [7]. HepG2-NTCP cells were grown in modified Eagle's medium-α containing 10% fetal bovine serum and 1% penicillin/streptomycin. The cells were grown at 37°C in 5% CO2. The chicken LMH hepatoma cell line (ATCC, Manassus, VA, USA) was grown in Waymouth's MB 752/1 medium (ATCC) with 10 fetal bovine serum. The Xenopus laevis A6 kidney cell line (ATCC) was grown in 75% NCTC 109 medium, 15% distilled water, and 10% fetal bovine serum at 26°C in 2% CO2.
Except as noted above, all media and media supplements for the HepG2 and A6 cell lines were obtained from Invitrogen (Carlsbad, CA, USA). Co-transfections and transactivations assays were performed as previously described [77]. The maximal activators and their concentrations were as follows: chVDR, xlVDR, and tnVDR - 200 nM 1α,25-(OH)2-vitamin D3; tnFXR - 5 μM GW4064; tnPXR - 50 μM 5α-androstan-3α-ol. All comparisons to maximal activators were done within the same microplate. Luciferase data were normalized to the internal β-galactosidase control and represent means ± SD of the assays.[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Common features pharmacophore development and database screening
*In the **Common features pharmacophore development and database screening** section:*
Computational molecular modeling studies were carried out using Catalyst™in Discovery Studio 2.5.5 (Accelrys, San Diego, CA, USA) by methods previously described [29]. Common features pharmacophores were developed based on the compounds tested against tnFXR, tnPXR and tnVDR using the HipHop method [78,79]. A HipHop pharmacophore attempts to describe the arrangement of key features that are important for biological activity.
## Molecular modeling studies
*In the **Molecular modeling studies** section:*
A structural model of the LBDs of the tnFXR and tnPXR were constructed using the Molecular Operating Environment (MOE; Chemical Computating Group, Montreal, Canada) Homology Model module. The modeling template used for FXR is the published crystal structure of rat FXR in complex with 6α-ethyl-chenodeoxycholic acid (PDB ID = 1OSV) [59] (this structure was chosen because it has bound bile acid unlike available crystal structures of hFXR). The template for PXR is hPXR co-crystallized with SR12813 (PDB ID = 1nrl)
[33]. Several energy minimization-based refinement procedures were implemented on the initial model, and the quality of the final model was confirmed by the WHATIF-Check program. Estimation of the volume of the LBP for the crystallographic structures and homology models described above were determined using CASTp http://sts.bioengr.uic.edu/castp/calculation.php[80].[](https://www.ncbi.nlm.nih.gov/mesh/C464660)
Molecular docking studies were performed by the GOLD docking program [81]. In addition to the docking studies with tnPXR and tnFXR, LCA was docked into the crystal structure for human VDR (PDB accession number 1DB1). During the docking process, the protein was held fixed while full conformational flexibility was allowed for ligands. For each ligand, 30 independent docking runs were performed to achieve the consensus orientation in the LBP.[](https://www.ncbi.nlm.nih.gov/mesh/D008095)
## Authors' contributions
*In the **Authors' contributions** section:*
NA and SE performed molecular modelling studies, including those that provided the basis for the site-directed mutagenesis experiments. LRH performed the analysis of bile salts of the green-spotted pufferfish. EMK and SWK performed the cloning, expression, and analysis of full-length VDRs in three actinopterygian fish species. EJR and MDK cloned Tetraodon FXR, Tetraodon PXR, chicken VDR, Xenopus laevis VDR, and Tetraodon VDR, as well as all site-directed mutants of mouse VDR. MDK performed the function assays of all GAL4/LBD chimeric receptors and drafted the manuscript.
All authors contributed to, read, and approved the final manuscript.[](https://www.ncbi.nlm.nih.gov/mesh/D001647)
## Supplementary Material
*In the **Supplementary Material** section:*
## Acknowledgements
*In the **Acknowledgements** section:*
MDK is supported by K08-GM074238 from the National Institutes of Health. SWK is supported by the National Science Foundation NSF0842510. SE kindly acknowledges Accelrys for providing Discovery Studio for use in this work.
|
# Introduction
Assessment of Acetylcholinesterase Activity Using [Indoxylacetate](https://www.ncbi.nlm.nih.gov/mesh/C053640) and Comparison with the Standard Ellman’s Method
# Abstract
*In the **Abstract** section:*
Assay of acetylcholinesterase (AChE) activity plays an important role in diagnostic, detection of pesticides and nerve agents, in vitro characterization of toxins and drugs including potential treatments for Alzheimer’s disease. These experiments were done in order to determine whether indoxylacetate could be an adequate chromogenic reactant for AChE assay evaluation. Moreover, the results were com[pared to the s](https://www.ncbi.nlm.nih.gov/mesh/C053640)tandard Ellman’s method. We calculated Michaelis constant Km (2.06 × 10−4 mol/L for acetylthiocholine and 3.21 × 10−3 mol/L for indoxylacetate) maximum reaction velocity Vmax (4.97 × 10−7 kat for ac[etylcholine and 7](https://www.ncbi.nlm.nih.gov/mesh/D000122).71 × 10−8 kat for indoxyla[cetate) for el](https://www.ncbi.nlm.nih.gov/mesh/C053640)ectric eel AChE. In a second part, inhibition values w[ere plotted f](https://www.ncbi.nlm.nih.gov/mesh/D000109)or paraoxon, and reactiva[tion efficacy ](https://www.ncbi.nlm.nih.gov/mesh/C053640)was measured for some standard oxime reactivators: obidoxime, pralidoxime (2-PAM) and HI-6. Though indoxylacetate is split with lower turnover ra[te, t](https://www.ncbi.nlm.nih.gov/mesh/D010091)his compound ap[pears as ](https://www.ncbi.nlm.nih.gov/mesh/D009768)a [very attrac](https://www.ncbi.nlm.nih.gov/mesh/C028797)ti[ve re](https://www.ncbi.nlm.nih.gov/mesh/C028797)actant[ sin](https://www.ncbi.nlm.nih.gov/mesh/C022870)ce it doe[s not show any](https://www.ncbi.nlm.nih.gov/mesh/C053640) chemical reactivity with oxime antidots and thiol used for the Ellman’s method. Thus it can be advantageously used for accurate measurement[ of A](https://www.ncbi.nlm.nih.gov/mesh/D010091)ChE activity. [Suita](https://www.ncbi.nlm.nih.gov/mesh/D013438)bility of assay for butyrylcholinesterase activity assessment is also discussed.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Acetylcholinesterase (AChE; EC 3.1.1.7.) is an enzyme participating in cholinergic neurotransmission. It breaks down acetylcholine which terminates the neurotransmission process [1]. AChE activity is inhibited by many compounds. The number of known inhibitors is rather extensive.
Two main types of inhibitors can be distinguished from a practical point of view: toxins and drugs [2]. From a mechanistic point of view, the inhibitors are compounds with different structural motives as they can bind to the esteratic part of the active site by esterification of serine hydroxyl, or interact with the alpha anionic part of the active site, the aromatic gorge and the peripheral anionic site [3].[](https://www.ncbi.nlm.nih.gov/mesh/D000109)
Assay of AChE activity can serve for diagnosis after potential exposure to organophosphorus or carbamate pesticides and nerve agents [4,5]. It can also be used for verification of treatment effectiveness, e.g., for Alzheimer’s disease therapy [6]. Novel drugs for Alzheimer’s disease or antidotal therapy are tested by in vitro methods when AChE is implicated in the treatment process [7,8]. Assay of nerve agents and selected pesticides by devices with AChE is another application of this enzyme [9].
Experimental protocols for AChE activity assay have been proposed. Unfortunately, the mechanism of AChE activity assay had limitations that can preclude its use in some pharmacological or toxicological experiments. The most common assay is based on Ellman’s method using an alternative substrate acetylthiocholine and 5,5’-dithio-bis-2-nitrobenzoic acid (DTNB). The reaction results in production of 5-thio-2-nitrobenzoate that has yellow color due to the shift of electrons to the sulfur atom. The method was developed by Ellman and coworkers in the early 1960s
[10] and it is still used up to now, generally with significant modifications [11]. The Ellman’s method is particularly limited for testing antidots against organophosphorus AChE inhibitors or for measuring AChE activity in samples of such treated individuals. The antidots contain reactive oxime group splitting DTNB and provide false positive reaction in a process called oximolysis [12]. In this work we present experiments to determine AChE activity assay using indoxylacetate as an alternative substrate. We introduce a new alternative protocol to the Ellman’s method, which could be of high interest when DTNB may generate unwanted side reactions.[](https://www.ncbi.nlm.nih.gov/mesh/D009943)
## Results
*In the **Results** section:*
In the first part of experiments, activity of AChE was assessed for different concentrations of substrates. The concentrations ranged from 10−2 to 10−7 mol/L and the assays were repeated four times. Saturations curves were constructed for the calculated enzyme activities. They are depicted in Figures 1 and 2. Acetylthiocholine above a concentration of 10−4 mol/L inhibited AChE activity.
For toxicological and pharmacological testing, concentration of substrate of 10−4 mol/L was chosen as optimal. Indoxylacetate did not inhibit AChE up to the highest tested concentration, i.e., 5 × 10−3 mol/L. However, it has to be mentioned that indoxylacetate and indigo are not highly soluble compounds. Octanol water partition coefficient KOW was calculated at 1.65 for indoxylacetate and 3.11 for indigo using EPI Suit software. The same software was used in order to estimate water solubility, 943 mg/L (5.39 mmol/L) for indoxylacetate and 109 mg/L (0.42 mmol/L) for indigo.[](https://www.ncbi.nlm.nih.gov/mesh/D000122)
Biochemical parameters were calculated in the second step using non-linear regression. The best fitting was found for no cooperativity model (n = 1). The calculated Km and Vmax are presented in Table 1. Km for acetylthiocholine and indoxylacetate were (2.06 ± 0.35) × 10−4 and (3.21 ± 0.31) × 10−3 mol/L, respectively.
The Vmax value was equal to (4.97 ± 0.42) × 10−7 kat for acetylthiocholine and (7.71 ± 0.56) × 10−8 kat for indoxylacetate.[](https://www.ncbi.nlm.nih.gov/mesh/D000122)
We also investigated potency of butyrycholinesterase (BuChE) to split indoxylacetate and substitute AChE in this way. We prepared a standard solution of BuChE and AChE with activity 5 × 10−7 kat in 25 μL. The activity of BuChE was assessed using butyrylthiocholine as a substrate and the protocol described in Section 4.2 for substrate level 1 mM. Activity of AChE and BuChE using 1 mM indoxylacetate were (2.96 ± 0.53) × 10−8 and (7.86 ± 0.65) × 10−8 kat, respectively.
Ratio of affinity AChE/BuChE toward indoxylacetate was calculated to be 2.65.[](https://www.ncbi.nlm.nih.gov/mesh/C053640)
Ethyl-paraoxon organophosphate was chosen as a model inhibitor. The final concentration of paraoxon in one cuvette ranged from 10−9 to 10−2 mol/L. The calibration curves (Figures 3, 4) were constructed for acetylthiocholine as well as indoxylacetate. The achieved enzyme activity and its decrease were re-calculated to percent of inhibition (I) due to better comparison of both methods. The limit of detection was calculated as signal to noise (control) ratio equal to three.
The reached limit of detection for paraoxon was 10−7 mol/L for both methods.[](https://www.ncbi.nlm.nih.gov/mesh/C121104)
2-PAM, obidoxime and HI-6 were tested as standard oxime reactivators. AChE was inhibited by ethyl-paraoxon up to 95% which means that 5% activity of AChE remained. The found return of activity was calculated as percent of reactivation i.e., percent of activity from the original AChE activity before inhibition. Oxime reactivators were evaluated in the two final concentrations in cuvette at 10−4 and 10−5 mol/L. The calculated reactivation efficacies are summarized in Table 2.
Comparing the reactivation efficacies, the achieved percent of reactivation were not significantly different on probability level 0.01 < P ≤ 0.05 or P ≤ 0.01 for both substrates.[](https://www.ncbi.nlm.nih.gov/mesh/C028797)
The spontaneous interaction between indoxylacetate and oxime reactivators was examined with the negative result. There was no significant increase of absorbance due to oxime reactivator, whereas the Ellman’s method was quite sensitive to interference by oxime reactivators. The tested drugs provided false positive results because they spontaneously reacted with DTNB providing yellow colored 5-thio-2-nitro benzoic acid. The false positive reaction was approximately equal to the absorbance shift provided by 5 × 10−7 kat of AChE.[](https://www.ncbi.nlm.nih.gov/mesh/C053640)
Potential interferences of biological matrices were tested using fresh rat blood. Blood was lysed using deionized water (blood:water—1:4). After spinning at 3,000 × g, absorbance at 670 nm was measured in supernatant. The found absorbance was >2.0. The measurement was repeated using supernatant diluted five-times (absorbance ∼0.50), ten-times (absorbance ∼0.25), twenty-times (absorbance ∼0.10), and forty-times (absorbance ∼0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Discussion
*In the **Discussion** section:*
Basic enzyme constants were calculated using both methods: Michaelis constant Km and maximum reaction velocity Vmax. This experiment was mandatory in order to assess the functionality of the method. The present literature contains significant variations for the evaluation of AChE enzymatic parameters. The reported parameters differ notably due to different methods of AChE isolation, AChE origins, and physico-chemical conditions of storage, as mentioned e.g., by Barteri et al. for electric eel AChE [13]. The Michaelis constant measured in this work is approximately ten-times lower than the Km reported for (Rattus novergicus) brain AChE: 2.65 mmol/L
[14] but it is in a very good agreement with the one found for electric eel AChE: 0.13–0.15 mmol/L [15] for the substrate acetylthiocholine iodide. Higher Km values are also known for common housefly (Musca domestica) AChE: 1.80 mmol/L [16]. Lower value than for rats and housefly are reported for human AChE: 0.98
[17]. The above mentioned Michaelis constants were achieved for acetylthiocholine as substrate. Unfortunately, the measured Michaelis constant using indoxylacetate cannot be compared with the literature as it is not available in the enzyme databases (Brenda) and papers. The fact that AChE is not inhibited in excess of indoxylacetate in contrast to acetylthiocholine is a novel very interesting output. Indoxylacetate was considered as a suitable chromogen and fluorogen in cholinesterase based assays
[18,19]; however, the biochemical parameters have not yet been assessed. Tests aimed at comparison of AChE and BuChE sensitivity toward indoxylacetate as a substrate proved suitability of indoxylacetate as a universal substrate for cholinesterases. BuChE have even nearly three times higher affinity to indoxylacetate when compared to AChE. In regard to this fact, selective inhibitor of BuChE, tetraisopropyl pyrophosphoramide (iso-OMPA), shall be added into cuvette when only AChE is demanded to be assayed and BuChE interference could occur. Thus organized experiment will not be affected by BuChE due to iso-OMPA action and it is reliable for e.g., AChE assay in blood samples [20].
Biochemistry of BuChE was not a primary objective in this experiment; however, work devoted to the issue would be perspective in the future.[](https://www.ncbi.nlm.nih.gov/mesh/C543539)
The comparison we performed with both indoxylacetate and acetylcholine as substrates shows similar calibration plots for paraoxon assay. The fact that evaluation of standard oxime reactivators shows the same reactivation potency is a very promising result. It raises the suitability of indoxylacetate as a new substrate for in vitro preliminary characterization of novel drugs. Moreover, contrary to the Ellman’s method reagents, the indoxylacetate does not directly react with oxime reactivators like DTNB [12,21]. In addition, indoxylacetate does not interact with thiols like DTNB with e.g., reduced glutathione [22].[](https://www.ncbi.nlm.nih.gov/mesh/C053640)
However, it has to be mentioned that beside significant advantages, indoxylacetate has some limitations that should be taken in consideration. First, the calculated maximum reaction velocity is nearly ten-times lower for indoxylacetate. The lower turnover rate has to be compensated by prolonging assay time or increasing AChE amount in suspension. The poor solubility in water of both indoxylacetate and indigo must also be pointed out. Indoxylacetate and indigo have rather high octanol water partition coefficient.[](https://www.ncbi.nlm.nih.gov/mesh/C053640)
The experimental data reported herein demonstrate that indoxylacetate can be recommended as a suitable substrate for toxicological or pharmacological characterization of new compounds implicated in AChE activity modulation. It could be a valuable alternative to the assay protocols based on Ellman’s method [23–25]. Pharmacological testing of anticholinersterase acting compounds is frequently based on blood AChE assay [13]. As seen in the results section, blood interference at 670 nm is quite low.
Even five-times diluted blood lysate had low interference at 670 nm to be spectrophotometrically assayed. This is unlike the Ellman method where blood interference is a serious problem [5,13]. Suitability of indoxylacetate for blood cholinesterases assay can be inferred. Moreover, the assay is transmittable into 96-wells microplates allowing the lowering of costs per assay.[](https://www.ncbi.nlm.nih.gov/mesh/C053640)
## Experimental Section
*In the **Experimental Section** section:*
## Chemicals
*In the **Chemicals** section:*
AChE (electric eel, Electrophorus electricus, origin; 3.33–16.7 μkat/mg i.e., μmol/s × mg), acetylthiocholine chloride, DTNB, ethyl-paraoxon, indoxylacetate and phosphate buffered saline (PBS) in tablets were purchased from Sigma-Aldrich. Organic solvents were of analytical grade. Sorbents and alkalines mixed with ethanol were used for paraoxon containing mixtures and disposable tools decontamination in a standard protocol. Oxime drugs including pralidoxime chloride (2-PAM; [(E)-(1-methylpyridin-2-ylidene)methyl]-oxoazanium chloride), obidoxime chloride (1,1’-[oxybis(methylene)]bis{4[(E)-(hydroxyimino)methyl]pyridinium}dichloride) and asoxime chloride (HI-6; [(Z)-[1-[(4-carbamoylpyridin-1-ium-1-yl)methoxymethyl]pyridin-2-ylidene]methyl]-oxoazanium dichloride) were previously synthesized at the Department of Toxicology, Faculty of Military Health Sciences, University of Defence, Hradec Kralove, Czech Republic. Purity was confirmed by thin layer chromatography.[](https://www.ncbi.nlm.nih.gov/mesh/D000122)
## Ellman’s Method for AChE
*In the **Ellman’s Method for AChE** section:*
Ellman’s method was done with slight modifications of the reference [26,27]. The chemical principle is depicted in Figure 5. A disposable cuvette was consequently filled with 0.4 mL of 0.4 mg/mL DTNB, 25 μL of AChE solution (0.5 μkat in 1 mM acetylthiocholine), 425 μL of PBS, 50 μL of paraoxon in isopropanol or isopropanol alone. The reaction was started by adding 100 μL of acetylthiocholine chloride in a given concentration for assessment of Km and Vmax or 1 mM for toxicological and pharmacological investigations. Absorbance at 412 nm was measured immediately and after one minute.
Enzyme activity was calculated estimating extinction coefficient ɛ = 14,150 M−1cm−1. The oxime drugs were tested in a similar protocol. 425 μL of PBS was reduced to 325 μL of PBS. Paraoxon was added in concentration providing 95% inhibition of AChE. Incubation time was set to 10 minutes.
After that, 100 μL of oxime reactivator suspended in PBS was injected into the cuvette and kept for another 10 minutes. The reaction was started again by addition of acetylthiocholine.[](https://www.ncbi.nlm.nih.gov/mesh/D004228)
## AChE Activity Assay Using Indoxylacetate
*In the **AChE Activity Assay Using Indoxylacetate** section:*
The reaction principle is depicted in Figure 6. The experiment was organized as a common spectrophotometric test using disposable cuvettes. One milliliter volume plastic cuvette was filled with 825 μL of PBS, 50 μL of paraoxon in isopropanol or isopropanol and 25 μL of AChE solution (the same activity as above) and gently shaken. The reaction was started by addition of indoxylacetate in 5% ethanol. This concentration was selected as a compromise between AChE inhibition and enabling of indoxylacetate solubility [9].
Absorbance of the mixture in the cuvette was measured at 670 nm shortly after mixture shaking and then after 30 minutes. Activity was calculated using extinction coefficient for which the value ɛ = 3,900 M−1cm−1 was adopted from literature [14]. Slight modification of the aforementioned protocol was made for oxime reactivators testing. 725 μL of PBS were placed into the cuvette instead of 825 μL. Paraoxon was injected in an amount providing 95% inhibition of AChE and the mixture was left for 10 minutes.
Reactivation was triggered by addition of 100 μL of tested oxime reactivator and colorimetric reaction was started by addition of indoxylacetate 10 minutes after the reactivator. All experiments were performed under standard laboratory conditions (SATP).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Partition Coefficient and Water Solubility Calculation
*In the **Partition Coefficient and Water Solubility Calculation** section:*
EPI Suite (Office of Pollution Toxics and Syracuse Research Corporation; US Environmental Protection Agency) software was used in order to calculate octanol water partition coefficient and water solubility.[](https://www.ncbi.nlm.nih.gov/mesh/D000442)
## Statistics and Data Processing
*In the **Statistics and Data Processing** section:*
Michaelis constant Km and maximum reaction velocity Vmax were calculated using non-linear regression and software Origin 8 SR2 (OriginLab Corporation, Northampton, MA, USA). The Hill equation was used throughout for variable cooperativity and non-cooperative model (n = 1). Origin 8 software was also used for data processing and inferential statistics. Significance of results was evaluated by one-way ANOVA with Tukey test. Both 0.01 < P ≤ 0.05 and P ≤
0.01 probability levels were calculated.
## Conclusions
*In the **Conclusions** section:*
Basic enzyme constants (Michaelis constant and maximum reaction velocity) were calculated for electric eel AChE and two substrates: acetylthiocholine and indoxylacetate. Performance of indoxylacetate in pharmacology and toxicology was evaluated, and the suitability to use indoxylacetate for cholinesterase based tests was demonstrated on ethylparaoxon as standard pesticide and 2-PAM, obidoxime and HI-6 as standard oxime reactivators drugs. We believe that this work could be useful in the development of alternative techniques to the Ellman’s method and that indoxylacetate could be of great interest in some new protocols.[](https://www.ncbi.nlm.nih.gov/mesh/D000122)
# References
*In the **References** section:*
Saturation curve for acetylcholinesterase (AChE) and indoxylacetate as a substrate. The plot was fitted by Hill equation. Error bars indicate standard deviation for n = 4.[](https://www.ncbi.nlm.nih.gov/mesh/C053640)
Saturation curve for AChE and acetylthiocholine chloride as a substrate. Semilagirmic is presented on the right. Plot on the left was fitted by the Hill equation. Error bars indicate standard deviation for n = 4.[](https://www.ncbi.nlm.nih.gov/mesh/D000122)
Calibration for ethyl paraoxon (logarithm of molar level) using AChE based assay with indoxylacetate as substrate. The absorbance shift is recalculated to percent of inhibition (I). The point in brackets was achieved by application of phosphate buffered saline instead of paraoxon. The error bars indicate standard deviation for four experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C121104)
Calibration for ethyl paraoxon using AChE based assay with acetylthiocholine as substrate. The description is the same as for Figure 5.[](https://www.ncbi.nlm.nih.gov/mesh/C121104)
Chemical mechanism of Ellman’s method.
Chemical mechanism of indoxylacetate performance as AChE chromogenic substrate.[](https://www.ncbi.nlm.nih.gov/mesh/C053640)
Biochemical parameters of electric eel acetylcholinesterase (AChE) calculated using non-linear regression analysis.
Km—Michaelis constant; Vmax—maximum reaction velocity.
Reactivation of ethyl-paraoxon inhibited AChE. The values indicate mean percent of reactivation for four experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C121104)
2-PAM—pralidoxime; HI-6—asoxime.[](https://www.ncbi.nlm.nih.gov/mesh/C028797)
|
# Introduction
The Chemical Composition and [Nitrogen](https://www.ncbi.nlm.nih.gov/mesh/D009584) Distribution of Chinese Yak (Maiwa) Milk
# Abstract
*In the **Abstract** section:*
The paper surveyed the chemical composition and nitrogen distribution of Maiwa yak milk, and compared the results with referenc[e compos](https://www.ncbi.nlm.nih.gov/mesh/D009584)ition of cow milk. Compared to cow milk, yak milk was richer in protein (especially whey protein), essential amino acids, fat, lactose and minerals (except phosphorus). The conte[nts of some nutrients](https://www.ncbi.nlm.nih.gov/mesh/D000601) (total[ protei](https://www.ncbi.nlm.nih.gov/mesh/D007785)n, lactose, essential [amino acid](https://www.ncbi.nlm.nih.gov/mesh/D010758)s and casein) were higher in the warm season than [in the ](https://www.ncbi.nlm.nih.gov/mesh/D007785)co[ld season. Higher rat](https://www.ncbi.nlm.nih.gov/mesh/D000601)ios of total essential amino acids/total amino acids (TEAA/TAA) and total essential amino ac[ids/total non essenti](https://www.ncbi.nlm.nih.gov/mesh/D000601)al amin[o acids (TE](https://www.ncbi.nlm.nih.gov/mesh/D000596)AA/TNEAA) were found i[n the yak milk from t](https://www.ncbi.nlm.nih.gov/mesh/D000601)he warm sea[son. However its annu](https://www.ncbi.nlm.nih.gov/mesh/D000601)al average ratio of EAA/TAA and that of EAA/NEAA were similar to those of cow milk. Yak milk was rich in calcium and iron (p < 0.05), and thus may serve as a nutritional ingredient wit[h a pot](https://www.ncbi.nlm.nih.gov/mesh/D002118)entia[l ap](https://www.ncbi.nlm.nih.gov/mesh/D007501)plication in industrial processing.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Yaks are found extensively on the plateau of the western Tibetan region of China in alpine and subalpine areas at altitudes from 2000–5000 m with a cold, semi-humid climate [1]. Domestic yak milk and yak meat is a vital part of the local economy in the Tibetan region of China. Yak milk is called natural concentrated milk because of its high fat (5.5–7.5%), protein (4.0–5.9%) and lactose (4.0–5.9%) content during the main lactating period [2,3]. Yak milk and milk products are the major ingredients of the daily diet of Tibetan herders, particularly for the weak, ill, elderly and young in the areas where yaks graze on the alpine meadows and mountain pastures.
Due to a shortage of fruit and vegetables, and limited food resources, yak milk and milk products (butter and cheese) are a vital source of vitamins and major sources of nutrition for Tibetan Herders [4]. However, large-scale industrialization of yak milk production is limited by low volume and seasonal cyclicity of individual milk production, around 150–500 kg of fresh milk per lactation, which depends on the breed, age, parity and body condition of the yak, pasture growth pasture quality, raising areas, milking time, milking methods, and other environmental factors [1,4].[](https://www.ncbi.nlm.nih.gov/mesh/D007785)
In recent years, production of yak milk has seen an increase to 40 million tons each year. However, industrially processed dairy products were no more than 25%, while the remainder was produced and consumed by traditional means [5]. Yak milk and milk products are gaining in popularity due to their special nutritional value. In China, Several dairy companies are being established to supply fresh pasteurized and UHT milk to consumers. Considering its economic potential, information on the composition as well as the chemical properties of yak milk is essential for the successful development of a yak product industry and for marketing.
There are differences in chemical characteristics between yak and cow milk. Yak milk is predominantly produced by seasonal breeding. Its composition varies with seasonal grass growth and climate change, as does milk production. The highest content of the composition, such as solids, lactose, protein and amino acids, are in the mid-lactation period but the fat content increases continuously into late lactation [6,7].
On the other hand, the bulk milk from cow herds varies little with the seasons because of year-round breeding, therefore the composition of cow’s milk shows minimal changes throughout the year [8].[](https://www.ncbi.nlm.nih.gov/mesh/D007785)
In China, According to the (Chinese) provincial annals of livestock breeds, there are 12 officially recognized breeds of domestic yak in China: the Jiulong yak and Maiwa yak in Sichuan province, Tianzhu White yak and Gannan yak in Gansu province, Pali yak, Jiali (“Alpine”) yak and Sibu yak in Tibet, Huanhu yak and Plateau yak in Qinghai province, Bazhou yak in Xinjiang and Zhongdian yak in Yunnan province, and one other, the “Long-hair-forehead yak” in Qinghai province. These 12 yak breeds belong to two main types, the Qinghai-Tibet Plateau type ('Plateau or Grassland type) and the Henduan Alpine type (Alpine or Valley type) [1]. Maiwa yak belongs to the Qinghai-Tibet Plateau type, numbering around 5.4 million animals, and ranking second in China’s yak population
[5]. Yak and yak milk have come to occupy a dominant position in the local economy.
Cow milk is a complex colloidal dispersion containing fat globules, casein micelles and whey proteins in an aqueous solution of lactose, minerals and a few other minor compounds. Its chemical properties depend on intrinsic compositional and structural factors, and on extrinsic factors such as temperature and post–milking treatments. Although the composition of yak milk has been studied in China [2,3,6,7,9], there is little information on the chemical composition, nitrogen distribution and mineral contents of yak milk produced from Maiwa breeds in the Qinghai-Tibet Plateau of China. The purpose of this paper is to investigate the chemical properties of yak milk from the Maiwa breed and to compare variation of these parameters in cold and warm season.
The results were compared with those obtained from cow’s milk. An understanding of these properties is important in the elucidation of complex chemical reactions that occur in yak milk and in the technological and engineering design and operation of the milk processes and processing equipment.[](https://www.ncbi.nlm.nih.gov/mesh/D007785)
## Results and Discussion
*In the **Results and Discussion** section:*
## Basic Chemical Composition of Yak Milk
*In the **Basic Chemical Composition of Yak Milk** section:*
The basic composition (fat, protein, lactose, ash and total solids) of yak milk is shown in Table 1. Except for ash content, the average content of fat, protein, lactose and total solid of the yak milk is higher than cow milk from the USA and the surrounding area [3,10]. The fat content of the yak milk was higher in the cold season than in the warm season, while the contents of total protein and lactose were higher in the warm season (p < 0.05). In general, the current data is similar to the proximate composition of other yak breeds published so far [1,11,12]. Because yak are allowed to graze under uncontrolled environmental conditions, the milk composition varies with seasonal grass growth and climate changes [11,12].
On the Qinghai-Tibet Plateau, the average annual air temperature is generally bellow 0 °C, while the average temperature in January drops below −10 °C. The average temperature in the hottest month (July) does not exceed 13 °C. During the warm season, the plentiful green grass is enough to feed yaks. The cold pastures on which the yak were grazed have predominantly short grass and rough grazing conditions, with sedges and shrubby plants [1]. With the growth of forage (from August to December), protein content in the swards declines from 115 g per kg DM (young grass) to 33 g per kg DM (mature grass), and crude fiber in the swards increases correspondingly [1,12].
The increased crude fiber can offer more acetic acid and butyric acid (the sources of fatty acids) for the mammary gland to synthesize more fat which may be the reason for the higher fat content of yak milk in the cold season [1]. The fat content in ovine milk showed similar seasonal changes from February to August (7.58% to 6.59%) [13].[](https://www.ncbi.nlm.nih.gov/mesh/D007785)
The protein content of Maiwa yak was far higher than that of other dairy breeds, but similar to that of buffalo milk [14]. The protein content of Maiwa yak was similar to that of yak breeds of Jiulong (4.9%), Tianzhu white (5.2%) and Jiali (5.0%), but lower than that of the yak breeds, like Pali (5.7%), Kyrgyzstan (5.3%), India (5.9%) and Nepa (5.4%) [1]. Sheng et al. [9] surveyed the protein content of Maiwa yak milk which was collected in October from the seven mid-lactating yaks.
The protein content (3.51%) seemed to be low compared with the average value (4.95%) of the milk from whole year. In this study, the average protein content of the milk was within the range of reported data (4.0–5.5%) [1].
The lactose content of Maiwa yak milk was 5.03% higher than that of cow milk. Higher contents of lactose in the yak milk will be beneficial to infants. Lactose in the distal bowel can help combat gastrointestinal disturbances resulting from undesirable putrefactive bacteria through promoting the growth of certain beneficial lactic-acid-producing bacteria [15].[](https://www.ncbi.nlm.nih.gov/mesh/D007785)
The ash content of the yak milk (0.79%) was similar to that of cow milk. In general, ash content of yak milk is around 0.7–0.9% during main lactating period. Furthermore, Weiner et al. [1] and Jiang et al. [11] analyzed the ash contents of Maiwa yak milk from June to September, and the ash content (0.82 ± 0.06%) did not show significant changes with season.
## Nitrogen Distribution in Yak Milk
*In the **Nitrogen Distribution in Yak Milk** section:*
The N-containing portions of milk can be divided into three broad fractions, including casein nitrogen (CN), whey protein nitrogen (WPN), and non protein nitrogen (NPN). Variations in the differences in nitrogen distribution of yak milk proteins between warm and cold seasons are shown in Table 2. Total nitrogen (TN) of yak milk from the warm season was higher than that from the cold season (p < 0.05). The protein contents may be affected by several factors, including breed, environmental temperature, diseases, and stage of lactation, parity and nutrition [16].
In general, high environmental temperature reduces the total protein content of milk [17], so the protein content of cow milk was higher during winters than during summers [16]. However, several papers have reported that energy intake (except fats and oils) may have a positive effect on milk protein content [18–20]. In the Qinghai-Tibet Plateau, grass and herbs cannot survive in a very cold winter. For natural grazing yaks, the malnutrition of feeds resulted in lower TN in winter.
Similarly, Walley et al. [21] demonstrated that the total protein content of milk was lowest when the energy intake of the cow was retarded.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
The NPN contents of the yak milk were 0.05% and 0.04% respectively in warm and cold seasons. The NPN constituted 5.56% and 5.30% of the total N in the yak milk, which did not vary much between seasons; and that the NPN proportion was consistent with that of cow’s milk [16]. The reason of above results is that the NPN content of milk is less variable among breeds, and the changes of milk NPN content with environmental temperature are similar in pattern to changes in protein content [16].[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
The contents of the whey protein nitrogen (WPN) did not change significantly between cold and warm seasons. The average WPN/TN of the yak milk was 19.63% which was higher than that of cow milk [16], but it was similar to that of sheep milk [22]. The average value of α-lactalbumin and β-lactoglobulin as a percentage of whey protein was upwards of 66% which was lower than that of cow whey protein [23]. The value found for α-lactalbumin is puzzling because it is not in line with the usual relationship between lactose and α-lactalbumin contents in mammalian milks.
Indeed, α-lactalbumin is half of the lactose synthetase enzyme and its content increases with lactose concentration [24]. In the collected samples of yak milk, the lactose concentrations were higher than that of bovine milk. However, the study on yak milk proteins has highlighted the lack of knowledge in the literature; deep further study is required.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
Although the level of CN was different between the warm and cold seasons (p < 0.05), the ratio of CN/TN was not significantly different. The highest level of CN was found in the warm season, which was closely consistent with the changes of TN [23,25]. For yak milk, the variation patterns of CN and CN/TN were consistent with those found in North American commingled goat milk [25]. The TN and CN showed significant variability between warm and cold seasons, while the ratios of nitrogen distribution, such as NPN/TN, WPN/TN, WPN/CN and CN/TN, remained constant throughout the seasons.
Therefore, this pattern of nitrogen distribution might be useful in deciding the end use for yak milk.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
The analysis results of amino acids, such as the contents of amino acids, total essential amino acids (TEAA), total non essential amino acids (TNEAA), and total amino acids (TAA), are given in Table 3. The results showed that the contents of some amino acids were unaffected by seasons. In the warm season, the TEAA was higher than that in the cold season (p < 0.05) which was consistent with the changes of the total protein content (Table 1); however, the TNEAA was not significantly different in the warm and cold seasons. In the warm season, the ratios of TEAA/TNEAA and TEAA/TAA were significantly higher than those in cold seasons, while the yearly average ratios of TEAA/TNEAA and TEAA/TAA were similar to those of cow milk [26]. The ratio of TEAA/TNEAA was attributed to the protein ratio of herbage grazed by the yaks [27].
Because the yaks were allowed to graze naturally at an average elevation of 3600 m on the Qinghai-Tibet Plateau, the important factors affecting milk quality are: pasture production and the quantity, growth status and nutritive value of the herbage. This means that all lactating yaks, irrespective of age, parity or breed type, or even location, tended to peak in yield in the summer season (June to August) when grass was at its best quality and quantity, while after August, as air temperature fell, the nutritive value declined [1]. Therefore, the TN, CN and TEAA of the yak milk were lower in the cold season.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## Mineral Contents of Yak Milk
*In the **Mineral Contents of Yak Milk** section:*
The major mineral contents of the yak milk from cold and warm seasons are given in Table 4. Yak milk and cow milk had similar ash content, around 0.8%. The major mineral contents of yak milk were much higher than those of cow milk while the content of phosphorous was in the range of cow’s milk [28]. The contents of these minerals did not show significant differences between warm and cold seasons, which kept the same trend as ash content. The average calcium content of the yak milk is 1545.45 mg/kg, while human milk has only one-fifth the amount of this mineral [22].
Dietary iron is required for a wide variety of biochemical processes. Human milk, as well as bovine milk and milk products, are poor sources of iron. To prevent iron deficiency and anemia in infants of 6–9 months, most infant formulas are supplemented with iron [29]. So the higher iron content (0.57 mg/kg) of yak milk can be a benefit in its nutritional value in infant foods. In general, the mineral content of yak milk seems to vary much more than that of cow milk due to the monthly differences in feeding.
The trace minerals in yak milk have not been extensively studied, even though they may be of considerable nutritional and health interest to humans.[](https://www.ncbi.nlm.nih.gov/mesh/D010758)
## Experimental Section
*In the **Experimental Section** section:*
## Collection of Milk Samples
*In the **Collection of Milk Samples** section:*
Sichuan province is the largest yak-raising province in China, there is more than four million yak and yak hybrids in the western and northern parts of Sichuan. Yak, in western parts of Sichuan, are found in all counties in Ganzi Tibetan autonomous prefecture and in Aba Tibetan and Qiang autonomous prefecture in the northeastern end of western Sichuan and most of counties in Liangshan Yi autonomous prefecture in the southern part of western Sichuan. Yak contributes 70 percent of the total milk production (180,000 tons annually) in Sichuan, much of it from Ganzi (Jiulong breed of yak) and Aba (Maiwa breed of yak). Numbers of yak have increased more in Aba than in Ganzi, most likely because of a better access to markets for yak products in Aba. Therefore the Maiwa breed of yak was selected for study.
One hundred and four pure Maiwa milk samples, in west part of Sichuan province, were collected monthly from Hongyuan County (31°51’–33°19’ N, 101°51’–103°23’ E) in May to December of 2009. All of the yak milk samples were collected from a family pasture. All of the yaks under study were grazing on natural pasture and did not have any supplementary feeding. About 20 percent of herds were less than 5 years old, with 2 parities. In the warm season (from May to September), yak was grazed in summer-autumn pasture where the altitude is about 4000 m, and in the cold season (from end of October to next April), yak was grazed in winter-spring pasture where the altitude is about 3200 m. 56 yak milk samples were collected in the warm season and 48 yak milk samples were collected in the cold season.
The fresh milk samples were collected in sterile plastic bottles, and immediately were transported to the nearest town by car. The milk samples were frozen in a −18 °C refrigerator. The frozen milk samples were transported to the laboratory by airplane. In the laboratory, the frozen milk samples were thawed and their chemical parameters were analyzed. The extra milk samples were stored at −20 °C.
The entire time from milking to beginning analysis was about 60 h.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
## The Determination of Basic Compositions
*In the **The Determination of Basic Compositions** section:*
The frozen yak milk samples were thawed to 38 ± 1 °C. The samples were gently thoroughly mixed by repeatedly inverting the sample bottle without causing frothing or churning. They were cooled to room temperature immediately before the analyses. The contents of total nitrogen (TN) and nonprotein nitrogen (NPN) were tested by Kjeldahl method according to the IDF 020-1 [30] and IDF 020-4 [31]. Non casein nitrogen (NCN) was tested according to describe by Wehr et al.
[32]. Casein was precipitated by acetic acid, NCN in the filtrate was detected by the Kjeldahl method. The nitrogen fraction were calculated as follows: protein nitrogen (PN) = TN-NPN, casein nitrogen (CN) = TN-NCN, and whey protein nitrogen = NCN-NPN. A nitrogen conversion factor of 6.38 was used to calculate protein contents of milk samples and various fractions. The fat content was tested with gravimetric method according to IDF 001D
[33]. Lactose was tested according to IDF 028A [34]. Ash content was tested after mineralisation of milk at 550 °C for 4 h according to IDF 027 [35]. The total solid (TS) was tested by drying 5 grams of milk sample at 100 ± 2 °C for 5 h in porcelain crucibles according to IDF 021B
[36]. Beta-lactoglobulin and alfa-lactalbumin were measured according to Bordin et al. (2001) [37]. Standard bovine whey proteins were purchased from Sigma. 10 mg α-La (lot L-5385 type I, ∼85%), 15 mg β-LgB (lot L-8005) and 10 mg β-LgA (lot L-7880) were dissolved in 5ml buffer solution (8M urea, 165 mM Tris, 44 mM sodium citrate and 0.3% (v/v) β-mercaptoethanol).
1 mL of skimmed yak milk was dissolved in 4 mL of buffer solution. The diluted samples were filtered through a 0.45 μm cellulose membrane and directly analyzed.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
## The Analysis of Amino Acids
*In the **The Analysis of Amino Acids** section:*
1 mL of milk samples were hydrolyzed with 6.0 M HCl in vacuum-sealed tubes at 110 °C for 22 h. The amino acids content was tested using an amino acid analyzer (L-8800, Hitachi, Japan). Amino acids were separated on a single-column ion-exchange chromatograph (Hitachi, 2622#, 4.6 mm × 60 mm) and were post-column derived with ninhydrin. Derivative amino acids were analyzed by spectrophotometer (570 nm and 440 nm) and quantified by comparing the area under the sample peak against that of an amino acid standard solution (Hitachi, Japan) of known content. The detailed procedure followed which was described by Zhang et al.
[38]. Tryptophan was lost during hydrolysis; therefore, tryptophan values are not reported. The results were given as means of triplicate analyses.[](https://www.ncbi.nlm.nih.gov/mesh/D006851)
## Mineral Assay of Samples
*In the **Mineral Assay of Samples** section:*
The contents of calcium, magnesium, copper, iron, manganese and zinc were analyzed by an Atomic Absorption Spectrometer (Atomic Spectormetry Analyst 800, Perkin Elmer, Vernon Hills, IL, USA) according to the standard method of GB 5413.21 [39]. The P content was tested by Spectrophotometer (WFZ754, Shanghai, China) according to IDF042 [40]. Each sample was analyzed in triplicate.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Statistical Analysis
*In the **Statistical Analysis** section:*
Data were analyzed by ANOVA method using SPSS software, version 10.0 (SPSS, Inc., Chicago, IL, USA). The differences among the means of the analysis data were compared at a significance level of p < 0.05.
## Conclusions
*In the **Conclusions** section:*
This work showed that the composition of most of the analyzed yak milk was affected by seasons, and differed in some aspects from cow’s milk. The yak milk has higher contents of fat, protein, lactose, and total solids than cow milk, and its contents changed with the seasons. The yak milk protein contained a higher ratio of WPN/TN. It is noteworthy that calcium and iron were richer in yak milk than in cow milk. These differences suggested that yak milk can serve as nutritional ingredients with the potential for industrial processing.
On the other hand, it will be interesting to further test yak milk properties, like acidification and heat stability. The knowledge will benefit the milk processing industry.[](https://www.ncbi.nlm.nih.gov/mesh/D007785)
# References
*In the **References** section:*
The basic chemical compositions of yak milk (g/100 g fresh milk).
Reference [10];
both a and b occurring in the same line means that the results are significantly different (p < 0.05); n: the number of samples analyzed and computed in the statistical analysis.
Concentration (%) of the nitrogen containing fractions in yak milk.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
Reference [10];
both a and b occurring in the same line means that the resultsare significantly different (p < 0.05); n: the number of samples analyzed and computed in the statistical analysis.
Amino acid contents of yak milk (g/100 g).[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Reference [26];
both a and b occurring in the same line means that the results are significantly different (p < 0.05); n: the number of samples analyzed and computed in the statistical analysis.
The mineral contents in yak milk (mg/kg).
Reference [28];
both a and b occurring in the same line means that the results are significantly different (p < 0.05).
|
# Introduction
Green Tea [Catechins](https://www.ncbi.nlm.nih.gov/mesh/D002392) Reduce Invasive Potential of Human Melanoma Cells by Targeting COX-2, PGE2 Receptors and Epithelial-to-Mesenchymal Transition
# Abstract
*In the **Abstract** section:*
Melanoma is the most serious type of skin disease and a leading cause of death from skin disease due to its highly metastatic ability. To develop more effective chemopreventive agents for the prevention of melanoma, we have determined the effect of green tea catechins on the invasive potential of human melanoma cells and the molecular mechanisms underlying these effects using A375 (BRAF-mutated) and Hs[294t (Non](https://www.ncbi.nlm.nih.gov/mesh/D002392)-BRAF-mutated) melanoma cell lines as an in vitro model. Employing cell invasion assays, we found that the inhibitory effects of green tea catechins on the cell migration were in the order of (-)-epigallocatechin-3-gallate (EGCG)>(-)-epigallocatechin>(-)-epicatechin-3-gallate>(-)-gall[ocatechin](https://www.ncbi.nlm.nih.gov/mesh/D002392)>(-)-epicatechin. Treatment of A375 and Hs29[4t cells with EGCG resulted in](https://www.ncbi.nlm.nih.gov/mesh/C045651) a[ dos](https://www.ncbi.nlm.nih.gov/mesh/C045651)e-[dependent inhibition](https://www.ncbi.nlm.nih.gov/mesh/C057580) [of cell migration or inva](https://www.ncbi.nlm.nih.gov/mesh/C062669)s[ion of these cell](https://www.ncbi.nlm.nih.gov/mesh/C057580)s[, which was ass](https://www.ncbi.nlm.nih.gov/mesh/D002392)ociated with a reduction in the levels of [cycl](https://www.ncbi.nlm.nih.gov/mesh/C045651)ooxygenase (COX)-2, prostaglandin (PG) E2 and PGE2 receptors (EP2 and EP4). Treatment of cells with celecoxib, a COX-2 inhibitor, also inhibited melanoma cell migrati[on. EGCG inhibits 12-](https://www.ncbi.nlm.nih.gov/mesh/D015232)O-tetradecanoylphorbol-13-acetate-, an inducer of COX-2, an[d PGE2-in](https://www.ncbi.nlm.nih.gov/mesh/D000068579)duced cell migration of cells. EGCG decreased EP2 agonist (bu[tapr](https://www.ncbi.nlm.nih.gov/mesh/C045651)ost)- and [EP4 agonist (Cay10580)-induced cell ](https://www.ncbi.nlm.nih.gov/mesh/D013755)migration ability. Moreover,[ EGC](https://www.ncbi.nlm.nih.gov/mesh/D015232)G inhibited the activation of NF-κ[B/p6](https://www.ncbi.nlm.nih.gov/mesh/C045651)5, an upstream regulator[ of COX-2](https://www.ncbi.nlm.nih.gov/mesh/C048491), in A375 melanoma c[ells, an](https://www.ncbi.nlm.nih.gov/mesh/C561931)d treatment of cells with caffeic acid phene[thyl](https://www.ncbi.nlm.nih.gov/mesh/C045651) ester, an inhibitor of NF-κB, also inhibited cell migration. Inhibition of melanoma cell migration by EGCG was associated w[ith transition of mesenchyma](https://www.ncbi.nlm.nih.gov/mesh/C055494)l stage to epithelial stage, which resulted in an increase in the levels of epithelial biomarkers[ (E-](https://www.ncbi.nlm.nih.gov/mesh/C045651)cadherin, cytokeratin and desmoglein 2) and a reduction in the levels of mesenchymal biomarkers (vimentin, fibronectin and N-cadherin) in A375 melanoma cells. Together, these results indicate that EGCG, a major green tea catechin, has the ability to inhibit melanoma cell invasion/migration, an essential step of metastasis, by targeting the e[ndog](https://www.ncbi.nlm.nih.gov/mesh/C045651)enous expression of [COX-2, P](https://www.ncbi.nlm.nih.gov/mesh/D002392)GE2 receptors and epithelial-to-mesenchymal transition.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
The melanoma remains the leading cause of death from skin diseases due to its propensity to metastasis. The statistical analysis from American Cancer Society indicated that in 2008, there were 8,420 melanoma-associated deaths in the U.S. and the number of new cases of invasive melanoma was estimated at 62,480 . The incidence of melanoma has increased in the past few decades in the United States and is increasing rapidly in children –. If recognized and treated early, melanoma is curable, but as the disease progresses its propensity to metastasize make it difficult to treat. Chronic exposure to solar ultraviolet (UV) radiation has been implicated in melanoma and non-melanoma skin cancers , .
Exposure of the skin to UV radiation induces an increase in the expression levels of cyclooxygenase-2 (COX-2), a rate-limiting enzyme that catalyzes the conversion of arachidonic acid to prostaglandins (PGs). The enhanced generation of PGs, specifically PGE2, plays a central role in orchestrating the multiple events involved in cancer invasion and metastasis. PGE2 is an important metabolite which has been implicated in these risks more than other PG metabolites, and has been shown to exert its effects through G protein-coupled receptors, EP1, EP2, EP3 and EP4, and has been implicated in angiogenesis, decreased host immunity and enhanced invasion and metastasis , . Although, melanoma is less common than other skin cancers, it causes the majority (75%) of skin cancer-related deaths. Once diagnosed with metastatic melanoma, most patients will die of their disease within 2 years , .
Since, melanoma is a highly malignant cancer with a potent capacity to metastasize distantly, an approach that reduces its metastatic ability may facilitate the development of an effective strategy for its treatment and/or prevention.[](https://www.ncbi.nlm.nih.gov/mesh/D016718)
Catechins isolated from the leaves of green tea (Camellia sinensis) have a number of beneficial health effects including anti-carcinogenic activity, which has been demonstrated in various tumor models , . In previous studies, we and others have shown that oral administration of an aqueous extract of green tea or green tea catechins, which are commonly called as polyphenols (a mixture of catechins), in drinking water inhibits UV radiation-induced non-melanoma skin cancer in mice in terms of tumor incidence, tumor multiplicity and tumor growth/size , . Multiple mechanisms or molecular targets have been reported by which green tea polyphenols protect the skin from skin tumors. These mechanisms include the DNA repair , , stimulation of immune system , , anti-inflammatory effects and anti-oxidant activity of green tea polyphenols in vitro and in vivo models. However, the beneficial effects of green tea polyphenols on melanoma are not well studied and less understood.
As green tea is commonly consumed as a beverage world-wide, we assessed the effect of its polyphenolic components on the invasive potential of melanoma cells using melanoma cell lines as an in vitro model. Beverage grade green tea leaves contains 5 major catechins or epicatechin derivatives: (-)-epicatechin, (-)-gallocatechin, (-)-epicatechin gallate, (-)-epigallaocatechin and (-)-epigallocatechin-3-gallate (EGCG) .[](https://www.ncbi.nlm.nih.gov /mesh/D002392)
In the present study, first we assessed the chemotherapeutic potential of various catechins on the migration capacity of human melanoma cells, as the migration of cancer cells is a major event in the metastatic cascade. For this purpose, two highly metastasis-specific melanoma cancer cell lines were selected: one is A375 which is BRAF mutated and activating mutations of the protooncogene BRAF have been observed in approximately 50% of malignant melanomas. Second cell line is Hs294t, which is also highly metastatic but not BRAF mutated. These two cell lines were used as an in vitro model for this study. In preliminary screening experiments, we identified that EGCG is a major active component of green tea polyphenols which significantly blocks the migration/invasion of melanoma cells compared to other catechins or epicatechin derivatives.
We further characterized the role of COX-2 and its metabolite PGE2 on the migration of human melanoma cancer cells and ascertained whether EGCG has any suppressive effects on the COX-2-mediated migration of these cells. Epithelial-to-mesenchymal transition (EMT), the process whereby epithelial cells transform into mesenchymal cells, has recently been shown to be relevant for cancer growth and cancer metastasis. During EMT, cells lose expression of proteins that promote cell-cell contact such as E-cadherin and acquire mesenchymal markers such as vimentin, fibronectin and N-cadherin, which promote cell invasion and metastasis . The EMT has also been associated with higher levels of inflammation and inflammatory mediators, and therefore we have also checked whether inhibition of COX-2 expression and PGE2 production by EGCG in melanoma cells is associated with reversal of EMT and that leads to inhibitory effect on melanoma cell migration. Here, we present evidence that EGCG inhibits the invasive potential of melanoma cells through transition of mesenchymal state to epithelial state in melanoma cells and that EGCG do so through a process that involves the reduction of COX-2 expression and lowering the levels of PGE2 and PGE2 receptors in melanoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/D002392)
## Materials and Methods
*In the **Materials and Methods** section:*
## Source of green tea catechins
*In the **Source of green tea catechins** section:*
Various purified tea catechins used in this study were obtained from Dr. Y. Hara of Mitsui Norin Company (Tokyo, Japan). These catechins or polyphenols are stable for at least two years when refrigerated at 4°C.[](https://www.ncbi.nlm.nih.gov/mesh/D002392)
## Antibodies, chemicals and reagents
*In the **Antibodies, chemicals and reagents** section:*
Celecoxib, PGE2, 12-O-tetradecanoylphorbol-13-acetate (TPA) and EP2 agonist were purchased from Sigma Chemical Co. (St. Louis, MO). Boyden Chambers and polycarbonate membranes (8 µm pore size) for cell invasion assays were obtained from Neuroprobe, Inc. (Gaithersburg, MD). The antibodies specific to N-cadherin, keratin-18 and fibronectin were obtained from Santa Cruz Biotechnology (Santa Cruz, CA), while antibodies for EP1, EP2, EP3, EP4, vimentin, E-cadherin, NF-κB, IKKα and IκBα and their secondary antibodies were purchased from Cell Signaling Technology (Beverly, MA). Desmoglein-2 was obtained from Abcam (Cambridge, MA). Antibodies specific for COX-2, EP4 agonist and an enzyme immunoassay kit for PGE2 analysis were obtained from Cayman Chemicals (Ann Arbor, MI).[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Cell lines and cell culture conditions
*In the **Cell lines and cell culture conditions** section:*
The normal human epidermal melanocytes (HEMa-LP, Catalogue #C-024-5C) were commercially obtained from Invitrogen (Carlsbad, CA), and were cultured in HMGS-2 medium supplemented with human melanocyte growth supplement provided by the supplier. The human melanoma cells lines, A375 and Hs294t, were purchased from the American Type Culture Collection (Manassas, VA). The cell lines were cultured as monolayers in RPMI 1640 culture medium supplemented with 10% heat-inactivated fetal bovine serum (Hyclone, Logan, UT), 100 µg/mL penicillin, and 100 µg/mL streptomycin and maintained in an incubator with 5% CO2 at 37°C. The EGCG was dissolved in a small amount of acetone, which was added to the complete cell culture medium [maximum concentration of acetone, 0.1% (v/v) in media] prior to addition to sub-confluent cells (60–70% confluent). Cells treated with acetone only served as a vehicle control.
To determine the effect of EGCG on TPA- or PGE2-mediated effects, EGCG was added in cell culture medium at least 30 minutes before the treatment of the cells with TPA, PGE2, PGE2 receptor or PGE2 receptor agonists.[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Cell invasion assay
*In the **Cell invasion assay** section:*
The invasion capacity of melanoma cells was determined in vitro using Boyden Chambers (Gaithersburg, MD) in which the two chambers were separated with matrigel coated Millipore membranes (6.5 mm diameter filters, 8 µM pore size), as detailed previously . Briefly, melanoma cells (1.5×104 cells/100 µL serum-reduced medium) were placed in the upper chamber of Boyden chambers, test agents were added alone, or in combination, to the upper chamber (200 µL), and the lower chamber contained the medium alone (150 µL). Chambers were assembled and kept in an incubator for 24 h. After incubation, cells from the upper surface of Millipore membranes were removed with gentle swabbing and the migrant cells on the lower surface of membranes were fixed and stained with crystal violet. Membranes were then washed with distilled water and mounted onto glass slides. The membranes were examined microscopically and cellular migration was determined by counting the number of stained cells on each membrane in at least 4–5 randomly selected fields using an Olympus BX41 microscope.
Representative photomicrographs were obtained using a Qcolor5 digital camera system fitted to an Olympus BX41 microscope. Resultant data are presented as a mean of migrating cells ± SD/microscopic field of three independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C059485)
## Scratch assay or wound healing assay
*In the **Scratch assay or wound healing assay** section:*
Scratch or wound healing assay was performed to detect melanoma cell migration, as detailed previously . Briefly, melanoma cells were grown to full confluency in six-well plates and incubated overnight in starvation medium. Cell monolayers were wounded with a sterile 100 µL pipette tip, washed with starvation medium to remove detached cells from the plates. Cells were left either untreated or treated with indicated doses of tea catechins in full medium and kept in a CO2 incubator for 48 h. After 48 h, medium was replaced with phosphate-buffered saline (PBS) buffer, the wound gap was observed and cells were photographed using an Olympus BX41 microscope fitted with digital camera.[](https://www.ncbi.nlm.nih.gov/mesh/D002392)
## Quantitation of prostaglandin E2 using PGE2 immunoassay kit
*In the **Quantitation of prostaglandin E2 using PGE2 immunoassay kit** section:*
The levels of PGE2 in cell homogenates were measured using the Cayman PGE2 Enzyme Immunoassay Kit (Ann Arbor, MI) following the manufacturer's protocol. Briefly, at indicated time point, cells were harvested and homogenized in 100 mM phosphate buffer, pH 7.4 containing 1 mM ethylenediamine tetraacetic acid and 10 µM indomethacin using a homogenizer. Homogenates were centrifuged and the supernatants were collected for the analysis of PGE2 concentrations following the manufacturer's protocol.[](https://www.ncbi.nlm.nih.gov/mesh/D015232)
## Western blot analysis
*In the **Western blot analysis** section:*
Cell lysates were prepared to analyze the expression levels of different proteins, as described previously . Briefly, following treatment of melanoma cells for the indicated time periods with or without EGCG or any other agent, the cells were harvested, washed with cold PBS and lysed with ice-cold lysis buffer supplemented with protease inhibitors. Equal amounts of proteins were resolved on 10% Tris-Glycine gels and transferred onto a nitrocellulose membrane. After blocking the non-specific binding sites, the membrane was incubated with the primary antibody at 4°C overnight. The membrane was then incubated with the appropriate peroxidase-conjugated secondary antibody and the immunoreactive bands were visualized using the enhanced chemiluminescence reagents.
Equal protein loading was verified on the membrane after stripping it and re-probed with anti-β actin antibody.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## Assay for NF-κB/p65 activity
*In the **Assay for NF-κB/p65 activity** section:*
Quantitative analysis of NF-κB/p65 activity was performed using the NF-κB TransAM Activity Assay Kit (Active Motif, Carlsbad, CA) following the manufacturer's protocol. Briefly, the nuclear extracts of cells were prepared using the Nuclear Extraction Kit (Active Motif, Carlsbad, CA) following the manufacturer's instructions. Absorbance was recorded at 450 nm using absorbance at 650 nm as the reference. The results are expressed as the percentage of the optical density of the non-EGCG-treated control group.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## Statistical analysis
*In the **Statistical analysis** section:*
For migration assays, the control and EGCG-, TPA- PGE2- or EP2- and EP4-agonists treatment groups or combined-treatment groups separately were compared using one-way analysis of variance (ANOVA) followed by post hoc Dunn's test using GraphPad Prism version 4.00 for Windows, GraphPad Software, San Diego, California, USA, www.graphpad.com. All quantitative data for cell migration are shown as mean ± SD/microscopic field. In each case P<0.05 was considered statistically significant.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## Results
*In the **Results** section:*
## Effect of green tea catechins on human melanoma cell migration/invasion
*In the **Effect of green tea catechins on human melanoma cell migration/invasion** section:*
In the present study, first we have assessed and compared the effect of various green tea catechins on the invasive potential of human melanoma cells in vitro using Boyden chamber. The molecular structures of five major tea catechins, EC, GC, ECG, EGC and EGCG are shown in Figure 1A. Using in vitro cell invasion assay, we found that treatment of A375 and Hs294t cells with equimolar concentration (25 µM) of catechins for 24 h resulted in inhibition of migration/invasion of these cells. The relative inhibitory effect of catechins on melanoma cell migration or invasion was in the order of: EGCG>EGC>ECG>GC>EC, as shown in Figure 1B. The experiment was repeated three times and resultant data of cell migration from each treatment group has been summarized in Figure 1C. These results also suggest that the cell migration ability of BRAF-mutated A375 cells was higher than non-BRAF-mutated Hs294t cells; however, the difference was not statistically significant.
This screening preliminary experiment revealed that EGCG has greater inhibitory effect on melanoma cell migration compared to other tea catechins; therefore, EGCG was selected for further studies of cell invasion behavior of human melanoma cells and the molecular mechanisms underlying these effects.[](https://www.ncbi.nlm.nih.gov/mesh/D002392)
Comparative effects of green tea catechins on the cell invasion capacity of melanoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/D002392)
(A) Chemical structures of major green tea catechins. EC, (-)-epicatechin; GC, (-)-gallocatechin; ECG, (-)-epicatechin-3-gallate; EGC, (-)-epigallocatechin, and EGCG, (-)-epigallocatechin-3-gallate. (B) Effect of equimolar concentration (25 µM) of various green tea catechins on the migration of human melanoma cells (A375 and Hs294t) after treatment for 24 h. (C) The migrating cells were counted and the results expressed as the mean number of migratory cells ± SD/microscopic field. Data collected from three independent experiments. Significant inhibition by EGC and EGCG versus non-tea catechins-treated controls, ¶
P<0.01; *
P<0.001.[](https://www.ncbi.nlm.nih.gov/mesh/D002392)
## EGCG inhibits human melanoma cell invasion
*In the **EGCG inhibits human melanoma cell invasion** section:*
Next, we determined dose-dependent effect of EGCG on the cell migration potential of A375 and Hs294t human melanoma cells using Boyden chamber cell migration assays. First, screening experiments were performed to determine the effects of lower concentrations of EGCG (µg/mL), which should not reduce the cell viability or induce apoptosis in these cells. As shown in Figure 2A, relative to untreated control cells, treatment of cells with EGCG at concentrations of 0, 10, 20 and 40 µg/mL reduced the invasive potential of A375 and Hs294t cells in a concentration-dependent manner. The density of the migrating cells on the membrane after staining with crystal violet is shown in Figure 2A, and the numbers of migrating cells/microscopic field are summarized in Figure 2B. The melanoma cell invasion potential was inhibited by 13% to 65% (P<0.01–0.001) in A375 cells and by 7% to 70% (P<0.001) in Hs294t cells in a concentration-dependent manner after treatment with EGCG for 24 h.
The density of cell migration was higher after 48 h, and the inhibitory effect of EGCG on melanoma cells migration was also comparatively higher than 24 h (data not shown). To verify that the inhibition of cancer cell migration by EGCG was a direct effect on migration ability, and that was not due to a reduction in cell viability, a trypan blue assay was performed using cells that were treated identically to those used in the migration assays. Treatment of A375 and Hs294t cells with various concentrations of EGCG (0, 10, 20 and 40 µg/mL) for 24 h had no significant inhibitory effect on cell viability or cell death (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
EGCG inhibits melanoma cell invasion/migration.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
(A) Treatment of human melanoma cells with EGCG for 24 h inhibits migration of A375 and Hs294t cells in a concentration-dependent manner. (B) The migrating cells were counted and the results expressed as the mean number of migratory cells ± SD/microscopic field. Significant inhibition by EGCG versus non-EGCG-treated controls, ¶
P<0.01; *
P<0.001. (C) Scratch or wound healing assay was performed to assess the effect of EGCG on the migration of A375 and Hs294t melanoma cells. Incubation of A375 or Hs294t cells with EGCG (10 and 20 µg/mL) for 48 h inhibits migration of cells compared to non-EGCG-treated control cells. Assay was repeated three times and representative pictures are shown. Space between dotted lines in each panel shows the space without or negligent number of migrating cells.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
We have further confirmed the inhibitory effect of EGCG on melanoma cell migration by employing scratch or wound healing assay, as described in Material and Methods. As shown in Figure 2C, relative to untreated control cells, treatment of cells with EGCG (10 and 20 µg/mL) reduced the migration capacity of A375 and Hs294t cells in a concentration-dependent manner after the treatment of cells for 48 h. The part of gap or wounding space between cell layers after making a wound was occupied by the migrating Hs294t cells which were not treated with EGCG. However, the healing of the wound or the empty space of the cells was not occupied by the migrating cells treated with EGCG and this effect was dose-dependent. The gap or wounding space between the cell layers is highlighted by broken white lines, as shown in Figure 2C.
Similar inhibitory effects of EGCG on cell migration using scratch assay was also found with A375 cells (Figure 2C, lower panels). These observations suggest that EGCG has the ability to inhibit the migration ability of melanoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## EGCG reduces endogenous COX-2 expression in melanoma cells
*In the **EGCG reduces endogenous COX-2 expression in melanoma cells** section:*
To examine whether the inhibitory effect of EGCG on the migration of the melanoma cells is associated with the reduction of endogenous expression of COX-2, we determined the levels of COX-2 in cell lysates of the various treatment groups using western blot analysis. Western blot analysis revealed that the treatment of A375 and Hs294t cells with EGCG reduced the levels of COX-2 expression in a concentration-dependent manner as compared to the expression in untreated controls (Figure 3A).[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
EGCG inhibits invasion of melanoma cells by reducing endogenous COX-2 expression.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
(A) Treatment of A375 and Hs294t cells with EGCG for 24 h resulted in down-regulation of COX-2 protein expression. (B) Treatment of A375 and Hs294t cells with celecoxib, an inhibitor of COX-2, for 24 h inhibits melanoma cell migration in a dose-dependent manner. The data are expressed as the mean number of migratory cells± SD/microscopic field. Significant difference versus non-celecoxib-treated control cells, *
P<0.001; †
P<0.05. (C) Treatment of A375 cells with EGCG (20 and 40 µg/mL) inhibits TPA (a COX-2 stimulator)-enhanced cell migration capacity. The data on cell migration capacity are summarized in terms of mean number of migrating cells/microscopic field ± SD, n = 3. Significant inhibition versus TPA treatment alone, *
P<0.001. (D) EGCG down regulates TPA-induced COX-2 expression in A375 cells. The levels of COX-2 were determined in cell lysates using western blot analysis.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## Celecoxib, a selective COX-2 inhibitor, inhibits melanoma cell migration
*In the **Celecoxib, a selective COX-2 inhibitor, inhibits melanoma cell migration** section:*
This experiment was performed to verify whether the inhibitory effect of EGCG on melanoma cell migration is mediated through its inhibitory effect on COX-2 expression. For this purpose, equal numbers of A375 and Hs294t cells were subjected to the cell invasion assay after treatment with various concentrations of celecoxib (0, 5, 10, 20 µM), a well known inhibitor of COX-2, for 24 h. As shown in Figure 3B, treatment of the A375 and Hs294t cells with celecoxib resulted in a significant reduction in the cell migration capacity of melanoma cells in a concentration-dependent manner as compared with non-celecoxib-treated controls (P<0.05–0.001). These data suggested that the inhibition of COX-2 expression by the treatment of cells with EGCG is associated with the inhibition of melanoma cell migration.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## EGCG inhibits TPA (an inducer of COX-2)-induced cell migration
*In the **EGCG inhibits TPA (an inducer of COX-2)-induced cell migration** section:*
TPA is a well known skin tumor promoter and has been shown to stimulate the expression of COX-2 in the skin cells ; therefore, the melanoma cells were treated with TPA for the stimulation of COX-2 expression in vitro, and thereafter determined the effect of TPA on the migration of melanoma cells. As shown in Figure 3C, treatment of A375 cells with TPA (40 ng/mL) for 24 h resulted in a significantly enhanced cell migration (P<0.001) compared to non-TPA-treated control cells. To determine whether EGCG inhibits TPA-induced cell migration in human melanoma cells, A375 cells were treated with TPA (40 ng/mL) with and without the treatment of EGCG for 24 h. The treatment of A375 cells with EGCG (20 and 40 µg/mL) resulted in a dose-dependent inhibition of TPA-induced cell migration. A summary of the cell migration data for the various treatment groups is provided in Figure 3C. Treatment of EGCG at the doses of 20 µg/mL and 40 µg/mL inhibited TPA-induced cell migration by 95% (P<0.001) and >100% (P<0.001) respectively.
To verify whether this inhibition of cell migration by EGCG is mediated through the inhibition of TPA-induced COX-2 expression, cell lysates were prepared and subjected to western blot analysis to estimate the levels of COX-2 expression. Western blot analysis data revealed that treatment of A375 cells with TPA for 24 h resulted in higher expression of COX-2 as compared to the expression in cells that were not treated with TPA (Figure 3D). Pretreatment of A375 cells with EGCG (20 and 40 µg/mL) for 24 h resulted in inhibition of TPA-induced COX-2 expression (Figure 3D). These data suggest that inhibition of TPA-induced cell migration by EGCG is mediated through the downregulation of COX-2 expression.[](https://www.ncbi.nlm.nih.gov/mesh/D013755)
## Inhibition of melanoma cell migration by EGCG is mediated through its inhibitory effects on PGE2 production
*In the **Inhibition of melanoma cell migration by EGCG is mediated through its inhibitory effects on PGE2 production** section:*
PGE2 is one of the metabolites of COX-2, and most of the biological activities of COX-2 are mediated through its metabolites, therefore, we examined whether treatment of melanoma cells with EGCG reduced the levels of PGE2 production. For this purpose, Hs294t melanoma cells were treated with EGCG for 24 h. Cells were harvested and the levels of PGE2 were determined using PGE2 immunoassay kit. As shown in Figure 4A, treatment of cells with EGCG resulted in a dose-dependent reduction in the levels of PGE2 in these cells. Next, we examined the effect of PGE2 on the Hs294t cell migration using in vitro cell invasion assay. Cell invasion data revealed that treatment of PGE2 significantly enhanced (P<0.01, P<0.001) the cell migration potential in a dose-dependent manner (Figure 4B).
Further, the effect of EGCG on PGE2-induced cell migration was evaluated. For this purpose, Hs294t cells were treated with PGE2 (10 µM) with and without the treatment with EGCG for 24 h and cell migration determined. We found that the treatment of melanoma cells with PGE2 resulted in a significant increase of cell migration (P<0.05) compared to the cells which were not treated with PGE2 (Figure 4C and 4D). Treatment of Hs294t cells with EGCG (20 or 40 µg/mL) for 24 h resulted in a dose-dependent inhibition of PGE2 (10 µM)-induced melanoma cell migration, as shown in Figure 4C and data are summarized in Figure 4D. Similar inhibitory effect of EGCG was observed on PGE2-induced cell migration in A375 melanoma cells (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D015232)
EGCG inhibits PGE2-induced melanoma cell migration.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
(A) Dose-dependent effect of EGCG on the levels of PGE2 in melanoma Hs294t cells. The levels of PGE2 are expressed in terms of pg/mg protein± SD, n = 3 independent experiments. Significant inhibition by EGCG versus non-EGCG-treated controls, *
P<0.001. (B) Treatment of Hs294t cells with PGE2 enhances cell migration in a concentration-dependent manner. Significant difference versus control, ¶
P<0.05, *
P<0.001. (C) Treatment of cells with EGCG (20 and 40 µg/mL) inhibits PGE2-enhanced cell migration capacity of Hs294t cells. Representative photomicrographs of cell migration were presented from three independent experiments. (D) The data on cell migration are summarized as a mean number of migratory cells ± SD/microscopic field. Significant inhibition versus PGE2 alone: *
P<0.001. Cell invasion assays under each experiment were repeated three times and in each case the migrating cells were counted and the results are expressed as a mean number of migratory cells ±SD/microscopic field.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## EGCG decreases the levels of PGE2 receptors in melanoma cells
*In the **EGCG decreases the levels of PGE2 receptors in melanoma cells** section:*
COX-2 metabolite PGE2 has been shown to manifest its biological activity via four known G-protein-coupled receptors (i.e., EP1-EP4) , . Therefore, we determined the effect of EGCG on the basal levels of PGE2 receptors in melanoma cells. Western blot analysis revealed that treatment of A375 cells with EGCG (0, 10, 20 and 40 µg/mL) for 24 h resulted in a dose-dependent reduction in the levels of EP2 and EP4 (Figure 5A). The inhibitory effect of EGCG was also observed on EP1 and EP3 but was less prominent than the effect on EP2 and EP4 (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D015232)
EGCG inhibits PGE2 receptors-induced human melanoma cell migration.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
(A) Treatment of A375 melanoma cells with EGCG for 24 h decreases the expression levels of PGE2 receptors EP2 and EP4 in a concentration-dependent manner. The cells were harvested 24 h after the treatment and cell lysates were prepared and subjected to western blot analysis, as detailed in Materials and Methods. (B) Treatment of A375 cells with EP2 agonist (butaprost) for 24 h enhances melanoma cell migration (left panels). The data on cell migration are summarized and expressed as a mean number of migratory cells ± SD/microscopic field, n = 3 (right panel). (C) Treatment of A375 cells with EGCG inhibits EP2 agonist (1.0 µM)-induced melanoma cell migration (left panels). The data on cell migration are summarized as a mean number of migratory cells ± SD/microscopic field, n = 3 (right panel). (D) Treatment of A375 cells with EP4 agonist (Cay10580) for 24 h enhances cell migration (left panels). The data on cell migration in each group are expressed as a mean number of migratory cells ± SD/microscopic field, n = 3 (right panel). (E) Treatment of A375 cells with EGCG inhibits EP4 agonist (1.0 µM)-induced cell migration (left panels). The data on cell migration are summarized as a mean number of migratory cells ± SD/microscopic field, n = 3 (right panel). Representative photomicrographs of cell migration were presented from three independent experiments in each case. Significant inhibition versus non-EGCG-treated controls, ¶
P<0.01; *
P<0.001; Significant increase versus non-EP2 agonist-treated or non-EP4 agonist-treated controls, †
P<0.001.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## An EP2 agonist and EP4 agonist promote the migration of melanoma cells while EGCG inhibits EP2 and EP4 agonists-stimulated cell migration
*In the **An EP2 agonist and EP4 agonist promote the migration of melanoma cells while EGCG inhibits EP2 and EP4 agonists-stimulated cell migration** section:*
To determine whether PGE2 receptor EP2 and EP4 has a role in melanoma cell migration, and whether EGCG inhibits their effects on cell migration, we further conducted cell invasion experiments with A375 melanoma cells. In vitro cell invasion experiments revealed that treatment of A375 cells with EP2 agonist (butaprost) significantly enhanced the migration ability of melanoma cells, as shown in Figure 5B. A summary of number of migrating cells/microscopic field is also shown (Figure 5B, right panel). Pretreatment of A375 cells with EGCG inhibits EP2 agonist (1 µM)-induced melanoma cell migration by 36% and 58% (P<0.001) at the dose of 20 and 40 µg/mL, respectively (Figure 5C right panel). Similarly, the effect of EP4 agonist was determined on the migration of A375 melanoma cells.
As shown in Figure 5D, EP4 agonist treatment of A375 cells for 24 h significantly enhanced (P<0.001) the cell migration in a dose-dependent manner. Treatment of cells with EGCG (20 and 40 µg/mL) significantly inhibited (41–66%, P<0.01 and P<0.001) EP4 agonist-induced cell migration (Figure 5E, left and right panels). These data suggest that the expressions of PGE2 receptors in melanoma cells have roles in cell migration, and that EGCG inhibits the migration of melanoma cells, at least in part, by decreasing the levels of PGE2 receptors.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## EGCG decreases the nuclear level and activity of NF-κB/p65 in melanoma cells: NF-κB acts as a mediator of melanoma cell invasion
*In the **EGCG decreases the nuclear level and activity of NF-κB/p65 in melanoma cells: NF-κB acts as a mediator of melanoma cell invasion** section:*
NF-κB is an upstream regulator of COX-2, therefore we assessed whether EGCG also affects the levels and activation of NF-κB in melanoma cells. To examine this effect, A375 cells were treated with various concentrations of EGCG (0, 10, 20 and 40 µg/mL) for 24 h, and thereafter cells were harvested and cell lysates prepared for western blot analysis. Western blot analysis revealed that treatment of cells with EGCG decreased the translocation of NF-κB/p65 in to the nucleus in a dose-dependent manner (Figure 6A). The activity of NF-κB also was significantly reduced (20–65%, P<0.001) after the treatment of cells with EGCG in a concentration-dependent manner (Figure 6B). The results also indicated that treatment of EGCG resulted in the downregulation of IKKα and degradation of IκBα (Figure 6A), which leads to the inactivation of NF-κB and its translocation to the nucleus.
Further, to assess whether NF-κB has a role in melanoma cell migration, A375 melanoma cells were treated with caffeic acid phenethyl ester (0, 5.0 and 10.0 µg/mL), a potent inhibitor of NF-κB, and cell migration was studied. As shown in Figures 6C and 6D, treatment of A375 cells with caffeic acid phenethyl ester resulted in a significant reduction of cell migration (26% and 57%; P<0.05, and P<0.001) compared to non-cafeic acid phenethyl ester-treated control cells, and these results are similar to that observed on treatment of the cells with EGCG (Figure 2).[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
Effect of EGCG on NF-κB activation in melanoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
(A) Treatment of A375 cells with EGCG decreases the basal level of NF-κB/p65 and IKKα while inhibiting the degradation of IκBα. After 24 h treatment with various concentrations of EGCG the cells were harvested and cytosolic and nuclear fractions were prepared and subjected to the analysis of NF-κB, IKKα and IκBα using western blot analysis. Representative blot is shown from three independent experiments with identical observations. (B) The activity of NF-κB/p65 in the nuclear fraction of the cells after treatment with and without EGCG for 24 h was measured using NF-κB/p65-specific activity assay kit, n = 3. Activity of NF-κB is expressed in terms of percent of control (non-EGCG-treated) group. Significant decrease versus control: *
P<0.001. (C) Treatment of A375 cells with caffeic acid phenethyl ester (CAPE), an inhibitor of NF-κB, inhibits melanoma cell migration. Representative photomicrographs are shown from three separate experiments. (D) Data on cell migration capacity are summarized as the mean number of migratory cells ± SD/microscopic field, n = 3. Significant inhibition versus non-CAPE-treated cells: ¶
P<0.05; *
P<0.001.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## EGCG reverses epithelial-to-mesenchymal transition in melanoma cells
*In the **EGCG reverses epithelial-to-mesenchymal transition in melanoma cells** section:*
Activation of NF-κB has been implicated in inflammation-induced tumor growth and progression, and has been identified as an important regulator of EMT in several cancer cell types –. As the inhibition of melanoma cell migration by EGCG is associated with the inactivation of NF-κB, we sought to examine whether EGCG targets EMT biomarkers or whether EGCG transforms mesenchymal biomarkers to epithelial biomarkers in melanoma cells and that is responsible for its inhibitory effect on cell migration. To examine this effect, A375 cells were treated with EGCG for 24 h, and cell lysates were prepared for the analyses of various epithelial and mesenchymal biomarkers using western blot analysis. As shown in Figure 7, western blot analyses revealed that EGCG increased the levels of the epithelial biomarkers, such as E-cadherin, keratin-18 and desmoglein 2 dose-dependently in melanoma cells compared to untreated controls. In contrast, the levels of mesenchymal biomarkers, such as vimentin, fibronectin and N-cadherin were reduced in melanoma cells after treatment with EGCG in a dose-dependent manner.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
Treatment of melanoma cells with EGCG results in mesenchymal-to-epithelial transition.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
Treatment of A375 cells with EGCG for 24 h enhances the levels of epithelial biomarkers in the cells, such as, the levels of E-cadherin, keratin-18 and desmoglein 2. Simultaneously the levels of mesenchymal biomarkers in melanoma cells, such as, vimentin, fibronectin and N-cadherin were decreased dose-dependently. Western blot analysis was performed as detailed under Materials and Methods. Representative blots are shown from three independent experiments with similar results.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
## Discussion
*In the **Discussion** section:*
Melanoma remains the leading cause of death from skin disease, in large part, due to its propensity to metastasize. Although melanoma is less common than non-melanoma skin cancers, which includes squamous cell and basal cell carcinoma, however, it causes approximately 75% skin cancer-related deaths. World Health Organization report indicated 48,000 melanoma-related deaths worldwide per year . Most patients suffering from malignant melanoma ultimately die of their disease within two years , . The development of new treatment options and novel strategies are required which can prevent the invasiveness of melanoma cells and that can inhibit the metastatic ability of melanoma cells.
Majority of cancers over-express COX-2, an enzyme responsible for the biosynthesis of PGs metabolites. Enhanced production of PGs, and particularly PGE2, has been linked with tumor progression, invasion and metastasis , . Because of its important role in tumor invasion and metastasis, COX-2 is considered as a promising target for cancer therapy , . Therefore, the search of novel and non-toxic inhibitors of COX-2 as well as the inhibitors of PGE2 may provide a better option for the treatment of malignant melanoma and that may prove a better strategy for its prevention or treatment.[](https://www.ncbi.nlm.nih.gov/mesh/D011453)
Green tea polyphenols/catechins have been shown to have anti-carcinogenic activities in various tumor models, including skin cancers , , . In the present study, we have found that EGCG has significantly greater anti-invasive activity in melanoma cells compared to EC, GC, ECG and EGC. The significant findings of the present study are that the treatment of melanoma cells with EGCG for 24 h inhibits cell migration in a dose-dependent manner, and that is associated with the inhibition of COX-2 expression and PGE2 production. Based on our experimental observations, cells will go under apoptosis or cell death if melanoma cells are treated with higher concentrations of EGCG or for more than 24 h. Under these conditions, cell migration will decrease, and this reduction in cell migration could be due to cell death and not due to changes in migrating behavior of cells. In this study, cell death or apoptosis is not a reason of EGCG-caused inhibition of melanoma cell migration.
The melanoma cells over-express COX-2, and the reduction in the levels of COX-2 by EGCG may be responsible for the inhibition of cell migration of melanoma cells. This notion is supported by the evidence that treatment of the melanoma cells with a potent COX-2 inhibitor (celecoxib) resulted in a reduction of cell migration. Studies have shown that TPA promotes COX-2 expression and subsequently enhances cell migration , and we have found that TPA-induced cell migration was blocked by the treatment of cells with EGCG. These observations further suggest that inhibition of melanoma cell migration by EGCG is mediated through the inhibition of COX-2 expression. PGE2 exerts its biologic functions through four G protein-coupled receptors, EP1, EP2, EP3 and EP4 , , , that can stimulate cell survival signals as well as invasive potential of cancer cells –.
PGE2 has been shown to promote lung cancer and melanoma cell migration, and that this effect of PGE2 is associated with the activation of PGE2 receptors , . Based on these investigations, we determined the involvement of the PGE2 receptors in EGCG-mediated inhibition of melanoma cell migration. It was found that the levels of EP2 and EP4 were decreased when melanoma cells were treated with EGCG. These data suggest that inhibition of the EP2 and EP4 levels by EGCG may contribute to the inhibition of melanoma cell migration. The inhibitory effect of EGCG on melanoma cell migration through the inhibitory effect on EP2 or EP4 was further verified by treating the cells with an EP2 agonist (butaprost) and an EP4 agonist (cay10580) with and without the treatment of cells with EGCG.
Our data revealed that the treatment of A375 cells with the EP2 agonist and EP4 agonist resulted in enhanced cell migration, and that EP2 agonist- and EP4 agonist-induced cell migration was significantly inhibited by the treatment of cells with EGCG using identical in vitro conditions. These observations support the evidence that inhibition of PGE2 receptors by green tea catechin EGCG may have contributed to the blocking of melanoma cell migration. These findings also demonstrate the feasibility of using EGCG as an alternative to COX-2 inhibitors, which show toxicity in some patients, given the fact that COX-2 remains an attractive target for cancer therapy. As EGCG acts by decreasing the expression of both COX-2 and EP receptors, this could be more effective because EGCG targets both ligand (PGE2) and receptor (EP). EGCG has also been shown to inhibit mammary cancer cell migration through the inhibition of nitric oxide and nitric oxide-mediated mechanisms .
Other phytochemicals also have been assessed for their inhibitory effect on cancer cell invasion and migration. Punathil and Katiyar have reported that treatment of non-small cell lung cancer cells with proanthocyanidins from grape seeds resulted in inhibition of cell migration following the inhibition of nitric oxide and guanylate cyclase pathways.[](https://www.ncbi.nlm.nih.gov/mesh/D059808)
As NF-κB is an upstream regulator of COX-2, we further checked the effect of EGCG on the levels of NF-κB/p65 in melanoma cells using western blot analysis. EGCG inhibits the activation of NF-κB/p65 in a dose-dependent manner. Caffeic acid phenethyl ester, an inhibitor of NF-κB, inhibits melanoma cell migration. These observations further support the hypothesis that the inhibitory effect of EGCG on melanoma cell migration is mediated, at least in part, through the downregulation of COX-2, PGE2 and PGE2 receptors. However, it remains to be examined whether down-regulation of other NF-κB target genes also contribute to the inhibition of invasive potential of melanoma cells.
NF-κB regulates a wide spectrum of biological processes, including inflammation, cell proliferation and apoptosis. Additionally, NF-κB has a role in tissue invasion, cancer cell migration and metastasis. Importantly, NF-κB has been identified as an important regulator of EMT in several cancer cell types –. EMT has been shown to play a major role in invasion and metastasis of epithelial tumors. EMT can render tumor cells migratory and invasive following its effect on all stages, which includes invasion, intravasation and extravasation .
During EMT, cells can change from an epithelial to a mesenchymal state. They lose their characteristic epithelial traits and instead acquire properties of mesenchymal cells. This process is primarily coordinated by the disappearance or loss of epithelial biomarkers such as E-cadherin and certain cytokeratins with the concomitant appearance of mesenchymal markers such as vimentin, fibronection and N-cadherin, etc. In the present study, we found that treatment of melanoma cells with EGCG resulted in suppression of mesenchymal biomarkers, such as vimentin, fibronectin and N-cadherin while restored the levels of epithelial biomarkers such as, E-cadherin, keratin and desmoglein 2, in melanoma cells which suggest that EGCG has the ability to transform mesenchymal characteristics to epithelial characteristics in melanoma cancer cells and this transition may also be one of the possible mechanisms through which EGCG reduce the invasiveness of melanoma cells and that lead to inhibition of migration of melanoma cells in our system.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
Together, the results from this study have identified for the first time that EGCG, a major component of green tea catechins or polyphenols, inhibit the invasive potential of melanoma cells and that involves: (i) the inhibitory effect of EGCG on endogenous COX-2 expression and successive down-regulation of PGE2 and PGE2 receptors, (ii) the inhibitory effect of EGCG on the activation of NF-κB/p65, which is the upstream regulator of COX-2, and (iii) the mesenchymal-to-epithelial transition. Further mechanism-based in vivo studies are required which can establish the importance of EGCG and its development as a pharmacologically safe non-toxic agent for the treatment of malignant melanoma by using either alone or in combination with other phytochemicals or anti-metastatic drugs.[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
Competing Interests: The authors have declared that no competing interests exist.
Funding: This work was supported by funds from the 5RO1 AT002536 by National Center for Complementary and Alternative Medicine, NCCAM/NIH (SKK) (http//www.cancer.gov), and Veterans Administration Merit Review Award (SKK). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
# References
*In the **References** section:*
|
# Introduction
Caveolin 3, Flotillin 1 and Influenza Virus Hemagglutinin Reside in Distinct Domains on the Sarcolemma of Skeletal Myofibers
# Abstract
*In the **Abstract** section:*
We examined the distribution of selected raft proteins on the sarcolemma of skeletal myofibers and the role of cholesterol environment in the distribution. Immunofluorescence staining showed that flotillin-1 and influenza hemagglutinin [exhibited r](https://www.ncbi.nlm.nih.gov/mesh/D002784)afts that located in the domains deficient of the dystrophin glycoprotein complex, but the distribution patterns of the two proteins were different. Cholesterol depletion from the sarcolemma by means of methyl-β-cyclodextrin resulted in distorted caveolar morphology and red[istribution](https://www.ncbi.nlm.nih.gov/mesh/D002784) of the caveolin 3 protein. Concomitantly, [the water permeabilit](https://www.ncbi.nlm.nih.gov/mesh/C108732)y of the sarcolemma increased significantly. However, cholesterol depletion did not reshuffle flotillin 1 or[ hema](https://www.ncbi.nlm.nih.gov/mesh/D014867)gglutinin. Furthermore, a hemagglutinin variant that lacked a raft[-targeting ](https://www.ncbi.nlm.nih.gov/mesh/D002784)signals exhibited a similar distribution pattern as the native raft protein. These findings indicate that each raft protein exhibits a strictly defined distribution in the sarcolemma. Only the distribution of caveolin 3 that binds cholesterol was exclusively dependent on cholesterol environment.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
## 1. Introduction
*In the **1. Introduction** section:*
Skeletal myofibers are unique cells that have a large plasma membrane separated into sarcolemma and transverse tubules. The dystrophin glycoprotein complex (DGC) distributes at the sarcolemma in an organized fashion forming a cross-striated pattern over the I-bands and M-lines of the underlying myofibrils and longitudinal stripes over the I-A junctional areas [1, 2]. DGC is thought to stabilize the sarcolemmal membrane that is subject to mechanical stress during muscle contractions. Accordingly, inherited defects in the components of the DGC and especially in dystrophin are manifested in various types of dystrophia diseases [3].
Interestingly, many endogenous sarcolemmal proteins such as Na,K-ATPase [2], transferrin receptor [4], caveolin 3 (cav 3) [4], chloride channel ClC-1 [5], and aquaporin 4 [6] occupy the areas covered by the DGC.
Caveolae pit structures that contain cav 3 protein also locate to the DGC regions. Several viral model glycoproteins also show a specific localization pattern in relation to the DGC mosaic [4].
Studies with mononucleated cells and giant plasma membrane vesicles have revealed that caveolins and several lipid-anchored proteins show a strong preference to cholesterol and sphingolipid-enriched lipid islets called rafts [7, 8]. These microdomains have been reported to exist as dispersed units of nanometer scale, but they may also form large clusters of several micrometers in size that lack nonraft proteins [9–12]. The integrity of rafts is crucially dependent on cholesterol.
Accordingly, cholesterol depletion by a cholesterol-sequestering drug, methyl-β-cyclodextrin (CDX), inhibits patching of antibody cross-linked raft proteins and increases the diffusion rates of certain raft-associated proteins [10, 13, 14].[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
Since the sarcolemma of myofibers comprises a regularly repeated two-domain mosaic and sarcolemmal nonraft proteins have been shown to comply with this pattern, we examined here whether the distribution patterns of raft proteins were domain specific. We found that contrasting the behavior of cav 3, flotillin 1, and the classical raft marker influenza virus hemagglutinin (HA) localized into the domains deficient of DGC. We next investigated whether the cholesterol environment affected the distribution of raft proteins between the sarcolemmal domains. Interestingly, the localization pattern of the caveolar protein cav 3 was altered upon cholesterol depletion, and the caveolar pits were deformed or destroyed. However, depletion of cholesterol did not reshuffle flotillin 1 or HA.
Furthermore, removal of the raft-targeting signals from HA indicated that the domain-specific localization was not dependent on the raft-targeting signals. While cholesterol was important to locate cav 3 into caveolae, it did not affect the distribution of flotillin 1 or HA on the sarcolemma.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
## 2. Methods
*In the **2. Methods** section:*
## 2.1. Isolation and Cultivation of Myofibers
*In the **2.1. Isolation and Cultivation of Myofibers** section:*
Myofibers were isolated from the flexor digitorum brevis (FDB) muscle of three-month-old female rats by using collagenase digestion as described [15]. The isolated myofibers were mounted on cell culture dishes coated with Matrigel (Becton-Dickinson Biosciences, Franklin Lake, NJ, USA) and cultivated in minimal essential medium (MEM) supplemented with 5% horse serum, L-glutamine and penicillin-streptomycin. For measurements of live cells, the isolated myofibers were mounted on Matrigel-coated glass-bottom dishes (WillcoWells, Amsterdam, Netherlands).[](https://www.ncbi.nlm.nih.gov/mesh/D005973)
## 2.2. CDX Treatments
*In the **2.2. CDX Treatments** section:*
Treatments with CDX (Sigma-Aldrich, St. Louis, Mo. USA) were performed in MEM containing L-glutamine, penicillin-streptomycin, and 0.05% (v/v) heat-inactivated and lipoprotein-depleted horse serum. The serum was delipidified by sequential ultracentrifugation prior to use to remove lipoprotein components, as described by Goldstein et al. [16]. Prior to the CDX treatments, the myofiber cultures were washed two times with MEM. The fibers were incubated in the presence of appropriate CDX concentrations in a volume of 0.5 mL for 1 h at 37°C.[](https://www.ncbi.nlm.nih.gov/mesh/C108732)
## 2.3. Detergent Extractions
*In the **2.3. Detergent Extractions** section:*
Isolated myofibers were scraped from the culture dishes and pelleted by centrifugation at 200 g for 2 min in a table top centrifuge. After that, the myofiber pellets were suspended in cold phosphate buffered saline (PBS) containing 1% Triton X-100 and complete protease inhibitor cocktail (Roche, Basel, Switzerland) and incubated for 10 min in an ice bath. The detergent-insoluble proteins were pelleted by centrifugation at 75,000 g for 2.5 h at 4°C. Proteins in the supernatants were precipitated with 10% TCA. Proteins in the pellets and the supernatants were separated with SDS/PAGE followed by western blotting using rabbit antiflotillin 1 (Sigma-Aldrich) as primary antibody and peroxidase-conjugated anti-rabbit IgG (Bio-Rad Laboratories) as secondary antibodies.
Detection was with chemiluminescence detection reagent (GE Healthcare) using Hyperfilm (GE Healthcare).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## 2.4. Hypotonic Swelling Experiments
*In the **2.4. Hypotonic Swelling Experiments** section:*
Freshly isolated myofibers on glass-bottom dishes were either treated or not treated with CDX and then incubated in the presence of 10 mM fluorophore Calcein-AM (Invitrogen, Eugene, OR, USA) for 5 min at 37°C. Thereafter, the culture was washed two times with MEM and covered with an additional layer of Matrigel to prevent detachment of myofibers during subsequent treatments as described in [6]. The myofibers were examined with Zeiss LSM510 confocal microscope (Carl Zeiss, Göttingen, Germany), and those showing a homogenous fluorescence were chosen for measurements. The region of interest was the entire myofiber, and the focal plane was set at the core region. A 60 s recording was performed before the hypotonic shock that was induced by changing isotonic PBS (300 mosM) to hypotonic (150 mosM).
Changing the medium resulted in a variable amount of movement of the myofiber being measured. Therefore, the region of interest was manually repositioned to correspond to the new location of the myofiber. All the recordings were performed at 23 ± 1°C by measuring the mean fluorescence intensity at 2 s intervals. The fluorescence intensity changes were analyzed by using Zeiss LSM510 Pascal software, and the measurement covered the entire fiber. Since a small fraction of the FDB myofibers contains aquaporin 4 [6], the presence of aquaporin 4 was determined by immunofluorescence staining after the recordings.
Measurements from myofibers containing aquaporin 4 were rejected since aquaporin 4 is a water channel.[](https://www.ncbi.nlm.nih.gov/mesh/C108732)
## 2.5. Electron Microscopy
*In the **2.5. Electron Microscopy** section:*
For immunolocalization studies, extensor digitorum longus (EDL) muscle was dissected and fixed with 4% paraformaldehyde in 0.1 M phosphate buffer for 1 h, sliced in small pieces and placed in 2.3 M sucrose overnight. Alternatively, isolated FDB myofibers were fixed with 4% paraformaldehyde for 1 h in 0.1 M phosphate buffer containing 73 mM sucrose, and then detached from the culture dishes and centrifuged at 2000 g for 1 min. After washings with PBS, the pellets were suspended in 12% gelatin at 37°C, centrifuged at 3000 g for 5 min, and chilled in an ice bath for 30 min. The congealed gelatin was sliced in small pieces and placed in 2.3 M sucrose overnight. The pieces of EDL or the gelatin-embedded FDB myofibers were frozen with liquid nitrogen, and thin sections (200 nm) were cut. The sections were first incubated in 50 mM glycine in PBS and then in 5% bovine serum albumin (BSA) supplemented with 0.1% cold water fish skin gelatin (Aurion, Wageningen, Netherlands) in PBS to block nonspecific binding.
Antibodies were diluted with 0.1% BSA-C (Aurion) in PBS, and the incubations were for 1 h at 37°C. Rabbit antimouse secondary antibodies (Zymed Laboratories) and protein A-gold complex [17] were used for detection. Sections were examined with Philips CM100 transmission electron microscope.[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
For conventional transmission electron microscopy, the isolated myofibers were fixed with 1.5% glutaraldehyde in PBS. The fixed fibers were scraped and pelleted by centrifugation and immersed in 12% agarose. Osmium tetroxide (1%) was used for postfixation, and the myofibers were embedded in Epon LX 112 (Ladd Research Industries, Burlington, Vt. USA). Thin sections (150 nm) were examined with the Philips CM100 electron microscope. The number of caveolae/μm of the sarcolemma was calculated using 4–7 photographed fields/fiber.
Image analysis was performed with UTHSCSA Image Tool for Windows version 3.[](https://www.ncbi.nlm.nih.gov/mesh/D005976)
## 2.6. Recombinant Viruses and In Vitro Mutagenesis
*In the **2.6. Recombinant Viruses and In Vitro Mutagenesis** section:*
Preparation of recombinant Semliki Forest Virus (SFV) particles encoding HA Japan/A/305/57 [Genbank: DQ508841.1] has been described previously [18]. To generate SFV particles encoding a mutant HA lacking all known raft-targeting signals, we used the mutant 2A511 HA described by Scheiffele et al. [19]. The cDNA of the mutant 2A511 HA in pSFV vector was subjected to in vitro mutagenesis to change the triplets encoding cysteine 536 at the C-terminal end of the transmembrane domain and cysteines 543 and 546 in the C-terminal tail into triplets encoding serines, to prevent palmitoylation of the protein [7]. The in vitro mutagenesis was performed by using the QuickChange site directed in vitro mutagenesis kit (Stratagene, La Jolla, CA, USA).
That the mutated product had the desired sequence was verified with ABI PRISM 3130XL sequencer and BigDye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems Inc., Foster City, CA, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
The isolated myofibers were infected with the recSFVs by applying viral stock medium into the culture medium at 1 : 3 dilution. The infection was allowed to proceed for 16–24 h at 37°C.
## 2.7. Immunohistochemistry
*In the **2.7. Immunohistochemistry** section:*
Isolated myofibers were fixed with 3% paraformaldehyde in PBS for 10 min. After permeabilization with 1% Triton X-100, the nonspecific binding was blocked with 1% BSA for 10 min. Primary antibodies were applied for 30 min at 37°C or 2 h at room temperature. The primary antibodies used were rabbit antiflotillin 1 (Sigma-Aldrich), mouse anti-β-dystroglycan (Novocastra Laboratories Ltd, Benton Lane, UK), mouse anti-cav 3 (Becton Dickinson, Franklin Lakes, NJ, USA), and mouse antidihydropyridine receptor (DHPR) (Affinity Bioreagents Inc., Golden, CO, USA). Secondary antibodies were Alexa 488-conjugated anti-rabbit IgG (Invitrogen) or Alexa 568-conjugated anti-mouse IgG (Invitrogen), and incubations lasted for 30 min at 37°C.
HA on the sarcolemma was detected by adding rabbit anti-HA antiserum [19] into the culture medium at 1 : 100 dilution. Incubation was for 1.5 h at 10–12°C followed by two washes with PBS and fixation with paraformaldehyde. Then the Alexa 488-conjugated anti-rabbit IgG was applied and incubated for 30 min at 37°C. The myofibers were next permeabilized by a 5-minute treatment with 1% Triton X-100 in PBS and processed for double immunofluorescence staining for β-dystroglycan.
Negative controls for all antibodies were performed by omitting the primary antibody, and these were blank. Samples were examined with Zeiss LSM510 confocal microscope.[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
## 2.8. Statistics
*In the **2.8. Statistics** section:*
All the data are expressed as mean ± SD, and n indicates the number of determinations. Two-sample t-test was used to compare two groups, and one-tailed hypothesis testing was used to determine P values. P < 0.05 was considered statistically significant.
## 3. Results
*In the **3. Results** section:*
## 3.1. Flotillin 1 and Cav 3 Reside in Separate Membrane Microdomains
*In the **3.1. Flotillin 1 and Cav 3 Reside in Separate Membrane Microdomains** section:*
The flotillin rafts are distinct from caveolae in mononucleated cells [20] in which the flotillin microdomains can exist in either flat or invaginated state [21]. Here, we examined whether flotillin 1 microdomains in skeletal muscle cells were distinct from the caveolae that contain cav 3. For this purpose, we performed double immunofluorescence staining for the two proteins in isolated myofibers that provide a view over the muscle cell surface. Figures 1(a)–1(c) show that flotillin 1 appeared as clusters at the A-band regions in the domains deficient of DGC. These domains are lacking cav 3 [4].
Since we found flotillin 1 in the regions deficient of DGC that are known to harbor transverse tubule openings [4], we next performed double staining for flotillin 1 and the transverse tubule marker DHPR. Confocal sectioning indicated that flotillin 1 staining flanked that of the DHPR staining (Figures 1(d)–1(f)), suggesting that the two proteins were close to each other but did not overlap. A longitudinal confocal section through a myofiber (Figures 1(g)–1(i)) suggests that flotillin 1 marked structures in the transverse tubule neck portions. Supporting this, immunoelectron microscopy studies consistently indicated that flotillin 1 located in structures 50–100 nm beneath the sarcolemmal membrane. Figure 1(j) shows an example.
## 3.2. Cholesterol Depletion Partially Reshuffles Cav 3 and Deforms Caveolae but Does Not Affect Flotillin 1 Distribution
*In the **3.2. Cholesterol Depletion Partially Reshuffles Cav 3 and Deforms Caveolae but Does Not Affect Flotillin 1 Distribution** section:*
We have previously reported that cav 3 disappears from the sarcolemma upon cholesterol depletion [6]. The finding that flotillin and cav 3 seem to reside in discrete structures prompted us to investigate whether depletion of cholesterol affected the distribution pattern of flotillin 1, too. We found that the immunofluorescence staining pattern of flotillin 1 remained unchanged after CDX treatment of the isolated myofibers. The intensity of flotillin immunostaining varied considerably from fiber to fiber; however, a systematic intensity change could not be observed upon CDX treatment. The intensity of the cav 3 staining on the sarcolemma was reduced in variable extent.
A salient feature after the CDX treatment was that rows of spots of cav 3 staining appeared beneath the sarcolemma (Figure 2(e)). The staining pattern mimicked that of DHPR suggesting localization of cav 3 in transverse tubules. Immunogold labeling verified that after cholesterol depletion cav 3 was abundant especially at the neck portions of the transverse tubules. Figures 2(h)–2(i) summarize these results.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
We next subjected cultured myofibers to various concentrations of CDX followed by extraction with cold Triton X-100. We found that flotillin 1 was sparingly soluble in Triton X-100 (soluble fraction was 21.5 ± 3.7%, n = 2), and surprisingly, CDX treatment only slightly increased its detergent solubility (3 mM CDX: 30.4 ± 5.6%, n = 3; 5 mM CDX: 31 ± 6.4%, n = 3). Similar analysis was also performed for cav 3, indicating that CDX treatment did not increase the solubility of the protein in Triton X-100. Figure 3 shows an example of the results. Both flotillin 1 as well as cav 3 floated in sucrose gradients, indicating that the insolubility was due to association with rafts.
These findings suggest that flotillin 1, like cav 3, resides in a very compactly packed lipid environment.[](https://www.ncbi.nlm.nih.gov/mesh/C108732)
Since cav 3 disappears from the sarcolemma upon CDX treatment, we next examined whether caveolae pits disappeared. Transmission electron microscopy studies of myofibers after CDX treatment indicated that, in comparison to the normal morphology of caveolae (Figure 4(a)), deformation occurred at 1 mM concentration of the drug (Figure 4(b)). Furthermore, the number of caveolae was reduced by about 50% in CDX-treated myofibers (2.9 ± 0.34 caveola/μm, n = 5 photographs) as compared to the controls without any drug treatment (5.9 ± 0.01 caveola/μm, n = 2). Increasing the CDX concentration to 5 mM resulted in destruction of the caveolar morphology (Figure 4(c)). These findings are compatible with those obtained with nonmuscle cells
[22].[](https://www.ncbi.nlm.nih.gov/mesh/C108732)
In addition to the morphology of caveolae, cholesterol has an impact on the water permeability of membranes [23]. This question is especially important with regard to the sarcolemma, owing to the fact that cholesterol-lowering medication has adverse effects on skeletal muscle. We therefore investigated whether the cholesterol depletion altered the swelling response of myofibers under hypotonic conditions. Interestingly, we found that after switching from isotonic to hypotonic osmolarity the intensity of calcein-AM reduced significantly more in the cholesterol-depleted myofibers as compared to control myofibers (Figure 5). This suggests that the sarcolemma became more permeable to water.
The effect was repeatable with 2 mM CDX, but at higher drug concentrations, a major fraction of the myofibers became leaky for the fluorophore as indicated by a slight decrease of the fluorescence intensity during the prerecording period before the hypotonic shock. The myofibers which remained intact after treatment with 3 mM CDX, however, did not show any additional increase in water permeability as compared with the myofibers treated with 2 mM drug concentration.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
## 3.3. The Existence and Localization of Influenza HA Microdomains Are Not Dependent on the Association with Rafts
*In the **3.3. The Existence and Localization of Influenza HA Microdomains Are Not Dependent on the Association with Rafts** section:*
The influenza HA is a viral model glycoprotein that was shown to acquire insolubility in detergents upon travelling from the Golgi apparatus to the plasma membrane [24]. Recent superresolution microscopy observations show that HA at the plasma membrane associates with microdomains ranging from 40 nm to irregularly shaped clusters of up to several microns [12]. Here, we utilized the influenza HA and its variants that are excluded from rafts, in order to further analyze whether raft association played a role in protein targeting to the sarcolemmal domains. We first expressed the raft-associated native HA in isolated myofibers and analyzed its localization on the sarcolemma in relation to the DGC.
In order to avoid interference caused by influenza virions that bind to the sarcolemma and the interactions of the HA with other viral proteins, we expressed HA by using recSFV instead of the parent influenza virus. We found that during the propagation of the infection, the HA protein appeared as small dots of uniform size. Some dots located in the domains devoid of DGC, but the majority located to the domain borderlines thus flanking the DGC domains (Figure 6(a)). Accordingly, this localization pattern is clearly distinct from that of caveolae, and it also seems different from that of flotillin 1 that exhibited clusters of irregular shapes and had a tendency to mark the central areas of the domains lacking DGC.
Since the anti-HA antibodies were applied to live myofibers, it is possible that the dots seen were induced by antibody cross-linking. To exclude this possibility, we performed fixing with paraformaldehyde followed by immunofluorescence staining without permeabilization. Fixation prior to the application of antibodies slightly permeabilized the sarcolemma and caused some background staining; however, dots of HA similar to those seen when antibodies were applied before fixation were observed (Figure 6(g)), locating to the DGC-deficient areas but flanking the DGC areas. Upon further propagation of the infection there appeared HA dots also within the domain occupied by the DGC (Figure 6(d)). This was only seen in myofibers exhibiting very intense fluorescence, suggesting overexpression conditions.[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
In order to approach the question whether it was the association with rafts that determined the localization of HA, we first treated the myofibers with CDX to deplete cholesterol. This treatment had no effect on the distribution of the protein (Figure 6(e)). Next we expressed the variant HA (2A511) that lacked the raft-targeting signals, namely specific hydrophobic amino acids in the transmembrane domain [19]. This mutant has been shown to be detergent soluble in myotubes [18]. As the wild-type HA, the variant HA was expressed using recSFV, and it was found to appear as dots and to localize to the domain borderlines.
Finally, we expressed the 2A511 mutant that also lacked palmitoylation sites [7] and found it to behave like the native HA (Figure 6(f)). We did not observe any difference in the dot size between the raft-associated and nonraft-associated forms of HA. Taken together, these findings imply that lipid environment had no crucial role in the targeting of HA.[](https://www.ncbi.nlm.nih.gov/mesh/C108732)
## 4. Discussion
*In the **4. Discussion** section:*
Previous studies have indicated that several sarcolemmal proteins reside within the DGC domains of the sarcolemmal mosaic [2, 4–6]. Here, we analyzed the distribution pattern of proteins which are associated with sphingolipid- and cholesterol-based microdomains, called rafts, and whether the raft association dictated their distribution patterns. When investigating the localization of three selected raft proteins on the sarcolemma, we found three different types of organization patterns which were dependent on the organization of the DGC. The caveolae rafts harboring cav 3 resided within the DGC domains, and their integrity was strictly dependent on cholesterol. There were clusters of raft proteins, namely flotillin 1 and HA also in the domains lacking DGC.
In this respect, the situation resembles that in the polarized epithelial cells in which there are rafts both in the apical as well as the basolateral domains [25–27]. It seems that the integrities of the flotillin and HA protein clusters were not dependent on the association with a lipid-based raft platform.[](https://www.ncbi.nlm.nih.gov/mesh/D013107)
The omega-shaped invaginations called caveolae comprise a special type of rafts. In skeletal myofibers, these caveolae are located at the sarcolemma and colocalize with the DGC at I-band regions [4]. That the caveolae structures and the localization of cav 3 were dependent on cholesterol is explained by the fact that cav 3 is a cholesterol-binding protein. Abolishing this interaction results in increased mobility of the protein, and it was shifted into the transverse tubules. Previous studies have localized small amounts of cav 3 in the transverse tubules [28]; however, the antibodies we used did not recognize this component.
We presume that relatively large amounts of cav 3 shifted to the transverse tubules upon cholesterol depletion whereby it became recognizable to the present antibodies. The fact that cholesterol depletion did not totally redistribute cav 3 is compatible with the idea that this protein is partially bound to DGC via β-dystroglycan [29]. A curved membrane may bring cav 3 in the vicinity of the membrane surrounding the caveolae, allowing the two proteins to interact in this restricted area. It is also important to note that the cholesterol depletion did not change the distribution pattern of β-dystroglycan although this protein has been previously suggested to reside in rafts [30].[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
A recent study has posed that the cav 3 protein residing in caveolae also localizes to the DGC-deficient areas at the necks of the transverse tubules [31]. Our results are at variety with this finding since cav 3 appeared to accumulate into transverse tubule necks upon cholesterol depletion only. Instead, we found that flotillin 1 that structurally resembles caveolins was exclusively located in the DGC-deficient membrane domains. Interestingly, there are caveolae-like flasks attached to the necks of the transverse tubules protruding to the sarcolemma
[4], and our results suggest that flotillin 1 is present in those structures. Accordingly, electron microscopic immunogold labeling experiments (Figure 1(j)), although suffering from low labeling intensity, support the idea of flotillin 1 locating to the flask structures at the transverse tubule openings. In mononucleated cells, flotillin 1 forms flask structures together with flotillin 2 [21]. Flotillin 2 is not expressed in skeletal muscle cells [32]; however, there may still be an unknown partner for flotillin 1 to form flask structures. It is notable that the glycophosphatidylinositol-anchored carbonic anhydrase IV was recently localized to transverse tubule openings in mouse skeletal myofibers
[33], and it is possible that this protein and flotillin 1 reside in the same lipid rafts.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
We found that the distribution patterns of flotillin 1 and HA did not change upon cholesterol removal. This is interesting since the organization of rafts and the spatial distribution of some raft-associated proteins are crucially dependent on cholesterol in mononucleated cells. Accordingly, removal of cholesterol has been shown to increase the mobility of raft-associated membrane proteins and to disrupt the organization of microdomains harboring proteins anchored by glycophosphatidyl inositol, resulting in intermixing of different types of rafts [13, 14, 34, 35]. It is notable that cholesterol depletion did not markedly render flotillin 1 soluble in Triton X-100, suggesting an especially rigid lipid organization such as suggested by Ilangumaran and Hoessli
[36]. Alternatively, flotillin-1 was bound to subsarcolemmal structures which retain the microdomains in position even upon cholesterol depletion. Our results show that, in addition to the changes in the morphology of caveolae, the barrier function of the plasma membrane is altered as manifested by increased water permeability when cholesterol level is lowered.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
Segregation of proteins in liquid-ordered microdomains takes place by means of several known raft-targeting signals [7, 8, 19]. It is possible that removal of these signals leads to diverse distribution patterns of raft-associated proteins on the cell surface. Indeed, it has been demonstrated that raft proteins are transported to the plasma membrane in specific transport vesicles in epithelial cells and also in yeast [37, 38].
We used the prototype raft protein influenza HA and found that HA appeared as small dots, the size of which apparently was below the resolution limit (<200 nm) of the confocal microscope. The dots located to the borderlines of the sarcolemmal domains, but during overexpression conditions, there were low-intensity dots also within the DGC domains. Compatible with this, a previous study showed that influenza virions were budding at the DGC domains during influenza virus infection that typically results in overexpression [4]. It is notable that removal of all known raft-targeting signals from HA did not change its distribution pattern. Importantly, the dot-like appearance of HA did not change, either.
It thus seems that HA molecules form small aggregates both in raft and nonraft environments, and these locate to the DGC borderlines, and their size is not dependent on the lipid environment. This idea is compatible with findings of the recent study in which the clustering of the transmembrane domains of HA was found to take place independently of raft association [39]. Regarding the localization, it is possible that anchoring to the subsarcolemmal cytoskeleton plays a role. It is notable that the native transmembrane proteins aquaporin 4 and Na,K-ATPase are associated with the DGC domains via binding to subsarcolemmal α-syntrophin and ankyrin-3, respectively [6, 40, 41]. Taken together, we have shown that raft proteins are not evenly distributed in the sarcolemma but locate in discrete regions within the domains defined by DGC.
However, the role of the lipid platform in dictating the domain-specific distribution of the proteins varied. It remains to be determined whether the binding of the cytoplasmic tails to the subsarcolemmal cytoskeleton was responsible for the domain-specific partitioning.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
Flotillin 1 resides in the DGC-deficient regions in structures near transverse tubule openings. A confocal section at the sarcolemma level indicates that flotillin 1 (a) appears as spots of irregular shape. Double staining for β-dystroglycan (b) and the merged image (c) show that dystroglycan and flotillin 1 do not colocalize. A confocal section at the sarcolemma level indicated that flotillin 1 (d) and DHPR (e) that mark transverse tubules are flanking each other, demonstrated in the merged image (f). A longitudinal confocal section in the middle of a myofiber indicates that flotillin 1 (g) exhibits intense staining that seems to be an extension of the DHPR staining (h) as demonstrated in the merged image (i). Immunoelectron microscopy for flotillin 1 in EDL section shows gold particles (arrowhead) in a structure beneath the sarcolemma near the transverse tubule neck (arrow) (j). Scale bars, 2 μm (a–i); 250 nm (j).[](https://www.ncbi.nlm.nih.gov/mesh/D006046)
Cav 3 and flotillin 1 reside in different types of rafts. Cav 3 (red) normally exhibits a cross-striated distribution pattern over the I bands, while flotillin 1 (green) locates to the A-band regions lacking cav 3 (a). A confocal section 5 μm beneath the sarcolemma shows neither flotillin 1 nor cav 3 (b). A confocal section at the middle of the myofiber shows flotillin and cav 3 on the sarcolemma exclusively (c). After treatment with 5 mM CDX, the sarcolemmal staining patterns retain (d), however, a confocal section 5 μm beneath the sarcolemma of the CDX-treated myofiber reveals double rows of dots of cav 3 (e). Some subsarcolemmal dots are also visible in the middle section (f). Scale bars, 2 μm. The flotillin and cav 3 staining patterns were abolished upon omitting the primary antibodies (g). EM immunogold labeling of FDB cryosections shows particles indicating the presence of cav 3 predominantly in the sarcolemma in myofibers not treated with CDX (h). In FDB myofiber treated with 5 mM CDX, the immunogold particles are associated with the transverse tubules, marked by dotted lines (i). Scale bar in (h, i) 250 nm.[](https://www.ncbi.nlm.nih.gov/mesh/C108732)
Flotillin 1 is more sparingly soluble in cold Triton X-100 than cav 3. Isolated myofibers were treated with 0, 3, and 5 mM CDX and then extracted with 1% Triton X-100. Soluble material (S) and pellets (Ps) were subjected to SDS/PAGE and western blotting using specific antibodies. Treatment of the myofibers with CDX only slightly increased the solubility of flotillin 1 to the detergent, whereas the solubility of cav 3 remained unaffected.[](https://www.ncbi.nlm.nih.gov/mesh/D017830)
Cholesterol depletion destroys the morphology of caveolae. (a) In an intact FDB myofiber, there are abundantly flask-shaped caveolae that form rosettes (marked by arrowheads) beneath the sarcolemma. (b) Treatment with 1 mM CDX reduces the number of caveolae, and the morphology of individual caveolae (arrowheads) is distorted. (c) Treatment with 5 mM CDX results in flattened caveolae without any distinguishable morphology. Scale bar, 500 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
Cholesterol depletion increases the water permeability of isolated myofibers. (a) Fluorescence micrographs of a calcein-loaded myofiber before (left) and 140 s after switching to the hypotonic buffer (right) that indicates decreased fluorescence. The myofiber was not treated with CDX. Scale bar, 50 μm. (b) Examples of the relative fluorescence (fluorescence at certain time point divided by the fluorescence at the beginning of the experiment) recordings for myofibers treated or not treated with CDX are shown. A sharp decline of the fluorescence intensity was often seen upon switching to the hypotonic buffer, owing to sample movement. (c) The change of fluorescence intensity is presented as a percentage between the value of fluorescence intensity in isotonic buffer just before and 140 s after switching from isotonic to hypotonic condition. N = 14 (control) and 16 (CDX-treated) myofibers from five different isolations. ***P < 0.001.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
HA localizes to the DGC-deficient domains independently of raft association. Double immunofluorescence staining for the sarcolemmal wild-type HA (green) and β-dystroglycan (red) is shown at the sarcolemma level (a) and at the middle of the fiber (b). Double immunofluorescence staining for a noninfected control shows no HA (c). At high expression level, HA occupies regions of β-dystroglycan (DGC-rich domain) (d). Treatment with 5 mM CDX does not change the distribution pattern of wild-type HA and β-dystroglycan (e). The distribution pattern of the nonraft-associated mutant HA lacking the palmitoylation sites and transmembrane raft-targeting signals also remained unchanged as shown by double staining for β-dystroglycan (f). Primary antibody incubation was performed at 10–12°C for 1.5 h before the fixation with 3% paraformaldehyde (a–f). Fixation of the myofiber before application of the anti-HA antibodies did not change the distribution pattern (g). The pictures have undergone brightness and contrast adjustment. Scale bar, 2 μm.[](https://www.ncbi.nlm.nih.gov/mesh/C108732)
|
# Introduction
Modulation Role of [Abscisic Acid](https://www.ncbi.nlm.nih.gov/mesh/D000040) ([ABA](https://www.ncbi.nlm.nih.gov/mesh/D000040)) on Growth, Water Relations and [Glycinebetaine](https://www.ncbi.nlm.nih.gov/mesh/D001622) Metabolism in Two Maize (Zea mays L.) Cultivars under Drought Stress
# Abstract
*In the **Abstract** section:*
The role of plant hormone abscisic acid (ABA) in plants under drought stress (DS) is crucial in modulating physiological responses that eventually lead to adaptation to an unfavora[ble environme](https://www.ncbi.nlm.nih.gov/mesh/D000040)nt[; h](https://www.ncbi.nlm.nih.gov/mesh/D000040)owever, the role of this hormone in modulation of glycinebetaine (GB) metabolism in maize particularly at the seedling stage is still poorly understood. Some hydroponic experiments were conducted to inves[tigate the mod](https://www.ncbi.nlm.nih.gov/mesh/D001622)ul[at](https://www.ncbi.nlm.nih.gov/mesh/D001622)ion role of ABA on plant growth, water relations and GB metabolism in the leaves of two maize cultivars, Zhengdan 958 (ZD958; drought tolerant), and Jundan 20 (JD20; [dro](https://www.ncbi.nlm.nih.gov/mesh/D000040)ught sensitive), subjected to integrat[ed](https://www.ncbi.nlm.nih.gov/mesh/D001622) root-zone drought stress (IR-DS) simulated by the addition of polyethylene glycol (PEG, 12% w/v, MW 6000). The IR-DS substantially resulted in increased betaine aldehyde dehydrogenase (BADH) activity and choline cont[ent which act as th](https://www.ncbi.nlm.nih.gov/mesh/D011092)e [key](https://www.ncbi.nlm.nih.gov/mesh/D011092) enzyme and initial substrate, respectively, in GB biosynthesis. Drought stress also induced accumulation of GB, where[as it c](https://www.ncbi.nlm.nih.gov/mesh/D002794)aused reduction in leaf relative water content (RWC) and dry matter (DM) in b[ot](https://www.ncbi.nlm.nih.gov/mesh/D001622)h cultivars. The contents of ABA and GB increased in drough[t-](https://www.ncbi.nlm.nih.gov/mesh/D001622)stressed maize seedlings, but ABA accumulated prior to GB accumulation under the drought treatment. These responses were m[ore](https://www.ncbi.nlm.nih.gov/mesh/D000040) pred[om](https://www.ncbi.nlm.nih.gov/mesh/D001622)inant in ZD958 than those in JD20. Addition of exoge[nou](https://www.ncbi.nlm.nih.gov/mesh/D000040)s ABA and fluridone (F[lu](https://www.ncbi.nlm.nih.gov/mesh/D001622)) (ABA synthesis inhibitor) applied separately increased and decreased BADH activity, respectively. Abscisic acid application enhanc[ed ](https://www.ncbi.nlm.nih.gov/mesh/D000040)GB ac[cumulatio](https://www.ncbi.nlm.nih.gov/mesh/C013351)n,[ le](https://www.ncbi.nlm.nih.gov/mesh/C013351)af [RWC](https://www.ncbi.nlm.nih.gov/mesh/D000040) and shoot DM production in both cultivars. However, of both maize cultivars, the drought sens[itive maize c](https://www.ncbi.nlm.nih.gov/mesh/D000040)ultivar (JD20) performed relatively better than the other maize cultivar ZD958 under both ABA and Flu application in view of all parameters appraised. It is, therefore, concluded that increase in both BADH activity and choline content possibly [res](https://www.ncbi.nlm.nih.gov/mesh/D000040)ulted[ in](https://www.ncbi.nlm.nih.gov/mesh/C013351) enhancement of GB accumulation under DS. The endogenous ABA was probably involved in the regulation of GB metabolism [by regu](https://www.ncbi.nlm.nih.gov/mesh/D002794)lating BADH activity, and resulting in modula[ti](https://www.ncbi.nlm.nih.gov/mesh/D001622)on of water relations and plant growth [und](https://www.ncbi.nlm.nih.gov/mesh/D000040)er drought, especially in the drought sensit[iv](https://www.ncbi.nlm.nih.gov/mesh/D001622)e maize cultivar JD20.
## 1. Introduction
*In the **1. Introduction** section:*
Plants are frequently exposed to a variety of abiotic stresses such as drought stress (DS), which hamper plant growth and crop productivity worldwide. Drought stress (DS) causes considerable yield reduction in most crops including maize. Maize (Zea mays L.) is an important cereal crop in northern China which is sensitive to drought. Understanding how plants tolerate these stresses is a prerequisite for developing strategies to improve plant stress tolerance.
Plants sense and adapt to different stresses by altering their physiological metabolism, and growth pattern, and mobilizing various defense mechanisms. Therefore, accumulation of osmolytes is a prerequisite for osmotic adjustment of all organisms under DS. It is well established that glycinebetaine (GB) accumulates in plants during their adaptation to various types of environmental stresses including drought. Glycinebetaine, a quaternary ammonium compound, is a very effective compatible solute which is found in a wide range of crops. In maize, one of GB accumulators, this compatible solute accumulates in leaves in response to water deficit.
Glycinebetaine has been reported to synthesize from its precursor choline by a two-step oxidation, via the intermediate betaine aldehyde. The first oxidation step is catalyzed by choline monooxygenase (CMO, EC 1.14.15.7), and the further oxidation to GB is catalyzed by betaine aldehyde dehydrogenase (BADH, EC 1.2.1.81), the enzymes involved in GB biosynthesis.[](https://www.ncbi.nlm.nih.gov/mesh/D001622)
Abscisic acid (ABA) plays an important role in physiological adaptation of plants to drought stress. It has been reported that ABA is not directly involved in modulation of cell enlargement and division, but it indirectly regulates plant growth by improving stomatal resistance to control transpiration and CO2 uptake. These ABA-induced adaptive changes can be of great importance for the survival and better growth of plants under unfavorable environmental conditions. Although varied roles of ABA are well documented, it remains unclear how this hormone coordinately regulates GB metabolism in relation to BADH activity and choline content, and in turn plant growth of different maize cultivars using both exogenous ABA and fluridone (Flu), a direct inhibitor of ABA synthesis.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
Keeping in view the above facts, we hypothesized that plant hormone ABA can compensate for drought-induced retardation in the growth of two maize cultivars i.e., Zhengdan 958 and Jundan 20 by employing up-regulation of constitutive GB metabolism to mediate plant adaptation to DS. Previously, the former cultivar showed relatively higher drought index as well as dry matter production and grain yield under drought. Thus, Zhengdan 958 was ranked as drought tolerant and Jundan 20 as drought sensitive. With this aim, we designed hydroponic experiments to clarify the responses of maize to exogenous ABA and the ABA synthesis inhibitor Flu with respect to growth and GB metabolism in maize plants subjected to integrated root-zone DS.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 2. Results and Discussion
*In the **2. Results and Discussion** section:*
## 2.1. Modulation Role of Abscisic Acid (ABA) on the Growth and Leaf Water Relations of Maize Seedlings under Drought Stress
*In the **2.1. Modulation Role of Abscisic Acid (ABA) on the Growth and Leaf Water Relations of Maize Seedlings under Drought Stress** section:*
Twelve days of integrated root-zone DS (IR-DS) induced by PEG significantly caused a decline in the growth of seedlings of both maize cultivars. The shoot dry weights (SDW) of Zhengdan 958 (ZD958) and Jundan 20 (JD20) under IR-DS were only 74% and 61% of those of the same two cultivars, respectively, under the controls in the absence of both exogenous ABA and fluridone (Flu), 89% and 73%, respectively, in the presence of ABA, and 67% and 65%, respectively, in the presence of Flu. The leaf relative content (RWC) of ZD958 and JD20 treated with PEG alone had 90% and 87% of the controls, 96% and 91% with both PEG and Flu, and 87% and 82% with both PEG and ABA. Regardless of ABA or Flu treatment, ZD958 showed higher values of SDW and leaf RWC than those of JD20 under IR-DS (Table 1).[](https://www.ncbi.nlm.nih.gov/mesh/D011092)
Exogenous ABA increased but Flu decreased the SDW of ZD958 and JD20 by 16% and 20%, and 17% and 14%, respectively under IR-DS treatments. As for leaf RWC, application of ABA caused increase while Flu induced decrease in ZD958 and JD20 by 4.8% and 5.1%, and 4.4% and 3.8%, respectively. However, under the controls, ABA or Flu application had no obvious impact on SDW and leaf RWC (Table 1).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 2.2. The Accumulation Pattern of Endogenous ABA and Glycinebetaine (GB) in Maize Seedlings under Drought Stress
*In the **2.2. The Accumulation Pattern of Endogenous ABA and Glycinebetaine (GB) in Maize Seedlings under Drought Stress** section:*
Endogenous ABA and GB in both cultivars accumulated with prolonged period of IR-DS treatment. These responses were more predominant in ZD958 than that in JD20. The ABA contents in ZD958 and JD20 reached their maximum after 24 and 36 h of the start of DS treatment, respectively, being 540% and 480% of those of the control plants. In contrast, GB contents in ZD958 and JD20 were maximum after 48 and 60 h of DS treatment, being 250% and 180% of those of the control plants, respectively. The maximum accumulation of ABA took place earlier than that of GB in the leaves of both drought-stressed maize cultivars (Figure 1).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 2.3. Modulation Role of ABA on Endogenous ABA and Glycinebetaine (GB) Accumulation in the Leaves of Maize Plants under Drought Stress
*In the **2.3. Modulation Role of ABA on Endogenous ABA and Glycinebetaine (GB) Accumulation in the Leaves of Maize Plants under Drought Stress** section:*
The accumulation of endogenous ABA and GB was affected by the exogenous application of ABA or Flu in maize plants subjected to IR-DS. In the absence of ABA or Flu, endogenous ABA and GB levels increased 210% and 140% in ZD958 and 190% and 90% in JD20, respectively under IR-DS. The corresponding values in the presence of ABA and Flu were 230% and 160% in ZD958 and 210% and 130% in JD20 for ABA level as well as 200% and 70% in ZD958 and 140% and 50% in JD20 for GB level, respectively (Figure 2A,B).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
Addition of ABA and Flu caused increases and decreases, respectively, in endogenous ABA level of drought-stressed plants by 26% and 34% in ZD958, and 29% and 31% in JD20 as well as in GB level by 23% and 25% in ZD958, and 28% and 21% in JD20, respectively, as compared with those in plants without ABA or Flu application. As for the controlled plants, application of ABA and Flu also increased and decreased the endogenous ABA level of both cultivars. However, the increase/decrease rates became less than those in drought-stresses plants. Additionally, ABA and Flu treatments had no significant impacts on GB level in the controlled plants (Figure 2A,B).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 2.4. Modulation Role of ABA on GB Metabolism in the Leaves of Maize Plants under Drought Stress
*In the **2.4. Modulation Role of ABA on GB Metabolism in the Leaves of Maize Plants under Drought Stress** section:*
Glycinebetaine metabolism determination can be done by assessing the amount of its precursor choline and the activity of the key enzyme BADH. The choline content and BADH activity in the leaves are shown in Figure 2C,D. The choline content and BADH activity increased by 59% and 156% in ZD958, and 37% and 99% in JD20 without ABA or Flu application, by 70% and 237% in ZD958, and 48% and 178% in JD20 with ABA application, and by 61% and 115% in ZD958, and 40% and 72% in JD20 with Flu application respectively, due to drought treatment.[](https://www.ncbi.nlm.nih.gov/mesh/D001622)
Exogenous ABA increased but Flu decreased BADH activity in both cultivars under IR-DS. The beneficial effects of ABA/negative effects of Flu on BADH activity were greater/less in JD20 than those in ZD958. However, these responses were not recorded under control conditions. As for choline content, there were no significant impacts of added ABA/Flu under both IR-DS and control treatments. However, choline content in plants treated with ABA was greater than that in plants treated with Flu under IR-DS.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 2.5. Correlations among All Parameters Measured
*In the **2.5. Correlations among All Parameters Measured** section:*
Correlation coefficients among all traits evaluated were significantly higher under DS than those under control. Importantly, correlations among DM, RWC and content of ABA and choline as well as BADH activity were evident under DS but not for those under control treatment (Table 2).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 2.6. Interaction of Exogenous ABA or Flu Treatment and Water Regimes as well as Correlation Coefficients for All Parameters Measured
*In the **2.6. Interaction of Exogenous ABA or Flu Treatment and Water Regimes as well as Correlation Coefficients for All Parameters Measured** section:*
Water regimes and exogenous ABA or Flu treatments had significant effects on all parameters (Table 3). The magnitudes of F values across the above parameters were in the order: water regime > exogenous ABA (Flu) > cultivars except choline content. The interaction effects among the above treatments were also mostly significant for all response variables except Cv × A and W × Cv × A as well as W × Cv × Flu for choline content and BADH activity.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 2.7. General Discussion
*In the **2.7. General Discussion** section:*
Drought stress (DS) is one of the most important abiotic stresses which severely affect crop growth and productivity. Relative water content (RWC) in plant leaves is contemplated as a potential indicator of plant water status, because it is involved in the metabolic activity in tissues. Decline in RWC reflects a loss of turgor that results in limited cell expansion and thereby reduced growth in crop plants. Drought stress has been widely reported to cause perturbance in water homeostasis. Decline in the availability of water in plant body leads to molecular damage, growth inhibition and even death.
It has already been reported that different crop cultivars show varying response to DS for different duration in view of water status and plant growth. The present studies have shown that integrated root-zone DS (IR-DS) caused a decrease in leaf RWC and shoot biomass in maize plants. However, improved RWC and shoot dry matter (DM) were obtained under exogenously applied ABA, while an opposite effect was recorded under exogenously applied fluridone (Flu). The positive impacts of ABA were more predominant in cv. Jundan 20 (JD20) while more negative effects of Flu in cv.
Zhengdan 958 (ZD958) as compared with their untreated counterparts (Table 1). Thus, exogenously applied ABA obviously enhanced the responses to drought especially in a drought sensitive cultivar (JD20), while the abscisic acid synthesis inhibitor Flu diminished the responses to drought, especially in the drought resistant cultivar (ZD958) in terms of plant growth and water relations, but these effects were non-significant under well-watered conditions. Perhaps ABA was involved in the regulation of water relations and plant growth in maize plants under drought.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
It is well known that glycinebetaine (GB) accumulates in many drought-stressed plants including maize. Glycinebetaine is synthesized from its precursor choline by a two-step oxidation via betaine aldehyde catalyzed by choline monooxygenase (CMO), a ferredoxindependent soluble Rieske-type protein, betaine aldehyde dehydrogenase (BADH), and soluble NAD+ dependent enzyme. Some researchers pointed out that their results were consistent with an adaptive value for betaine accumulation in barley (Hordeum vulgare L.) during prolonged water stress and in sugar beet (Beta vulgaris L.) under salinity. Considerable genetic variation in betaine-accumulating potential in them was also reported. However, there are not sufficient data to prove a clear GB metabolism in drought-stressed maize plants.
The current experiments have shown that the IR-DS increased BADH activity and choline content, as well as induced high accumulation of GB, especially in a drought resistant cultivar (ZD958). Such increases in both BADH activity and choline content were correlated with enhanced accumulation of GB in the maize plants in response to DS (Figure 2C,D).[](https://www.ncbi.nlm.nih.gov/mesh/D001622)
A widely reported adaptation in plants to counteract abiotic stress is high accumulation of stress hormones and compatible organic solutes. Among these substances, ABA is a commonly occurring plant growth regulator in plants actively involved in the control of plant growth and development under drought conditions. For example, the role of ABA in closing stomata of drought-stressed plants has been widely reported. This effect is suggested to be vital for fast growth resumption and recovery of water content of plants. Additionally, GB is a major organic osmolyte that accumulates in a variety of plant species in response to different stresses.
The accumulation of GB has been found in many organisms, including higher plants. GB has been widely reported to play a part in the tolerance mechanism against a stress. Since accumulation of GB in plant tissues serves as an index of the internal water status of plants, increased leaf RWC in crop plants under drought stress suggests that GB may play a protective role in preventing cell damage from stress-induced dehydration. However, a relationship between ABA and drought stress in promoting GB accumulation in maize is not well understood. In the current study, the contents of both ABA and GB were found to be increased in drought-stressed maize seedlings, but the peak of ABA accumulation occurred earlier than GB under the drought regime.
These responses in a drought-resistant cultivar, ZD958, were stronger than those in the drought-sensitive JD20 (Figure 1).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
Additionally, exogenous ABA imposed greater positive impacts of ABA and GB accumulation in the drought-sensitive cultivar, and the abscisic acid synthesis inhibitor Flu had negative effects on the drought resistant cultivar (Figure 2A,B).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
Plant growth inhibition and relative adaptive physiological processes such as osmotic regulation caused by environmental stress might be regulated by the application of plant hormones. There are many reports demonstrating positive effects of exogenous ABA and the ABA synthesis inhibitor Flu on water relations, plant growth and final crop yield under water deficit conditions [,] (Table 1). The ABA treatments increased GB accumulation and the BADH amount, activity as well as its expression, e.g., in barley under osmotic stress, in pear (Pyrus pyrifolia) under drought stress and sorghum (Sorghum bicolor) under salinity treatment. However, some previous reports showed that ABA may not be associated with osmotic stress-induced betaine accumulation in oat (Avena sterilis) and periwinkle (Vinca major). The different response to ABA treatment with different plant species may be dependent on their cell signaling pathway in response to environmental stress.
Thus, the drought-induced betaine accumulation may involve probably either in ABA-dependent or -independent pathways, which may differ among plant species. Up to now, differential effects of ABA and Flu on modulation of GB metabolism in relation to BADH activity and choline content in maize are still not clear. In our experiments, it has been found that exogenously applied ABA and Flu increased and decreased BADH activity (no effect on choline content), respectively, thereby regulating GB accumulation under DS. The leaf water relations and shoot DM production were correspondingly improved in the maize plants. The greater positive/less negative responses due to exogenous ABA/Flu occurred in JD20 as compared with ZD958.
It is, therefore, concluded that endogenous ABA was probably involved in the regulation of GB metabolism by increasing BADH activity only, as well as improving water relations and plant growth under drought, especially in the drought sensitive cultivar (Jundan 20). Moreover, significant correlations among GB metabolism parameters, DM production and leaf RWC were evident in maize plants under DS, but no or less significant under control treatments (Table 2). These results show that contents of ABA, GB and choline as well as BADH activity could be used as potential selection criteria for drought tolerance in maize. Additionally, the effects of exogenously applied ABA and Flu on the above mentioned parameters were highly significant. It is, therefore, suggested that optimal application of plant hormone dose can benefit plant growth under water deficit conditions, but this response is cultivar-specific as is evident from data reported in Table 3.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 3. Materials and Methods
*In the **3. Materials and Methods** section:*
## 3.1. Plant Material and Trial Location
*In the **3.1. Plant Material and Trial Location** section:*
Hydroponic experiments were performed in a controlled growth chamber at the College of Life Sciences of Northwest A & F University, Yangling, China. The seeds of two maize (Zea mays L.) cultivars Zhengdan 958 and Jundan 20 were supplied for the present experiments by the Agronomy College of the same University. According to the previous hydroponic experiment and a field experiment carried out in the farm belonging to the same university, dry matter production and grain yield of the Zhengdan 958 cultivar were relatively little affected by drought.
## 3.2. Plant Growth and Experimental Design
*In the **3.2. Plant Growth and Experimental Design** section:*
The seeds of both maize cultivars were germinated at 28 °C for 72 h in the dark. The young seedlings were inserted into holes of styrofoam boards placed in plastic containers (inner length: 26 cm; width 18 cm; height 12 cm) containing treatment solutions. The experimental units were placed in the growth chamber under the following environmental conditions: average day/night temperature 25/18 °C, relative humidity 60–70%, light intensity 350 μmol/m2/s and 16 h of photoperiod. The containers were wrapped with black plastic to protect roots from light. Four and eight days after placement of the seedlings in deionized water, the deionized water was replaced by one-half-strength and complete nutrient solution, respectively, which contained all essential mineral nutrients for plant growth.
The pH of the nutrient solution was adjusted to 6.30 (±0.05) every day.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
When seedlings were at the stage of three-leaf, drought stress (DS) treatment started by adding 120 g/kg (w/w) polyethylene glycol (PEG-6000) to achieve drought (osmotic) stress level of approximately −0.23 MPa, dissolved in full strength nutrient solution. The nutrient solution without PEG-6000 served as the control (CK). For each water stress treatment, the seedlings were fed with nutrient solution containing 100 μmol/L ABA or 10 μmol/L fluridone (Flu). The experimental design consisted of six treatments: (1) Control; (2) 12% (w/v) PEG-6000 (PEG); (3) control plus 100 μmol/L ABA (Control + ABA); (4) 12% (w/v) PEG-6000 plus 100 μmol/L ABA (PEG + ABA); (5) control plus 10 μmol/L fluridone (Control + Flu); (6) 12% (w/v) PEG-6000 plus 10 μmol/L fluridone (PEG + Flu).[](https://www.ncbi.nlm.nih.gov/mesh/D011092)
To explore the relationship between ABA and drought stress (DS) in promoting glycinebetaine (GB) accumulation in maize plants, a second experiment was carried out using both maize cultivars Zhengdan 958 and Jundan 20 under the two treatments, control and 12% (w/v) PEG-6000 in nutrient solution.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
Plants were grown in a growth chamber in 3.4 L plastic pots, which were sealed carefully to avoid evaporation, with a sponge wrapped around the interface of the roots and the shoots. All treatment units were replicated four times, with a completely randomized design. The treatment solutions were aerated twelve hours a day.
All experiments were repeated twice. Data presented here are means of four replicates of the two experiments (n = 8).
## 3.3. Sampling and Recording of Data
*In the **3.3. Sampling and Recording of Data** section:*
The maize plants were harvested after 12 days of the onset of drought or ABA or Flu treatments in the first experiment and after 0, 12, 24, 36, 48, 60 and 72 h after the onset of drought in the second experiment, respectively. Fresh mass of the shoots was recorded. Then all fresh samples were placed in an oven at 105 °C for 30 min, and then dried to a constant weight at 75 °C.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
Relative water content (RWC) was determined using fresh weight (FW), dry weight (DW) and the turgid weight (TW). RWC was calculated from the equation described by Gao: RWC (%) = (FW − DW)/(TW − DW)× 100.
Glycinebetaine (GB) content was determined following Grieve and Grattan with some modifications. Dried and finely powdered plant materials (0.5 g) were shaken with 20 mL of deionized water for 24 h at 25 °C. The extracts were diluted 1:1 with 2 N H2SO4. Aliquots of 0.5 mL were put in test tubes and cooled in ice water for 1 h, before a cold KI-I2 reagent (200 μL) was added. The tubes were stored at 0–4 °C for 16 h and then centrifuged at 12,000× g for 15 min at 4 °C.
The supernatant was aspirated. The periodite crystals were dissolved in 5 mL of 1, 2-dichloroethane. After 2–2.5 h, the absorbance was read at 365 nm with a UV–visible spectrophotometer. Reference standards of GB (50–200 g/mL) were used for calibration and estimation of GB concentration in unknown samples. The GB content was expressed as nmol/g DW.[](https://www.ncbi.nlm.nih.gov/mesh/D001622)
BADH activity was assayed as described by Daniell et al. with some modifications. To obtain crude protein extracts, plant materials were homogenized in 250 μL homogenization buffer containing 50 mM HEPES-KOH, pH 8.0, 1 mM EDTA, 20 mM sodium metabisulfite, 10 mM sodium borate, 5 mM ascorbic acid, 5 mM dithiothreitol, and 2% (w/v) PVPP. The homogenates were then centrifuged at 12,000× g for 15 min at 4 °C and the supernatants used for determination of BADH activity. The BADH activity was assayed by monitoring the absorbance at 340 nm with 0.05 mM betaine aldehyde chloride as a substrate. The activity was calculated using the extinction coefficient of 6220/M/cm for NADH.
The BADH activity was expressed as μmol/min/mg protein. Protein concentration of the crude extract was measured by the method of Gao using bovine serum albumin as a standard.[](https://www.ncbi.nlm.nih.gov/mesh/D006531)
Choline content was determined following Feng and Ren and Richard and Emily with some modifications. Dried and finely powdered plant materials (0.5 g) were added with 70 mL of deionized water in a triangle vase (100 mL volume) and shaken up. The vase was bathed in hot water at 80 °C for 15 min, shaken properly with an oscillator for 30 min, set to 100 mL volume scale, and then filtered through a filter paper. Aliquots of 10 mL filtered solution were poured into a triangle vase, cooled in ice water to −5 °C before adding 15 mL Reinecke salt- methanol solution (4 g Reinecke salt was dissolved in 100 mL methanol). The mixed solution was stirred for 30 min and placed in a refrigerator for 12 h.
The red water insoluble substance was filtered out, washed with 10 mL propylalcohol three times, dissolved with acetone, and set to 25 mL volume scale. The choline content was assayed by monitoring the absorbance at 520 nm with UV-visible spectrophotometer using acetone as a control. Reference standards of choline (9–27 mg/mL) were used for calibration and estimation of choline concentration in the unknown samples. The choline content was expressed as nmol/g DW.[](https://www.ncbi.nlm.nih.gov/mesh/D002794)
The content of ABA was assayed in plant leaves by ELISA (enzyme-linked immunosorbent assay) using a monoclonal antibody (MAB) raised against ABA. The polyclonal antibody (RAMIG) was raised against the mouse immunoglobulins and ABA labeled with alkaline phosphatase, according to Weiler. The ABA content was expressed as nmol/g DW.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 3.4. Statistical Analysis
*In the **3.4. Statistical Analysis** section:*
The data for all attributes were analyzed statistically to work out analysis of variance using the SAS software package. The analysis of variance (ANOVA) was followed by the least significance test (LSD) to determine the significant difference among the mean values at the 0.05 level.
## 4. Conclusions
*In the **4. Conclusions** section:*
The results presented in this paper provide evidence that both BADH activity and choline content were involved for enhanced accumulation of GB in maize plants under drought stress. The endogenous ABA seemed to be involved in modulating GB accumulation by enhancing BADH activity, thereby improving leaf RWC and enhancing shoot DM in drought-stressed maize plants, especially in the drought sensitive cultivar (JD20). The peak of ABA content reached earlier than that of GB in the leaves of drought-stressed maize plants. Such, endogenous ABA probably played a positive role as a signal in the regulation of GB metabolism, water relations and plant growth by regulating BADH activity, but not choline content. The above results have proved that the drought-induced GB accumulation in maize may be involved probably in ABA-dependent pathway.
The exogenous ABA provided some protection against the DS effects on maize plants by regulating endogenous ABA level and GB metabolism.[](https://www.ncbi.nlm.nih.gov/mesh/D002794)
# References
*In the **References** section:*
Effects of drought treatment on contents of endogenous ABA and glycinebetaine (GB) in the seedlings of Zhengdan 858 (A) and Jundan 20 (B) cultivars.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
Effects of exogenous ABA and fluridone (Flu) on endogenous ABA content (A), glycinebetaine (GB) content (B), choline content (C) and betaine aldehyde dehydrogenase (BADH) activity (D) in the seedlings of two maize cultivars under drought stress induced by PEG and control (C). The same letters on the top of the column within each variable are not significantly different among twelve treatments at the 0.05 level.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
Mean values in the same column followed by the same letters within variables are not significantly different among six treatments, and the same row followed by the same letters in parentheses between the two cultivars with the same water and ABA or Flu treatment. Letters (a–c) showing least significant difference between treatments and cultivars at the 0.05 level.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
Modulation effect of abscisic acid (ABA) on dry matter (DM) and leaf relative water content (RWC) of maize seedlings under drought stress.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
significant at P < 0.05, 0.01, 0.001, respectively.
Correlation coefficients of dry matter (DM), leaf relative water content (RWC), ABA content (ABAC), GB content (GBC), choline content (CC) and BADH activity (BADHA) of both cultivars under drought stress (above diagonal) and control condition (below diagonal).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
significance at P < 0.05, 0.01, 0.001, respectively.
F values of the effects of exogenous ABA (A) or fluridone (Flu) treatment, cultivar (Cv), water regime (W) and their interactions on dry matter (DM), leaf relative water content (RWC), ABA content (ABAC), GB content (GBC), choline content (CC) and BADH activity (BADHA).[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
|
# Introduction
Trappin-2/Elafin Modulate Innate Immune Responses of Human Endometrial Epithelial Cells to [PolyI∶C](https://www.ncbi.nlm.nih.gov/mesh/D011070)
# Abstract
*In the **Abstract** section:*
Background
Upon viral recognition, innate and adaptive antiviral immune responses are initiated by genital epithelial cells (ECs) to eradicate or contain viral infection. Such responses, however, are often accompanied by inflammation that contributes to acquisition and progression of sexually transmitted infections (STIs). Hence, interventions/factors enhancing antiviral protection while reducing inflammation may prove beneficial in controlling the spread of STIs. Serine antiprotease trappin-2 (Tr) and its cleaved form, elafin (E), are alarm antimicrobials secreted by multiple cells, including genital epithelia.
Methodology and Principal Findings
We investigated whether and how each Tr and E (Tr/E) contribute to antiviral defenses against a synthetic mimic of viral dsRNA, polyinosine-polycytidylic acid (polyI∶C) and vesicular stomatitis virus. We show that delivery of a replication-deficient adenovector expressing Tr gene (Ad/Tr) to human endometrial epithelial cells, HEC-1A, resulted in secretion of functional Tr, whereas both Tr/E were detected in response to polyI∶C. Moreover, Tr/E were found to significantly reduce viral replication by either acting directly on virus or through enhancing polyI∶C-driven antiviral protection. The latter was associated with reduced levels of pro-inflammatory factors IL-8, IL-6, TNFα, lowered expression of RIG-I, MDA5 and attenuated NF-κB activation. Interestingly, enhanced polyI∶C-driven antiviral protection of HEC-Ad/Tr cells was partially mediated through IRF3 activation, but[ not associated with higher in](https://www.ncbi.nlm.nih.gov/mesh/D011070)du[ction o](https://www.ncbi.nlm.nih.gov/mesh/D011070)f IFNβ, suggesting multiple antiviral mechanisms of Tr/E and the involvement of alternative factors or pathways.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Conclusions and Significance
This is the first evidence of both Tr/E altering viral binding/entry, innate recognition and mounting of antiviral and inflammatory responses in genital ECs that could have significant implications for homeostasis of the female genital tract.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Genital epithelial cells (ECs) provide the first line of defense against sexually-transmitted infections (STIs). Upon viral sensing through pattern-recognition receptors (PRRs), ECs initiate innate and adaptive immune responses that serve to eradicate or contain viral pathogens. ECs can directly respond to viruses and viral pathogen-associated molecular patterns (PAMPS) by secreting innate protective factors, including defensins and cathelicidins as well as members of the whey-acidic protein (WAP) family. Of the 18 human WAP proteins, only a few have been well characterized to date, and among the better understood are serine antiproteases elafin (E) with its precursor, trappin-2 (Tr), as well as secretory leukocyte protease inhibitor (SLPI), and prostate stromal protein 20 kDa (ps20).
The physiological role of serine antiproteases has been extensively studied over the past two decades, mainly due to their contribution to homeostatic equilibrium through the control of proteases, inflammation, and infections. Together with other proteins, such as snake venom neurotoxins and whey acidic protein, serine antiproteases share an evolutionary conserved canonical cysteine-rich four-disulfide core (FDC) domain, or the WAP domain, involved in protease inhibition. Trappin-2 (9.9 kDa) (or pre-elafin) is a secreted and unglycosylated protein of 95-amino acids (aa) that contains an N-terminal cementoin domain (38-aa) and elafin (5.9 kDa), a C-terminal inhibitory WAP (57-aa) domain. Elafin is released from the N-terminus of Tr by proteolysis, arguably most efficiently by mast cell tryptase. Antiprotease activity and wound repair were the first described properties of Tr and E (Tr/E), similar to SLPI.
Unlike ps20, SLPI along with Tr/E are functional neutrophil serine protease inhibitors. Inhibition of human neutrophil elastase (HNE) and proteinase 3 by the inhibitory loop on a WAP domain allows Tr/E to control tissues proteolysis associated with excessive inflammation in a neutrophil-rich environment. In turn, cross-linking between repeated hexapeptide motifs (GQDPVK) on the N-terminal portion of each Tr/E and extracellular matrix proteins arguably allows Tr/E to repair compromised tissue integrity. In addition, due to their cationic nature, but not exclusively, Tr/E were shown to possess antimicrobial activity against Gram-negative and Gram-positive bacteria and certain fungal infections. Worth mentioning is that similar to SLPI, antibacterial activity of Tr/E appeared to be independent of their antiprotease function.
Later, anti-inflammatory features of the antiproteases were also described, showing that Tr/E and SLPI were capable of reducing activation of NF-κB and AP-1 by altering IκB activation and proteosomal degradation, respectively, in response to inflammatory and bacterial stimulation. More recent studies, however, also reported immunomodulatory properties of Tr/E. Indeed, depending on the environment, Tr/E can either dampen inflammation or promote immunostimulatory events and prime the immune system. Both Tr/E are found at mucosal surfaces, in tissues and multiple cell types, including genital ECs and regarded as alarm antiproteases, as they are mainly produced in response to pro-inflammatory stimuli like LPS, TNFα, and IL-1β. Interestingly, ECs from the female genital tract (FGT) produce Tr/E constitutively, with uterine cells capable of producing even greater amounts of Tr/E in response to a viral ligand, polyinosine-polycytidylic acid (polyI∶C), indicating the significance of these molecules in controlling the local milieu in the FGT.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Viral double-stranded RNA (dsRNA) is a PAMP generated during the life cycle of most, if not all, viruses. Double-stranded RNA, including viral dsRNA and its synthetic mimic polyI∶C, are recognized by at least two families of PRRs: Toll-like receptors (TLRs), including TLR 3, and RNA helicases, namely retinoic acid inducible gene-I (RIG-I) and melanoma differentiation associated gene 5 (MDA5). Following recognition of dsRNA, activated PRRs initiate a series of signaling events, triggering phosphorylation, homodimerization and translocation into the nucleus of a set of transcription factors like interferon (IFN) regulatory factor 3 (IRF3), IRF7, NF-κB, and ATF2/c-Jun (c-Jun). Inside the nucleus, these transcription factors either work independently or interact with each other in triggering the transcription of antiviral and inflammatory gene products, such as type I IFNs and IL-8, IL-6, and TNFα. Specifically, IRF3 alone can directly bind to the IFN-stimulated response element in the promoter region of interferon-stimulated genes (ISGs) and activate a set of ISGs and their products in the absence of type I IFN production.
Such antiviral cascade is devoid of excessive inflammatory responses and is induced when a low viral stimulation is detected. Alternatively, in response to a high viral load, IRF3 associates with NF-κB and c-Jun to trigger the production of IFNβ and the induction of ISGs in an IFN-dependent mode, which also triggers robust inflammation, target cell recruitment, and/or tissue damage due to the production of pro-inflammatory cytokines and chemokines along with antiviral type I IFNs molecules.[](https://www.ncbi.nlm.nih.gov/mesh/C077118)
Treatment with polyI∶C has been shown to induce potent antiviral protection in vitro and in vivo, making polyI∶C an attractive candidate for microbicide or vaccine adjuvant trials against STIs. However, in addition to antiviral activity, polyI∶C also triggers the release of pro-inflammatory mediators, therefore potentially negating its beneficial effects in the FGT. While pro-inflammatory factors are important for immune cell recruitment and activation, if poorly controlled, they may also be detrimental in FGT, since acquisition and pathogenesis of common STIs are associated with immune activation and inflammation. Hence, interventions leading to better control of inflammatory responses may prove to be more beneficial for increased antiviral protection and overall health in the FGT. Recently, new evidence has accumulated on the role of Tr/E in protection against viruses.
Indeed, Tr/E have been associated with resistance to HIV mucosal transmission in commercial sex workers (CSWs) in Kenya. Later, Ghosh et al. reported an anti-HIV feature of E in in vitro study. Furthermore, Roghanian et al. documented that Tr expression increased Ad/LacZ viral clearance, as well as secondary immune responses, in a murine model in vivo .[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Collectively, these data clearly demonstrate the importance of Tr/E in antiviral protection, although specifically how Tr/E contribute to antiviral immune-inflammatory responses is still unknown. Thus, the objective of this study was to elucidate whether and how Tr/E contribute to innate antiviral and inflammatory responses in uterine ECs elicited by a viral ligand, polyI∶C. Here, we describe novel antiviral and immunomodulatory properties of Tr/E. To the best of our knowledge, this is the first study to present evidence of Tr/E affecting host innate recognition and modulating antiviral and inflammatory responses in genital ECs against polyI∶C.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Materials and Methods
*In the **Materials and Methods** section:*
## Reagents
*In the **Reagents** section:*
PolyI∶C and lipopolysaccharide (LPS) were reconstituted in the phosphate-buffered saline (PBS) and used at concentrations shown in figures (Sigma-Aldrich, Oakville, ON, Canada). Human recombinant (r) proteins Tr (rTr) (R&D Systems, Burlington, ON, Canada) and in-house recombinant E (rE) (described below) were used in in vitro experiments and as reference markers for Western blotting. The amount of Tr/E being used in this study ranges from 0.2 to 5 µg/ml to cover the physiological levels (within 1 µg/ml) reported previously as well as concentrations achieved in supernatants of human endometrial carcinoma (HEC)-1A cells infected with a replication-deficient adenovirus (Ad) expressing human Tr gene (Ad/Tr) (over 1 µg/ml).[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Cell lines
*In the **Cell lines** section:*
HEC-1A, Caco-2 (human colonic epithelial cells), and A549 (derived from a type II human alveolar cell carcinoma) cells were obtained from American Type Culture Collection (Rockville, MD). HEC-1A and A549 cells were cultured in McCoy's 5A Medium Modified (Invitrogen Life Technologies, Burlington, Ontario, Canada) and DMEM, respectively, supplemented with 10% fetal bovine serum (FBS), 1% HEPES, 1% l-glutamine (Invitrogen Life Technologies), and 1% penicillin-streptomycin (Sigma-Aldrich) at 37°C in 5% CO2. Caco-2 cells were cultured in DMEM growth medium containing 5% FBS, 1% l-glutamine, 1% penicillin-streptomycin, 5 mL NEAA, and 9.6 mL NaHCO3.[](https://www.ncbi.nlm.nih.gov/mesh/C113109)
## Adenoviral constructs and delivery in cell culture
*In the **Adenoviral constructs and delivery in cell culture** section:*
The Ad constructs used in this study have been described in detail elsewhere. To express human Tr, the Ad/Tr vector, encoding gene for 95-aa human Tr, was used. This Ad construct was previously called Ad/E. E1, E3-deleted empty adenovirus Ad-dl703 (Ad/dl), coding for no transgene, was used as a control for Ad/Tr. Both Ad vectors were prepared at the Centre for Gene Therapeutics at McMaster University (Hamilton, ON, Canada). To generate supernatants containing Tr, HEC-1A cells were infected with MOI 0–50 plaque-forming units (PFU) of Ad/Tr (Ad/Tr-cells) or Ad/dl (Ad/dl-cells) overnight at 37°C in Opti-MEM®
I Reduced Serum Medium (Invitrogen Life Technologies), washed with PBS and incubated for 12 h in serum-containing medium. Cells were washed again and incubated for additional 24 h in serum-free cell culture medium. Cell-free supernatants were used either for protein measurement by ELISA or for antiprotease activity against HNE. Another aliquot of supernatants was further concentrated with 3–30 kDa MWCO centrifugal filter units (Amicon, Millipore, Billerica, MA, USA) as per supplier's instructions and used in in vitro studies.
For routine experiments, epithelial cells were either treated with Opti-MEM medium alone (UT) or with MOI 50 PFU of Ad/dl or Ad/Tr at 37°C overnight. After PBS washes and rest for 4 h, cells were incubated in serum-containing medium alone or with polyI∶C for additional 24 h.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Vesicular stomatitis virus (VSV) plaque reduction assay and antiviral assay
*In the **Vesicular stomatitis virus (VSV) plaque reduction assay and antiviral assay** section:*
Vesicular stomatitis virus (VSV-GFP), a lytic IFN-sensitive virus expressing green fluorescent protein under the viral promoter (a kind gift from Dr. Brian Lichty, McMaster University), was used in a plaque reduction assay to assess the role of Tr/E in antiviral protection. This method is based on determining the ability of VSV-GFP to replicate in cell cultures in presence of biologically active antiviral factors, e.g., IFNs. Briefly, HEC-1A cells were seeded in 96-well culture plates and infected with Ad (Ad-cells) as described above, followed by treatment with medium alone or polyI∶C for 24 h. Induction of antiviral response was assessed by subsequently challenging cell monolayers in serum-free medium with MOI 1 PFU of VSV-GFP. GFP fluorescence intensity was visualized 24 h later on a Typhoon Trio (Amersham Bioscience, GE Healthcare Bio-Sciences Corp., Piscataway, NJ, USA) and quantified using Image Quant 5.2 software.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## MTT viability assay
*In the **MTT viability assay** section:*
MTT assay, described elsewhere, was used as per supplier's instructions to determine viability of Ad-exposed and polyI∶C-treated HEC-1A cells (Biotium Inc., Hayward, CA, USA).[](https://www.ncbi.nlm.nih.gov/mesh/C022616)
## ELISA assays
*In the **ELISA assays** section:*
Cell-free supernatants were stored at −70°C until assayed for human Tr/E, IL-8, TNFα, IL-6 with ELISA Duoset kit (R&D Systems), human IFNβ by ELISA kit from Antigenic America Inc. (Huntington Station, NY, USA), and IFNα subtypes by VeriKine™ Human Interferon-Alpha Multi-Subtype ELISA Kit (Piscataway, NJ, USA) according to the supplier's protocol. Analytes were quantified based on standard curves obtained using an ELISA reader Tecan Safire ELISA reader (MTX Labs Systems Inc.). Cut off limit for Tr/E and IL-8 was 31.25 pg/ml; for TNFα and IFNβ was 15.6 pg/ml; for IFNα subtypes was 12.5 pg/ml, and levels detected below these limits were considered as undetectable.
## Generation of recombinant human elafin
*In the **Generation of recombinant human elafin** section:*
To prepare rE protein, cDNA fragments encoding the relevant part of Tr (amino acid residues A61–Q117) were amplified from HEC-1A cells cDNA using the following primers: 5′-ACAGGATCCGCGCAAGAGCCAGTCAAAGGTCCA-3′ and: 5′-CAGGAATTCTCACTGGGGAACGAAACAGGCCATC-3′. The amplified elafin-cDNA was gel purified, restriction digested, and directionally subcloned into BamH I and EcoR I site of bacteria expression vector pHAT10 (BD Biosciences, Rockville, MD, USA) in frame to the HAT-tag. The clone was confirmed by restriction digestion and nucleic acid sequencing of both strands. The plasmid was transformed into Escherichia coli BL21 strain (Codon Plus, Palo Alto, CA, USA). When the cells grew in Luria-Bertani broth containing 100 µg/ml ampicillin to an A600 of 0.55–0.60 at 37°C, protein expression was induced by adding isopropyl-β-D-thiogalactopyranoside (IPTG) to a final concentration of 1 mM for 6 h.
The cells were harvested by centrifugation at 2,000×g for 5 min, and cell pellets were washed once with ice-cold PBS (pH 7.4) and resuspended in wash buffer (50 mM Na2PO4, 300 mM NaCl, pH 7.0). The desired recombinant protein, rE, was found in the soluble fraction in pilot experiment. After disrupting the cells by freeze-thaw and subsequently by sonication, the lysates were centrifuged at 20,000×g for 20 min, and the soluble fraction was collected. The soluble recombinant protein was purified from this fraction by BD TALON metal affinity resin (BD Biosciences, Mississauga, ON, Canada) according to the manufacturer's instructions and dialyzed against PBS at 4°C. The expressed human rE was confirmed by Western blotting targeting the HAT-tag as well as by specific antibody for human E.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## Preparation of cell extracts and Western blot (WB) analysis
*In the **Preparation of cell extracts and Western blot (WB) analysis** section:*
Whole-cell extracts were prepared by using whole-cell extract buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1%NP-40, 1%SDS, 1×protease inhibitor (Roche, Mississauga, ON, Canada)) as per standard protocol. Protein amount was quantified using Bradford assay with bovine serum albumin (BSA) (Sigma-Aldrich) as a standard and Bio-Rad Dye Reagent Concentrate as a protein stain (Bio-Rad Laboratories, Mississauga, ON, Canada). WB was performed on a 10% polyacrylamide denaturing SDS-PAGE gel and PVDF membranes (Amersham, Arlington Heights, IL, USA) as per standard protocol, using the following primary antibodies: anti-human Tr/E TRAB2O (Hycult Biotech, Uden, Netherlands), tIRF3 FL-425 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), pIRF3 Ser396 (Cell Signaling, Danvers, MA, USA), RIG-I #4520 (Cell Signaling), MDA-5 R470 (Cell Signaling), c-Jun (60A8) (Cell Signaling), p-c-Jun (Ser73) (Cell Signaling), TLR3 IMG-5631-2 (Imgenex, San Diego, CA, USA), NF-κB p65 (C-20): sc-372 (Santa Cruz), GAPDH ab9485 (Abcam) antibodies at a dilution of 1∶1000, except for GAPDH (1∶5000) and p-NF-κB p65 (Ser 536): sc-33020 (Santa Cruz) (1∶100). After incubation with corresponding horse-radish peroxidase (HRP)-conjugated secondary antibody (Bio-Rad Laboratories), blots were visualized using a SuperSignal West Femto or Pico Chemiluminescent Substrate kit (Thermo Scientific, Rockford, IL, USA). GAPDH was used as internal loading control.
Quantification of band intensities was done using MBF_ImageJ for Microscopy Software.[](https://www.ncbi.nlm.nih.gov/mesh/D006851)
## Transfection and luciferase assay
*In the **Transfection and luciferase assay** section:*
HEC-1A cells (5×105 per well) were transfected in 0.5 mL of Opti-MEM® medium (Invitrogen Life Technologies) in a 24-well plate with 20 ng of DNA of each of pgkβ-Gal and pNF-κB-Luc or pAP-1-Luc (Stratagene, La Jolla, CA, USA) plasmids (total DNA 40 ng per well) and 50 MOI of Ad/dl or Ad/Tr overnight at 37°C utilizing Lipofectamine 2000 (Invitrogen Life Technologies) as per supplier's instructions. Pgk-β-Galactosidase plasmid was used as a normalization control. After transfection, cells were extensively washed with PBS and allowed to rest for 4 h followed by stimulation with polyI∶C and LPS for 2–4 h. After ligand stimulation, cells were washed twice with cold PBS, lysed with 1× reporter buffer (Enhanced Luciferase Assay Kit (Cat # 556866, BD Pharmingen) and subjected to one freeze-thaw cycle, after which cell lysates were collected and assayed in a 96-well plate for luciferase (Enhanced Luciferase Assay Kit, BD Pharmingen) and β-galactosidase (Luminescent β-Gal assay, Promega, Madison, WI, USA) activities separately as per supplier's protocol, using Opticomp I luminometer (MGM instruments).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## IRF3 knockdown by RNA interference
*In the **IRF3 knockdown by RNA interference** section:*
Small interfering (siRNA) molecule (Invitrogen Life Technologies) targeting IRF3 (GenBank accession number NM_001571) within 247–1530 ORF through the following sequences: IRF3-498 (start) CAACCGCAAAGAAGGGTTGCGTTTA, or non-targeting siRNA, RNAi Negative Control (medium GC content), 12935-300, (Invitrogen Life Technologies) were used to specifically knockdown IRF3. Transfections of siRNA (8 pmol) were done using Lipofectamine RNAiMAX and Opti-MEM® I Reduced Serum Medium (Invitrogen Life Technologies) as per supplier's instructions. HEC-1A cells, 3×104 in a 100 µl of total volume of complete growth medium, were transfected in a 96-well BD Falcon culture plate (BD Biosciences) for 48 h before adding MOI 50 PFU of Ad/dl or Ad/Tr. Knockdown efficiency was monitored using WB.
## RNA extraction and real-time quantitative PCR analysis
*In the **RNA extraction and real-time quantitative PCR analysis** section:*
The protocol for RNA isolation was described elsewhere. Briefly, total RNA was isolated from Ad-cells cultured with medium alone or polyI∶C for 6 h, using TRIzol reagent (Invitrogen Life Technologies) according to the supplier's protocol. RNA was DNase-treated with DNA-free (Ambion, Austin, TX, USA) and complementary (cDNA) was synthesized from total RNA using SuperScript reverse transcriptase III (Invitrogen Life Technologies) as per supplier's protocol. Real-time quantitative PCR was performed in a total volume of 25 µl using 1× Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA), 5 µl of diluted cDNA, 500 nmol forward primer, 500 nmol reverse primer (Mobix; McMaster University, ON, Canada), and 200 nmol probe in a 96-well plate. TaqMan oligonucleotide primers and probes labeled with 6FAM at 5′-end and a non-florescent quencher at 3′-end were designed using Primer Express 1.5 (Applied Biosystems) and selected following the TaqMan rules of Applied Biosystems.
The sequences were as follows: RIG-I: 5′-AGGGCTTTACAAATCCTGCTCTCTTCA-3′ (probe), 5′-GGTGTTCCAGATGCCAGAC-3′ (forward), 5′-TTCCGCAAATGTGAAGTGTATAA-3′(reverse); MDA5: 5′-TTTGGCTTGCTTCGTGGCCC-3′(probe), 5′-TGATTCCCCTTCCTCAGATAC-3′(forward), 5′-TGCATCAAGATTGGCACATAGT-3′(reverse); TLR3: 5′-TGTGGATAGCTCTCC-3′(probe), 5′-CCGAAGGGTGGCCCTTA-3′(forward), 5′-AAGTTACGAAGAGGCTGGAATGG-3′ (reverse); 18S rRNA: 5′-CGGAATTAACCAGACAAATCGCTCCA -3′ (probe), 5′-GTGCATGGCCGTTCTTAGTT-3′ (forward), 5′-TGCCAGAGTCTCGTTCGTTAT-3′ (reverse). The expression of 18S ribosomal RNA (rRNA) was used as an internal control. PCR was run with the standard program: 95°C 10 min, 40 times of cycling 95°C 15 sec and 60°C 1 min in a 96-well plate with an ABI PRIZM 7900HT Sequence Detection System using the Sequence Detector Software 2.2 (Applied Biosystems). To determine the expression of ISG56, a semi-quantitative RT-PCR was performed with oligonucleotide primer sequences (Mobix): 5″-GACAGGAAGCTGAAGGAGAAA-3′ (445-bp product) (forward), 5′-TcTTGCATTGTTTCTTCTACCACT-3′ (reverse).
PCR program was as follows: 94°C for 2.5 min, 30 cycles of 94°C for 20 sec, 55°C for 30 sec, 72°C fir 1 min, 72°C for 5 min. PCR products were electrophoresed on 2% agarose gel using a 100-bp DNA ladder (Invitrogen Life Technologies) as a marker to identify PCR products. The gel was stained with a loading fluorescent dye EZ-vision N472-Q (AMRESCO Inc., Solon, OH, USA) and visualized with UV transilluminator (Gel Doc 2000, BioRad, Mississauga, ON, Canada). Densities of DNA bands were quantified to signal volumes using ImageQuant 5.0.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Immunofluorescence staining
*In the **Immunofluorescence staining** section:*
Immunofluorescence staining was performed as described earlier, but with minor modifications. Ad-cells, grown on an 8-well BD Falcon culture slides (BD Biosciences), were medium- or 25 µg/ml polyI∶C-treated for 4 h, fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 in PBS for 20 min, and blocked for 1 h at ambient temperature in a blocking solution (0.1% Triton X-100, containing 5% goat serum/BSA). IRF3 was detected using 1∶100 dilution in blocking solution of IBL18781 (IBL, Gunma, Japan) antibody for 1 h. Negative control rabbit immunoglobulin fraction (DakoCytomation, Glostrup, Denmark) served as an isotype control and was diluted to match the protein content of the primary anti-IRF3 antibody. Secondary antibody, Alexa Fluor 488 conjugated goat anti-rabbit IgG (Molecular Probes, Eugene, OR, USA) was added to cells in a blocking solution for 1 h. The nuclei were visualized by staining with propidium iodide.
Images were acquired using an inverted laser-scanning confocal microscope (LSM 510, Zeiss, Oberkochen, Germany).[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Protease inhibition assay (PIA)
*In the **Protease inhibition assay (PIA)** section:*
This assay was performed by measuring inhibition of HNE activity by Tr/E as was described earlier, but with minor modifications. Elastase-inhibitory activity was measured in a 96-well plate by combining cell-free undiluted supernatant (generated as described earlier in Materials and Methods and before or after polyI∶C treatment) (final volume 10 µl/well), serially diluted each rTr and rE, or medium alone, to a known quantity of purified HNE (50 ng in 10 µl/well) or diluent alone (negative control), and incubating for 30 min at 37°C. Subsequently, 50 µl of HNE substrate, N-methoxysuccinyl-Ala-Ala-Pro-Val p-nitroanilide (Sigma-Aldrich), diluted to 50 µg/ml in 50 mM Tris, 0.1% Triton, 0.5 M sodium chloride, pH 8 buffer were added and the hydrolysis was recorded by monitoring the increase of absorbance at 405 for 15 min using a Tecan Safire ELISA reader (MTX Labs Systems).[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Statistical analysis
*In the **Statistical analysis** section:*
Data were expressed as means ± standard deviation (SD). Statistical analysis was performed with either unpaired Student's t test or a one-way analysis of variance (ANOVA) using Sigma Stat 2.03.
## Results
*In the **Results** section:*
## Ad/Tr-cells secrete Tr and both Tr/E are detected in Ad/Tr-cells supernatants in response to polyI∶C
*In the **Ad/Tr-cells secrete Tr and both Tr/E are detected in Ad/Tr-cells supernatants in response to polyI∶C** section:*
Initial studies evaluated the expression of the secreted Tr from Ad/Tr-cells by ELISA that detects both Tr/E. We report that Ad/Tr-cells secreted significant amounts of Tr/E in an Ad-dose-dependent manner (Figure 1A). In contrast, untreated HEC-1A (UT) and Ad/dl-cells expressed very low to no Tr/E (Figure 1A). Since the antibodies used in the ELISA did not distinguish Tr from E in Ad/Tr-cells supernatants (Ad/Tr-sups), WB was performed to clarify the presence of both Tr/E. Two recombinant reference proteins, namely commercial 6×His-Tr (rTr) and in-house HAT-E (rE), were used as comparative markers for Tr and E, respectively. Due to tag insertion, both reference proteins appeared 3–4 kDa higher than the appearance of untagged proteins would be expected. Thus, rE band appeared at ∼10–11 kDa, and rTr at ∼13–15 kDa.
Additionally, a smaller band ∼10–11 kDa was also observed in the rTr reference protein (Figure 1B, lane 8), being most likely E present in the protein preparation. This smaller band also appeared higher than would be expected for E, due to the His-tag insertion. Hence, we conclude that the commercial rTr is a mixture and contains both Tr/E that are indicated as such on WB by arrows (Figure 1B, lane 8). WB data demonstrate that supernatants of UT and Ad/dl-cells supernatants (Ad/dl-sups) did not show any detectable forms of endogenous Tr/E in the blot (Figure 1B, lane 1–4). In contrast, WB of Ad/Tr-sups in the absence of polyI∶C stimulation revealed protein bands that appeared between Tr/E bands of the reference proteins and thus were considered as Tr (Figure 1B, lane 5).
The bands in lane 5 represent secreted Tr that resulted from Ad/Tr infection of HEC-1A cells. Further, following polyI∶C treatment of Ad/Tr-cells, the intensity of the Tr bands in Ad/Tr-sups was greatly increased, indicating a significantly higher amount of Tr (Figure 1B, lane 6) being produced. Interestingly, a smaller protein band also appeared (Figure 1B, lane 6) that was below the level of rE-reference marker protein, as well as the E band in the rTr reference protein (Figure 1B, lane 7), and thus was considered as E. Taken together, these results indicate that Ad/Tr-cells secrete Tr, and both Tr/E are detected in the supernatants following Ad/Tr-cells stimulation with polyI∶C.[](https://www.ncbi.nlm.nih.gov/mesh/D006639)
Ad/Tr-cells secrete Tr that is functionally active against HNE and both Tr/E are detected in Ad/Tr-cells supernatants in response to polyI∶C.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
HEC-1A cells were either treated with medium alone or with MOI 10–50 of Ad/dl or Ad/Tr. Supernatants were collected 36 h after Ad removal and tested for Tr/E expression by ELISA (A) or for antiprotease activity (C), where results are expressed as percent reduction over a positive control (media alone plus HNE and a substrate). The data are representative of two independent experiments performed in triplicate and are shown as the mean ± SD. Statistical analysis was performed using Student's t test with * representing significant difference between the groups, p<0.05. (B) Immunoblotting analysis of supernatants from HEC-1A cells developed using TRAB20 antibodies. Cells were either left untreated (lanes 1,2) or treated with MOI 50 of Ad/dl (lanes 3,4) or Ad/Tr (lanes 5,6) and then incubated for 24 h in presence of medium alone (−) or 25 µg/ml polyI∶C (+). Two recombinant reference proteins, namely in-house HAT-E (rE) (lane 7) and commercial 6×His-Tr (rTr) (lane 8), were used as comparative markers for E and Tr. Bands corresponding to the forms of Ad-induced Tr and E are indicated on the blot by arrows.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Ad/Tr-cells secrete Tr that is functionally active against HNE
*In the **Ad/Tr-cells secrete Tr that is functionally active against HNE** section:*
Being foremost a protease inhibitor, Tr in Ad/Tr-sups was tested in protease inhibition assay. The results of anti-protease assay showed that Ad/Tr-sups significantly inhibited HNE activity in a dose-dependent manner, compared to Ad/dl-sups (Figure 1C). Following polyI∶C stimulation, Ad/Tr-sups, containing presumably both Tr/E, were also tested and found functional against HNE (data not shown), comparable to before polyI∶C treatment activity.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Tr/E significantly reduce VSV-GFP infection and enhance polyI∶C-driven antiviral protection in Ad/Tr-cells
*In the **Tr/E significantly reduce VSV-GFP infection and enhance polyI∶C-driven antiviral protection in Ad/Tr-cells** section:*
Since polyI∶C stimulation of ECs triggers the induction of antiviral protection, we next determined whether Tr/E could modulate antiviral protection. Results of standard VSV-GFP plaque reduction assays show that the delivery of Ad/Tr significantly reduced VSV-GFP replication in polyI∶C-untreated cells (Figure 2A and 2B) that was beyond the anti-viral protection induced by Ad delivery alone (i.e., in Ad/dl group). Ad delivery is known to activate innate immune responses, and the fact that susceptibility of HEC-1A cells to VSV-GFP infection (Figure 2B) was already reduced by treating the cells with Ad/dl alone confirms the Ad-induced innate activation. Since polyI∶C-untreated Ad/Tr-cells secrete only Tr (Figure 1B), we conclude that the presence of Tr was associated with significantly increased antiviral protection of Ad/Tr-cells in the absence of polyI∶C. Following polyI∶C treatment, VSV-GFP replication was further reduced across the groups in Ad-cells (Figure 2A and 2B); viral replication in Ad/Tr-cells also remained significantly reduced (with up to 50% reduction, p<0.05) and antiviral protection increased, compared to polyI∶C-treated Ad/dl-cells (Figure 2A and 2B). Based on these results and those shown in Figure 1B, we deduce that the presence of both Tr/E was associated with significantly increased cellular antiviral protection in Ad/Tr-cells after polyI∶C treatment.
Altogether, these observations clearly indicate that exogenous Tr expression in Ad/Tr-cells before VSV-GFP challenge has a separate antiviral protective mechanism in addition to the Ad- and polyI∶C-induced responses; furthermore, both Tr/E appear to mediate enhanced polyI∶C-induced cellular antiviral protection in Ad/Tr-cells.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Tr/E significantly reduce VSV-GFP infection and enhance polyI∶C-driven antiviral protection in Ad/Tr-cells.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Untreated (UT) or treated with MOI 50 of Ad/dl or Ad/Tr HEC-1A cells were medium- or polyI∶C 0.1–25 µg/ml-treated for 24 h, followed by MOI 1 VSV-GFP. GFP fluorescence intensity was visualized 24 h later using a Typhoon scanner (A), and relative fluorescence intensity to virus positive control was determined and presented as % of UT control (B). (C) VSV-GFP load after pre-incubation of 5 µg/ml of rTr either with MOI 1 of VSV-GFP (rTr+v) or cells (rTr+c) for 1 h, followed by their washing and addition of VSV-GFP. Medium-treated VSV-GFP (v) served as a positive control. (D) HEC-1A cells were treated with 5 µg/ml of rTr for 1 h in serum-free medium, after which medium or 0.1 µg/ml polyI∶C was added for 24 h followed by MOI 1 of VSV-GFP. This polyI∶C dose was chosen because anti-inflammatory effects of secreted/soluble proteins, compared to Ad/Tr-cells, were less potent. (E) HEC-1A cells were pretreated with equal volumes of supernatants from Ad/dl-cells (Ad/dl-sups) and Ad/Tr-cells (Ad/Tr-sups, 5 µg/ml final concentration) for 1 h before medium or 0.1 or 5 µg/ml polyI∶C was added for 24 h followed by MOI 1 VSV-GFP. Lower polyI∶C doses were used to better visualize VSV-GFP viral replication 24 h later, using a Typhoon scanner. Insert in panel (E) depicts VSV-GFP load after pre-incubation of same volumes and concentrations of Ad/dl- and Ad/Tr-sups with either MOI 1 VSV-GFP (dl+v) or (Tr+v), respectively, or cells (dl+c) or (Tr+c) for 1 h, followed by washing and addition of MOI 1 VSV-GFP. Medium-treated VSV-GFP (v) served as an untreated control. The data are representative of at least two independent experiments performed in triplicate and shown as mean ± SD. Statistical analysis was performed using ANOVA or Student's t test in (C), and * representing significant difference, when p<0.05. (F and G) Cell growth and metabolic activity of HEC-1A cells were determined by standard MTT Cell Viability Assay Kit following infection with MOI 10–50 of Ad/dl or Ad/Tr (F) or after stimulation with increasing doses of polyI∶C (G). Cell viability was expressed as % of untreated cells, which served as a negative control group, and was designated 100%; the results are expressed as % of negative control.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
We next attempted to elucidate specific mode(s) of Tr/E antiviral activity. Given the possibility of additional Tr/E being released from Ad/Tr-cells after polyI∶C stimulation and upon VSV-GFP challenge and thus potentially acting on both virus and cells, we tested if both or each Tr/E individually were acting through two separate mechanisms: (i) direct antiviral activity exerted during the virus/cell encounter and viral binding/entry, targeting either virus or cells; and (ii) indirect cell-associated immunomodulatory activity that targets polyI∶C-triggered cellular antiviral responses and protection. To minimize the Ad-associated effects, we used secreted/soluble proteins delivered to Ad-uninfected HEC-1A cells. Because we could not use Ad/Tr-sups after polyI∶C stimulation as a source of both Tr/E, since they would contain residual polyI∶C and thus mask the effect of Tr/E alone, a commercial rTr (with a C-terminus His-tag) was used as a source of both Tr/E, based on results shown in Figure 1B. For comparative assessment of each Tr/E individually, we used Ad/Tr-sups before polyI∶C stimulation as a source of secreted Tr (no tag), and rE (with an N-terminus HAT-tag) was used as a source of E. All the proteins were initially compared for their antiprotease activity and found equally potent against HNE at around 10 µg/ml (data not shown).
Figure 2C shows that when virus, but not cells, was pre-treated with recombinant rTr for 1 h before addition onto cells, viral replication was significantly reduced by about 20% (p<0.05), compared to media alone. Of note, no further reduction in viral replication was noticed when a higher dose of rTr was used. These results clearly indicate that rTr (a mixture of Tr/E) has a statistically significant, albeit modest, antiviral effect, which appears to be due to direct, or virus-mediated, activity of the proteins. We further show that pretreatment of HEC-1A cells with rTr, followed by co-culturing with polyI∶C for 24 h before VSV-GFP challenge, significantly reduced viral replication (up to 30%, p<0.05) and enhanced polyI∶C-induced antiviral protection, compared to polyI∶C alone (Figure 2D). Interestingly, when cells were pretreated with rTr, then washed and challenged with VSV-GFP 24 h later without prior polyI∶C stimulation, no significant decrease in viral replication was observed (data not shown), suggesting that rTr does not induce potent antiviral cellular responses without polyI∶C stimulation and reduces VSV-GFP replication only following direct contact with virus.
Collectively, these results suggest that secreted/soluble rTr, as a mixture of both Tr/ E, demonstrated two distinct properties: a virus-mediated antiviral activity and the modulation/enhancement of polyI∶C-induced cellular antiviral responses, suggesting that the presence of both Tr/E is required for both of these activities.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Next we assessed antiviral properties of individual secreted/soluble Tr/E preparations. We show that VSV-GFP replication was reduced (up to 40%, p<0.05) in Ad/Tr-sups-treated cells (Figure 2E), compared to controls, but only after polyI∶C treatment (Figure 2E, main graph and insert), suggesting that Tr from Ad/Tr-sups does not exhibit potent direct antiviral activity, but is capable of enhancing polyI∶C-induced cellular responses. Insert in Figure 2E demonstrates that, although a trend toward a decreased viral replication was noted when Ad/Tr-sups were pre-incubated with the virus, there was no significant inhibitory effect observed from Ad/Tr-sups, compared to their controls. Furthermore, surprisingly, in contrast to rTr or Ad/Tr-sups, rE did not exhibit any antiviral properties (data not shown), indicating differential antiviral properties of the tested proteins that could be potentially related to their structural differences. Collectively, these results clearly indicate that both Tr/E appear to be required for antiviral protection mediated via both mechanisms.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Exogenous expression of Tr and polyI∶C stimulation do not lead to impaired cell viability or metabolic activity of Ad/Tr-cells
*In the **Exogenous expression of Tr and polyI∶C stimulation do not lead to impaired cell viability or metabolic activity of Ad/Tr-cells** section:*
Considering that polyI∶C can induce apoptosis, we determined whether Ad/Tr-cells had impaired cell viability and metabolic activity, using a standard MTT Cell Viability Assay. Our results determined no significant changes in viability and metabolic activity among the groups following either Ad, with recovery period of 4 h versus 24 h (data not shown), or polyI∶C treatment (Figure 2F and 2G). Therefore, we can exclude the impairment in the treated cells as the cause of differences between Ad/Tr- and Ad/dl-cells.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Ad/Tr-cells respond to polyI∶C treatment with modulated production of IFNβ
*In the **Ad/Tr-cells respond to polyI∶C treatment with modulated production of IFNβ** section:*
We next elucidated the immunomodulatory effect of Tr/E in Ad/Tr-cells in context of polyI∶C stimulation. Type I IFNs are considered a hallmark of antiviral response, with IFNβ being a key correlate of polyI∶C-induced antiviral protection. We determined whether enhanced antiviral protection in presence of Tr/E was associated with higher levels of IFNβ, using quantitative real-time RT-PCR and commercial ELISA. Figure 3 demonstrates that polyI∶C treatment of Ad-cells triggered a significant induction of IFNβ expression/secretion in contrast to untreated Ad-cells. Interestingly, Ad/Tr-cells responded to polyI∶C treatment with significantly dampened, but not completely abrogated, levels of IFNβ, compared to Ad/dl-cells.
Since IFNβ levels were reduced at both mRNA (Figure 3A) and protein (Figure 3B) levels, it is likely that IFNβ expression was affected primarily at the transcription level. A similar dampening effect was observed when IFNβ protein was assessed at earlier time point (6 h) or in response to a lower dose of polyI∶C (0.1 µg/ml) (data not shown). We have also attempted to measure other members of type I IFNs as well. Namely, multi-subtypes of IFNα were measured by commercial ELISA; however, levels of proteins detected in all supernatants were below the sensitivity of the ELISA and thus considered undetectable (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Ad/Tr-cells respond to polyI∶C treatment with reduced production of IFNβ.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
HEC-1A cells were either treated with medium alone (UT) or with MOI 50 of Ad/dl or Ad/Tr and incubated for 6–24 h in presence of media or 25 µg/ml of polyI∶C. (A) At 6 h post treatment, IFNβ mRNA from total RNA was determined by real-time quantitative RT-PCR. Values are normalized to a housekeeping gene 18S in the same sample and presented as fold induction over UT cells in absence of polyI∶C treatment. (B) At 24 h of polyI∶C stimulation, supernatants were tested for IFNβ expression by ELISA. Data are representative of at least two independent experiments performed in triplicate and expressed as the mean ± SD, shown in pg/ml. Statistical analysis was performed using Student's t test with * representing significant difference between the groups, p<0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Enhanced polyI∶C-driven antiviral protection is associated with hyperactivation of IRF3 in Ad/Tr-cells
*In the **Enhanced polyI∶C-driven antiviral protection is associated with hyperactivation of IRF3 in Ad/Tr-cells** section:*
Phosphorylation and nuclear translocation of IRF3 are key events in the transcriptional activation of inducible ISG cellular genes, and polyI∶C was previously shown to induce activation of IRF3 in vitro . Qualitative and quantitative analyses of phosphorylated IRF3 (pIRF3) showed that, compared to control cells, IRF3 phosphorylation in Ad/Tr-polyI∶C-treated cells was initially modestly reduced at 1 h, but then subsequently increased after 2 h (Figure 4A, pIRF3 WB panel and quantifying histogram) and remained increased for up to 24 h (data not shown). In contrast, no significant changes were evident in the total IRF3 (tIRF3) protein amount for Ad/Tr-cells, compared to controls (Figure 4A, tIRF3 WB panel). Confocal imaging at 4 h post polyI∶C stimulation (Figure 4B) corroborated WB findings of increased IRF3 phosphorylation in Ad/Tr-cells, since significantly more Ad/Tr-cells appeared with IRF3 translocated into the nucleus, compared to Ad/dl-cells (Figure 4B, polyI∶C panel). Taken together, these results suggest that exogenous expression of Tr/E promotes hyperactivation of IRF3 that is associated with increased antiviral protection of Ad/Tr-cells against VSV-GFP challenge.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Enhanced polyI∶C-driven antiviral protection in Ad/Tr-cells is associated with hyperactivation of IRF3.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
(A) Western blots and their analyses of phosphorylated IRF3, total IRF3, and GAPDH proteins were performed from whole-cell extracts of HEC-Ad cells that were either left untreated or treated with 25 µg/ml polyI∶C during indicated time points. (B) Immunofluorescence analysis of IRF3 nuclear translocation following either medium alone or polyI∶C 25 µg/ml treatment for 4 h. Representative staining is shown for IRF3 (green), nuclear stain (PI) (red), and composite (yellow) at magnification 2520×. The data are representative of three independent experiments with similar results.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Tr/E significantly decrease mRNA expression of ISG56 in Ad/Tr-cells following polyI∶C stimulation
*In the **Tr/E significantly decrease mRNA expression of ISG56 in Ad/Tr-cells following polyI∶C stimulation** section:*
ISG15 and ISG56 have been implicated in antiviral protection and linked to activation of IRF3 in response to polyI∶C. Results of RT-PCR demonstrated no difference in mRNA levels of ISG15 between Ad/Tr and Ad/dl groups after polyI∶C stimulation (data not shown). In contrast, mRNA expression of ISG56 (Figure 5A and 5B) appeared significantly reduced in Ad/Tr-cells in response to polyI∶C.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Exogenous expression of Tr/E significantly decreases mRNA expression of ISG56 in Ad/Tr-cells following polyI∶C stimulation.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
HEC-1A cells were either treated with medium alone or with MOI 50 of Ad/dl or Ad/Tr and incubated in presence of media or 25 µg/ml of polyI∶C. (A) At 6 h post treatment, total RNA was harvested and mRNA levels of ISG56 were assessed by conventional RT-PCR. (B) Quantification of ISG56 expression using ImageQuant software. Values are normalized to a housekeeping gene 18S in the same sample and presented as relative fold induction over untreated cells, shown in arbitrary units. The data are representative of at least two independent experiments performed in triplicate and are shown as the mean ± SD. Statistical analysis was performed using Student's t test with * representing significant difference between the groups, p<0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Tr/E alter phosphorylation and transcriptional activity of NF-κB in Ad/Tr-cells in response to polyI∶C
*In the **Tr/E alter phosphorylation and transcriptional activity of NF-κB in Ad/Tr-cells in response to polyI∶C** section:*
We next elucidated phosphorylation and activation of other transcription factors, including NF-κB and c-Jun, that have also been shown to contribute to antiviral responses. Qualitative and quantitative results show that NF-κB p65 phosphorylation in Ad/Tr-cells was attenuated after 1 h post-polyI∶C exposure and remained decreased until 8 h, compared to the control group (Figure 6A, NF-κB WB panel and Figure 6B, NF-κB quantifying histogram). Our results further revealed that overall c-Jun phosphorylation was only transiently reduced in Ad/Tr-cells between 1 h and 4 h and returned to levels comparable to such of controls at 8 h post polyI∶C treatment (Figure 6A, pc-Jun WB panel and Figure 6B, pc-Jun quantifying histogram). These results prompted us to next evaluate the effect of Tr expression on transcriptional activity of NF-κB and c-Jun, which was assessed by luciferase reporter gene assay. Data shown in Figure 6C, NF-κB/Luc panel, demonstrate that transcriptional activity of NF-κB in Ad/dl-polyI∶C-treated cells was significantly induced at 4 h, compared to untreated and LPS-treated Ad/dl-cells.
However, polyI∶C-induced NF-κB transcriptional activity in Ad/Tr-cells was significantly reduced, compared to Ad/dl-cells. In contrast, transcriptional activity of AP-1 was not different between the groups following stimulation with either of the ligands (Figure 6C, AP-1/Luc panel). Collectively, our data show that Tr/E expression attenuated NF-κB activation not only at the phosphorylation level, but also at the level of its transcriptional activity, which was more pronounced compared to changes observed for c-Jun/AP-1.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Altered NF-κB phosphorylation and transcriptional activity and reduced RIG-I and MDA5 expression in polyI∶C-treated Ad/Tr-cells.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Western blots (A) and their quantification (B) of phosphorylated p65 subunit of NF-κB (pNF-κB) as well as c-Jun (pc-Jun) and GAPDH proteins were performed from whole-cell extracts of Ad-cells that were either left untreated or treated with 25 µg/ml polyI∶C during indicated time points. Transcription activity (C) of NF-κB (NF-κB/Luc panel) and (AP-1/Luc panel) was assessed by luciferase reporter assay by transfecting HEC-1A cells with pgkβ-Gal and pNF-κB-Luc or pAP-1-Luc plasmids (total DNA 40 ng per well) (p) alone or together with 50 MOI of Ad/dl or Ad/Tr overnight, washing, allowing to rest for 4 h and stimulating with 25 µg/ml polyI∶C and 1 µg/ml LPS for 4 h. Then luciferase and β-galactosidase activities were determined in cell lysates and expressed as relative luciferase units using galactosidase plasmid as normalization control. Data are presented as mean ± SD and are representative of three experiments for NF-κB and two - for AP-1. (D–F, left panels) Total RNA was harvested and mRNA expression of RIG-I, MDA5, and TLR3 was determined by real-time quantitative RT-PCR at 6 h post treatment with polyI∶C 25 µg/ml. Values are normalized to a housekeeping gene 18S in the same sample and presented as fold induction over untreated cells. (D–F, right panels) Western blot analysis of RIG-I, MDA5, TLR3, and GAPDH protein was performed from whole-cell extracts of Ad-cells that were either treated or not with 25 µg/ml polyI∶C during indicated time points. The data are representative of at least two independent experiments performed in triplicate and are shown as the mean ± SD. Statistical analysis was performed using Student's t test with * representing significant difference between the groups, p<0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Tr/E significantly reduce levels of dsRNA sensors RIG-I and MDA5 in polyI∶C-treated Ad/Tr-cells
*In the **Tr/E significantly reduce levels of dsRNA sensors RIG-I and MDA5 in polyI∶C-treated Ad/Tr-cells** section:*
PolyI∶C-induced antiviral protection of primary genital ECs was previously associated with heightened expression of TLR3. Thus, we evaluated the expression levels of dsRNA sensors in Ad/Tr-cells following polyI∶C treatment. Figures 6D–F show that in contrast to TLR3, polyI∶C stimulation induced a significant (over 40 times) increase in mRNA expression of RIG-I and MDA5 (Figure 6D–E, left panels) in all UT and Ad-cells. Further, compared to UT and Ad/dl-cells, Ad/Tr-cells surprisingly had significantly attenuated expression of RIG-I and MDA5 at both mRNA and protein levels (Figure 6D–E, left and right panels). RIG-I expression in Ad/Tr-cells appeared to be affected at earlier time point, around 1 h, compared to MDA5 expression that was attenuated at 8 h after polyI∶C stimulation.
Moreover, reduced expression levels of RIG-I and MDA5 were sustained for up to 24 h following polyI∶C treatment (data not shown). The expression of TLR3, however, was not different between the groups (Figure 6F). Collectively, these data indicate that mRNA and protein expression of RIG-I and MDA5 are increased in response to polyI∶C stimulation, but the magnitude of expression is reduced in Ad/Tr-cells. Further, that we observed differential pattern of expression among RIG-I, MDA5, and TLR3 could indicate either a different kinetics of responses of these sensors, or that each RIG-I, MDA5, and TLR3 respond differentially to polyI∶C, known to be a mixture of various lengths of dsRNA.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## IRF3 is required for polyI∶C-driven antiviral protection in Ad/Tr-cells
*In the **IRF3 is required for polyI∶C-driven antiviral protection in Ad/Tr-cells** section:*
We next determined whether IRF3 was required for enhanced polyI∶C-driven antiviral state in Ad/Tr-cells. IRF3 was knocked down by utilizing IRF3-specific small interfering RNA and cells were subsequently treated, or not, with polyI∶C before challenging with VSV-GFP. The greatest knockdown efficiency was observed between 72 h and 96 h post transfection (data not shown). Results of WB demonstrated no apparent expression of tIRF3 after siRNA treatment (Figure 7A). Further, VSV-GFP replication in polyI∶C-untreated Ad/dl- and Ad/Tr-cells was not significantly altered when IRF3 was knocked down (Figure 7B).
This observation suggests that IRF3 is dispensable for antiviral defense against VSV-GFP infection in both Ad/dl and Ad/Tr groups in the absence of polyI∶C treatment and that markedly reduced viral replication in Ad/Tr-cells was attributed to other, yet unidentified, factor(s). In contrast, when polyI∶C was added, VSV-GFP replication was initially markedly reduced in both Ad/dl and Ad/Tr groups in the presence of IRF3, but then significantly increased in the absence of IRF3; yet, Ad/Tr-cells still remained more protected than Ad/dl group. This observation implies that enhancement of polyI∶C-triggered antiviral protection in Ad/Tr-cells was only partially dependent on IRF3. Figure 7B also showed that VSV-GFP replication was restored about 50% of the original viral load detected in each of the groups in absence of IRF3 and polyI∶C stimulation, suggesting that additional mechanisms/factors were contributing to polyI∶C-induced antiviral protection. Collectively, these findings indicate that IRF3 plays an equally important role in antiviral protection in both Ad/dl and Ad/Tr groups, and that the presence of IRF3 is important, but not essential, for enhanced polyI∶C-induced antiviral protection in Ad/Tr-cell, as other factors, perhaps upstream of IRF3, may also be contributing to this protection.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
IRF3 is required for polyI∶C-driven antiviral protection in Ad/Tr-cells.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
(A) HEC-1A cells were left untreated or transfected with a non-targeting control siRNA (ctrl siRNA), or IRF3 siRNA (IRF3 siRNA) for 48 h. Two days after siRNA delivery, HEC-1A received MOI 50 of Ad/dl or Ad/Tr followed by 25 µg/ml of polyI∶C treatment for 24 h and Western blot analyses of total IRF3 (tIRF3) and GAPDH proteins were performed from whole-cell extracts 96 h post-transfection. (B) Twenty four hours following polyI∶C treatment, cells were infected with MOI 1 of VSV-GFP for another 24 h. Levels of GFP fluorescence were visualized and quantified using a Typhoon scanner. The fluorescence reading of treated cultures was normalized to untreated (control) cultures and presented as percentage relative fluorescence. The data are representative of two independent experiments performed in triplicate and are shown as the mean ± SD. Statistical analysis was performed using Student's t test with * representing significant difference between the groups, p<0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Tr/E significantly reduce pro-inflammatory cytokines in ECs following polyI∶C and LPS stimulation
*In the **Tr/E significantly reduce pro-inflammatory cytokines in ECs following polyI∶C and LPS stimulation** section:*
PolyI∶C stimulation induces not only antiviral, but also pro-inflammatory factors that are regulated by viral sensors and transcription factors, including NF-κB and c-Jun. We therefore assessed levels of pro-inflammatory mediators IL-8, TNFα, and IL-6 in Ad/Tr-cells treated with polyI∶C. Figures 8A–C demonstrate that Ad/Tr-cells, regardless of their origin (i.e., genital HEC-1A or gut Caco-2, etc.), secreted significantly lower levels of IL-8 24 h after polyI∶C and LPS treatment, compared to controls; TNFα and IL-6 were similarly reduced in Ad-cells in response to polyI∶C (data not shown). Interestingly, stimulation of HEC-1A with LPS did not produce a significant increase in IL-8, compared to untreated cells, possibly due to a low baseline expression of TLR4 in the genital EC. These results clearly indicate that Tr/E controlled the release of pro-inflammatory mediators in ligand-treated cells. Interestingly, Tr/E secreted by Ad/Tr-cells appeared to significantly contribute to reduced IL-8 levels.
Indeed, Tr/E neutralization with specific anti-Tr/E TRAB20 (HyCult Biotech) antibodies (under the control of antiprotease assay) that were added to Ad/Tr-cells 1 h prior to 0.1 µg/ml of polyI∶C treatment and subsequently co-cultured with polyI∶C for additional 24 h to neutralize any secreted Tr/E, led to significantly higher polyI∶C-induced IL-8 secretion (up to 40%, p = 0.03) in Ad/Tr-cells (data not shown). Figures 8D and 8E further show that HEC-1A cells pre-treated with either Ad/Tr-sups, rTr or rE before subsequent co-culture with polyI∶C also released lower levels of IL-8 in response to polyI∶C. Of note, anti-inflammatory effects of secreted/soluble proteins, compared to Ad/Tr-cells, appeared to be less potent and context-dependent, which forced us to use a lower dose of 0.1 µg/ml of polyI∶C. Additionally, either Ad/Tr-sups or rTr/rE alone did not trigger any significant IL-8 production in absence of polyI∶C. Collectively, our findings indicate that Tr/E are capable of inhibiting inflammatory responses in ECs from various sources and against both viral and bacterial PAMPs, and that secreted/soluble Tr/E significantly contribute to reduction in polyI∶C-induced IL-8 secretion. However, stimulation of immune responses can also be observed depending on experimental conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Tr/E significantly reduce protein secretion of IL-8 in ECs following polyI∶C and LPS stimulation.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Secretion of IL-8 in supernatants from HEC-1A (A), A549 (B), and Caco-2 (C) cells initially infected with MOI 50 of Ad/dl or Ad/Tr and subsequently stimulated with either 25 µg/ml of polyI∶C or 1 µg/ml of LPS for 24 h. Following stimulation, supernatants were tested for IL-8 secretion by ELISA. Data are representative of at least two independent experiments performed in triplicate and expressed as the mean ± SD, shown in pg/ml. Statistical analysis was performed using Student's t test with * representing significant difference between the groups, p<0.05. (D) Secretion of IL-8 in supernatants from HEC-1A cells that were pre-treated with HEC-Ad/dl and HEC-Ad/Tr supernatants before polyI∶C stimulation. HEC-1A received 50 µl of concentrated supernatants containing around 10 µg/ml of Tr/E in Ad/Tr supernatants and 0.004 µg/ml of Tr/E in Ad/dl sups for 1 h, to which additional 50 µl of medium alone or polyI∶C to the final concentration of 0.1 µg/ml were added for 24 h. (E) Secretion of IL-8 in supernatants from HEC-1A cells treated with commercial 6×His-Tr (rTr) or in-house HAT-E (rE) for 1 h and then stimulated with medium alone or 0.1 µg/ml of polyI∶C for 24 h. A lower dose of polyI∶C was used in (D) and (E), since anti-inflammatory effects of secreted/soluble proteins, compared to Ad/Tr-cells, were less potent. Statistical analysis was performed using ANOVA, and * representing significant difference between the polyI∶C group and rTr groups; † representing significant difference between the polyI∶C and rE groups, p<0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
## Discussion
*In the **Discussion** section:*
WAP proteins, including antiproteases Tr/E, SLPI, and ps20, are pleiotropic molecules known to play multiple and significant roles in health and disease. Indeed, ps20 has been reported as a potential diagnostic marker in prostate cancer and as a novel negative signature protein in HIV infection. In contrast, SLPI and Tr/E show significant therapeutic potential in atherosclerosis as well as cardiovascular, lung and gut disorders. Additionally, higher levels of Tr/E in CVLs of HIV-resistant CSWs and the testing of the Lactobacilli-based elafin delivery system for combating STIs in the FGT would further support this notion. Our data showed that delivery of Ad/Tr to HEC-1A cells resulted in secretion of functional Tr, while both Tr/E were detected following treatment of these cells with polyI∶C. Moreover, polyI∶C treatment further resulted in Tr/E-enhanced antiviral protection and significantly reduced pro-inflammatory IL-8, IL-6, TNFα that were associated with lower expression of viral innate sensors RIG-I and MDA5 and altered NF-κB activation in Ad/Tr-cells.
Notably, increased antiviral protection was due in part to Tr/E ability to act directly on virus or by modulating polyI∶C-driven cellular antiviral responses. Interestingly, such Tr/E-augmented cellular responses triggered by polyI∶C were partially mediated through IRF3 activation, but not higher induction of IFNβ, thus suggesting multiple antiviral mechanisms of Tr/E and the involvement of alternative and still unidentified factors or pathways.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
This is the first study that comparatively assessed the presence and potential mechanisms of antiviral activity of each Tr/E. Here, we presented evidence showing two distinct, but likely complimentary antiviral properties of Tr/E: (i) direct antiviral activity exerted during the virus/cell interaction and targeting virus, but not cells; and (ii) indirect and cell-associated immunomodulatory activity, targeting polyI∶C-triggered cellular antiviral responses. The virus-mediated activity was observed in the absence of polyI∶C stimulation and in the presence of Tr in Ad/Tr-cells (presumably present in both Ad/Tr-sups and Ad/Tr-cells). Because only Tr was detected in polyI∶C-untreated Ad/Tr-cells, and because the expression of IFNβ was low and not different between untreated Ad/dl and Ad/Tr groups, we conclude that in the absence of polyI∶C, Tr alone was mediating antiviral activity in Ad/Tr-cells by acting either directly on virus or indirectly through cells. However, we failed to transfer this direct antiviral effect of Tr via Ad/Tr-sups, possibly due to the absence of an additive protective effect from augmented intracellular Tr (as would be expected in Ad/Tr-cells),or due to an inefficient delivery of Tr in supernatants and a “diluting" effect from other antiviral factors released in response to Ad/dl delivery (Fig. 2E, insert).[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Additionally, the presence of both Tr/E, as in rTr preparation, was also protective against VSV-GFP challenge in the absence of polyI∶C treatment. Although no reports describing direct antiviral activity of Tr are published to date and no precise mechanisms of Tr/E direct antiviral effect have been identified, our observation with rTr is in line with Ghosh et al. findings, describing a direct antiviral effect of E as a mode of action against HIV. Interestingly, rTr also appeared to have only virus- and not cell-mediated protective effects, similar to E mentioned earlier. In contrast, a close WAP member SLPI, was shown to have only cell-mediated antiviral effects, at least against HIV and herpes simplex virus (HSV). Examples of same-family members having differential antiviral mechanisms have also been described for other innate molecules, such as human defensins against HIV and HSV.
These reports suggest that molecules even from the same group, may possess their own, potentially different in potency and targets, exquisite antiviral activities and yet still uniquely contribute to overall mucosal protection against STIs.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Our results further suggest that the presence of both Tr/E might be required for each virus- (direct) and cell-associated (indirect) antiviral effect, as was evident from our data using Ad/Tr-cells, Ad/Tr-sups, and rTr, indicating that perhaps the most efficient antiviral protection depends on collaborative work of both Tr/E. Interestingly, when HIV-susceptible, but uninfected, CSWs were followed prospectively, those who remained HIV-negative had elevated levels of both Tr/E detected in CVLs. Additionally, when characterizing the specificity of proteins secreted after Ad/Tr infection, our ELISA and WB results also showed that Ad/Tr-cells secreted both Tr/E independently in response to polyI∶C, while only Tr was detected without polyI∶C stimulation. Although Ghosh et al. also reported that primary uterine EC produced Tr/E in response to polyI∶C, the independent production of E was never demonstrated. Further, only Tr (13–16 kDa) was previously identified in supernatants from LPS-stimulated alveolar ECs. On conjuncture, these results suggest that expression of each Tr/E could be a tissue/cell- or ligand-specific defense mechanism against an unknown protease that was potentially activated in response to a viral ligand.
It might be important in the future to clarify whether primary genital EC from the FGT produce each Tr/E independently in response to polyI∶C, similarly to Ad/Tr-cells.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
It is unclear why tested rE failed to show antiviral activity against VSV-GFP; it could be attributed, however, to a HAT-tag insertion at the N-terminus of rE. Indeed, all rTr, rE, and secreted Tr were equally functional against HNE (data not shown) and capable of inhibiting IL-8 production in response to polyI∶C (Figure 8). Yet, while rTr with a His-C-terminus tag exhibited antiviral activity, rE with a HAT-N-terminus tag did not. This observation suggests that blocking N-terminus, but not C-terminus, appears to be critical for antiviral activity of rE. An earlier study by McMichael et al. supports this argument, since they showed that the N-terminus of Tr had a better affinity for LPS than its C-terminus end. Additionally, it is unclear why we observed increased levels of IL-8 with higher concentrations of rTr and rE. But one possible explanation could be that the proteins were initially delivered and left on cells in serum-free conditions, thus promoting the activation of pro-inflammatory events as was previously shown for Tr/E in response to LPS.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
The second antiviral property of Tr/E observed in our study was an indirect cell-mediated immunomodulatory activity of Tr/E, targeting polyI∶C-induced antiviral cellular responses. In contrast to responses to bacterial or pro-inflammatory stimuli, the scope and specific mechanism(s) of viral ligand-triggered immunomodulatory activity of Tr/E have never been fully investigated. Our data demonstrate that this indirect cell-associated activity is targeting viral recognition through modulation of RNA helicase expression as well as the induction of key inflammatory and antiviral innate signaling pathways and mediators.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Our results showed that polyI∶C-triggered antiviral cellular protection was significantly enhanced in the presence of Tr (in Ad/Tr-sups) and Tr/E (in Ad/Tr-cells and in rTr). We also showed that polyI∶C-mediated activation of IRF3 was further induced in Ad/Tr-cells, compared to controls, whereas IFNβ expression was dampened. Interestingly, human β defensin 3 and cathelicidin LL37 that were previously shown to have antiviral, including anti-HIV, activity, were also reported to inhibit IFNβ production in vitro in response to LPS and polyI∶C, respectively. These observations further support our results and strengthen the earlier argument of Tr/E acting either directly against VSV-GFP or through cells and additional factors/pathways in Ad/Tr-cells. Furthermore, moderation of immune-inflammatory responses and thus curbing undesirable immune activation might be one of the protective mechanisms of innate antimicrobials at mucosal sites.
Additionally, in searching for ISGs typically associated with antiviral protection and IRF3 activation, we found that expression of ISG15 was not significantly changed, unlike ISG56 being reduced and in agreement with IFNβ data. It is not entirely understood why such discordance was observed; however, it could be due to the fact that ISG15 was shown to be regulated by either IRF3 or IFNβ, unlike ISG56 that was shown to be under the regulation of IFNβ or viruses and thus following IFNβ pattern of induction as shown in our study. The alternative explanation could be that these two genes follow a different temporal pattern of activation that was overlooked here.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
This is the first report on the involvement of serine antiproteases, Tr/E in particular, in antiviral signaling pathways. As no prior data are available on the role of Tr/E in IFNβ and IRF3 induction, further and more detailed investigations might be required to explain why in Ad/Tr-cells IFNβ and ISG56 levels were reduced while IRF3 activation was increased. We hypothesize, however, that this phenomenon could be an attempt of Tr/E to control antiviral inflammatory events through RIG-I/MDA5 and NF-κB downregulation while increasing cellular protection through activation of IRF3 and/or alternative factors or pathways. Although most of the studies show ISG56 to be associated with upregulated IRF3, our finding is in line with data from Li et al. showing that a knockdown of ISG56 was associated with increased IRF3 activation and inhibition of VSV-GFP replication as a result of ISG56 mediating MITA-TBK1 interaction and subsequent downstream activation of IRF3.
It is also possible that alternative factors/pathways, in addition to IRF3, regulate IFNβ and ISG56 expression and contribute to Tr/E-enhanced antiviral protection, which is also supported by our IRF3 siRNA data. Collectively, these data indicate that in the presence of Tr/E, antiviral protection is increased and that direct or indirect antiviral effect(s) of Tr/E depend, but not exclusively, on IRF3 and other factors, perhaps upstream of IRF3.
Inflammation is one of the leading factors predisposing to acquisition and disease progression of STIs in the FGT. This notion is supported by the fact that “immune quiescence" and reduced immune activation are crucial for resistance against STIs, while dysregulated TLR expression and immune-inflammatory responses are detrimental. Here, we showed that Tr/E individually or as a mixture, as well as in Ad/Tr-cells and as secreted/soluble proteins in Ad/Tr-sups, were capable of reducing IL-8, IL-6, and TNFα expression in response to polyI∶C. Moreover, in Ad/Tr-cells we also observed significantly reduced activation and transcriptional activity of NF-κB. The IL-8 inhibitory effect was not specific to human endometrial ECs, or to polyI∶C, indicating that similar effects could be observed at other mucosal surfaces and in response to different microbial ligands. We further showed that, compared to controls, mRNA and protein levels of RIG-I and MDA5 (mainly at a later time point), but not TLR3, were significantly diminished in response to polyI∶C and in presence of Tr/E. Immunomodulatory properties of both Tr/E demonstrated in models of pro-inflammatory and bacterial (LPS) stimulations were shown to depend on inhibition of NF-κB and AP-1 activation, thus further supporting our NF-κB data. However, Tr/E inhibitory effect targeting antiviral immune responses, including viral sensing, has not been previously reported.
Hence, modulation of expression of RIG-I, MDA5, and pro-inflammatory mediators shown here could represent novel antiviral functions of Tr/E, possibly even executed at different levels, namely receptors and transcription factors. That we observed differential pattern of RIG-I, MDA5, and TLR3 expression could indicate either different temporal kinetics of responses of these sensors, or that each RIG-I, MDA5, and TLR3 respond differentially to polyI∶C, being a mixture of variable lengths of dsRNA. Further, the lack of polyI∶C-triggered TLR3 induction in HEC-1A compared to primary genital ECs, also likely reflects tissue or structure-dependent differences between the cells, suggesting that primary genital ECs may exhibit distinct results. Collectively, these observations suggest that Tr/E can alter innate viral recognition and mounting of antiviral immune-inflammatory responses.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
The precise mechanism(s) of immunomodulatory effects of Tr/E, as both secreted and/or intracellularly expressed proteins, on viral sensors, cytokines, and IFNβ in response to polyI∶C treatment is still largely unknown. This is partly because the existence of the cognate receptor for Tr/E remains elusive, and it is equally unknown whether Tr/E require a receptor to function. We propose that reduced levels of IL-8, IL-6, TNFα and IFNβ in Ad/Tr cells in response to polyI∶C are likely a result of overall attenuation of RIG-I and MDA5 levels, as they are known to regulate the expression of pro-inflammatory and antiviral mediators through activation of main signaling pathways, such as NF-κB that is downregulated in our study. It remains to be elucidated, however, how Tr/E specifically inhibit RIG-I and MDA5 expression. It is plausible that Tr/E directly bind to polyI∶C, as was shown for binding of LL37 to polyI∶C, as well as Tr/E binding to LPS.
Such an interaction may alter binding/recognition of polyI∶C by its cognate receptors, including cell-surface scavenger receptor A or intracellular sensors RIG-I, MDA5, and TLR3, which in turn could explain our reduced expression levels of RIG-I and MDA5. Another possible site of inhibition by Tr/E could be downstream of receptors/viral sensors and involve Tr/E binding to DNA and competing for specific DNA binding sites with transcription factors including NF-κB, as was shown for SLPI as one of its anti-inflammatory mechanisms in response to LPS.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
A noteworthy observation of this study is that while Tr from Ad/Tr-sups and rE were found functional against HNE and able to inhibit polyI∶C-induced IL-8 production, they did not show any antiviral activity, suggesting that antiprotease, anti-inflammatory, and antiviral activities of the tested proteins may not necessarily be co-dependent or predictive of each other; nonetheless, they can be complimentary. This observation is supported by earlier reports, showing both a protease non-inhibitory N-terminus and an inhibitory C-terminus of Tr exhibiting comparable antibacterial and antifungal functions. In contrast, Mulligan et al. showed that SLPI Gly(72) mutant, unlike other mutants tested in that study, lost its in vivo immunosuppressive activity against NF-κB activation and neutrophil recruitment in the lungs that appeared to be most closely related to SLPI's trypsin-inhibiting activity. Although being an important property of both Tr/E, the inhibition of HNE activity is not considered a critical function for our studies, since epithelial cells do not make neutrophil elastase and thus, the earlier discussed Tr/E-mediated changes are most unlikely attributed to antielastase activity of the proteins. The above observations indicate that perhaps additional structure-function studies might be warranted in the future to specifically address the cross-talk between antiprotease, anti-inflammatory, and antiviral properties of Tr/E and their specific roles in defense against viruses.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Overall, our data support and further extend earlier observations on immunomodulatory effects of Tr/E. This work demonstrates that in genital ECs and in response to polyI∶C, Tr/E antiviral effects are mediated through direct or virus targeting activity and indirect or cell-associated immunomodulatory function(s) that target host innate recognition and mounting of antiviral and inflammatory responses. While dampening of IFNβ, a key antiviral mediator, may seem counterintuitive and detrimental to antiviral defenses, our findings suggest that directly or indirectly increased antiviral protection and moderated, or finely-tuned, inflammation, might be more advantageous to a host in the context of viral exposure. In conclusion, this study clearly demonstrates the importance of Tr/E in antiviral protection. Our findings also propose the existence of multiple targets and potentially several and unique modes of action for each of the proteins, which warrant additional research in the future.[](https://www.ncbi.nlm.nih.gov/mesh/D011070)
Competing Interests: The authors have declared that no competing interests exist.
Funding: This work was supported by a grant as part of the Comprehensive T Cell Vaccine Immune Monitoring Consortium (CTC-VIMC), a key component of the Collaboration for AIDS Vaccine Development (CAVD) funded by the Bill & Melinda Gates Foundation. A.G.D. was supported by a Studentship and K.L.R. by a Career Scientist Award from the Ontario HIV Treatment Network (OHTN). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
# References
*In the **References** section:*
|
# Introduction
Cortactin Is Involved in the Entry of Coxiella burnetii into Non-Phagocytic Cells
# Abstract
*In the **Abstract** section:*
Background
Cortactin is a key regulator of the actin cytoskeleton and is involved in pathogen-host cell interactions. Numerous pathogens exploit the phagocytic process and actin cytoskeleton to infect host cells. Coxiella burnetii, the etiologic agent of Q fever, is internalized by host cells through a molecular mechanism that is poorly understood.
Methodology/Principal Finding
Here we analyzed the role of different cortactin motifs in the internalization of C. burnetii by non-phagocytic cells. C. burnetii internalization into HeLa cells was significantly reduced when the cells expressed GFP-cortactin W525K, which carries a mutation in the SH3 domain that renders the protein unable to bind targets such as N-WASP. However, internalization was unaffected when the cells expressed the W22A mutant, which has a mutation in the N-terminal acidic region that destroys the protein’s ability to bind and activate Arp2/3. We also determined whether the phosphorylation status of cortactin is important for internalization. Expression of GFP-cortactin 3F, which lacks phosphorylatable tyrosines, significantly increased internalization of C. burnetii, while expression of GFP-cortactin 3D, a phosphotyrosine mimic, did not affect it. In contrast, expression of GFP-cortactin 2A, which lacks phosphorylatable serines, inhibited C. burnetii internalization, while expression of GFP-cortactin SD, a phosphoserine mimic, did not affect it. Interestingly, inhibitors of Src kinase and the MEK-ERK kinase pathway blocked internalization. In fact, both ki[nases rea](https://www.ncbi.nlm.nih.gov/mesh/D014443)ched maximal activity at 15 min of C. burnetii infection, after which activity decreased to basal [levels. Despite](https://www.ncbi.nlm.nih.gov/mesh/D019000) the decrease in kinase activity, cortactin phosphorylation at Tyr421 reached a peak at 1 h of infect[ion.](https://www.ncbi.nlm.nih.gov/mesh/D012694)https://www.ncbi.nlm.nih.gov/mesh/D010768)
Conclusions/Significance
Our results suggest that the SH3 domain of cortactin is implicated in C. burnetii entry into HeLa cells. Furthermore, cortactin phosphorylation at serine and dephosphorylation at tyrosine favor C. burnetii internalization. We present evidence that ERK and Src kinases play a role early in infection by this pathogen.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Phagocytosis is the process that cells have developed for the engulfment of particulate material such as apoptotic cells, cell debris and, also, inert particles. Furthermore, phagocytosis represents a crucial event that triggers host defense mechanisms against invading pathogens. Nevertheless, several pathogens have acquired different strategies to alter these mechanisms to survive and multiply within host cell, causing infectious diseases , . The phagocytic process is initiated by a recognition step in which ligands on the particle surface bind receptors on the membrane of host cells . The ligand-receptor interaction leads to actin cytoskeleton and membrane rearrangements that permit, first, particle engulfment and, later, particle sequestration into a phagosome which precedes phagosome maturation into a phagolysosome , .
Dynamic remodeling of the actin cytoskeleton is not only intimately involved in phagocytosis but also in other essential cellular processes, including cell adhesion and motility , vesicle transport , , apoptosis and endocytosis , all of which require dynamic remodeling of the actin cytoskeleton. There are numerous actin-associated proteins and several upstream signaling molecules that work in a coordinated way to control with exquisite precision the spatial and temporal assembly of actin structures, which can rapidly change in response to internal and external signals , . Proteins of the Arp2/3 complex that function as nucleators of branched actin filaments are activated by interaction with members of the Wiskott-Aldrich syndrome protein (WASP) family and cortactin , . Initial activation of WASP depends on its interaction with Rho family GTPases . These multicomponent complexes of Arp2/3-WASP-cortactin are involved in cellular processes such as cell motility , endocytosis and phagocytosis , .
Interestingly, some pathogens can regulate the host actin cytoskeleton during infection , .
Cortactin is a key regulator of the actin cytoskeleton, and it plays a crucial role in tumor cell invasion , ruffles and lamellipodium formation during integrin-mediated cell adhesion , and podosome formation . Cortactin is also an important component of the endocytic machinery . It has emerged as a common target of pathogen-host cell interactions. For example, cortactin has been implicated in the adhesion of Escherichia coli and in invasion by Shigella, Neisseria, Chlamydia, Staphylococcus and Listeria. The phosphorylation status of cortactin has been proposed to differentially regulate the invasion of many microbial pathogens.
Cortactin is also involved in actin-based motility of many pathogens during their intracellular trafficking .
Cortactin possesses an N-terminal acidic domain (NTA) and F-actin-binding repeats that activate the Arp2/3 complex to initiate actin polymerization . Cortactin also has a proline-serine-threonine-rich region (PST) that contains tyrosine residues critical for cortactin function. The C-terminal SH3 domain of cortactin binds various proteins, such as N-WASP proteins , . The Verprolin Cofilin Acidic domain (VCA) of WASP members can also activate the Arp2/3 complex . Theoretically N-WASP, cortactin and the Arp2/3 complex can form ternary complexes .
Cortactin is phosphorylated by tyrosine kinases (Src, Fer, Syk and Abl) and serine/threonine kinases (ERK and Pak) in response to a wide range of stimuli that induce cytoskeletal rearrangement, including growth factor stimulation, cell adhesion and hyperosmotic stress .[](https://www.ncbi.nlm.nih.gov/mesh/D011392)
Coxiella burnetii, the causative agent of human Q fever, is an obligate intracellular bacterium found in a wide range of hosts, including livestock and humans. The primary route of infection in humans is inhalation of contaminated aerosols , . Infected animals shed C. burnetii in their milk, urine and feces, and the bacteria are dispersed together with amniotic fluids and the placenta during birthing. These bacteria can survive for long periods in the environment, since they are highly resistant to heat, desiccation and common disinfectants.
C. burnetii inhabits mainly monocytes/macrophages but can infect a wide variety of cultured cell lines in vitro . This bacterium resides in an acidic parasitophorous vacuole (PV) with late endosome-lysosome characteristics –. The PV also interacts with the autophagic pathway, acquiring autophagosomal features , , . Interestingly, we have shown that PV biogenesis is regulated by actin and Rho family GTPases .
In this report we describe the involvement of cortactin in C. burnetii internalization into HeLa cells, a non-professional phagocyte cell line. We investigated the role of the Arp2/3-activating DDW motif in the N-terminal acidic region and of the SH3 domain at the C-terminus of cortactin during C. burnetii internalization. We observed that overexpression of cortactin mutated in the SH3 domain inhibits uptake of the bacterium, suggesting that the SH3 domain is important for internalization. We also analyzed the role of cortactin phosphorylation in internalization. By overexpressing cortactin mutants that are non-phosphorylatable and that mimic phosphorylation, we show that cortactin favors C. burnetii internalization in a tyrosine dephosphorylation- and/or serine phosphorylation-dependent manner.
Furthermore, pharmacological inhibition of Src and ERK kinases reduce C. burnetii uptake. Our results indicate that phosphorylation status of cortactin affect internalization of C. burnetii.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
The SH3 domain of cortactin is important for C. burnetii internalization.
(A) HeLa cells were transfected with pGFP-cortactin WT, GFP-cortactin W22A, a mutant that does not interact with Arp 2/3, or GFP-W525K, a mutant unable to bind and activate targets such as N-WASP. Transfected cells were infected for 4 h with C. burnetii. Cells were fixed and processed for immunofluorescence with a specific anti-C. burnetii antibody (see Methods). Cells were analyzed by confocal microscopy. In the merged images (panels d, h, l and p), extracellular C. burnetii are shown in green and red pseudocolors while intracellular C. burnetii are shown in red pseudocolor. Bars, 10 µm. (B) Quantification of C. burnetii internalized by transfected HeLa cells. Results are expressed as means ± SE of at least three independent experiments. *, P<0.05. (%), percentage of total number of bacteria.
## Methods
*In the **Methods** section:*
## Materials
*In the **Materials** section:*
Dulbecco’s Modified Eagle's Medium (D-MEM), fetal bovine serum (FBS), penicillin and streptomycin were obtained from Gibco BRL/Life Technologies (Buenos Aires, Argentina). Vectors encoding a fusion of green fluorescent protein (GFP) with cortactin WT (full-length cortactin) or GFP-cortactin 3F (cortactin mutated in the three tyrosine-phosphorylation sites recognized by Src) were kindly provided by S. Bourdoulous (Département de Biologie Cellulaire, Institut Cochin, Université Paris, Paris, France). Rabbit polyclonal anti-Coxiella antibody against Nine Mile phase II, clone 4 (RSA439) was generously provided by Dr. Robert Heinzen (Rocky Mountain Laboratories, NIAID, NIH, Hamilton, MT, USA). Secondary antibodies were purchased from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA, USA). Rabbit monoclonal anti-phosphocortactin (Tyr421) antibody was purchased from Abcam (MA, USA), mouse monoclonal anti-actin Ab-5 antibody was purchased from BD (Buenos Aires, Argentina), rabbit anti-phosphoSrc (Tyr416) (Cell Signaling Inc., MA, USA) and anti-Src antibodies were generously provided by Arlinet Kierbel (Montevideo Pasteur Institute, Montevideo, Uruguay), and mouse monoclonal anti-phosphoERK1/2 (Tyr204) and rabbit polyclonal anti-ERK antibodies were purchased from Santa Cruz Biotechnology (California, USA).
The inhibitors PD98059 and SU6656 were from Invitrogen (Buenos Aires, Argentina) and CalBiochem (Darmstadt, Germany), respectively. Protease inhibitor cocktail was from Sigma (Buenos Aires, Argentina).[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Cell Culture
*In the **Cell Culture** section:*
HeLa cells were grown in DMEM supplemented with 10% heat-inactivated FBS, 2.2 g/l sodium bicarbonate, 2 mM glutamine and 0.1% penicillin/streptomycin at 37°C under 5% CO2.[](https://www.ncbi.nlm.nih.gov/mesh/D017693)
## Propagation of Phase II Coxiella burnetii
*In the **Propagation of Phase II Coxiella burnetii** section:*
Clone 4 phase II Nine Mile strain of C. burnetii bacteria were provided by Ted Hackstadt (Rocky Mountain Laboratories, NIAID, NIH, Hamilton, MT, USA) and handled in a biosafety level II facility. Non-confluent Vero cells were cultured in T25 flasks at 37°C under 5% CO2 in DMEM medium supplemented with 5% FBS, 0.22 g/l sodium bicarbonate and 20 mM Hepes, pH 7 (MfbH). Cultures were infected with C. burnetii phase II suspensions for 6 days at 37°C under 5% CO2. After freezing at −70°C, the flasks were thawed, and the cells scraped and passed 20 times through a 27-gauge needle connected to a syringe. Cell lysates were centrifuged at 800×g for 10 min at 4°C.
The supernatants were centrifuged at 24,000×g for 30 min at 4°C, and pellets containing C. burnetii were resuspended in phosphate-buffered saline (PBS; 10 mM sodium phosphate, 0.9% NaCl), aliquoted and frozen at −70°C.[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
## Infection of HeLa Cells with Coxiella burnetii
*In the **Infection of HeLa Cells with Coxiella burnetii** section:*
Cells (5×105) were seeded on sterile glass coverslips placed in 24-well plates and grown overnight in MfbH medium. For infection, a 5-µl aliquot of C. burnetii suspension was added per well (Multiplicity of infection: ∼20). Cells were incubated for different lengths of time at 37°C under 5% CO2.[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
## Quantification of Internalized Bacteria by Indirect Immunofluorescence
*In the **Quantification of Internalized Bacteria by Indirect Immunofluorescence** section:*
To determine the number of internalized bacteria a double cycle antibody staining protocol was used . Briefly, HeLa cells were fixed with 2% paraformaldehyde in PBS for 10 min at 37°C, washed with PBS and blocked with 50 mM NH4Cl in PBS. After washing, cells were incubated with rabbit antibody against C. burnetii (1∶1000) and donkey anti-rabbit secondary antibody conjugated with Cy3 (1∶600) in PBS containing 0.5% BSA (nonpermeabilizing conditions to label extracellular bacteria). After washing, cells were incubated with the same rabbit antibody against C. burnetii (1∶1000) and a donkey anti-rabbit secondary antibody conjugated with Cy5 (1∶600) in PBS containing 0.5% BSA and 0.05% saponin (permeabilizing conditions to label total bacteria: intracellular and extracellular bacteria). Coverslips were mounted with Mowiol and examined by confocal microscopy.
The intracellular bacteria are expressed as a percentage of the total number of bacteria per cell.[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
## Cell Transfection
*In the **Cell Transfection** section:*
Cells were transfected for 6 h with 2 µg/ml pGFP empty vector or pGFP plasmids expressing fusions of GFP with wild-type cortactin (WT) or one of the following mutants: single point mutants, W22A and W525K; double mutants, S405/418D (SD) and S405/418A (2A); and triple mutants, Y421/466/482D (3D) and Y421/466/482F (3F). Cell transfection was carried out using Lipofectamine™ 2000 (Invitrogen, Buenos Aires, Argentina), according to the manufacturer’s instructions. After 6 h of transfection, the cells were washed and incubated for 18 h in MfbH medium at 37°C under 5% CO2.[](https://www.ncbi.nlm.nih.gov/mesh/C086724)
## Western Blotting
*In the **Western Blotting** section:*
Hela cells were cultured on 60-mm dishes and infected as described above for different lengths of time. After the indicated infection periods, cells were washed with PBS, scraped into ice-cold lysis buffer (50 mM Tris-HCl, pH 7.2, 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS, 50 mM NaCl, 10 mM MgCl2, 2 mM Na3VO4, 10 mM NaF, 0.5 mg/ml DTT, 2 mM EDTA) supplemented with a protease inhibitor cocktail and kept on ice for 20 min. Lysates were clarified by centrifugation at 2000×g for 15 min at 4°C. Clarified lysates were transferred to clean tubes, mixed with Laemmli buffer and boiled for 5 min. The samples were resolved by SDS-PAGE and the proteins transferred to nitrocellulose membranes using standard procedures. Membranes were blocked for 2 h at 4°C in Tween-Tris-buffered saline (TTBS; 0.1% Tween 20, 100 mM Tris/HCl, 0.9% NaCl) supplemented with 5% BSA, then incubated overnight at 4°C with the appropriate primary antibodies.
The membranes were washed three times with TTBS, then, incubated for 2 h at room temperature with appropriate peroxidase-conjugated secondary antibodies. Membranes were washed again with TTBS and developed using the ECL Western blotting system (GE Healthcare) according to the supplier’s recommendations. Blotting with anti-GAPDH or anti-actin antibody was carried out to provide loading controls. Band densitometry was carried out using ImageJ software (NIH, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Fluorescence Microscopy
*In the **Fluorescence Microscopy** section:*
HeLa cells were analyzed by confocal microscopy using an FV1000 Olympus Confocal Microscope and FV 10-ASW 1.7 Software (Olympus, Japan). Images were processed using ImageJ software.
## Statistical Analysis
*In the **Statistical Analysis** section:*
Results were analyzed by the ANOVA test in conjunction with Tuckey and Dunnett tests.
Cortactin mutants not phosphorylatable by Src and ERK stimulated and inhibited C. burnetii internalization, respectively.
HeLa cells were transfected with pGFP, pGFP-cortactin WT (wild type) or plasmids encoding GFP fusions with one of the following cortactin mutants: (A) pGFP-cortactin 3F, a Src non-phosphorylatable mutant; (B) pGFP-cortactin 3D, which mimics cortactin phosphorylated by Src; or (C) pGFP-cortactin 2A, an ERK non-phosphorylatable mutant, or pGFP-cortactin SD, which mimics cortactin phosphorylated by ERK. Transfected cells were infected for 4 h with C. burnetii, fixed and processed for indirect immunofluorescence to determine intracellular C. burnetii (see Methods). Cells were analyzed by confocal microscopy. Results are expressed as means ± SE of at least three independent experiments. *, P<0.05; **, P<0.01; ***, P<0.001. (%), percentage of total number of bacteria.
Src and ERK kinases are involved in C. burnetii internalization.
(A) HeLa cells were incubated for 1 h at room temperature with 0.05% DMSO (control), 15 µM PD 98059 (MEK-ERK inhibitor) or 5 µM SU6656 (Src inhibitor). Then the cells were infected for 4 h with C. burnetii in the presence of the inhibitors. Cells were fixed and processed for indirect immunofluorescence using a specific anti-C. burnetii antibody (see Methods). Bars, 10 µm. (B) Quantification of C. burnetii internalized by treated HeLa cells. Results are expressed as means ± SE of three independent experiments. **, P<0.01. (%), percentage of the total number of bacteria.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
## Results
*In the **Results** section:*
## The SH3 Domain of Cortactin is Important for C. burnetii Internalization
*In the **The SH3 Domain of Cortactin is Important for C. burnetii Internalization** section:*
Cortactin is an F-actin regulatory protein that plays an important function in various cellular processes such as cell adhesion, motility and endocytosis. However, its role in phagocytosis has been poorly characterized. Interestingly, cortactin is recruited to the contact sites made by several pathogens with the host plasma membrane during infection .
In its N-terminal acidic region (NTA), cortactin contains a short motif called DDW that binds and activates the Arp2/3 complex . This motif is followed by 6.5 tandem repeats of a 37-residue sequence responsible for F-actin binding. It has a Src homology 3 (SH3) domain at the C-terminus that mediates the interaction with various proteins, including the Arp2/3-stimulating Wiscott-Aldrich protein N-WASP . These interactions link actin remodeling to several specific processes. Mutations in cortactin that abrogate Arp2/3 activation (W22A) or SH3 domain binding function (W525K) have been described , .
To analyze the role of the different cortactin motifs in C. burnetii internalization, we tested two cortactin mutants: W22A (20DDW22 motif mutated to 20DDA22), an NTA mutant that has lost its ability to bind and activate Arp2/3 ; and W525K, an SH3 mutant that is unable to bind certain targets such as N-WASP . HeLa cells were transfected with plasmids encoding GFP-cortactin WT, GFP-cortactin W22A or GFP-cortactin W525K, infected with C. burnetii at 37°C for 4 h, processed for indirect immunofluorescence and analyzed by fluorescence microscopy (see Methods). We decided to allow 4 h for C. burnetii internalization in order to detect the intracellular bacteria with sufficient resolution.
To quantify C. burnetii internalization by immunofluorescence we used the conventional double cycle antibody staining protocol for discriminate between extra- and intracellular bacteria (see Methods). In the Fig. 1A (panels d, h, l and p), extracellular bacteria present double staining (green and red pseudocolors) while intracellular ones present single staining (red pseudocolor). The intracellular bacteria are expressed as a percentage of the total number of bacteria per cell (Fig 1B and 1A, panels c, g, k and o). As shown in Fig.
1B, in cells expressing GFP-cortactin W525K, C. burnetii internalization was significantly lower than that observed in cells expressing GFP alone. In contrast, expression of the W22A mutant did not affect C. burnetii internalization. These results suggest that the SH3 domain of cortactin is critical for C. burnetii entry into HeLa cells.
Tyrosine phosphorylation of cortactin during C. burnetii infection.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
Lysates of HeLa cells infected for different lengths of time with live (A) or heat-killed (B) C. burnetii were analyzed by SDS-PAGE and Western blot using an antibody against phosphoTyr421-cortactin (P-cortactin) or an anti-GAPDH antibody. 0 min: control HeLa cells incubated in the absence of C. burnetii. The data were analyzed with ImageJ software. The ratio between phosphorylated cortactin and GAPDH levels are shown. Results are expressed as means ± SE from at least three independent experiments. *, P<0.05. (RU), relative units.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## A Non-tyrosinephosphorylatable Cortactin Mutant Stimulates C. burnetii Internalization
*In the **A Non-tyrosinephosphorylatable Cortactin Mutant Stimulates C. burnetii Internalization** section:*
Post-translational modifications such as phosphorylation at tyrosine and serine residues at PST region of cortactin regulate its cellular function. Src kinases and other Tyr kinases phosphorylate human cortactin predominantly at three sites in vitro, Tyr421, Tyr470 and Tyr486 (corresponding to Tyr421, Tyr466 and Tyr482 in murine cortactin), while ERK and PAK phosphorylate Ser405/Ser418 and Ser113, respectively , , , –. In addition to that, c-Met and Fer kinases can phosphorylate cortactin on tyrosine residues , . The combined mutation of Tyr421, Tyr466, and Tyr482 abolishes tyrosine phosphorylation of cortactin in cells under various conditions , . Thus, these Src phosphorylation sites have been the focus of functional characterization.
At the same time, several mass spectrometry-phosphoproteomic studies have identified additional phosphorylated tyrosine residues , . A number of individual phosphotyrosine sites have been reported independently in different cell types and in response to diverse stimuli, but their regulation and function remain to be investigated. A validated tool to study the role of phosphorylation is to use non-phosphorylatable and phospho-mimetic mutants. To determine whether tyrosine phosphorylation plays a role in internalization, HeLa cells were transfected with a plasmid encoding GFP-cortactin 3F (Y421,466,482F), which encodes a cortactin that cannot be phosphorylated by Src on tyrosines. Transfected HeLa cells were infected with C. burnetii at 37°C for 4 h, processed for indirect immunofluorescence and analyzed by fluorescence microscopy.
As shown in Fig. 2A, the levels of C. burnetii internalization were similar in cells expressing either GFP (control) or GFP-cortactin WT, but significantly higher in cells expressing GFP-cortactin 3F. In contrast, when HeLa cells were transfected with a plasmid encoding pGFP-cortactin 3D (Y421,466,482D), which mimics cortactin phosphorylated by Src, internalization of C. burnetii was similar to that in cells expressing GFP-cortactin WT (Fig. 2B). These results suggest that C. burnetii internalization is favored when cortactin is dephosphorylated.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
Tyrosine phosphorylation of Src kinase during the C. burnetii infection.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
Lysates of HeLa cells infected with C. burnetii for different lengths of time were analyzed by SDS-PAGE and Western blot using anti-phospho-Src (P-Src) and anti-Src antibodies. 0 min: control HeLa cells incubated in the absence of C. burnetii. Data were analyzed with ImageJ software. The ratio between phosphorylated and total Src levels is shown. Results are expressed as means ± SE of at least three independent experiments. *, P<0.05. (RU), relative units.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## A Non-serinephosphorylatable Cortactin Mutant Inhibits C. burnetii Internalization
*In the **A Non-serinephosphorylatable Cortactin Mutant Inhibits C. burnetii Internalization** section:*
Since the cellular function of cortactin is also regulated by phosphorylation of its serine residues, we studied whether serine phosphorylation is important for C. burnetii internalization. HeLa cells were transfected with pGFP, pGFP-cortactin WT, pGFP-cortactin 2A or pGFP-cortactin SD. Cortactin 2A (S405,418A) cannot be phosphorylated by ERK, while cortactin SD (S405,418D) mimics the effects of ERK phosphorylation. As shown in Fig. 2C, expression of the ERK-phosphorylation mimic of cortactin led to internalization similar to that in cells expressing GFP or cortactin WT.
Interestingly, the cortactin mutant not phosphorylatable by ERK inhibited internalization, indicating that cortactin phosphorylation on serine residues is important for internalization. These results suggest that serine phosphorylation of cortactin regulates internalization of C. burnetii during infection.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
Tyrosine phosphorylation of ERK kinase during C. burnetii infection.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
Lysates of HeLa cells infected with C. burnetii for different lengths of time were analyzed by SDS-PAGE and Western blot using anti-phospho-ERK (P-ERK) and anti-ERK antibodies. 0 min: control HeLa cells incubated in the absence of C. burnetii. Data were analyzed with ImageJ software. The ratio between phosphorylated and total ERK levels is shown. Results are expressed as means ± SE of at least two independent experiments. *, P<0.05. (RU), relative units.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## Src and ERK Kinases are Involved in C. burnetii Internalization
*In the **Src and ERK Kinases are Involved in C. burnetii Internalization** section:*
Given that dynamic phosphorylation of tyrosine and serine residues in cortactin is involved in C. burnetii uptake and that cortactin is a substrate of kinases of the Src and ERK families, we analyzed directly whether these kinases are involved in C. burnetii internalization. HeLa cells were treated during infection with Src kinase inhibitor SU6656 or with PD98059, an inhibitor of MEK kinases which are upstream activators of ERK kinases , . Fig. 3A shows extracellular bacteria (panels a, c and e) and total bacteria (panels b, d and f) in untreated or inhibitor-treated cells. Quantification of internalized bacteria is shown in Fig.
3B. Both inhibitors significantly blocked internalization. Similar inhibition of the C. burnetii uptake was observed during infection of a macrophage cell line treated with the MEK kinase inhibitor (Fig. S1A and B). These results suggest that kinases of the Src and ERK families are involved in C. burnetii internalization.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
## C. burnetii Induces Phosphorylation of Cortactin on Tyr421 During Infection
*In the **C. burnetii Induces Phosphorylation of Cortactin on Tyr421 During Infection** section:*
Several pathogens have been found to modify the phosphorylation status of cortactin during their interaction with host cells . To determine whether C. burnetii causes cortactin phosphorylation during infection, we examined the phosphorylation status of Tyr421 after different infection periods. HeLa cells were infected with C. burnetii for different periods of time, and clarified lysates were analyzed by SDS-PAGE and Western blotting using an antibody that specifically recognizes cortactin phosphorylated on Tyr421. Fig. 4A shows that the maximal level of phospho-Tyr421 cortactin was observed at 1 h of infection with live bacteria.
Similar cortactin phosphorylation levels were observed during infection of a phagocytic cell (Fig. S2B). In contrast, the level of this form of phosphorylated cortactin did not change significantly when the infection was carried out with heat-killed C. burnetii (Fig. 4B). These results suggest that C. burnetii induces phosphorylation of cortactin on Tyr421 early during infection.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
## Src and ERK Kinases are Activated During C. burnetii Infection
*In the **Src and ERK Kinases are Activated During C. burnetii Infection** section:*
Considering that cortactin is phosphorylated on tyrosine residues and that it is a substrate of Src kinase, we reasoned that this enzyme may be activated during C. burnetii infection. To investigate this possibility, clarified lysates of HeLa cells infected with C. burnetii were analyzed by SDS-PAGE and Western blotting using an antibody that specifically recognizes phospho-Src (pTyr416), the activated form of the kinase. Fig. 5 shows the levels of activated Src and total Src. Src was activated early during infection, within 15 min, after which the level of activated enzyme decreased to basal levels.
This result suggests that Src kinase is activated early during C. burnetii-host cell interaction.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
Cortactin can also be regulated by phosphorylation of serine residues and ERK is one kinase involved in this reaction. To investigate whether ERK is activated during C. burnetii infection, lysates of infected HeLa cells were analyzed by SDS-PAGE and Western blotting using an antibody that specifically recognizes phosphorylated ERK. As shown in Fig. 6, ERK was activated at 15 min of infection. Similarly, the phosphorylation in ERK was observed during infection of a phagocytic cell (Fig.
S2A). This result suggests that ERK, similar to Src kinase, is activated early during C. burnetii-host cell interaction.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
## Discussion
*In the **Discussion** section:*
Bacterial pathogens manipulate the host cell cytoskeleton to avoid phagocytosis, to invade and/or to become mobile in the host cell cytoplasm. They often interact with actin filaments by modulating the activity of different actin-interacting effectors in the host. One such effector is cortactin, an actin-binding protein that plays a crucial role in the regulation of actin dynamics. Cortactin has been implicated in the infection process of several microbial pathogens .
The present study contributes to understanding the role of cortactin in bacterial pathogenesis. We provide evidence that the SH3 domain and serine phosphorylation of cortactin are involved in signal transduction pathways that support the internalization of avirulent C. burnetti into non-phagocytic cells, whereas tyrosine phosphorylation of cortactin suppresses this internalization. We also show that Src and ERK kinases are activated during the initial stages of C. burnetti infection.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
Cortactin stimulates actin polymerization by binding, via its N-terminal domain, the Arp2/3 complex, or by binding, through its C-terminal SH3 domain, N-WASP, a well-known activator of Arp2/3 , . We report here that expression in HeLa cells of the cortactin mutant W525K, which carries a mutation in the C-terminal SH3 domain, significantly inhibited C. burnetii internalization (Fig. 1), suggesting an important role for this domain in bacterial entry. We hypothesize that avirulent C. burnetii requires minor modification of the actin cytoskeleton at the plasma membrane to be internalized, so we propose that SH3 domain-mediated recruitment of N-WASP (and then Arp2/3) is sufficient to stimulate actin assembly and bacterial uptake. Our results are similar to those observed in the formation of the pedestal-like structure during E. coli infection. Indeed, expression of the cortactin mutant W525K significantly reduced the number of pedestals induced by enteropathogenic Escherichia coli (EPEC) or enterohemorrhagic E. coli (EHEC) in HeLa cells , .
In addition, cell motility can be regulated by cortactin through its C-terminal SH3 domain independently of the presence of the N-terminal portion which is consistent with the ability of the SH3 domain on its own to stimulate N-WASP and actin polymerization in vitro . Cortactin cleavage by calpain has also been shown to be important for cell migration . Cells expressing a calpain-resistant cortactin showed reduced migration and increased membrane protrusion. This phenotype was reverted by expression of a calpain-resistant cortactin with the W525K mutation, which suggests that the SH3 domain of cortactin is required for the stimulation of membrane protrusions.
N-WASP activation depends on the phosphorylation status of serine and tyrosine residues in the C-terminal domain of cortactin . Expression of a non-serine-phosphorylatable cortactin mutant impairs pedestal formation in cells infected with EPEC or EHEC . These results suggest that ERK phosphorylation of cortactin contributes to pedestal formation. In HeLa cells expressing the same cortactin mutant we observed significant inhibition of bacterial uptake (Fig. 2). Thus, similarly to pedestal formation induced by E. coli, entry of C. burnetii into HeLa cells requires serine phosphorylation of cortactin.
Our conclusion is also supported by the observations that ERK1/2 kinases were transiently activated early during HeLa cell infection and then later deactivated (Fig. 6), and that HeLa cell treated with the inhibitor of MEK-dependent ERK1/2 activation showed a reduction in C. burnetii internalization (Fig. 3).[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
In our experimental model, expression of a non-tyrosine-phosphorylatable cortactin mutant increased C. burnetii internalization (Fig. 2). However, pedestal formation induced by EPEC was reduced in HeLa cells expressing the same mutant , . Moreover, we observed cortactin phosphorylation on Tyr421 at 1 h of infection with C. burnetii, after which the phosphorylation returned to basal levels (Fig. 4). Similar kinetics of tyrosine phosphorylation-dephosphorylation were observed in HeLa cells infected with pre-activated EHEC . In vitro experiments have shown that cortactin phosphorylated by Src does not interact with and activate N-WASP, which leads to inhibition of pedestal formation , .
Based on these results, we can speculate that during C. burnetii infection of HeLa cells, cortactin must be dephosphorylated on its tyrosine residues in order to interact with N-WASP, leading to actin remodeling and bacterial internalization. At the same time, tyrosine dephosphorylation of cortactin increases its actin-crosslinking activity in vitro . Therefore, cortactin dephosphorylated on its tyrosine residues may cross-link small actin filaments, forming a discrete actin meshwork close to the bacterial attachment site, which then allows C. burnetii internalization.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
Cortactin can be tyrosine-phosphorylated not only by Src kinases but also by Abl kinases . Abl kinases can be activated by autophosphorylation and by phosphorylation by Src family kinases , . We observed that Src kinase was activated at 15 min of infection and then the levels of activated enzyme decreased to basal levels (Fig. 5), while cortactin was phosphorylated on Tyr421 at 1 h of infection. We also observed that the pharmacological Src inhibitor decreased C. burnetii internalization (Fig. 3). While these observations may be due to the direct action of Src kinase on cortactin, they may also be due to the action of Abl kinase, activated by Src.
Internalization of Chlamydia trachomatis also involves cortactin, and this protein is phosphorylated at 1 h of infection by Abl kinases but not Src . Shigella entry into host cells also requires activation of Abl kinases . Therefore we speculate that during C. burnetii infection, Abl kinases are activated to phosphorylate cortactin. On the other hand, the strong effect of the chemical inhibition of Src kinase could indicate that other Src substrates apart from cortactin might participate in C. burnetii entry.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
We show here that during C. burnetii infection, Src is transiently activated and then inactivated, and cortactin is tyrosine-phosphorylated and then dephosphorylated. We detected that C. burnetii entry induces the tyrosine phosphorylation of cortactin at early time points, with a maximum peak around 60 min (Fig. 4). In addition to that, we found that the 3F cortactin mutant with non-phosphorylatable tyrosines enhances C. burnetii entry at 4 h after infection (Fig. 2). This seems to indicate that the tyrosine phosphorylation of cortactin is required at the initial steps while it would inhibit entry at later time points. These processes are similar to those observed during infection of gastric epithelial cells by Helicobacter pylori.
This pathogen promotes an early but transient phosphorylation of cortactin. The infected cells become scattered and elongated, and this phenotype depends on Src phosphorylation of CagA (a protein secreted by a type IV secretion system), which inactivates c-Src and leads to cortactin dephosphorylation by an unknown mechanism . It is tempting to speculate that C. burnetii internalization occurs by a mechanism similar to that of H. pylori. To our knowledge, C. burnetii, H. pylori and EHEC are the only three pathogens known to induce dephosphorylation of cortactin during host cell infection. Recently, H. pylori has been shown to induce cortactin phosphorylation on serines in a CagA-independent manner, and this form of cortactin stimulates actin rearrangement and cell elongation .
We show here that expression of a cortactin mutant lacking phosphorylatable serines inhibited C. burnetii internalization, which suggests that serine phosphorylation of cortactin is necessary for C. burnetii entry.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
THP-1 monocytes infected with virulent C. burnetii exhibit intense membrane protrusions associated with major actin cytoskeleton reorganization, while infection with avirulent bacteria induced a few membrane folds without significantly affecting cell morphology . Although it was not the focus of the present study, we think that the membrane folds stimulated by avirulent C. burnetii result from a modest actin cytoskeleton rearrangement that facilitates bacterial uptake. Meconi and collaborators also showed that actin cytoskeleton reorganization is associated with tyrosine phosphorylation of the Src family kinases Hck and Lyn very early during infection of THP-1 cells with virulent, but not avirulent, C. burnetii . In the present study, using an anti-phospho-Src antibody that recognizes several members of the Src family, including Hck and Lyn, we found that Src was activated early during infection with avirulent bacteria. Meconi et al. also found, using an anti-phosphoTyr monoclonal Ab, that virulent C. burnetii, but not avirulent bacteria, stimulate the tyrosine phosphorylation of several proteins.
Using a similar experimental strategy, we observed a significant labeling of proteins with masses around 85 kDa at 1 h of HeLa infection with avirulent C. burnetii (data not shown). Cortactin migrates as a doublet of 80 and 85 kDa in SDS-PAGE , . We cannot rule out the possibility that the differences between our results and those of Meconi et al. are due to the different cell types used.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
In conclusion, our results indicate that serine phosphorylation of cortactin and its SH3 domain are involved in a signal transduction mechanism that favors C. burnetii uptake, while tyrosine phosphorylation suppresses this uptake. Our results suggest that a complex series of events occurs during C. burnetii internalization into non-phagocytic cells. Early after infection, Src and ERK kinases may phosphorylate unknown substrates, perhaps other kinases such as Abl that in turn phosphorylate tyrosine residues in cortactin. This may regulate an early step in internalization. At a later stage, tyrosine dephosphorylation and serine phosphorylation of cortactin take place, which regulates a later step of internalization.
In this way, tyrosine phosphatases and serine kinases from the host cell and/or C. burnetti regulate the phosphorylation status of cortactin to favor C. burnetti entry into the host cell. Thus, the results reported here indicate that dynamic phosphorylation of cortactin is important for C. burnetii internalization during infection.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
## Supporting Information
*In the **Supporting Information** section:*
Competing Interests: The authors have declared that no competing interests exist.
Funding: This work was partly supported by grants from the Science and Innovation Ministry of Spain (PS-09/00080) to NM-Q and from Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET, PIP 5794) and Secretaría de Ciencia, Técnica y Posgrado (SECTyP, 06/J108), Universidad Nacional de Cuyo to WB. This work was also supported by the Agencia Nacional de Promocioen Cientifica y Tecnologica (PICT # 02336). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. No additional external funding was received for this study.
# References
*In the **References** section:*
|
# Introduction
[MC-12](https://www.ncbi.nlm.nih.gov/mesh/D009842), an Annexin A1-Based Peptide, Is Effective in the Treatment of Experimental Colitis
# Abstract
*In the **Abstract** section:*
Annexin A1 (ANXA1) inhibits NF-κB, a key regulator of inflammation, the common pathophysiological mechanism of inflammatory bowel diseases (IBD). MC-12, an ANXA1-based tripeptide, suppresses NF-κB activation. Here, we determined the eff[icacy](https://www.ncbi.nlm.nih.gov/mesh/D009842) of MC-12 in the [control of](https://www.ncbi.nlm.nih.gov/mesh/D009842) IBD. Mice with colitis induced by dextran sodium sulfate (DSS) or [2,4,6](https://www.ncbi.nlm.nih.gov/mesh/D009842)-trinitro benzene sulfonic acid (TNBS) were treated w[ith various doses of M](https://www.ncbi.nlm.nih.gov/mesh/D016264)C-[12 ](https://www.ncbi.nlm.nih.gov/mesh/D016264)admin[istered intraperitoneally, orally or](https://www.ncbi.nlm.nih.gov/mesh/D014302) i[ntra](https://www.ncbi.nlm.nih.gov/mesh/D014302)rectally. We determined colon length [and t](https://www.ncbi.nlm.nih.gov/mesh/D009842)he histological score of colitis, and assayed: in colon tissue the levels of TNF-α, IFN-γ, IL-1β, IL-6 and IL-10 by RT-PCR; prostaglandin E2 (PGE2), cytoplasmic phospholipase A2 (cPLA2) and myeloperoxidase by immun[oassay; and COX-](https://www.ncbi.nlm.nih.gov/mesh/D015232)2 [and ](https://www.ncbi.nlm.nih.gov/mesh/D015232)NF- κB by immunohistochemistry; and in serum the levels of various cytokines by immunoassay. In both models MC-12: reversed dose-dependently colonic inflammation; inhibited by up to 47% myeloperoxid[ase a](https://www.ncbi.nlm.nih.gov/mesh/D009842)ctivity; had a minimal effect on cytoplasmic phospholipase A2; reduced significantly the induced levels of TNF-α, IFN-γ, IL-1β, IL-6 and IL-10, returning them to baseline. DSS and TNBS markedly activated NF-κB in colonic epithelial cells and MC-12 decreased this[ ef](https://www.ncbi.nlm.nih.gov/mesh/D016264)fect [by 8](https://www.ncbi.nlm.nih.gov/mesh/D014302)5.8% and 72.5%, respectively. MC-12 had a similar effect i[n cul](https://www.ncbi.nlm.nih.gov/mesh/D009842)tured NCM460 normal colon epithelial cells. Finally, MC-1[2 sup](https://www.ncbi.nlm.nih.gov/mesh/D009842)pressed the induction of COX-2 expression, the level of PGE2 in the colon and PGE[2 met](https://www.ncbi.nlm.nih.gov/mesh/D009842)abolite in serum. In conclusion, MC-12, representing a novel[ cla](https://www.ncbi.nlm.nih.gov/mesh/D015232)ss of short peptid[e in](https://www.ncbi.nlm.nih.gov/mesh/D015232)hibitors of NF-κB, has a strong effec[t aga](https://www.ncbi.nlm.nih.gov/mesh/D009842)inst colitis in two preclinical models recapitulating features of human IBD. Its mechanism of action is complex and includes pronounced inhibition of NF-κB. MC-12 merits further development as an agent for the control of IBD.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Inflammatory bowel diseases (IBD) are a set of complex, life-long and for some patients devastating diseases, for which there is no satisfactory treatment. There are two distinct clinical entities, ulcerative colitis and Crohn’s disease. A shared clinical manifestation of them is ulcerations in the intestinal mucosa; Crohn’s disease can affect the entire digestive tract while ulcerative colitis affects only the colon. The treatment of IBD has been based on anti-inflammatory medications, with steroids having been the mainstay of treatment for years. The prime limitations of anti-inflammatory medications are variable efficacy and side effects, which, in the case of steroids, can limit dosing or duration of treatment or force physicians to altogether discontinue them.
The recently introduced biological agents have also significant limitations, and, in addition, high cost. It is clear that there is a pressing need for new agents for the control of the clinical manifestations of IBD.[](https://www.ncbi.nlm.nih.gov/mesh/D013256)
Inflammation is the underlying theme in IBD (hence the word inflammatory in their name). A key regulator of inflammation is NF-κB, a transcription factor that is normally sequestered in the cytoplasm. When activated, NF-κB translocates into the nucleus where it regulates the expression of a multitude of genes related to inflammation. We have recently unraveled the connection between glucocorticoids and NF-κB and have proposed a novel mechanism by which they act. Briefly, we have demonstrated that glucocorticoids induce the expression of annexin A1 (ANXA1), which then binds to the p65 subunit of NF-κB, inhibiting its activation.
There is a nearly perfect correlation between the anti-inflammatory potency of the various steroids and the induction of ANXA1, on one hand, and the suppression of NF-κB on the other. Short (around 20 amino acids) C-terminal fragments of ANXA1 are known to have many of its biological activities.[](https://www.ncbi.nlm.nih.gov/mesh/D005938)
We have synthesized MC-12, a tripeptide based on the structure of ANXA1 that is as effective as ANXA1 in suppressing NF-κB activation. This peptide is representative of several such ANXA1 peptides that inhibit NF-κB. We assessed its potential efficacy in colitis using two mouse models of IBD, one based on the administration of dextran sulfate sodium (DSS) and the other on trinitrobenzene sulfonic acid (TNBS). Our data demonstrate that administration of MC-12 reverses in a dose-dependent manner the inflammatory reaction of the colon and prevents the development of ulcerations. There were no apparent side effects from the administration of MC-12 to mice. MC-12 modulates several inflammatory mediators, including NF-κB and several cytokines.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Results
*In the **Results** section:*
## MC-12 Reduces Experimental Colitis in Mice
*In the **MC-12 Reduces Experimental Colitis in Mice** section:*
By day 8, mice receiving DSS lost on average 8.4% of their baseline body weight (Fig. 1A). MC-12 given by IP and PO at both doses and by IR at the higher dose prevented such weight loss in a dose-dependent manner, with those receiving the highest IP dose of MC-12 (25 mg/kg), showing 7.7% increase in their body weight compared to baseline. Of note, MC-12 was more effective in preventing weight loss when given IP than PO. Fig. 1B shows the length of the colon of these mice.
Compared to normal control, the colon of DSS-treated mice was shorter by 2.3 cm (8.7±0.29 vs. 6.4±0.42 cm; mean ± SEM, for this and all subsequent values; p<0.01). Administration of MC-12 prevented dose-dependently most of this reduction in colon length, with IP, PO and IR administration producing essentially identical results; under treatment, the length of the colon ranged between 7.3±0.29 and 7.7±0.20 cm (p<0.05 for all differences from vehicle control). Macroscopically, shortened colons showed wall edema and fewer feces in the lumen. We also studied colitis induced by TNBS in SJL/J mice. The optimal dose of TNBS for our study was 100 µl of a 2.5% ethanolic solution instilled intracolonically.
As shown in Fig. 1C, TNBS reduced the body weight of the animals on day 3 by 18.2% and shortened the length of the colon by 25% (Fig. 1D), compared to baseline. MC-12 25 mg/kg failed to prevent the weight loss (84.7±1.1 vs. 81.8±1.3% in vehicle-treated controls, p>0.05) and the shortening of the colon (6.4±0.20 vs. 6.4±0.07 in vehicle-treated controls, p>0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
The effect of MC-12 on DSS- and TNBS-induced colitis in mice.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Mice (7/group) received 2% DSS in drinking water or 2.5% TNBS intra-colonically to induce colitis and were treated with vehicle or MC-12 given IP, PO or IR. A: The body weight of mice of DSS model during treatment by IP, PO or IR, expressed as percentage of baseline (day 0). B: The colon length of DSS model with MC-12 treatments by IP, PO or IR. C: The body weight of mice of TNBS model during treatment, expressed as percentage of baseline (day 0). D: The colonic length of mice of TNBS model in three treatment groups. These studies were repeated at least once giving similar results. Values are mean ± SEM. *, statistically significant difference from the vehicle-treated group.[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
As expected, DSS induced colitis in these mice. The histological sections shown in Fig. 2A–F demonstrate changes in the colonic mucosa by DSS, including significant inflammation, accumulation of mucus and development of ulcers. The granulocytes present in the mucosa establish the development of acute inflammation. Treatment with MC-12 reduced the degree of inflammation, essentially restoring the integrity of the mucosa.
After treatment with DSS the histological colitis score became 25.8±2.01 from 0 in normal controls (Fig. 2I). MC-12 reduced the histological colitis score dose-dependently, regardless of its route of administration. Compared to vehicle-treated controls, MC-12 given IP reduced the histological score by 48.9% at 5 mg/kg (13.2±1.62 vs. 25.8±2.01, p<0.01) and 66.8% at 25 mg/kg (8.6±1.43 vs. 25.8±2.01, p<0.01). The dose of 25 mg/kg showed a lower histological score than the dose of 5 mg/kg (8.6±1.43 vs. 13.2±1.62; p<0.05). Given orally, MC-12 was slightly less effective, reducing this score by 50.2% at 5 mg/kg (12.9±1.01 vs. 25.8±2.01, p<0.01) and 48.4% at 25 mg/kg (13.3±1.05 vs. 25.8±2.01, p<0.01).
Similarly, the IR administration of MC-12 decreased the histological score by 33.3% (p<0.05) at the dose of 5 mg/kg and 58.1% (p<0.01) at the dose of 25 mg/kg compared to DSS control group. Regardless of its lacking effect on body weight and colon length in the TNBS model, MC-12 had a significant anti-inflammatory effect on the colonic mucosa as shown in Fig. 2G–H, reducing the histological score by 39.1%, compared to TNBS control as shown in Fig. 2J (38.3±1.7 vs. 23.3±3.3; p<0.01).[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
MC-12 ameliorates colitis induced by both DSS and TNBS.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Paraffin sections of colonic tissues were stained with hematoxylin & eosin and their histological scores were determined as in Methods. A: Normal mucosa of the C57BL/6 mouse. B: Severe inflammation including infiltration by inflammatory cells, edema, loss of crypts and ulcerations are seen in a DSS + vehicle-treated mouse. C, D: MC-12, 5 or 25 mg/kg, IP, significantly decreased DSS-induced colonic inflammation. E: Treatment with MC-12, 25 mg/kg PO, decreased DSS-induced colonic inflammation. F: Treatment with MC-12, 25 mg/kg, IR, markedly reduced DSS-induced colitis. G: Colonic mucosa from a TNBS vehicle-treated mouse, showing severe crypt loss and inflammatory cell infiltration (arrow). Inset: normal mucosa of a healthy SJL/J mouse. H: Treatment with MC-12 25 mg/kg for 2 days decreased the inflammatory cell infiltration and crypt loss (arrow indicates remaining crypts). I, J: The histological score of the various study groups of both DSS- and TNBS-induced colitis. Values are mean ± SEM. H&E staining; magnification 100x.[](https://www.ncbi.nlm.nih.gov/mesh/D010232)
## MC-12 is not Cytotoxic to Normal Colonic Epithelial Cells
*In the **MC-12 is not Cytotoxic to Normal Colonic Epithelial Cells** section:*
MC-12 failed to induce cytotoxicity to the NCM460 normal colonic epithelial cell line. The 24-h IC50 value of MC-12 was higher than 5 mM; higher concentrations were not studied. This concentration far exceeds concentrations of MC-12 that inhibited, for example, the activation of NF-κB (low µM range; shown below).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## The Constituent Amino Acids of MC-12 Have No Effect on Colitis
*In the **The Constituent Amino Acids of MC-12 Have No Effect on Colitis** section:*
Since peptides are subject to hydrolytic cleavage of their peptide bonds, especially when administered orally, we examined whether MC-12 acts against colitis after its potential degradation to its three constituent amino acids. To this end, we treated mice with DSS-induced colitis using a solution containing the three amino acids of MC-12 (Ac-Gln, Ala and Trp) at equimolar concentrations. We treated mice with DSS-colitis following the same protocol as in the previous studies. The amino acid solution was given ip at a dose equivalent to 25 mg/kg of intact MC-12. The amino acid solution had no significant effect on any of the three parameters that we evaluated: body weight (98.8% vs. 96.7%), colon length (6.5±0.13 vs. 6.2±0.17) and histological score (19.9±2.0 vs. 24.8±1.9); all differences were statistically not significant.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## MC-12 Reduces DSS- and TNBS-induced Inflammation in Colonic Mucosa: Effects on MPO, cPLA2 and Cytokines
*In the **MC-12 Reduces DSS- and TNBS-induced Inflammation in Colonic Mucosa: Effects on MPO, cPLA2 and Cytokines** section:*
To assess the effect of MC-12 on the inflammatory changes associated with experimental colitis, we determined in colon tissue samples the activity of myeloperoxidase and cytosolic phospholipase A2 (cPLA2) as well as the response of five inflammatory cytokines. MPO activity is an indicator of the degree of acute inflammation in a given tissue. cPLA2, a phospholipase recognizing the sn-2 acyl bond of phospholipids, releases lysophospholipid and arachidonic acid, which can then be converted to prostaglandins and leukotrienes, both inflammatory mediators. Finally, the pro-inflammatory cytokines TNF-α, IFN-γ, IL-1β and IL-6, and anti-inflammatory cytokine IL-10 have been implicated in experimental and human colitis. Indeed, cytokines are thought to orchestrate the development, recurrence and exacerbation of IBD.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Fig. 3 summarizes our findings. DSS and TNBS increased MPO activity 2.4- and 3.5-fold compared to normal controls. In both models, MC-12 inhibited the enhanced MPO activity significantly and in a dose-dependent manner. This reduction was 30% (p<0.05) and 47% (p<0.01) at 5 mg/kg and 25 mg/kg, respectively, in the DSS model and 39% (p<0.01) in the TNBS model.[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
The effect of MC-12 on parameters of inflammation.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
MPO (A, B), cPLA2 (C, D) activity and the mRNA levels of proinflammatory and anti-inflammatory cytokines (E, F) were measured in colon tissue samples of both DSS- and TNBS-induced colitis mouse models. In both models MC-12 significantly inhibited MPO activity and the mRNA levels of all cytokines but not cPLA2 activity. Values are mean ± SEM. *p<0.05 compared to vehicle-treated group, **p<0.01 compared to vehicle-treated group.[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
Similar to MPO, the activity of cPLA2 in colon mucosa was increased 2.9-fold by DSS and 4.5-fold by TNBS compared to normal mice. The effect of MC-12 on cPLA2 activity was modest and statistically not significant; in the three MC-12 treatment groups, the reduction in cPLA2 activity ranged between 18% and 24%.[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
We also determined the response of the cytokines TNF-α, IFN-γ, IL-1β, IL-6 and IL-10 in the colon by measuring their corresponding mRNA levels using real-time PCR. As shown in Fig. 4E–F, compared to controls, DSS increased the mRNA levels of all these cytokines by 9- to 17-fold and TNBS by 4- to 10-fold. Treatment with MC-12 at either dose significantly reduced the mRNA levels of TNF-α, IL-1β, IFN-γ, IL-6 and IL-10,[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
MC-12 reduces the serum levels of cytokines.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
The serum levels of TNF-α, IFN-γ, IL-1β, IL-6 and IL-10 were measured by ELISA as in Methods. Both DSS and TNBS induced their levels and treatments with MC-12, 25 mg/kg administered PO, IP or IR reduced these levels 1.6- to 5.1-fold. *p<0.01, #p<0.05 compared to vehicle-treated controls.[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
Next, we measured the levels of these cytokines in the serum of the experimental animals. As shown in Fig. 4, DSS or TNBS induced the TNF-α, IFN-γ, IL-1β, IL-6, and IL-10 by 2.4- to 19-fold. Treatment with MC-12 administered PO, IP or IR significantly reduced by 1.7- to 6.4-fold the levels of all these cytokines.[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
## MC-12 Inhibits the Activation of NF-κB in vivo and in vitro
*In the **MC-12 Inhibits the Activation of NF-κB in vivo and in vitro** section:*
The chronic mucosal inflammation in IBD is characterized by hyperactivation of effector immune cells, which produce high levels of pro-inflammatory cytokines, resulting in colonic tissue damage. NF-κB, the master regulator of all inflammation, has been identified as a key regulator in this immunological setting. Its activation, markedly induced in IBD patients, strongly influences the course of mucosal inflammation. NF-κB is also the molecular target of the activity of MC-12. Therefore, we determined by immunohistochemistry the level of NF-κB activation in the colonic mucosa of our mice, using an antibody recognizing the phosphorylation of ser276 of NF-κB’s p65 subunit.
As a methodological control, we also showed marked activation of NF-κB in colon samples from patients with ulcerative colitis using the same primary antibody and method (Figure S1).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
As shown in Fig. 5A–E, there was minimal baseline activation of NF-κB in the colon of normal animals. DSS activated NF-κB in colonic epithelial cells (0.7±0.20 vs. 8.8±1.45, p<0.01), in agreement with previous reports. MC-12 reversed this effect nearly completely (Fig. 5I), regardless of its route of administration: at the highest dose (25 mg/kg) MC-12 decreased this effect of DSS by 85.8% (8.8±1.45 vs. 1.2±0.32; p<0.01). Fig.
5F–H demonstrates that TNBS activated NF-κB in the colonic epithelium; as shown in Fig. 5J, the percentage of cells with activated NF-κB is 8.4±1.32 in TNBS-treated mice vs. 1.1±0.25 in controls (p<0.01). Treatment with MC-12 inhibited the TNBS-induced NF-κB activation by 72.5% (2.3±0.70 vs. 8.4±1.32, p<0.01).[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
MC-12 inhibits NF-κB activation induced by DSS and TNBS in vivo and in vitro.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Colon tissue sections were immunohistochemically stained with anti-phospho-NF-κB p65 antibody. Normal colon mucosa from C57BL/6 mice (A) or from SJL/J mice (F) shows a few cells with p-p65 nuclear positive staining. Colon mucosa with DSS (B) or TNBS (G) shows increased NF-κB nuclear-positive cells, most of which are crypt epithelial cells. Markedly less p-p65 nuclear translocation was shown in colon mucosa from DSS model treated with MC-12, 5 mg/kg ip (C), 25 mg/kg ip (D) or 25 mg/kg po (E) or from TNBS model treated with MC-12 25 mg/kg ip (H). Changes in p-p65 nuclear positive, evident in the photos are quantified (I, J). DSS increased NF-κB-DNA binding in NCM460 cells determined by EMSA, and MC-12 30 and 300 µM significantly blocked this effect (K). Values are mean ± SEM. **p<0.01 compared to vehicle-treated group. IHC staining; magnification 200x.[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
The inhibitory effect of MC-12 on NF-κB was also documented in cultured NCM460 cells (normal colon epithelial cells). As expected, DSS significantly increased NF-κB-DNA binding activity in the NCM460 cells. MC-12 at concentrations of 30 µM and 300 µM essentially eliminated this NF-κB activation (Fig. 5K).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## MC-12 Inhibits the Induction of COX-2 and Decreases PGE2 Levels in vivo
*In the **MC-12 Inhibits the Induction of COX-2 and Decreases PGE2 Levels in vivo** section:*
The eicosanoid cascade seems to be involved in the pathogenesis of IBD but there is some controversy regarding its specific role. COX-2 is induced in experimental colitis whereas NSAIDs are thought to exacerbate colitis in humans. Thus we evaluated the effect of MC-12 on COX-2 expression and PGE2 levels in colonic tissue and its metabolite 13,14-dihydro-15-keto prostaglandin E2 in the serum of mice.[](https://www.ncbi.nlm.nih.gov/mesh/D015777)
As shown in Fig. 6, in both models of experimental colitis MC-12 markedly reduced the expression of COX-2 in the colonic mucosa (Fig. 6A and B). In addition it decreased the levels of PGE2 in colonic mucosa (87.9±15.8 or 127.0±12.5 vs. 241.4±73.7, p<0.05, Fig. 6C) and PGE2 metabolite in serum (4.3±0.9 or 4.2±2.4 vs. 16.8±4.9, p<0.01, Fig. 6D).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
MC-12 inhibits COX-2 induction and decreases PGE2 levels in DSS- and TNBS-induced colitis. A:[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Representative photomicrographs of tissue sections with COX-2 immunohistochemical staining on colon tissue from normal (no treatment), vehicle- or MC-12-treated mice from the DSS (upper panels) or TNBS (lower panel) groups. MC-12 treatment: 25 mg/kg ip. B: The COX-2 expression scores in the various groups of animals (n = 8/group). Differences were evaluated by Pearson’s χ2 method. C, D: MC-12 decreased the levels of PGE2 in colonic mucosa and of PGE2 metabolite in serum. Values are mean ± SEM. **p<0.01 compared to vehicle-treated group, *p<0.05 compared to vehicle-treated group.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Discussion
*In the **Discussion** section:*
Our data demonstrate the strong anti-inflammatory effect of the novel tripeptide MC-12 in two models of experimental colitis. MC-12, designed as an inhibitor of NF-κB based on a recently unraveled mechanism of NF-κB control, had a profound inhibitory effect on NF-κB activation and on a circuitry of dependent inflammatory mediators.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Experimental models of colitis enable us to not only study its pathogenetic components during the various phases of colitis but also to assess the therapeutic efficacy of experimental agents. The two murine models of colitis that we employed in these studies represent two different entities. The DSS model, technically a model of chemical injury to the colon, recapitulates features of ulcerative colitis, including severe leukocyte infiltration, crypt damage, and tissue edema often accompanied by severe ulceration. The TNBS model, on the other hand, has many of the characteristic features of Crohn’s disease in humans such as severe transmural inflammation. Both are associated with weight loss and gastrointestinal manifestations.[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
As expected, in both models we observed significant weight loss and decreased colon length, both of which were reversed by MC-12 in DSS-treated mice, either totally or mostly, depending on MC-12’s dose. Interestingly, MC-12 failed to reverse these changes in the TNBS model. Regardless, however, of its effects on body weight and colon length, MC-12 displayed a strong anti-inflammatory effect in both models. The inhibition of inflammation reached nearly 70% in DDS-treated mice and almost 40% in TNBS-treated mice. The dose of MC-12 was relevant as was its route of administration.
In the DSS model in which two drug doses and two routes of administration were evaluated, the dose effect was unmistakable and the ip route was modestly better than the oral route. The latter may well reflect a greater chance to hydrolyze MC-12 as it travels down the alimentary canal. Our data confirmed that fully hydrolyzed, MC-12 is rendered ineffective; in a reconstitution experiment the three amino acids of MC-12 were administered orally but had no effect on colitis. Indirect as this result may be, one cannot fail to surmise that properly formulated the MC-12 tripeptide may have even higher efficacy in colitis.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Even simple inspection of the tissue sections of the colon makes it clear that MC-12 acts as an anti-inflammatory agent; our data have objectively and quantitatively confirmed this. In addition to histological scoring, in both models MC-12 significantly reduced MPO activity in the colonic mucosa. MPO, an enzyme contained in lysosomes of neutrophils and, to a much lesser extent, of monocytes and tissue macrophages, has been used as a neutrophil and monocyte marker to evaluate the presence and extent of inflammation in colonic mucosa.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
We explored the mechanism of the anti-inflammatory action of MC-12. The a priori molecular target of MC-12 is NF-κB. MC-12 is, however, a peptide derived from ANXA1. Since ANXA1 inhibits phospholipase A2 activity as part of its anti-inflammatory action and PLA2 activity is increased in IBD colonic mucosa, we studied the potential effect of MC-12 on PLA2. Although we demonstrated 3–5 fold enhanced activity of PLA2 in the colitic mucosa, MC-12 did not affect it, ruling out such a mechanism of action. It is likely that the anti-PLA2 property of ANXA1 lies outside the area corresponding to the tripeptide.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
NF-κB plays a critical role in the pathophysiology of IBD. NF-κB is markedly activated in the inflamed gut, especially in macrophages and epithelial cells, to the point that the degree of NF-κB activation correlates with the severity of intestinal inflammation. Through its extensive transcriptional activity, NF-κB is thought to trigger proinflammatory loops involving cytokines and eicosanoids. Indeed, given its strategic position in the inflammatory process of IBD, NF-κB is considered an excellent target for the development of pharmacological agents for IBD.[](https://www.ncbi.nlm.nih.gov/mesh/D015777)
Our data, showing profound inhibition of NF-κB by MC-12 confirmed our initial expectation. Indeed, MC-12 inhibited the activation of NF-κB by 86% in the DSS model and by 72% in the TNBS model. In addition, the in vitro study using normal colonocytes in which NF-κB was activated by DSS generated congruent results, leaving little doubt as to the specificity of the effect of MC-12. NF-κB is apparently very relevant to the pathogenesis if IBD. Several studies have documented in both animal models of colitis and in humans with IBD the brisk activation of NF-κB; our control experiment with human samples generated equally impressive results.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
MC-12 had a dual effect on the eicosanoid cascade. First, it suppressed the expression of COX-2, which was induced in both models of colitis. And, second, it reduced the production of PGE2; it is unclear if the reduced PGE2 levels reflect exclusively the suppressive effect of MC-12 on COX-2 expression, the enzyme that catalyzes a crucial step in prostaglandin biosynthesis. The effect of MC-12 on PGE2 is consistent with its general role in inflammation. Whether the effect of MC-12 on COX-2 is a result of its effect on NF-κB or a direct effect is uncertain; NF-κB is known to transcriptionally regulate the expression of COX-2.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
In summary, MC-12 markedly suppressed inflammation in both DSS- and TNBS-induced colitis in mice predominantly or exclusively by inhibiting NF-κB activation. Given the magnitude of its effect, the plausibility of its proposed mechanism of action and early evidence for safety, MC-12 merits further development as an agent for the control of IBD.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Materials and Methods
*In the **Materials and Methods** section:*
## Reagents and Cell Culture
*In the **Reagents and Cell Culture** section:*
MC-12 (Ac-Gln-Ala-Trp) was custom-synthesized by GenScript (Piscataway, NJ). NCM460 cell line (purchased from Incell Corporation, LLC, San Antonio, TX), which was derived from normal human colon mucosal epithelium, was grown in M3:10A media (Incell Corporation, LLC, San Antonio, TX). Cells (3×106) were seeded in 100 mm dishes and after 24 h test compounds were added into the media for another 24 h, when cell nuclei were isolated for EMSA analysis.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Cell Viability
*In the **Cell Viability** section:*
NCM460 cells were seeded into 96-well plates at 2×104 cells per well. After overnight incubation, they were treated for 24 h with various concentrations of MC-12 (25 µM – 5 mM). Cell viability was determined by a modified colorimetric assay using 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT). Briefly, the culture medium was removed and replaced with 100 µl of complete medium containing 0.5 mg/ml MTT. After a 4-h incubation at 37°C, 100 µl of a solution containing 10% SDS and 0.01 N HCl were added.
The plate was incubated until MTT formazan crystals were dissolved. Absorbance at 570 nm was measured on a microplate reader and the IC50 was calculated after subtraction of blank values.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Experimental Animals
*In the **Experimental Animals** section:*
Female C57BL/6 and SJL/J mice (Taconic, Hudson, NY), 7–9 weeks old, were kept under controlled temperature (25°C) with a 12/12-hour light-dark cycle and free access to a standard diet and drinking water. The mice were allowed to acclimate for 7 days before the start of experiments. Our studies were approved by the Institutional Animal Care and Use Committee of Stony Brook University.
## DSS- and TNBS-induced Colitis, MC-12 Treatment, and Histological Evaluation
*In the **DSS- and TNBS-induced Colitis, MC-12 Treatment, and Histological Evaluation** section:*
## DSS model
*In the **DSS model** section:*
Forty nine C57BL/6 mice were divided into 7 groups (7 mice per group). The mice received 2% dextran sulfate sodium (DSS, MW 36,000 to 50,000, MP Biomedicals, Solon, OH) in drinking water for 8 days; control mice received regular drinking water. During the period when DSS was administered, mice were given MC-12, 5 or 25 mg/kg, by intraperitoneally (IP), oral gavage (PO) or intrarectally (IR), whereas the control group was given normal saline. The mice were weighed and monitored daily for rectal bleeding or prolapse. All mice were euthanized at the end of the study period.
Blood samples were collected and colons were dissected and their length was measured. The colon was frozen for molecular analyses except for the distal 3 cm, which were fixed in 4% neutralized formalin, cut into six equal fragments, dehydrated and embedded into paraffin. Cross sections of the colon were stained with hematoxylin and eosin (H&E). In these sections, we determined the histological score by the degree (0–3) and extent (0–3) of inflammation, crypt damage (0–4) and the area involved (0–4) as described by Dieleman et al . The score of each of the first three parameters was multiplied by the fourth and the sum of these three multiples was the final score (ranging from 0 to 40).[](https://www.ncbi.nlm.nih.gov/mesh/D016264)
## TNBS model
*In the **TNBS model** section:*
Total 21 SJL were divided into 3 groups, 7 mice per group. The mice received 100 µl of 2.5% TNBS solution in 50% ethanol by intra-colonic instillation using a 3.5 F catheter, which was inserted 4 cm into the colon under mild ketamine/xylazine anesthesia. Mice received MC-12 25 mg/kg or vehicle ip once a day for two days. Body weight was monitored daily and mice were euthanized on the third day when blood and colon tissues were collected and processed as described above. Cross sections of colon tissue were stained with H&E, and the histological score was determined as in the DSS model.[](https://www.ncbi.nlm.nih.gov/mesh/D014302)
## ELISA
*In the **ELISA** section:*
The serum levels of TNF-α, IFN-γ, IL-1β, IL-6 and IL-10 were measured using the Milliplex Map kit (EMD Millipore, Billerica, MA) that following the manufacturer’s instructions. Briefly, 25 µl frozen serum were incubated with 25 µl magnetic beads and 25 µl assay buffer overnight at 4°C. After washing with a Hand-Held Mag Plate Washer (Affymetrics, Santa Clara, CA), the beads were incubated at room temperature with 25 µl biotinylated secondary antibody and 25 µl streptavidin-phycoerythrin for 1 hr each. Fluorescence intensity was determined using the Bio-Plex 200 System (Bio-Rad, Hercules, CA) with a low PMT calibration.
All samples were run in duplicate and analyzed on the same day.
## Myeloperoxidase Activity
*In the **Myeloperoxidase Activity** section:*
MPO activity was measured using a commercial kit and following the instructions of the manufacturer (Invitrogen, Eugene, OR). Briefly, a portion of colon tissue was homogenized in PBS, centrifuged at 10,000×g for 15 min and 50 µl of the supernatant from each sample were added into a 96-well microplate. Then we added 50 µl of 2X APF working solution to all samples and standard wells and incubated the plate at room temperature for 30 min. We stopped the reaction by adding 10 µl of 10X chlorination inhibitor. We measured the fluorescence intensity using a Multiplate Reader (Molecular Devices, Sunnyvale, CA) with excitation at 485 nm and emission at 530 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Cytosolic Phospholipase A2 (cPLA2) Activity
*In the **Cytosolic Phospholipase A2 (cPLA2) Activity** section:*
We determined cPLA2 activity using the cPLA2 assay kit and following the instructions of the manufacturer (Cayman Chemical, Ann Arbor, MI). Briefly, a portion of colon tissue was homogenized in cold PBS and centrifuged at 10,000×g for 15 min and 10 µl of supernatant from each sample and 5 µl assay buffer were added into the wells of a 96-well microplate. We initiated the reaction by adding 200 µl substrate solution to all wells and incubating for 5 min at room temperature. The fluorescence intensity was measured using a Multiplate Reader (Molecular Devices, Sunnyvale, CA) with excitation at 485 nm and emission at 530 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## PGE2 Competitive Enzyme Immunoassay (EIA)
*In the **PGE2 Competitive Enzyme Immunoassay (EIA)** section:*
PGE2 level in colon tissue and PGE2 metabolites in serum were extracted according to manufacturer’s protocol and measured using the PGE2 EIA Kit (#514010, Cayman Chemical, Ann Arbor, MI) and PGE metabolite EIA kit (#514531, Cayman Chemical, Ann Arbor, MI), respectively. Concentrations (pg/ml) were estimated from the absorbance of the calculated standard curve. Results were calculated using the Cayman Chemical computer spreadsheet.[](https://www.ncbi.nlm.nih.gov/mesh/D015232)
## Immunohistochemistry
*In the **Immunohistochemistry** section:*
Immunohistochemical staining for phospho-NF-κB (activated form of NF-κB) and cyclooxygenase-2 (COX-2) was performed on colon tissue samples. Briefly, paraffin-embedded sections (4 µm thick) were deparaffinized, rehydrated, and microwave-heated for 15 min in 0.01 mol/L citric buffer (pH 6.0) for antigen retrieval. Then, 3% hydrogen peroxide was applied to block endogenous peroxidase activity. After 30 min of blocking with normal serum (Invitrogen, Carlsbad, CA), the primary rabbit anti-phospho-NF-κB p65 ser276 antibody (henceforth p-p65, Cell Signaling, Danvers, MA) or rabbit anti-COX-2 polyclonal antibody (Cayman Chemical, Ann Arbor, MI) or the corresponding control isotype IgG were applied and incubated overnight at 4°C. Slides were washed thrice with PBS, each for 5 min.
The biotinylated secondary antibody and the streptavidin-biotin complex were applied, each for a 60 min incubation at room temperature. After rinsing with PBS, the slides were immersed for 10 min in 3,3′-diaminobenzidine (Sigma, St. Louis, MO) solution (0.4 mg/mL, with 0.003% hydrogen peroxide), monitored under the microscope and the reaction was terminated with distilled water. Slides were then counterstained with hematoxylin, dehydrated, and coverslipped. Five fields per section were photographed and the percentage of positive cells in colonic epithelium was determined. The intensity of cytoplasmic staining of COX-2 was assessed semi-quantitatively as previously described: negative: no staining or <10% positive cells; 1+: weak staining or 10–25% positive cells; 2+: moderate staining or 25–50% positive cells; 3+: strong staining or >50% positive cells.[](https://www.ncbi.nlm.nih.gov/mesh/D010232)
## Electrophoretic Mobility Shift Assay (EMSA)
*In the **Electrophoretic Mobility Shift Assay (EMSA)** section:*
After the indicated treatment, cell nuclear fractions were isolated from 3×106 cells as previously described. The NF-κB activity was determined using the LightShift chemiluminescent EMSA kit (Thermo Fisher Scientific, Rockford, IL) and following the manufacturer’s instructions. Briefly, the nuclear extracts were incubated with biotin-labeled DNA probes (5′-AGTTGAGGGGACTTTCCCAGGC-3′) at 37°C for 20 min, then loaded onto the polyacrylamide gel and transferred to a nylon membrane. The membrane was exposed to UV-light for 10 min for cross-linking of the transferred DNA, incubated with stabilized streptavidin-horseradish peroxide conjugate in blocking buffer for 15 min and covered with substrate working solution, followed by exposure to X-ray film.[](https://www.ncbi.nlm.nih.gov/mesh/D001710)
## Real-time Quantitative PCR
*In the **Real-time Quantitative PCR** section:*
About 100 mg of colon tissue were placed in 1 ml of cold Trizol (Invitrogen Life Technologies, Carlsbad, CA), homogenized immediately with a rotor power homogenizer, and RNA was extracted following the manufacturer’s instructions. Total RNA was retrotranscribed with M-MLV reverse transcriptase (Invitrogen Life Technologies, Carlsbad, CA) using random primers. Real-time quantitative PCR was performed in a CFD-3200 Opticon detector (BioRad, Hercules, CA) using QuantiTect SYBR Green PCR Kit (Qiagen, Valencia CA). The PCR cycling conditions were: 40 cycles of 60 seconds at 94°C, 30 seconds at 51.4°C and 30 seconds at 72°C. PCR primers (forward and reverse primers) were designed based on published sequences: TNF-α: AGGCTGCCCCGACTACGT and GACTTTCTCCTGGTATGAGATAGCAAA; IFN-γ: CAGCAACAGCAAGGCGAAA and CTGGACCTGTGG GTTGTTGAC; IL-1β: TCGCTCAGGGTCACAAGAAA and CATCAGAGGCAAGGAGGAAAAC; IL-6:
ACAAGTCGGAGGCTTAATTACACAT and ATGTGTAATTAAGCCTCCGACTTGT; IL-10: ATGCTGCCT GCTCTTACTGACTG and TTGCCATTGCACAACTCTTTTC; β-actin: AGATTACTGCTCTGGCTCCTA and CAAAGAAAGGGTGTAAAACG. Relative expression levels of mRNA were normalized to β-actin.[](https://www.ncbi.nlm.nih.gov/mesh/C411644)
## Statistics
*In the **Statistics** section:*
Data were expressed as mean ± SEM and analyzed by comparing means with One-Way ANOVA using the SPSS program (Version 11.5.0). An additional post hoc test for multiple comparisons was performed. p<0.05 denotes statistically significant differences. The comparison between groups for COX-2 IHC staining was performed using Pearson’s χ2 method.
## Supporting Information
*In the **Supporting Information** section:*
Competing Interests: The authors have declared that no competing interests exist.
Funding: This work was supported by National Institutes of Health grants: R01CA139454 and R01CA09242308. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
# References
*In the **References** section:*
|
# Introduction
Differential roles of galanin on mechanical and cooling responses at the primary afferent nociceptor
# Abstract
*In the **Abstract** section:*
Background
Galanin is expressed in a small percentage of intact small diameter sensory neurons of the dorsal root ganglia and in the afferent terminals of the superficial lamina of the dorsal horn of the spinal cord. The neuropeptide modulates nociception demonstrating dose-dependent pro- and anti-nociceptive actions in the naïve animal. Galanin also plays an important role in chronic pain, with the anti-nociceptive actions enhanced in rodent neuropathic pain models. In this study we compared the role played by galanin and its receptors in mechanical and cold allodynia by identifying individual rat C-fibre nociceptors and characterising their responses to mechanical or acetone stimulation.[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
Results
Mechanically evoked responses in C-fibre nociceptors from naive rats were sensitised after close intra-arterial infusion of galanin or Gal2-11 (a galanin receptor-2/3 agonist) confirming previous data that galanin modulates nociception via activation of GalR2. In contrast, the same dose and route of administration of galanin, but not Gal2-11, inhibited acetone and menthol cooling evoked responses, demonstrating that this inhibitory mechanism is not mediated by activation of GalR2. We then used the partial saphenous nerve ligation injury model of neuropathic pain (PSNI) and the complete Freund’s adjuvant model of inflammation in the rat and demonstrated that close intra-arterial infusion of galanin, but not Gal2-11, reduced cooling evoked nociceptor activity and cooling allodynia in both paradigms, whilst galanin and Gal2-11 both decreased mechanical activation thresholds. A previously described transgenic mouse line which inducib[ly over](https://www.ncbi.nlm.nih.gov/mesh/C434734)-expresses galanin (Gal-OE) after nerve injury was then used to investigate whether manipulating the levels of endogenous galanin also modulates cooling evoked nociceptive behaviours after PSNI.[ Aceton](https://www.ncbi.nlm.nih.gov/mesh/C434734)e withdrawal[ behavi](https://www.ncbi.nlm.nih.gov/mesh/D000096)ours [in naiv](https://www.ncbi.nlm.nih.gov/mesh/D008610)e mice showed no differences between Gal-OE and wildtype (WT) mice. 7-days after PSNI Gal-OE mice demonstrated a significant reduction in the duration of acetone-induced nociceptive behaviours compared to WT mi[ce.](https://www.ncbi.nlm.nih.gov/mesh/D005620).nih.gov/mesh/C434734)
Conclusions
These data identify a novel galaninergic mechanism that inhibits cooling evoked neuronal activity and nociceptive behaviours via a putative GalR1 mode of action that would also be consistent with a TRP channel-dependent mechanism.
## Background
*In the **Background** section:*
Galanin is expressed at low levels in less than 5% of adult rodent dorsal root ganglia (DRG) neurons; those expressing galanin are predominantly small diameter C-fibre nociceptors. Higher levels of the neuropeptide are expressed in the primary afferent terminals in the superficial layers of the dorsal horn in naïve animals [1]. Levels of galanin in the rodent, primate and human DRG markedly rise after peripheral nerve injury [1,2]. Administration of galanin to either the peripheral [3] or central [4] nervous systems results in a significant alteration in mechanosensory nociceptive behaviours. These actions are dose-dependent with facilitation occurring at low concentrations and inhibition at higher concentrations [3,4].
After nerve injury, when a subset of mechano-nociceptors are sensitised to mechanical stimulation [5] and the endogenous levels of galanin in the DRG are high, there is good agreement in the literature that the anti-nociceptive actions of galanin are enhanced in rodent models of neuropathic pain [6]. In contrast to the extensive dataset on galanin modulation of mechanosensory behaviours, there have been few studies on galanin and cold behaviours. Work to date has shown that in the naïve uninjured animal intrathecal infusion of low dose galanin facilitates cold nociceptive behavioural responses [7]. In contrast, in a model of neuropathic pain galanin is anti-nociceptive, attenuating cold pain [8].
The nociceptive effects of galanin are mediated by the activation of one or more of three G-protein coupled galanin receptor subtypes, designated GalR1, GalR2 and GalR3. Studies using in-situ hybridization have shown that GalR1 and GalR2 mRNAs are expressed by 51% and 83% of adult rat DRG neurons respectively [9], and the levels of both sub-types decrease after nerve injury [10,11]. In contrast, no change in the expression of either subtype was observed in the dorsal horn of the spinal cord after axotomy [12]. Expression of GalR3 in the spinal cord and DRG in both rat and mouse is very low as determined by RT-70 PCR [13,14], and in situ hybridization [15].
We and others have shown that the central and peripheral anti-nociceptive effects of galanin on mechanosensory thresholds are principally mediated by activation of GalR1 [7] and GalR2 [3], respectively. Only two papers have studied which of the galanin receptors mediate the effects of galanin on cold nociception. Liu et al. showed that intrathecal infusion of low dose Gal2-11 (a GalR2/3-specific agonist) increased nociceptive responses to acetone in the naive rat, similar to that seen with galanin
[7]. In contrast, Blakeman et al. demonstrated an increase in cold pain scores in naive GalR1-KO mice compared to strain-matched wild-type (WT) controls [16].[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
Here we investigated the role that galanin and its receptors play at the primary afferent with particular reference to cold sensitivity, demonstrating that close intra-arterial administration of galanin inhibits C-fibre nociceptor cooling activity in normal, nerve-injured and inflamed animals. These actions appear to be independent of GalR2 activation. These data identify a novel galaninergic mechanism that inhibits cooling-evoked neuronal activity and nociceptive behaviours consistent with a TRP channel-dependent mechanism.
## Results
*In the **Results** section:*
Using multi-unit recordings in naive rats and models of neuropathic and inflammatory pain, we studied the differential effects of galanin and the GalR2/3-specific agonist Gal2-11 on primary afferent nociceptor activities to mechanical and cooling stimulation.[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
## Differential effects of galanin and Gal2-11 on naïve primary afferent nociceptor activities to mechanical and cooling stimulation
*In the **Differential effects of galanin and Gal2-11 on naïve primary afferent nociceptor activities to mechanical and cooling stimulation** section:*
Galanin 100 μM was administered via c.i.a. (close intra-arterial) as previously described [17]. Mechanical activation thresholds were significantly reduced following galanin administration (Figure 1A, ***p < 0.001) and mechanically evoked activity was enhanced (Figure 1B, ** p < 0.01). Similarly, c.i.a. administration of 100 μM Gal2-11 led to a significant reduction in activation thresholds of mechanically sensitive C-fibre afferents (Figure 1C, **p < 0.01) and an increase in mechanically evoked activity (Figure 1D, *p < 0.05).[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
Facilitatory actions of galaninviaactivation of GalR2 on mechano-nociceptor afferents [A] Mechanical activation thresholds were reduced by c.i.a 100 μM galanin in naïve rats (*** p < 0.001 Mann Whitney test, n = 24) and [B] mechanically evoked activity was increased by c.i.a 100 μM galanin in response to a 10 g vF hair mechanical stimulus (**p < 0.01 paired t-test). [C] Mechanical activation thresholds were reduced following c.i.a 100 μM Gal2-11 administration (** p < 0.01 Mann Whitney test, n = 14) and [D] mechanically evoked activity was increased by c.i.a 100 μM Gal2-11 in response to a 10 g vF hair mechanical stimulus (* p < 0.01 paired t-test, n = 14).[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
Application of acetone, which leads to mild nociceptive behaviours in naïve rodents [18], reduced skin temperature for approximately 30–60 seconds by ~5°C from a resting temperature of ~25°C (Figure 2A). Room temperature saline was used to confirm that the mechanical application of a drop of fluid to the receptive field did not alter skin temperature (Figure 2A) or neuronal responses (Figure 2B). On-going neuronal activity in cooling primary afferents (as previously described [19]), was identified from the application of room temperature saline (Figures 2B and 3A). All cooling sensitive afferents had CVs less than 1 m/s and were thus classified as cooling sensitive C-fibres.
Application of acetone to the receptive field significantly increased afferent evoked activity compared to room temperature saline (Figures 2B &3A *p < 0.05, **p < 0.01 one way ANOVA). Acetone responses following c.i.a. 100 μM galanin were significantly reduced (Figures 2B &2D, * p < 0.05 Paired t-test). In contrast, 100 μM Gal2-11 had no effect on acetone evoked responses (Figures 3A and 3C).[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
Exogenous galanin attenuates cooling responses. [A] Digitised trace of recording of subcutaneous skin temperature of the dorsal surface of the rat hindpaw. Room temperature saline did not alter skin temperature. Application of acetone led to a reduction in skin temperature. This returned to normal within 100 seconds. A second application of acetone again resulted in a similar drop in temperature. [B] Acetone application to the receptive field of an identified cooling sensitive primary afferent led to a significant increase in neuronal firing, which was significantly greater than the control application of room temperature saline (**p < 0.01 one way ANOVA with post-hoc Bonferroni test, n = 6). Cooling evoked activity was significantly inhibited following administration of c.i.a. 100 μM galanin (*p < 0.05 Paired t-test, n = 6). [C] Example trace of an acetone/cooling sensitive C fibre in response to room temperature saline. [D] Example trace of an acetone/cooling sensitive C fibre. [E] Acetone evoked activity was attenuated in the same C fibre by c.i.a. 100 μM galanin. AP = action potential, AP number events = total number of action potentials occurring per second.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
Acetone evoked activity is not inhibited by GalR2/3 activation. [A] c.i.a. 100 μM Gal2-11 did not alter acetone-induced cooling responses (n = 6). [B] Example trace of an acetone/cooling sensitive C fibre. [C] The acetone evoked activity was not altered by Gal2-11. AP = action potential, AP number events = total number of action potentials occurring per second.[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
We then investigated the effects of galanin and Gal2-11 on the neuronal activity elicited after application of the cooling agent menthol; an activator of the cold sensitive channel TRPM8 [20]. Menthol evoked action potentials in 8 out of 10 acetone/cooling sensitive afferents. A single c.i.a bolus dose of 100 mM menthol led to a robust increase in evoked activity (Figure 4A) which was significantly reduced when preceded by c.i.a administration of 100 μM galanin (Figure 4B, *p < 0.05 one way ANOVA) but not by c.i.a 100 μM Gal2-11 (Figure 4B).[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
Menthol evoked activity is not inhibited by GalR2/3 activation. [A] The concentration response relationship demonstrates that acetone sensitive C fibres responds only to c.i.a. 100 mM menthol (***p < 0.001, one way ANOVA with post-hoc Bonferroni test, n = 4). [B] Menthol evoked activity was reduced by c.i.a. 100 μM galanin but not by c.i.a. 100 μM Gal2-11 (* p < 0.05, one way ANOVA with post-hoc Bonferroni test, n = 7).[](https://www.ncbi.nlm.nih.gov/mesh/D008610)
## Differential effects of galanin and Gal2-11 on primary afferent nociceptor activities to mechanical and cooling stimulation in chronic pain models
*In the **Differential effects of galanin and Gal2-11 on primary afferent nociceptor activities to mechanical and cooling stimulation in chronic pain models** section:*
In rats that had undergone PSNI that results in mechanical and cold allodynia, c.i.a. 100 μM galanin and c.i.a 100 μM Gal2-11 both led to a significant reduction in mechanical activation thresholds of C-fibres (Figure 5A, ***p < 0.001). In contrast, acetone-evoked activity in C-fibre afferents was significantly attenuated following c.i.a 100 μM galanin (Figure 5B, *p < 0.05, ***p < 0.001) but not by c.i.a 100 μM Gal2-11 (Figure 5B).[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
Differential role of galanin in a model of neuropathic pain [A] Exogenous c.i.a 100 μM galanin and c.i.a 100 μM Gal2-11 both facilitated mechano-nociceptor responses in the partial saphenous nerve injury model of neuropathic pain (** p < 0.01, *** p < 0.001, Kruskal Wallis test with Dunn’s multiple comparison test, galanin n = 15, Gal2-11 n = 6). [B] In the PSNI neuropathic pain model acetone application to the receptive field of cold sensitive C fibres led to a significant increase in evoked activity (***p < 0.001, Kruskal-Wallis with Dunn’s multiple comparison test). Acetone-induced evoked activity was inhibited by c.i.a 100 μM galanin but was not altered by c.i.a 100 μM Gal2-11 (***p < 0.001, one way ANOVA with post-hoc Bonferroni test, n = 10) [C] Acetone-induced withdrawal behaviours in naïve Gal-OE transgenic mice did not differ from WT controls. 7-days after nerve injury the Gal-OE mice demonstrated a significant reduction in cooling pain like behaviours compared to matched WT mice (*p < 0.01, two way ANOVA with post-hoc Bonferroni test, n = 7).[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
A previously described transgenic mouse line which inducibly over-expresses galanin (Gal-OE) after nerve injury [21] was used to investigate whether manipulating the levels of endogenous galanin also modulates cooling evoked nociceptive behaviours after PSNI. We have previously shown a marked attenuation of mechanical allodynia in these mice after PSNI [3,21]. Acetone withdrawal behaviours in naive mice showed no differences between Gal-OE and WT mice. 7-days after PSNI Gal-OE mice demonstrated a significant reduction in the duration of acetone-induced nociceptive behaviours compared to WT mice (Figure 5C, *p < 0.01 two way ANOVA).[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
Similar responses were also observed following CFA treatment in the rat. The degree of inflammation at 5 days was confirmed by measuring ankle joint swelling (Figure 6A, ***p < 0.001, paired t-test). 100 μM galanin and 100 μM Gal2-11 by c.i.a administration both led to a significant reduction in mechanical activation thresholds of mechanically sensitive C-fibres (Figure 6B, **p < 0.01, ***p < 0.001). In contrast, c.i.a 100 μM galanin, but not c.i.a 100 μM Gal2-11, attenuated the acetone evoked responses in cooling sensitive C-fibres (Figure 6C, *p < 0.05 one way ANOVA).[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
Differential role of galanin in CFA inflammation [A] CFA inflammation was confirmed through significant ipsilateral ankle joint swelling compared to the contralateral joint (***p < 0.001, paired t-test, n = 6). [B] Exogenous c.i.a. 100 μM galanin and c.i.a. 100 μM Gal2-11 both significantly facilitated mechano-nociceptor mechanical responses in CFA inflamed rats (** p < 0.01, *** p < 0.001, Kruskal-Wallis with Dunns multiple comparison test, n = 9). [C] Exogenous c.i.a. 100 μM galanin, but not c.i.a. 100 μM Gal2-11, reduced acetone evoked activity (* p < 0.05, one way ANOVA with post-hoc Bonferroni, n = 6).[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
## Discussion
*In the **Discussion** section:*
This study describes the differential roles played by galanin in the peripheral nervous system with respect to mechanical- and cold-pain behaviours in the naive state and following models of neuropathic and inflammatory pain. Here we show opposing actions of galanin, with inhibition of cold- and facilitation of mechanical-responses, in C-fibre nociceptors.
Cold allodynia (the sensation of pain to a normally innocuous stimulus) is a common complaint of patients with neuropathic pain, irrespective of the underlying cause [22-24] and similar findings are found in a range of rodent models of neuropathic and inflammatory pain behaviours following stimulation with acetone or menthol [25-28]. Cooling detection in humans is highly variable. It occurs after a drop in temperature of only a few degrees [29,30] and is highly correlated with the activation temperature of TRPM8
[31-33] (31-27°C). Cold pain thresholds are also highly variable, but are usually reported at lower temperatures [31,34] (<10°C). Patients with neuropathic pain report pain after only a small reduction in skin temperature with thresholds reported to be >20°C [35]. At these temperatures, symptoms are described as painful with intense pain occurring at lower temperatures [23] (~15°C).
Recent work has made considerable progress in defining the molecular transducers of cutaneous thermo-sensation and the mechanisms that underlie cold allodynia. A number of thermal sensors have been identified which all belong to the Transient Receptor Potential (TRP) ion channel family. The best characterised cold receptor is TRPM8 [36-38]. TRPA1 was proposed as a noxious cold sensor but more recent evidence has disputed a specific contribution for TRPA1 in cold transduction
[28,39,40]. TRPM8 is expressed by a subpopulation of small diameter C-fibre neurons [41]. TRPM8 responds to cooling agents such as acetone with an activation temperature threshold of ~26°C, with activity increasing in magnitude down to 8°C [36,38]. Menthol activates TRPM8
[42] and does not influence TRPA1 activity in non-human mammals [20,43,44]. Consistent with these findings, behavioural nociceptive responses over a range of innocuous and noxious temperatures and to the application of menthol are attenuated in TRPM8-KO mice, whereas these responses remain intact in the TRPA1-KO [39].[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
In the present study we have shown that galanin, but not Gal2-11, inhibits acetone and menthol responses in the naive rodent and following models of neuropathic and inflammatory pain. Similarly, overexpression of galanin after nerve injury in the DRG of Gal-OE mice led to the inhibition of cooling pain responses compared to those of wild type mice, confirming that high levels of endogenous galanin also inhibit cooling behaviours. The modulatory effects of galanin on cooling are independent of GalR2 and GalR3 activation since Gal2-11, which is equally selective for GalR2 and GalR3 [45], does not inhibit cooling pain behaviours. Further, since GalR3 is expressed at very low levels in the primary nociceptor in the intact and injured state
[14], these findings imply the effects of galanin on cold responses are mediated by activation of GalR1. These findings are consistent with our work on the Gal-KO [46] and that of Blakeman et al. which demonstrate an increase in cold pain scores in naive GalR1-KO mice compared to strain-matched wild-type (WT) controls [16].[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
Since galanin suppresses the effects of acetone and menthol and the majority of acetone-sensitive C-fibres were also activated by c.i.a administration of menthol, this implies a mechanism via the modulation of the TRPM8 cold sensor. To date, no publication has investigated the effects of galanin and its receptors on TRPM8 activity. However, peripheral administration of galanin has been shown to modulate TRPV1 (Transient Receptor Potential Vanilloid) activity following inflammation, as GalR1-activation inhibits capsaicin-evoked pain behaviours [47], whilst GalR2-activation facilitates the same nociceptive responses [48].
Of note, all three galanin receptors couple to Gi/o and inhibit adenylyl cyclase [49,50], whilst GalR2 also signals via Gq/11 to activate phospholipase C (PLC) and protein kinase C (PKC) [51]. Previous work has shown that TRPM8 activity in the DRG is inhibited by the Gi/o adenylyl cyclase pathway [52], which is again consistent with a GalR1-dependent modulation of the channel.[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
In contrast to the above inhibitory effects of peripheral administration of galanin on cooling nociception, here we show that c.i.a administration of the same dose (100 μM) of galanin and Gal2-11 both facilitate mechano-sensory responses. The nature and magnitude of these changes (Figure 1) are very similar to our previously published dataset using injection of 100nM galanin or 100nM Gal2-11 directly into the afferent receptive field in naïve adult male rats [3] (Figures 15A and 6B in that paper). This implies that the local concentration of galanin or Gal2-11 at the nociceptor is ~1000-fold lower than that administered by c.i.a bolus injection which may be explained by a combination of dilution of the peptides by the blood volume of the hind limb and only partial tissue penetration of the peptides. Our current and previous findings confirm that the peripheral mechano-nociceptive effects of galanin appear to be predominantly dependent on activation of GalR2, in contrast to the effects of galanin administered intrathecally which appear to be mediated by activation of GalR1
[7].[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
In conclusion, peripheral administration of galanin, but not Gal2-11, inhibits acetone- and menthol-evoked activity in C-fibre nociceptors in the normal animal and in models of neuropathic and inflammatory pain. These data identify a novel putative GalR1-dependent pathway that inhibits cooling-evoked neuronal activity and nociceptive behaviours, consistent with a TRPM8-dependent mechanism. Taken together with the known peripheral GalR2-dependent modulation of mechano-nociception, these findings imply that activation of peripheral GalR1 and GalR2 will both be important to treat cold and mechanical allodynia that occurs in various neuropathic pain conditions.[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
## Materials and methods
*In the **Materials and methods** section:*
## Animals
*In the **Animals** section:*
Experiments were performed on adult male Wistar rats (250 g-350 g) and mice (~25 g). 47 Wistar rats were used for electrophysiological experiments. 10 transgenic mice that inducibly over-express galanin in the DRG as previously described [3], and 10 matched WT controls were used in nociceptive behavioural testing. Animals were fed standard chow and water ad libitum and all experiments were carried out in accordance with the United Kingdom Animals (Scientific Procedures) Act 1986.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Electrophysiological recordings
*In the **Electrophysiological recordings** section:*
Anaesthesia was induced using sodium pentobarbital (60 mg/kg i.p. rats, Sigma-Aldrich, UK) and were maintained deeply anaesthetised and areflexive for the duration of the experiment (~20 mg/kg/hr, sodium pentobarbital delivered intravenously via cannulation of the external jugular vein). A tracheotomy was performed to maintain the airway. Monitoring of blood pressure was carried out via a femoral artery cannulation of the left hind limb. Rodent body temperature was maintained at ~37.5°C by means of a feedback controlled heater and rectal thermistor. At the end of all experiments, rats were overdosed with a single large bolus of (60 mg/ml) sodium pentobarbital.[](https://www.ncbi.nlm.nih.gov/mesh/D010424)
An incision was made along the inguinal fossa of the right hind leg to expose the saphenous nerve which runs superficially under the skin adjacent to the femoral vasculature. Using blunt dissection, connective tissue was removed and the skin was freed. Using the skin a pool was formed by attaching the skin to a metal pool ring, which was filled with warmed mineral oil (Sigma-Aldrich, UK). Fine forceps were used to dissect out a sufficient length of the nerve that was placed onto a dental mirror. The main saphenous nerve trunk was further prepared by removing the surrounding epineurium.
Fine filaments were dissected out and cut proximally. These nerve filaments were placed onto bipolar platinum recording electrodes to record from individual afferent responses. Identified fine nerve filaments were ensured to have a low number of responsive receptive fields in the hindpaw (<3), with these receptive fields not overlapping. The neuronal activity of individual afferents was captured and spike sorted offline using micro1401 and Spike2 software (C.E.D. Cambridge, UK).[](https://www.ncbi.nlm.nih.gov/mesh/D008899)
## Afferent characterisation
*In the **Afferent characterisation** section:*
Primary afferents were identified using search stimuli of both blunt/pinch mechanical stimulation of the medial surface of the hindpaw, and electrical stimulation. Each afferent identified was characterised with the same protocol as described below. Upon identification of a receptive field the conduction velocity (CV) for the afferent was recorded. Electrical stimulation of the receptive field (0.5 ms stimulus duration, up to 100 V, every 3 seconds) was used to calculate the conduction velocity of the afferent. Three or more robust CVs from the receptive field were required to confirm reproducibility and confirmation of afferent excitation.
CVs less than 1 m/s were classed as C fibres, those with greater than 1 m/s as A fibres [53]. On-going neuronal activity was monitored for 100 seconds following CV characterisation. This 100 second period recorded any neuronal activity from the afferent that occurred without any stimulus being applied to the receptive field. Mechanical responses were recorded using calibrated plastic filaments; von Frey hairs (Linton Instruments, UK) and brush stimulation. Von Frey hairs (vF) were used to elicit mechanical responses, with each vF hair applied three times for a maximum of five seconds.
The mechanical activation threshold was determined using an adapted up and down method, and the lowest vF hair that elicited a robust response (response >3 action potentials) was noted as the activation threshold [54]. Afferents were also tested for innocuous cooling responses. Acetone was applied to the receptive field to test for cooling responses. This stimulus leads to withdrawal behaviours
[55] and is a standard tool to excite cooling neuronal responses [56,57]. Prior to acetone, a single drop of room temperature saline was applied (using a Gilson and 1 ml pipette tip) to the receptive field to identify any mechanical response occurring due to the application of a drop of fluid. Neuronal activity was recorded for 30 seconds post stimulation. Room temperature saline and acetone were each applied a total of 3 times with 5 minutes between each application to allow skin temperature and neuronal activity to recover to baseline levels. Mean acetone and saline evoked responses were calculated.
To note, cooling afferent groups studied with drug administration include both CCs (cold sensitive C fibres) and CMCs (mechano-cold sensitive C fibres) as previously described [58,59]. Drug responses were not different between these two groups.[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
## Experimental protocol and drug administration
*In the **Experimental protocol and drug administration** section:*
Following baseline mechanical and cooling characterisation of primary afferent fibres, drugs were administered via the close intra-arterial (c.i.a) route as previously described [17,60]. The left femoral artery was cannulated in the groin and the cannula was advanced proximally, approximately ~2.5 cm, towards the bifurcation of the descending aorta. Cannula positioning was confirmed by visual inspection upon termination of experiment. Menthol (Sigma-Aldrich), 100 μM galanin (Bachem) or 100 μM Gal2-11 (Sigma-Aldrich, UK), were made up in saline and injected and washed through the c.i.a. cannula with 400ul of saline (containing heparin 50units/ml).
5 minutes after galanin or Gal2-11 injection either mechanical, acetone and menthol responses were tested. Since galanin and Gal2-11 have short half-life of <10 minutes [46,61], in some experiments galanin and Gal2-11 were tested sequentially in the same animal to reduce the number of animals used. 30 minutes was allowed between peptides to allow for metabolism and removal. 100 mM menthol (Sigma-Aldrich, UK) was administered via c.i.a cannula as previously described [28,55,60]. Menthol was dissolved in 10% ethanol, 10% Tween80 and 80% saline and vehicle alone had no effect on neuronal activity (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D008610)
## Temperature recordings
*In the **Temperature recordings** section:*
A thermocouple was advanced under the skin of the medial surface of the hindpaw using a 25-gauge needle. The thermocouple was used to identify changes in skin temperature following application of various stimuli. Saline and acetone were applied and temperature changes monitored and recorded using a CED micro 1401 and Spike2 software as described previously [56].[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## Chronic pain models
*In the **Chronic pain models** section:*
For all recovery surgeries; partial saphenous nerve ligation injury (PSNI) and complete Freund’s adjuvant (CFA), anaesthesia was induced using 4% halothane in oxygen and maintained with 2-3% for all procedures.[](https://www.ncbi.nlm.nih.gov/mesh/D005620)
## PSNI surgery
*In the **PSNI surgery** section:*
The PSNI surgery was performed as previously described [57,62] on 19 male wistar rats and 20 male mice, leading to the characteristic neuropathic pain phenotypes of mechanical allodynia and cooling allodynia [57,62,63]. An incision was made along the inguinal fossa of the right hind leg and the saphenous nerve was isolated using blunt dissection. A sterile silk suture (mouse: 7.0 suture and rat: 4.0 suture) was used to tightly ligate 50% of the saphenous nerve. The skin incision was closed using size 4.0 sutures. The animals were allowed to recover and monitored for the duration of the experiment.
Electrophysiological experimentation was performed 7 days after PSNI surgery.[](https://www.ncbi.nlm.nih.gov/mesh/D047011)
## Complete freunds adjuvant (CFA) inflammation
*In the **Complete freunds adjuvant (CFA) inflammation** section:*
Two 50 μl injections of CFA (1 mg/ml, Sigma-Aldrich, UK), one on either side of the tibio-tarsal ankle joint, were used to induce inflammation in nine male Wistar rats as previously described [60] for electrophysiological experimentation and left to develop for five days. This intervention leads to significant swelling of the ankle and hindpaw [64].[](https://www.ncbi.nlm.nih.gov/mesh/D005620)
## Animal behaviour
*In the **Animal behaviour** section:*
Mice were habituated to behavioural testing enclosures (mesh floored holding chambers) the day prior to testing for 15 minutes. For testing days animals were placed into behavioural enclosures 10 minutes prior to testing. Using a 1 ml syringe, acetone was applied to the plantar surface of the hindpaw. The duration of flinching/pain like behaviour (seconds) was recorded immediately following acetone application for a total period of five minutes. This was performed three times for each hindpaw, from which the mean was calculated.
Stimulations of the right and left hindpaw were alternated to prevent sensitisation and allow recovery of skin temperature. Baselines were taken for two separate recording sessions and were then used to generate the baseline timepoint. Seven days post nerve injury animals were re-tested for acetone responses.[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
## Statistical analysis
*In the **Statistical analysis** section:*
Paired t-test and one way ANOVA with post-hoc Bonferroni tests were used to determine primary afferent evoked activity to mechanical or cooling stimulation before and after drug, as appropriate (galanin, Gal2-11 and menthol, *p < 0.05, **p < 0.01, ***p < 0.001). Mechanical activation threshold was determined by Mann Whitney Test or Kruskal-Wallis test with Dunns multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001). To determine if galanin or Gal2-11 facilitated mechanical responses (increased number of action potentials) to an evoked stimulus, a cut off was set as two standard deviations above and below the mean of the evoked activity. In addition, cooling sensitive primary afferent evoked activity in response to acetone was calculated through a fold change in activity compared to that following room temperature saline application. Percentage change of cold induced pain like behavioural responses following PSNI were calculated from baseline naïve animal responses.
Two way ANOVA with post bonferroni tests was used to determine differences in the Gal-OE cooling behaviour analysis. All data are presented as mean ± SEM.[](https://www.ncbi.nlm.nih.gov/mesh/C434734)
## Abbreviations
*In the **Abbreviations** section:*
CFA: Complete Freunds adjuvant; c.i.a: Close intra arterial; DRG: Dorsal root ganglia; Gal: Galanin; GalR: Galanin receptor; Gal-OE: Galanin over-expressing transgenic mice; i.t: Intrathecal; KO: Knockout; PSNI: Partial saphenous nerve injury; WT: Wild-type mice.[](https://www.ncbi.nlm.nih.gov/mesh/D005620)
## Competing interest
*In the **Competing interest** section:*
The authors declare no competing interest.
## Authors’ contributions
*In the **Authors’ contributions** section:*
RH formulated the hypothesis, initiated and organized the study. DW and LD provided the funding. RH performed the experimental work and analysed the data. RH, LD and DW drafted the manuscript. All authors read and approved the final manuscript.
## Acknowledgements
*In the **Acknowledgements** section:*
Work was supported by the Medical Research Council, The Wellcome Trust and National Institute on Aging (AG10668). We also thank Prof Bruce Matthews for his invaluable advice and assistance.
|
# Introduction
Trp RNA-Binding Attenuation Protein: Modifying Symmetry and Stability of a Circular Oligomer
# Abstract
*In the **Abstract** section:*
Background
Subunit number is amongst the most important structural parameters that determine size, symmetry and geometry of a circular protein oligomer. The L-tryptophan biosynthesis regulator, TRAP, present in several Bacilli, is a good model system for investi[gating deter](https://www.ncbi.nlm.nih.gov/mesh/D014364)minants of the oligomeric state. A short segment of C-terminal residues defines whether TRAP forms an 11-mer or 12-mer assembly. To understand which oligomeric state is more stable, we examine the stability of several wild type and mutant TRAP proteins.
Methodology/Principal Findings
Among the wild type B. stearothermophilus, B. halodurans and B. subtilis TRAP, we find that the former is the most stable whilst the latter is the least. Thermal stability of all TRAP is shown to increase with L-tryptophan concentration. We also find that mutant TRAP molecules that are truncated at the C-terminus - and hence induced to form 12-mers, distinct from their 11-mer wild type counterparts - have increased melting temperatures. We show that the same effect can be achieved by a point mutation S72N at a subunit interface, which leads to exclusion of C-terminal residues from the interface. Our findings are supported by dye-based scanning fluorimetry, CD spectroscopy, and by crystal structure and mass spectrometry analysis of the B. subtilis S72N TRAP.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Conclusions/Significance
We conclude that the oligomeric state of a circular protein can be changed by introducing a point mutation at a subunit interface. Exclusion (or deletion) of the C-terminus from the subunit interface has a major impact on properties of TRAP oligomers, making them more stable, and we argue that the cause of these changes is the altered oligomeric state. The more stable TRAP oligomers could be used in potential applications of TRAP in bionanotechnology.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Cyclic protein oligomers pose promise as the basis for engineering molecular machines, including those that operate by rotation of one or more components about a principal axis. Their suitability, owing to their unique geometry, is evident by their frequent occurrence in key bio-mechanical and regulatory roles. Examples include circular protein oligomers found in flagellar motors , the ATP synthase and the molecular motor of tailed bacteriophages which translocate DNA into the capsid during viral particle assembly –. The stoichiometry of the MS-ring and C-ring in flagellar motors are variable . Likewise, the central component of the bacteriophage DNA-translocating motor can also exist in different oligomeric states .
This demonstrates a general problem associated with instability of circular oligomers and variation in their oligomeric states when they are extracted from their natural environment. Engineering stable molecular devices of suitable size, geometry, and hence subunit number is of interest for applications in molecular biology and medicine. For example, applications of viral DNA-translocating devices in the transfer of genetic information across biological membranes are attractive. Recent studies aimed at the incorporation of a portal protein into lipid bilayers look particularly promising .[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
The tryptophan RNA-binding attenuation protein, TRAP, possesses similar geometric features to other circular assemblies, being assembled of multiple subunits and containing a central tunnel. Typically composed of 11 or 12 identical subunits, with a molecular weight of 90.6 to 101.9 kDa , , TRAP regulates L-tryptophan biosynthesis genes by attenuation in many Bacilli . When activated by increased levels of L-tryptophan – bound to receptor sites nestled between adjacent subunits – TRAP binds single stranded RNA at the leader region of the trpEDCFBA operon transcript. This modulates terminator hairpin formation and so downregulates L-tryptophan biosynthesis . In this study TRAP is employed as a model for investigating the influences of oligomeric state and stability of such assemblies.
To date TRAP has already alluded to its value in bionanotechnology, showing that mutations to residues protruding into the central cavity allow gold nanoparticle binding . Nanotubes consisting of stacked TRAP rings linked by disulfide bonds have also been shown to self-assemble, in response to the introduction of cysteine residues on the ring-faces orthogonal to the principal axis .[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
TRAP oligomers have previously been reported to be highly thermostable . Predictably, we find that the melting temperature is highest in the case of TRAP from B. stearothermohphilus, a hyperthermophile. The presence of L-tryptophan – and hence the bound holo-TRAP state – has been associated with a greater structural rigidity and integrity . Prior to the recent determination of the B. stearothermophilus apo-TRAP crystal structure TRAP was crystallized in the presence of L-tryptophan , . We find that L-tryptophan binds with a Kd in the 2.8 µM to 6.9 µM range, with small variation between TRAP from different species.
We explore how the presence of L-tryptophan and its’ binding affects thermal stability in the context of the three wild type and three mutant TRAP molecules. We also explore the influence of the oligomeric state on thermal stability by comparing wild type 11-subunit and mutant 12-subunit molecules from the same species. Finally, we show how the 11-subunit to 12-subunit switch could be induced by the point mutation S72N. The analysis includes crystal structure and native mass spectrometry data on B. subtilis S72N TRAP. We find that the increase in subunit number increases the stability of the assembly.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## Results
*In the **Results** section:*
## Crystal Structure and Native Mass Spectrometry of B. subtilis S72N TRAP
*In the **Crystal Structure and Native Mass Spectrometry of B. subtilis S72N TRAP** section:*
We previously demonstrated the importance of the C-terminus in oligomer formation during B. subtilis TRAP assembly . However, this work required deletion of the C-terminus to increase the subunit stoichiometry and we questioned if a more subtle mutation at the C-terminus could induce formation of a 12-subunit assembly. Based on the previous structural work, we chose the S72N mutation because we expected that the bulky side chain at this position would result in the exclusion of N72 and following C-terminal residues from the subunit interface. The crystal structure of B. subtilis S72N TRAP, determined at 2.7 Å, reveals a circular 12-mer assembly ( Figure 1 ), which is similar to previously reported crystal structures of TRAP. In the crystal the 12 subunit assembly is generated by combination of six TRAP subunits, present in the asymmetric unit, with the crystallographic 2-fold axis of the P21212 space group ( Table 1 ).
The final electron density maps allowed positioning of all residues, with the exception of the four N-terminal residues and the five C-terminal residues of most subunits, for which no clear electron density has been observed, indicating their flexibility. The structure of the S72N TRAP was compared against the wild type TRAP ( Figure 2A ), which highlighted change in the conformation of the C-terminus in the mutant protein, starting from position 69. The weighted Fo – Fc omit electron density maps for residues E69 and M70 ( Figure 2B ) demonstrate the absence of interpretable electron density beyond position 70 in all chains, indicating that five C-terminal residues are disordered. The crystal structure compares well with our predictions on the effect of the point mutation, in inducing the removal of the C-terminus from the subunit interface through steric hindrance, thus facilitating rigid-body rotations of subunits for the accommodation of one additional subunit into the ring.
Ribbon diagram of the B. subtilis S72N TRAP viewed along the 12-fold axis.
Each subunit is shown in a different colour. L-tryptophan molecules are shown as van der Waals models with oxygen atoms in red, nitrogen atoms in blue and carbon atoms in yellow.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Values in parentheses are for the highest resolution shell.
Rmerge = ∑hkl∑i|Ii(h) - <I(h)>|/∑hkl∑i Ii(h), where I(h) is intensity of reflection h, <I(h)> is average value of intensity, the sum ∑hkl is over all measured reflections and the sum ∑i is over i measurements of a reflection.
Crystallographic R = ∑hkl||Fobs - Fcalc||/∑hkl|Fobs|, Rfree was calculated using a randomly chosen set of reflections that were excluded from the refinement.
Data collection and refinement statistics for B. subtilis S72N TRAP.
To understand if the 12-mer state of the S72N TRAP observed in the X-ray structure, predominates in solution, we analyzed this protein by native mass spectrometry (Text S1). The 12-mer state was identified as the dominant species in the mass spectra (Figure S1), as opposed to the wild type protein which forms 11-mers .
Structural differences between 11-subunit and 12-subunit TRAP.
(A) Comparison of the B. subtilis wild type (red) and S72N mutant TRAP (blue). Two neighboring subunits were least-square fitted using main chain atoms of the subunit on the left. The C-terminal residues that are pivoted out of the subunit interface in S72N TRAP are highlighted in white on the wild type TRAP, starting from residue 69. (B) C-terminal residues E69 and M70 are shown in sticks with main chain in yellow and side chains in turquoise, the rest of each subunit is shown in ribbons. The weighted F
o – F
c omit maps, contoured at 2σ, were calculated after omitting residues 69 and 70 from the final model and 10 cycles of refinement.
## 12-mer TRAP Assemblies are More Stable
*In the **12-mer TRAP Assemblies are More Stable** section:*
Melting temperatures for six TRAP molecules, including three wild type and three mutant TRAP listed in Table 2, were determined at varied L-tryptophan concentrations (0–100 µM) using dye-based scanning fluorimetry as described in the methods. Fluorescence versus temperature melting curve data (Figure S2, S3) was averaged between four observations and the midpoints of fitted sigmoidal curves were taken as the melting temperatures, Tm. Melting temperatures (Table S1) were then plotted against L-tryptophan concentration ( Figure 3 ).[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Melting temperatures of different TRAP oligomers.
The dependence of melting temperature is shown for six TRAP oligomers, as a function of L-tryptophan concentration, determined by dye-based scanning fluorimetry. From the top: B. stearothermophilus E71stop 12-mer (triangle), B. stearothermophilus wild type 11-mer (square), B. halodurans wild type 12-mer (diamond), B. subtilis K71stop 12-mer (open cross), B. subtilis S72N 12-mer (circle), B. subtilis wild type 11-mer (crossed square).[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
L-tryptophan binding constants (Kd) of three wild type and three mutant TRAP.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
As the fluorescent dye could potentially bind in the tryptophan-binding pocket, we employed CD spectroscopy (Text S2) to verify that melting temperatures were not perturbed by interaction with the dye, Figure S2. We note that there is coherence between melting temperatures derived from the two respective techniques.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Increasing the L-tryptophan concentration increases the thermal stability of all the TRAP proteins except for wild type B. subtilis TRAP. B. halodurans TRAP was included as a reference, as it is naturally a 12-mer, without engineering. Throughout the L-tryptophan concentrations for which the Tm was measured, the wild type B. subilis TRAP was least stable, while the B. stearothermophilus was most stable. Significantly, the mutant 12 subunit assemblies were at least as stable, but usually more stable than the 11 subunit wild-type, at all L-tryptophan concentrations measured. This stabilization was particularly marked at high L-tryptophan concentrations as the mutant 12-mer subunit TRAPs have a larger increase in Tm with increasing L-tryptophan concentration than their wild type 11-mer counterparts.
For example, from 0 to 100 µM L-tryptophan, B. subtilis wild type increases in stability by 0.8°C, while the Tm of the K71stop and S72N 12-mer mutants increases by 23.2°C and 13.3°C respectively. The B. stearothermophilus wild type TRAP increases in stability by 7.1°C, whereas the Tm of E71stop TRAP increases by 20.2°C. The thermal stability of B. subtilis wild type TRAP appeared to show little dependence on L-tryptophan concentration up to 100 µM. We observed that the stability of the wild type 12-mer B. halodurans TRAP lies midway between that of mutant 12-mers, B. subtilis K71stop and B. stearothermophilus E71stop TRAP (
Figure 3 ).[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## L-Tryptophan Binding
*In the **L-Tryptophan Binding** section:*
The increased stabilization by L-tryptophan of the 12-subunit assemblies, as compared to 11-mers, was unexpected. To investigate this further we measured the affinity of each TRAP for L-tryptophan ( Table 2 ). A comparison of the dissociation constants of the mutant 12-mers to their 11-mer wild type counterparts showed no obvious correlation. Although an 11-mer to 12-mer transition in B. subtilis TRAP increases the affinity of binding, this is not observed in B. stearothermophilus TRAP. This suggests stabilization by L-tryptophan is not directly related to affinity towards L-tryptophan.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## Discussion
*In the **Discussion** section:*
The conformation of the C-terminus has previously been shown to play a major role in determining whether the 11-mer or the 12-mer oligomeric state assembles , and the new structure of the S72N mutant reinforces this. The relationship between subunit number and thermal stability has for the first time been probed, revealing the 12-mer configuration is more stable than the 11-mer counterpart.
## Comparison of Thermal Stability in Wild Type TRAP
*In the **Comparison of Thermal Stability in Wild Type TRAP** section:*
We compared the thermal stability of B. stearothermophilus, B. halodurans, and B. subtilis wild type TRAP. The relative thermal stabilities of the oligomers reflect the temperature environments of the respective species’ habitats, as might be expected. For example, at 50 µM L-tryptophan, the melting temperatures for the respective wild type TRAP are: B. stearothermophilus, 81.4°C; B. halodurans, 75.6°C; B. subtilis, 44.3°C. These melting temperatures reflect the respective species’ hyperthermophilic, thermophilic and mesophilic origins respectively. This relationship appears to hold across all of the L-tryptophan concentration range ( Figure 3 ).[](https://www.ncbi.nlm.nih.gov
/mesh/D014364)
Within these wild type TRAP oligomers, B. stearothermophilus and B. subtilis both have 11-mer TRAP, and B. halodurans has 12-mer TRAP. It is apparent therefore that the temperature environment in which the species have evolved is the overriding factor in determining the thermal stability of the wild type TRAP, as we have an 11-mer as both the most and least stable. However, we find that one effect of inducing a change of oligomeric state from 11-mer to 12-mer by mutations is increased thermal stability, implying that species-specific structural characteristics unrelated to subunit number are not the only determinant of thermal stability.
## C-terminal Truncation Induces 11- to 12-mer Transition and Increases Stability
*In the **C-terminal Truncation Induces 11- to 12-mer Transition and Increases Stability** section:*
Comparing the structure of the 12-mer B. halodurans TRAP with that of the 11-mer structures of B. subtilis TRAP and B. stearothermophilus TRAP, the importance of the C-terminus (residues 71 to 75) becomes apparent. In B. halodurans TRAP, the nature of the last four amino acids dictate that the C-terminal segment is excluded from the subunit-subunit interface, thereby generating favorable interactions at the outer surface of the oligomer. This is considered as the crucial factor in permitting the higher oligomeric state, which is related to the 11-mer state by a 2.7o rigid-body rotation of adjacent subunits . Hence what can be considered macroscopically as the result of the extraction of a thin “wedge” between subunits on the outer side of the ring, has the effect of decreasing the angle that a single subunit contributes towards the full 360°.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Taking our investigation a step further, we focused further studies at determining whether there is a general relationship between the oligomeric state of TRAP and its thermal stability, hence exploring factors that are not species-specific. The scenario found in the 12-mer B. halodurans TRAP, regarding the exclusion of the C-terminus, could logically be mimicked by truncating the 4 or 5 C-terminal residues, removing this segment from the subunit interface. This is observed in the previously determined structures of B. stearothermophilus E71stop TRAP and B. subtilis K71stop TRAP . With respect to thermal stability accompanying the 11-mer to 12-mer transition, we find that this truncation results in an increase in melting temperature, so that at 50 µM L-tryptophan the increase in melting temperature between B. stearothermophilus wild type and E71stop is 6°C, from 81.4°C to 87.4°C. The equivalent increase between B. subtilis wild type and K71stop is 18.7°C, 44.3°C to 63.0°C.
This hints at an increase in thermal stability in response to incorporation of an additional subunit into the ring. However, this difference in melting temperature could also be a result of the absence of the C-terminal residues themselves. For instance, we note that both B. subtilis and B. stearothermophilus wild type TRAP contain two flexible lysine residues at their C-termini that can contribute electrostatically to the overall stability of TRAP. To understand which of these two factors – C-terminus or oligomeric state – has the greater influence we compared thermal stability of wild type B. subtilis TRAP 11-mer with that of the B. subtilis S72N TRAP 12-mer, which still possesses the C-terminal segment but has it excluded from the subunit interface for steric reasons.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## S72N Mutation also Increases Subunit Number and Stability
*In the **S72N Mutation also Increases Subunit Number and Stability** section:*
Exclusion of the C-terminus from the subunit interface can be achieved by introducing steric strain at a key region towards the C-terminus that causes the segment to pivot out. In this respect, exclusion of the C-terminus in S72N is achieved in a manner more similar to how exclusion is achieved in B. halodurans TRAP. Residue 72 in particular (Ser 72 in B. subtilis TRAP) – which sits at the subunit-subunit interface in both wild type 11-mer structures studied – serves as a critical determinant in the conformation of the C-terminus. This is evident from the crystal structure presented in this study. From the structure of the S72N TRAP we see that substituting serine 72 for a bulkier amino acid, asparagine, does indeed introduce the necessary constraints to exclude the subsequent segment from the subunit interface.
The thermal stability of this mutant 12-mer TRAP, S72N, is higher than that of its wild type 11-mer counterpart. For example at 50 µM L-tryptophan the melting temperature of the wild type and mutant oligomers are 44.°C and 54.5°C respectively, yielding a stabilization of 10.2°C. This is significant, albeit it is less compared to the 18.7°C stabilization we see from C-terminal truncation (i.e. B. subtilis K71stop).[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
We conclude that by inducing an expansion in the oligomeric state from 11 to 12 subunits the thermal stability of the oligomer is increased. This increase is predominantly a result of the oligomeric state transition, and to a much lesser extent the influence of C-terminal or other specific residues. A potential, general relationship in TRAP oligomers therefore exists, in which melting temperature depends on subunit number. Further studies should be focused on evaluation of influence of the oligomeric state of TRAP on its activity in vivo.
## Binding Affinity Towards L-tryptophan
*In the **Binding Affinity Towards L-tryptophan** section:*
To explore whether the stabilization by L-tryptophan was directly related to the affinity of TRAP oligomers to the binding of L-tryptophan, dissociation constants were measured. Results showed no obvious correlation with thermal stability or subunit number ( Table 2 ). Although between B. subtilis wild type and S72N mutant TRAP, the 11-mer to 12-mer transition does coincide with a slight increase in affinity for L-tryptophan, these changes are only twofold, implying there are other more significant influences.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## Increasing L-tryptophan Concentration Increases Thermal Stability
*In the **Increasing L-tryptophan Concentration Increases Thermal Stability** section:*
For most TRAP oligomers studied we observe an increase in melting temperature in response to an increase in L-tryptophan concentration from 0 µM to 100 µM added. The trend does not appear in B. subtilis wild type TRAP however, over the L-tryptophan concentration range studied. Most concentrations in the monitored range represent many times the typical Kd value, hence L-tryptophan binding-sites in TRAP are expected to be saturated. This therefore suggests a degree of non-specific binding of L-tryptophan to TRAP as the principal cause of the stabilization, over a thermodynamic or kinetic influence involving the L-tryptophan binding sites. It is likely that non-specific binding, beyond the 1∶1 stoichiometry, would hence increase with increasing concentration of L-tryptophan.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## Conclusions
*In the **Conclusions** section:*
This study demonstrates how the oligomeric state of a circular protein can be changed by introducing a point mutation at a subunit interface. Importantly, we demonstrate that a change in subunit number changes the stability of a circular assembly in vitro. In the case of TRAP, used in our studies, an increase in subunit number increases the stability for the oligomeric assembly. Predictably, this stability is dependent on solvent conditions, specifically L-tryptophan concentration. We conclude that a similarly subtle approach could be used during protein engineering of other ring-like structures, for stabilizing a particular oligomeric state thereby modifying dimensions, symmetry and stability.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## Materials and Methods
*In the **Materials and Methods** section:*
## Gene Cloning, Protein Purification, Crystallization and Data Collection
*In the **Gene Cloning, Protein Purification, Crystallization and Data Collection** section:*
The B. subtilis S72N mutant TRAP was generated using the QuikChange kit (Stratagene, US) and a pET9a plasmid containing the wild type gene. S72N TRAP proteins were produced and purified as described previously . Before crystallization, protein samples were transferred into solution containing 2 mM Tris (pH 8.5), 300 mM NaCl and purified by size-exclusion chromatography using Superdex 200 Column (GE healthcare, UK). TRAP samples were initially analysed by size-exclusion chromatography, employing a Superdex 200 10/300 column (GE healthcare, UK) for the preliminary oligomeric assembly analysis. A 200–300 µg TRAP sample was injected and eluted by a buffer containing 20 mM Tris-HCl, 30 mM NaCl, and 0.1 mM L-tryptophan (pH 8.5).
TRAP samples with an elution volume less than 14.5 ml were selected for further analysis by native mass spectrometry analysis (Figure S1) and X-ray crystallographic studies.[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
Crystallization was carried out at 18°C by hanging drop vapor diffusion. For crystallization, B. subtilis S72N TRAP was transferred into solution containing 10 mM triethanolamine (pH 8.0), 100 mM NaCl, 15 mM L-tryptophan and concentrated to 33.5 mg/ml. The reservoir contained 100 mM Bis-Tris propane (pH 8.5), 200 mM KSCN and 13% PEG 3350 (v/v). Protein crystals were frozen using solutions containing all the crystallization ingredients with addition of 20% glycerol (v/v). The X-ray data were collected at 120 K at I04 beamline at the Diamond Light Source.
Data were processed using HKL2000 , ( Table 1 ).[](https://www.ncbi.nlm.nih.gov/mesh/C009546)
## Structure Determination and Refinement
*In the **Structure Determination and Refinement** section:*
All crystallographic calculations were carried out using the CCP4 program package . The structure was determined by molecular replacement using MOLREP with three adjacent subunits of B. stearothermophilus TRAP as a search model. Refinement was performed by REFMAC and model rebuilding was done using COOT . Water molecules were added automatically with the program ARP/wARP and further corrected using maximum likelihood-weighted 2|Fo| - |Fc| and |Fo| - |Fc| electron density maps. Molecular contacts between adjacent monomers of TRAP were examined by CONTACT .
Figure 1 was generated using Bobscript , Figure 2A using UCSF Chimera , and Figure 2B using CCP4 mg .[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Dye-based Scanning Fluorimetry
*In the **Dye-based Scanning Fluorimetry** section:*
Dye-based scanning fluorimetry, or the Thermofluor method, was performed as previously described , . Protein samples were prepared by transferring to 50 mM Tris pH 8, 150 mM NaCl, 10 mM MgCl2, with variable L-tryptophan concentration, in 30 kDa cut-off Vivaspin concentrators, and diluting to 1 mg/ml, monitoring concentration by the Comassie-blue Bradford method. L-tryptophan concentration was varied from 0 to 0.1 mM (final well concentration). Samples were transferred to 96-well thin-walled clear PCR plates. Each well contained 30 µl of solution, consisting of 15 µl of the 1 mg/ml protein (yielding a final concentration of 0.5 mg/ml), 1/1000 diluted SYPRO Orange (Sigma-Aldrich) and 50 mM Tris pH 8 with a specified L-tryptophan concentration.
A Stratagene Mx3005P QPCR instrument was programmed to increase temperature in 1°C increments at 30 second intervals, from 25°C to 95°C with 470 nm excitation, measuring fluorescence at 570 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
## Measurement of L-tryptophan Dissociation Constants
*In the **Measurement of L-tryptophan Dissociation Constants** section:*
L-tryptophan binding to TRAP was measured using a fluorescence assay based on competition between L-tryptophan and 1-anilionaphthalene-8-sulfonic acid (ANS) as described previously . ANS fluoresces weakly in solution but fluoresces strongly at 460 nm with 372 nm excitation upon binding to TRAP. L-tryptophan has a greater affinity than ANS for TRAP, and displaces ANS thus reducing the fluorescence. The change in fluorescence, ΔF, was monitored using a LS-50B fluorometer (Perkin Elmer). A 100 µL sample of 15 µM TRAP was prepared in 60 µM ANS and 0.5 mM sodium phosphate at pH 8.0.
Readings of fluorescence intensity were taken from 0 to 2000 µM tryptophan. At each concentration of tryptophan, fluorescence intensity was collected at room temperature for one minute to allow the reaction to reach equilibrium. Fluorescence intensity was averaged for each concentration of L-tryptophan and then used to calculate the change in fluorescence (ΔF). The maximal change in fluorescence was set to 100% and values of ΔF were normalised accordingly. GraphPad Prizm 4 was used to plot % ΔF as a function of tryptophan concentration to obtain L-tryptophan dissociation constants and binding curves.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## Data Deposition
*In the **Data Deposition** section:*
Refined coordinates of the B. subtilis S72N TRAP and structure factors have been deposited with the Protein Data Bank under accession code 4B27.
## Supporting Information
*In the **Supporting Information** section:*
# References
*In the **References** section:*
|
# Introduction
Sex- and age-specific trends in antibiotic resistance patterns of Escherichia coli urinary isolates from outpatients
# Abstract
*In the **Abstract** section:*
Background
Urinary tract infections (UTIs) are one of the most common infections treated in ambulatory care settings, however the epidemiology differs by age and sex. The incidence of UTI is far greater in females than males, and infection in pediatric patients is more often due to anatomical abnormalities. The purpose of this research was to describe age- and sex-specific trends in antibiotic susceptibility to common urinary anti-infectives among urinary isolates of Escherichia coli from ambulatory primary care patients in a regional health maintenance organization.
Methods
Clinical microbiology data were collected for all urine cultures from patients with visits to primary care clinics in a regional health maintenance organization between 2005 and 2010. The first positive culture for E. coli tested for antibiotic susceptibilities per patient per year was included in the analysis dataset. The frequency of susceptibility to ampicillin, amoxicillin-clavulanate, ciprofloxacin, nitrofurantoin, and trimethoprim/sulfamethoxazole (TMP/SMX) was calculated for male and female patients. The Cochrane-Mantel-Haenzel test was used to test for differences in age-stratified susceptibility to each antibiotic between males and females.[](https://www.ncbi.nlm.nih.gov/mesh/D000667)
Results
A total of 43,493 E. coli isolates from 34,539 unique patients were identified for study inclusion. After stratifying by age, E. coli susceptibility to ampicillin, amoxicillin-clavulanate, ciprofloxacin, and nitrofurantoin differed significantly between males and females. However, the magnitude of the differences was less than 10% for all strata except amoxicillin-clavulanate susceptibility in E. coli isolated from males age 18–64 compared to females of the same age.[](https://www.ncbi.nlm.nih.gov/mesh/D000667)
Conclusions
We did not observe clinically meaningful differences in antibiotic susceptibility to common urinary anti-infectives among E. coli isolated from males versus females. These data suggest that male sex alone should not be used as an indication for empiric use of second-line broad-spectrum antibiotic agents for the treatment of UTIs.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Urinary tract infections (UTIs) are one of the most commonly treated bacterial infections and account for over 10 million ambulatory care visits annually in the United States. Antibiotic treatment is typically selected empirically, based on the patient clinical presentation, medical history and local patterns of antibiotic susceptibility. Because the incidence of UTIs is significantly greater among women, much of the research has focused on women. Thus there exists a paucity of research on UTIs in men. As such, guidelines on the diagnosis, treatment, and management of UTIs focus largely on infection in women.
The incidence, presentation, and course of infection for UTIs in men and women differ in large part due to anatomical differences. For the treatment of acute uncomplicated cystitis in women, trimethoprim/sulfamethoxazole (TMP/SMX), nitrofurantoin, fosfomycin, and pivmecillinam are recommended first-line empiric therapies. In men, because prostate involvement occurs in roughly 90% of cases, an empiric agent should be selected that achieves therapeutic concentrations in prostatic tissues (e.g., trimethoprim or ciprofloxacin).[](https://www.ncbi.nlm.nih.gov/mesh/D015662)
While historically it was believed that the causative organism in UTIs differed between men and women, more recent data has shown that for both sexes the primary causative pathogen is Escherichia coli, which accounts for 75-90% of UTIs. With the increases in antibiotic resistance among E. coli and other Enterobacteriaceae over the past several decades, surveillance data have become critical for appropriate empiric selection of antibiotic therapy. U.S. guidelines specify that TMP/SMX should be avoided for empiric treatment of uncomplicated acute cystitis or pyelonephritis in populations where non-susceptibility to this agent exceeds 20% in uropathogens.[](https://www.ncbi.nlm.nih.gov/mesh/D015662)
While some surveillance studies have identified significant differences in the frequency of susceptibility to common urinary anti-infectives between isolates collected from male and female patients, this has not been consistently observed. The objective of this study was to describe age- and sex-specific antibiotic susceptibility patterns for common urinary anti-infectives among E. coli urine isolates using six years of data from a large ambulatory primary care patient population.
## Methods
*In the **Methods** section:*
## Study design and patient population
*In the **Study design and patient population** section:*
We conducted a cross-sectional study of urinary E. coli isolates from outpatients of Kaiser Permanente Northwest (KPNW) primary care clinics. KPNW is a regional health maintenance organization that serves over 485,000 members in northwest Oregon and southwest Washington. Primary care clinics were defined as non-specialty care clinics within the Family Practice, Internal Medicine, or Pediatrics departments; all primary care clinics were included. Urine cultures positive for E. coli that were drawn from patients with visits in the primary care clinics between January 1, 2005 and December 31, 2010 were eligible for inclusion in the analysis. Cultures in which less than 10,000 colonies/mL were identified or 3 or more organisms were isolated were excluded.
The analysis dataset was then limited to the first isolate tested for antibiotic susceptibilities per patient and year to minimize potential bias resulting from repeat culturing.
## Data collection
*In the **Data collection** section:*
Data were electronically extracted from the virtual data warehouse maintained by the Kaiser Permanente Center for Health Research. Collected data included patient demographics, department in which the clinic visit occurred, and all clinical microbiology data for urine cultures. Approval for this study was obtained from the KPNW institutional review board.
## Data analysis
*In the **Data analysis** section:*
The frequency of susceptibility to ampicillin, amoxicillin-clavulanate, ciprofloxacin, nitrofurantoin, and trimethoprim/sulfamethoxazole (TMP/SMX) was calculated for E. coli isolates stratified by patient sex and year. The Cochrane-Armitage test for trend was used to identify changes in the frequency of susceptibility to each antibiotic over time independently among males and females. Patient age at the time of culture was categorized as less than 18 years, 18–64 years, and 65 years and older. The frequency of susceptibility to each antibiotic was also compared across age categories among males and females using the Cochran-Mantel-Haenszel test. An alpha level less than or equal to 0.05 was the statistical significance level for all analyses and data were analyzed with SAS (version 9.2, SAS Corporation).
Clinically significant differences were defined as differences of 10% or greater.[](https://www.ncbi.nlm.nih.gov/mesh/D000667)
## Results
*In the **Results** section:*
## Description of study sample
*In the **Description of study sample** section:*
During the study time frame 190,396 urine cultures were performed at primary care clinic visits; 70,180 were positive for E. coli and 57,550 of those were tested for antibiotic susceptibilities. After restricting the data to the first isolate per patient per year, 43,493 isolates from 34,539 unique patients remained in the final analysis data set. Table 1 describes the demographics of these patients. Of the included isolates, 2,520 (5.8%) were from male patients.
Patient characteristics based on first E. coli isolate from urine specimen in KPNW outpatients 2005-2010a
## Antibiotic susceptibility of urinary Escherichia coli isolates by patient sex
*In the **Antibiotic susceptibility of urinary Escherichia coli isolates by patient sex** section:*
Overall, 66.0% (1,664/2,520) of E. coli isolated from males and 66.3% (27,175/40,971) from females were susceptible to ampicillin. Amoxicillin-clavulanate susceptibility was 56.9% (484/850) among males and 67.3% (9,291/13,798) among females; however, it should be noted that testing for susceptibility to amoxicillin-clavulanate was routinely performed and reported only for ampicillin non-susceptible E. coli. Ciprofloxacin susceptibility for the E. coli isolates was 93.2% (2,292/2,458) among males and 95.9% (37,900/39,524) among females, and nitrofurantoin susceptibility was 96.4% (2,430/2,520) among males and 97.6% (39,979/40,968) among females. For TMP/SMX, 86.3% (2,176/2,520) of E. coli isolated from males were susceptible compared to 84.7% (34,674/40,958) among females.[](https://www.ncbi.nlm.nih.gov/mesh/D000667)
## Trends in antibiotic susceptibility of urinary Escherichia coli isolates over time
*In the **Trends in antibiotic susceptibility of urinary Escherichia coli isolates over time** section:*
Figure 1 presents the susceptibility of E. coli to each antibiotic by sex and year. Susceptibility to amoxicillin-clavulanate and TMP/SMX decreased significantly over time among both males and females. Ciprofloxacin susceptibility also decreased significantly over time among females, but not males. Table 2 presents the age stratified antibiotic susceptibilities for E. coli isolated from males and females. The age-specific susceptibilities differed significantly between males and females for all antibiotics except TMP/SMX.[](https://www.ncbi.nlm.nih.gov/mesh/D019980)
Susceptibility to urinary anti-Infectives by sex and year for E. coli cultured from urine. For all figures, males shown as blue solid line, females as red dashed line. a, ampicillin susceptibility (test for trend: males, p=0.058; females, p=0.6256); b, amoxicillin clavulanate susceptibility (test for trend: males, p=0.0006; females, p<0.0001); c, ciprofloxacin susceptibility (test for trend: males, p=0.1445; females, p<0.0001); d, nitrofurantoin susceptibility (test for trend: males, p=0.4967; females, p=0.5508); e, trimethoprim/sulfamethoxazole susceptibility (test for trend: males, p=0.0004; females, p<0.0001).[](https://www.ncbi.nlm.nih.gov/mesh/D000667)
Susceptibility to urinary anti-infectives by sex and age for E. coli cultured from urine*
## Discussion
*In the **Discussion** section:*
In this study, we observed statistically significant differences between males and females in the age-specific susceptibilities of E. coli to ampicillin, amoxicillin-clavulanate, ciprofloxacin, and nitrofurantoin. Urinary E. coli isolates from male patients tended to exhibit increased antibiotic resistance than isolates from female patients. Despite the statistical significance of time trends and differences in age-specific susceptibilities, the magnitude of these differences was generally less than 5% and thus may not represent clinically meaningful differences. The exception was susceptibility to amoxicillin-clavulanate, where susceptibility was roughly 10% lower in males age 18 to 64 years than females in the same age group. Yet these differences should be interpreted cautiously because susceptibility to amoxicillin-clavulanate was only provided for ampicillin non-susceptible E. coli isolates in the current analysis.
If all ampicillin susceptible isolates are assumed to be amoxicillin-clavulanate susceptible, then the magnitude of the differences between males and females would be less than 5% in all age categories (difference not statistically significant).[](https://www.ncbi.nlm.nih.gov/mesh/D000667)
While few other studies have explored the differences in antibiotic susceptibility of uropathogens isolated from male and female ambulatory patients, our findings are consistent with the trends observed in the literature. A recent 10-year study of community UTI in Portuguese patients also identified differences in antibiotic susceptibility by patient sex. The authors reported that urinary isolates of E. coli were significantly more resistant to fluoroquinolones, penicillins, nitrofurantoin, and first and second generation cephalosporins among men compared to women. Another study focused on pediatric patients, also identified significantly higher resistance to TMP/SMX and ciprofloxacin in male versus female patients. The NAUTICA surveillance study of outpatient UTIs reported greater antibiotic resistance to ciprofloxacin, levofloxacin, and TMP/SMX among all urinary isolates from U.S. and Canadian male patients.
In the CANWARD study, antibiotic susceptibility among all E. coli isolates (not limited to urine isolates) collected from Canadian tertiary medical centers were compared and resistance was also observed to be significantly higher to ciprofloxacin, levofloxacin, and TMP/SMX in isolates collected from male patients versus female patients.[](https://www.ncbi.nlm.nih.gov/mesh/D024841)
The prevalence and susceptibilities of antibiotic-resistant bacteria varies widely by geographic region, thus these data may not be generalizable to all regions. Local surveillance data of antibiotic susceptibilities would be needed to validate these findings in different patient populations. It should be noted that in our study resistance to TMP/SMX did not exceed 20%, as it does in many other regions of the U.S., thus in this patient setting TMP/SMX may be a viable empiric treatment option in both males and female patients. Additionally, antibiotic susceptibilities may differ by other patient or infection characteristics (e.g., type of UTI). These data were not available for analysis in this study and further investigation in this area is needed.
Also, because clinical signs and symptoms of infection were not available and because not all suspected infections are cultured, the use of all urinary microbiology from this population cultures may not represent all true infections.[](https://www.ncbi.nlm.nih.gov/mesh/D015662)
Currently there exists insufficient data to inform the development of evidence-based guidelines for the treatment of community UTIs in men. UTIs in adult males are considered complicated infections according to most clinical definitions. Treatment recommendations vary from longer durations of therapy to the selection of broad-spectrum agents such as fluoroquinolones. These recommendations are driven by the high proportion of UTIs in men with prostate involvement and concerns surrounding antibiotic resistance to TMP/SMX. Previous research among Swiss outpatients has demonstrated the selection of fluoroquinolones versus TMP/SMX for the treatment of UTI is often influenced by nonclinical factors.
In this study, TMP/SMX, a recommended first-line agent for UTIs in both children and adults, the differences in susceptibility between E. coli isolated from males and females were neither statistically nor clinically significant and resistance rates are below 20%. Consequently, in this population, there is no evidence that male sex alone should be an indication for empiric selection of a second-line broad-spectrum antibiotic agent for the treatment of UTI. While more research is needed on treatment effectiveness of different regimens used to treat community UTIs in men, population-specific antibiotic susceptibility data are a necessary component in the empiric antibiotic selection process.[](https://www.ncbi.nlm.nih.gov/mesh/D024841)
## Conclusions
*In the **Conclusions** section:*
In the era of increasing antibiotic resistance, prudent use of antibiotics is critical to prolong the clinical effectiveness of existing agents. Excessive use of broad-spectrum agents, such as fluoroquinolones, increases the evolutionary selective pressures that drive the increasing prevalence of resistance. These data suggest that first-line urinary anti-infectives such as TMP/SMX may be effective agents for treating UTIs in men. While more data are needed, clinicians should use local surveillance data to guide the prudent, empiric selection of antibiotic therapy for UTIs.[](https://www.ncbi.nlm.nih.gov/mesh/D024841)
## Abbreviations
*In the **Abbreviations** section:*
UTI: Urinary Tract Infection; KPNW: Kaiser Permanente Northwest; TMP/SMX: trimethoprim/sulfamethoxazole[](https://www.ncbi.nlm.nih.gov/mesh/D015662)
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interests.
## Authors’ contributions
*In the **Authors’ contributions** section:*
JCM participated in the conception and design of the study; the acquisition, statistical analysis, and interpretation of the data; and the drafting of the manuscript. JCM, MRE, DTB, and DHS assisted with the interpretation of the data and participated in the critical revision of the manuscript. All authors read and approved the final manuscript.
## Pre-publication history
*In the **Pre-publication history** section:*
The pre-publication history for this paper can be accessed here:
http://www.biomedcentral.com/1471-2296/14/25/prepub
|
# Introduction
An unexpected role of neuroligin-2 in regulating KCC2 and [GABA](https://www.ncbi.nlm.nih.gov/mesh/D005680) functional switch
# Abstract
*In the **Abstract** section:*
Background
GABAA receptors are ligand-gated Cl- channels, and the intracellular Cl- concentration governs whether GABA function is excitatory or inhibitory. During early br[ain](https://www.ncbi.nlm.nih.gov/mesh/D002712) development, GABA undergoes fu[ncti](https://www.ncbi.nlm.nih.gov/mesh/D005680)onal switch from excitation to inhibition: GABA depolarizes immature ne[uron](https://www.ncbi.nlm.nih.gov/mesh/D005680)s but hyperpolarizes mature neurons due to a developmental d[ecre](https://www.ncbi.nlm.nih.gov/mesh/D005680)ase of intracellular Cl- concentration. This GABA functional switch is mainly mediated by the up-regulation of KC[C2,](https://www.ncbi.nlm.nih.gov/mesh/D002712) a potassium-chloride[ cot](https://www.ncbi.nlm.nih.gov/mesh/D005680)ransporter that pumps Cl- outside neurons. However, the upstream factor that regulates KCC2 expression is unclear.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
Results
We report here that KCC2 is unexpectedly regulated by neuroligin-2 (NL2), a cell adhesion molecule specifically localized at GABAergic synapses. The expression of NL2 precedes that of KCC2 in early postnatal development. Upon knockdown of NL2, the expression level of KCC2 is significantly decreased, and GABA functional switch is significantly delayed during early development. Overexpression of shRNA-proof NL2 rescues both KCC2 reduction and delayed GABA functional switch induced by NL2 shRNAs. Moreover, NL2 appears to be required to maintain GABA inhibitory function even in mature neurons, because knockdown NL2 reverses GABA action to excitatory. Gramicidin-perforated patch clamp recordings confirm that NL2 directly regulates the GABA equilibrium potential. We further demonstrate that knockdown of NL2 decreases dendritic spines through down-regulating KCC2.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Conclusions
Our data suggest that in addition to its conventional role as a cell adhesion molecule to regulate GABAergic synaptogenesis, NL2 also regulates KCC2 to modulate GABA functional switch and even glutamatergic synapses. Therefore, NL2 may serve as a master regulator in balancing excitation and inhibition in the brain.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Background
*In the **Background** section:*
A delicate balance between excitatory and inhibitory neurotransmission is critical for brain functions. GABA is the principle inhibitory neurotransmitter in the adult brain, and dysfunction of GABAergic transmission may contribute to the onset of many neurological disorders including epilepsy, schizophrenia, autism spectrum disorders, and major depressive disorders. Because GABAA receptors are ligand-gated Cl- channels, the efficacy of GABAergic transmission is modulated by changes in intracellular Cl- concentration. Two chloride transporters, NKCC1 and KCC2, import and export Cl- across neuronal membranes correspondingly. In early neural development, the expression level of NKCC1 is initially high and gradually down-regulated, while KCC2 expression is up-regulated.
Such developmental changes of NKCC1 and KCC2 result in a shift of intracellular Cl- concentration from high to low and a corresponding shift of GABAA receptor reversal potential from depolarizing to hyperpolarizing. Therefore, GABA is not a simple inhibitory neurotransmitter, but rather undergoes a functional switch from excitation to inhibition during brain development. GABA-mediated excitation regulates neural differentiation, migration, and synaptogenesis. Our previous work found that GABAergic synaptogenesis precedes glutamatergic synaptogenesis due to the earlier expression of GABAA receptors than that of glutamate receptors in embryonic neurons. So far, it is unclear whether there is any coordination between GABA functional switch and GABAergic synapse formation.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Neuroligins (NLs) are a family of postsynaptic cell adhesion molecules that interact with presynaptic neurexins. NL1 and NL2 are selectively localized at glutamatergic and GABAergic synapses, and manipulations of NL1 and NL2 expression level have been shown to regulate glutamatergic and GABAergic synaptogenesis, respectively. Transgenic mice overexpressing NL2 showed enhanced GABAergic transmission, whereas NL2 knockout mice showed decreased GABAergic transmission. We have previously shown that NL2 is a critical cell adhesion molecule capable of inducing functional GABAergic synapses in neuron-HEK cell hetero-cocultures. Our recent work further identified a loss-of-function mutation of NL2 in schizophrenia patients, suggesting an indispensable role of NL2 in regulating GABAergic functions.
Here, we uncover a novel function of NL2 in regulating KCC2 expression and GABA functional switch from excitation to inhibition during neurodevelopment. Knockdown of NL2 also induces an unexpected reduction in glutamatergic events and dendritic spines. Therefore, in addition to its role as a cell adhesion molecule at GABAergic synapses, NL2 may serve as a master regulator of the delicate balance between glutamatergic and GABAergic functions in neural networks.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Results
*In the **Results** section:*
## Neuroligin-2 unexpectedly regulates KCC2 expression
*In the **Neuroligin-2 unexpectedly regulates KCC2 expression** section:*
Neuroligin-2 (NL2) is a cell adhesion molecule mainly localized at GABAergic synapses. We have previously demonstrated that overexpression of NL2 and GABAA receptors in HEK cells can induce fully functional GABAergic synapses when cocultured with neurons. Recently we also identified a mutant NL2 from human schizophrenia patient that is defective in promoting GABAergic synapse formation. During our continued study of NL2 in synapse formation and plasticity, we made an unexpected finding that NL2 regulates KCC2, a K+-Cl- cotransporter that is critical in controlling intracellular Cl- concentration and the polarity of GABA action. We investigated the function of NL2 by using small-hairpin RNA (shRNA) mediated knockdown in cultured mouse cortical neurons.
Two previously well characterized shRNAs were used to knockdown NL2 expression level: one is a chained shRNA targeting all NL1-3 (NLmiR) and the other is a NL2-specific shRNA (NL2shRNA). When HA-tagged wild type (WT) NL2 was coexpressed with NL2shRNA or NLmiR, the expression level of HA-NL2 was reduced by more than 80% compared to the coexpression with mCherry alone as a control (Figure 1A, quantified in Figure 1C left columns). In contrast, the expression of a shRNA-proof mutant version of NL2 (HA-NL2*) was not affected by NL2shRNA or NLmiR (Figure 1B, quantified in Figure 1C right columns). Therefore, both NLmiR and NL2shRNA can efficiently knockdown NL2, and the mutant NL2* was suitable for rescue experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
Knockdown of neuroligin-2 decreases KCC2 expression. A, Efficient knockdown of NL2 by NL2shRNA or NLmiR in mouse cortical neurons. Neurons were co-transfected with mCherry at 2 DIV and HA-NL2 expression was assayed by HA-immunostaining at 8 DIV. B, shRNA-proof mutant HA-NL2* was resistant to the knockdown of NL2shRNA or NLmiR. Scale bar, 10 μm. C, Bar graphs showing quantified somatic HA-immunostaining intensity (normalized by mCherry controls). ***p < 0.001 (one-way ANOVA). D, Representative images showing reduced KCC2 immunostaining (green) in NLmiR-transfected neurons (arrowhead) compared to non-transfected neurons (arrow, 12 DIV). Scale bar, 20 μm. E, NL2*, but not NL1* rescued the KCC2 level when coexpressed with NLmiR. Inlets showed HA-immunostaining (gray) to confirm the expression of HA-NL1* and HA-NL2*. Scale bar, 20 μm. F, NL2shRNA also decreased KCC2 expression level. Scale bar, 20 μm. G, Bar graphs showing quantified somatic KCC2 immunofluorescence intensity in non-transfected (nonTF) neurons (173 ± 4 a.u.) and neurons transfected with mCherry (172 ± 10), NLmiR (73 ± 8), NLmiR + NL1* (93 ± 9), NLmiR + NL2* (120 ± 12), or NL2shRNA (80 ± 11). a.u., arbitrary unit. **p < 0.01, ***p < 0.001, n.s., not significant, one-way ANOVA. H, NL2shRNA specifically knock down NL2 but not KCC2 in HEK 293T cells, which were co-transfected with NL2 + GFP, NL2 + NL2shRNA, KCC2 + GFP, or KCC2 + NL2shRNA. Total protein lysate was analyzed by immunoblot. Actin was used as loading control. I, NKCC1 immunostaining signal (red) was not altered in NL2shRNA-transfected neurons (green, arrowhead, 9 DIV). Scale bar, 20 μm. Bar graphs show the quantified somatic NKCC1 signal intensity (p > 0.7, unpaired Student’s t-test).
When we transfected mouse cortical neurons with NLmiR, we observed a significant reduction of the KCC2 expression level compared to that of non-transfected (nonTF) or mCherry-transfected control neurons (Figure 1D, quantified in Figure 1G). This is unexpected because no previous study reported any connection between NL2 and KCC2. To find out the relative contribution of NL1 versus NL2 to the reduced KCC2 expression, shRNA-proof HA-NL1* or HA-NL2* was coexpressed with NLmiR to test which one can rescue the KCC2 expression. HA-immunostaining confirmed the expression of shRNA-proof NL1* and NL2* in the presence of NLmiR (Figure 1E inlets). Coexpression of HA-NL2*, but not HA-NL1*, with NLmiR significantly rescued the KCC2 expression level (Figure 1E, quantified in Figure 1G), suggesting that NL2 is a potential regulator of KCC2.
This was confirmed by the detection of a similar reduction of KCC2 level after knocking down NL2 specifically with NL2shRNA (Figure 1F, quantified in Figure 1G). One concern regarding shRNAs is whether they might have off-target effect directly on KCC2. To exclude this possibility, we co-transfected NL2shRNA together with KCC2 in HEK 293T cells to examine whether KCC2 expression might be altered. Immunoblot demonstrated that while WT NL2 expression was significantly reduced by NL2shRNA (Figure 1H, top panel), KCC2 expression was not affected by NL2shRNA (Figure 1H, bottom panel). We next examined the effect of NL2shRNA on NKCC1, which imports Cl- and counteracts the action of KCC2.
Immunostaining with antibody specific for NKCC1 showed that NL2 knockdown did not change NKCC1 expression level (Figure 1I). Together, our data demonstrated a novel function of NL2 in regulating KCC2, both of which were found previously playing critical roles in GABA function but not yet linked together.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
## Neuroligin-2 regulates GABA excitation-inhibition switch
*In the **Neuroligin-2 regulates GABA excitation-inhibition switch** section:*
KCC2 is a key player in controlling intracellular Cl- concentration and driving GABA functional switch from excitation to inhibition during early development. Therefore, if KCC2 is regulated by NL2, we predict that GABA functional switch might be affected accordingly. To test this hypothesis, cortical neurons were co-transfected at 2 DIV by mCherry with NLmiR, and Fura-2 Ca2+ ratio imaging was employed to monitor GABA-evoked Ca2+ influx to determine whether GABA action is excitatory or inhibitory. When neurons were analyzed 6 days later (at 8 DIV), we found that GABA (100 μM) application evoked small Ca2+ responses in less than 50% of non-transfected control neurons but large Ca2+ responses in over 90% of NLmiR-transfected neurons (Figure 2A-B). As a control, high potassium (90 mM) stimulation induced robust Ca2+ responses in both non-transfected and NLmiR-transfected neurons (Figure 2B).
GABA-evoked Ca2+ increase was completely blocked by GABAA receptor antagonist bicuculline (BIC, 20 μM) (Figure 2C), suggesting that Ca2+ influx was mediated by GABAA receptor activation.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
Knockdown of neuroligin-2 abolishes GABA functional switch in cortical neurons. A, Representative images showing 8 DIV NLmiR-transfected neurons loaded with Fura-2. Somata of both nonTF and transfected neurons were selected (Fura-2 panel, white circles) to measure 340/380 ratio signal. Scale bar, 40 μm. B, Sample traces showing somatic Ca2+ responses after stimulation with 100 μM GABA and 90 mM KCl in nonTF neurons (gray, 8 DIV) and NLmiR-transfected neurons (black). C, Averaged sample traces showing that BIC (20 μM) blocked GABA-evoked Ca2+ responses in both nonTF (gray, n = 6) and NLmiR-transfected neurons (black, n = 17). D, The time courses of GABA functional switch in nonTF neurons and neurons transfected with mCherry, NLmiR, NLmiR + NL1*, NLmiR + NL2*, or NL2shRNA (3–4 independent cultures; n = 2291, 272, 433, 350, 322, 152 neurons, respectively). E, Bar graphs showing the amplitude of GABA-evoked Ca2+ increases in neurons transfected with mCherry, NLmiR, NLmiR + NL1*, NLmiR + NL2*, or NL2shRNA at 4 and 12 DIV. There was no difference between all groups at 4 DIV (p > 0.6). ***p < 0.001, one-way ANOVA. F, Percentage of mature neurons showing GABA-evoked Ca2+ increase after NLmiR or NL2shRNA transfection. Arrows in D and F indicate the time of transfection (TF).[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Using Ca2+ imaging approach, we further delineated the time course of GABA functional switch by monitoring the gradual decrease of GABA-evoked Ca2+ responses in developing neurons. We found that mouse cortical neurons typically complete their GABA functional switch around two weeks in dissociated cultures (Figure 2D), similar to previous reports. Quantitatively, for control neurons transfected with mCherry, GABA evoked Ca2+ responses in 80 ± 5% (n = 96) neurons at 4 DIV, but only 8 ± 3% (n = 61) at 12 DIV (Figure 2D), suggesting that the majority of neurons have finished the GABA excitation-inhibition transition by 12 DIV. However, after NL knockdown, even at 12 DIV, GABA still evoked Ca2+ responses in more than 80% of transfected neurons (Figure 2D; NLmiR, 86 ± 5%, n = 86; NL2shRNA, 86 ± 8%, n = 41; p < 0.001 for transfection of mCherry vs. NLmiR or NL2shRNA, two-way ANOVA). Importantly, coexpression of NLmiR with NL2*, but not NL1*, promoted GABA functional switch by 12 DIV (Figure 2D; NLmiR vs. NLmiR+NL2*, p < 0.001, one-way ANOVA at 12 DIV), suggesting that NL2 may regulate GABA excitation-inhibition switch.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Besides quantifying the percentage of neurons responding to GABA, we also compared the amplitude of GABA-evoked Ca2+ responses in individual neurons. At 4 DIV, most neurons in all groups showed significant GABA-evoked Ca2+ responses, suggesting an excitatory action of GABA (Figure 2E). At 12 DIV, while control neurons transfected with mCherry showed very small Ca2+ responses, neurons transfected with NLmiR or NL2shRNA still showed large GABA-evoked Ca2+ responses with the amplitudes similar to those at 4 DIV (Figure 2E), suggesting no GABA functional switch occurred after knockdown of NL2. Coexpression of NLmiR with NL2*, but not NL1*, resulted in significantly smaller GABA-evoked Ca2+ responses at 12 DIV (Figure 2E), suggesting that NL2 may restore GABA functional switch. Collectively, our Ca2+ imaging data demonstrate that NL2 plays a critical role in regulating GABA functional switch during early development.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
We further investigated whether NL2 is required for maintaining GABAergic inhibition in mature neurons after the completion of GABA functional switch. To address this question, we transfected neurons at 12–14 DIV with NLmiR or NL2shRNA and analyzed GABA-evoked Ca2+ responses at 16 and 21 DIV. Mature neurons in control group rarely showed any GABA-evoked Ca2+ responses (Figure 2F; non-transfected, only 4 out 264 neurons; mCherry-transfected, 0/7), but more than 50% mature neurons transfected with NLmiR (n = 59) or NL2shRNA (n = 12) showed significant GABA-evoked Ca2+ responses (Figure 2F). Therefore, NL2 is not only required for GABA functional switch in immature neurons, but also required for the maintenance of GABA inhibition in mature neurons.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Neuroligin-2 regulates GABA equilibrium potential
*In the **Neuroligin-2 regulates GABA equilibrium potential** section:*
The observed KCC2 reduction and large GABA-evoked Ca2+ responses after NL2 knockdown suggest an excitatory action of GABA due to depolarized GABAA receptor reversal potential (EGABA). To directly examine EGABA, we performed gramicidin-perforated patch clamp recordings to keep intracellular Cl- intact. In control neurons at 10–13 DIV, GABA application (40 μM, 50 ms) typically evoked small depolarizing or hyperpolarizing membrane potential changes (Figure 3A, top traces). In contrast, in neurons transfected with NL2shRNA, GABA reliably evoked action potentials on top of large depolarizing responses (Figure 3A, bottom traces), confirming that GABA function remains excitatory after NL2 knockdown. Changing holding membrane potentials under voltage-clamp condition revealed a significant depolarizing shift in EGABA after NL2 knockdown (Figure 3B).
Quantitatively, knockdown of NL2 alone resulted in a depolarizing shift of 16 mV in EGABA and knockdown of NL1-3 induced a shift of 22 mV (Figure 3C-D; EGABA: Control, -56 ± 3 mV; NL2shRNA, -40 ± 2 mV; NLmiR, -34 ± 2 mV). Interestingly, overexpression of NL2 caused an opposite change: a significant hyperpolarizing shift of 12 mV in EGABA (Figure 3C-D; NL2, -68 ± 3 mV). On the other hand, the resting membrane potential was not significantly altered by NL2 manipulations (Control, -60 ± 2 mV; NLmiR, -58 ± 4 mV; NL2shRNA, -63 ± 2 mV; NL2, -67 ± 2 mV; p > 0.08). These results suggest that NL2 plays an active role in controlling the functional polarity of GABA action.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Neuroligin-2 regulates GABA equilibrium potential. A, Representative current clamp recordings showing differential changes in membrane potentials evoked by GABA in 10–13 DIV control neurons (gray) and NL2shRNA-transfected neurons (black). Note that GABA evoked action potentials in most NL2shRNA-transfected neurons (4 out of 5 neurons) but rarely in control neurons (1/6). B, Representative voltage-clamp recordings showing GABA-evoked currents at different holding potentials in the presence of TTX (0.5 μM) and DNQX (10 μM). C, I-V plot of the mean GABA-evoked peak currents in control neurons and neurons transfected with NLmiR, NL2shRNA, or NL2. Note the opposite shift in EGABA between NL2 overexpression and NL2 knockdown groups. D, Bar graphs showing mean EGABA of control neurons and neurons transfected with NLmiR, NL2shRNA, or NL2. Dashed line indicates the mean EGABA of control neurons. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Regulation of KCC2 by neuroligin-2 is independent of GABAA receptor activation and neuronal activity
*In the **Regulation of KCC2 by neuroligin-2 is independent of GABAA receptor activation and neuronal activity** section:*
Since NL2 regulates GABAergic synaptogenesis, we wondered whether the regulation of KCC2 by NL2 might be mediated by an alteration in GABAA receptor activation or neuronal activity. To test this idea, cortical neurons were transfected with NL2shRNA at 2 DIV and chronically treated for 9 days with BIC (20 μM) to block GABAA receptors or with tetrodotoxin (TTX, 1 μM) to block action potential firing. Blocking the activation of GABAA receptors appeared to have no effect on the KCC2 level because the neurons transfected with NL2shRNA still showed a significant reduction of KCC2 compared to adjacent non-transfected controls (Figure 4A middle panel, quantified in Figure 4B). Similarly, inhibiting action potential firing also had no effect on the reduction of KCC2 induced by NL2 knockdown (Figure 4A right panel, quantified in Figure 4B). Therefore, NL2 regulation of KCC2 is not mediated by GABAA receptors or neuronal activity.[](https://www.ncbi.nlm.nih.gov/mesh/D001640)
Regulation of KCC2 by NL2 is independent of GABAA receptor function or neuronal activity. A, Representative images showing reduced KCC2 immunostaining (red) in neurons transfected with NL2shRNA (green, arrowheads, 11 DIV), regardless whether treated with BIC or TTX. DMSO (0.1%, Control), BIC (20 μM) or TTX (1 μM) was added into culture medium after transfection at 2 DIV and replenished every two days. Scale bar, 20 μm. B, Bar graphs showing somatic KCC2 immunofluorescence intensity in neurons with (+) and without (-) NL2shRNA transfection (n = 11–12 neurons per group). Cultures were treated with DMSO (Control), BIC or TTX after transfection. ***p < 0.001, unpaired Student’s t-test. C, Representative images showing MAP2 (green) and KCC2 (red) immunostaining at 4 and 12 DIV in neurons treated with DMSO, BIC or TTX, starting at 2 DIV. Scale bar, 20 μm. D, Bar graphs showing developmental increase of the somatic KCC2 immunofluorescence intensity (n = 12 neurons per group). a.u., arbitrary unit. E, Representative immunoblot showing NL2 expression in 4 and 12 DIV neurons treated with DMSO, BIC or TTX, starting at 2 DIV. Actin was used as loading control. F, Quantification of NL2 expression level as measured by immunoblot (n = 3 independent cultures). NL2 expression was normalized to the expression level of 4 DIV control. G, Calcium imaging showing the time courses of GABA functional switch in control neurons and neurons treated with BIC or TTX (3 independent cultures; n = 1141, 717, 737 neurons, respectively).[](https://www.ncbi.nlm.nih.gov/mesh/D001640)
We then investigated whether developmental expression of KCC2 and NL2 rely upon GABAA receptor activation or neuronal activity. As expected, KCC2 expression showed a significant increase from 4 to 12 DIV (Figure 4C left panel). After chronic treatment with BIC or TTX, KCC2 level was similarly increased compared to the control (Figure 4C middle and right panels, quantified in Figure 4D; n =12 neurons per group; p < 0.001 for developmental increase; p > 0.5 for drug treatment, Two-way ANOVA). Immunoblot analysis also found that NL2 expression level increased significantly from 4 to 12 DIV and were not affected by BIC or TTX treatment (Figure 4E-F; n = 3; p < 0.05 for developmental increase; p > 0.3 for drug treatment, Two-way ANOVA). Functionally, Ca2+ imaging experiments further showed comparable time courses of GABA functional switch between control and BIC- or TTX-treated neurons (Figure 4G; p > 0.7 for drug treatment, Two-way ANOVA).
Together, our data suggest that the developmental up-regulation of NL2 and KCC2 as well as GABA functional switch are likely regulated by cell-intrinsic mechanisms, independent of GABAA receptor activation or neuronal activity.[](https://www.ncbi.nlm.nih.gov/mesh/D001640)
## Neuroligin-2 expression precedes KCC2 in vivo
*In the **Neuroligin-2 expression precedes KCC2 in vivo** section:*
We reasoned that if NL2 regulates KCC2, the onset of NL2 expression should precede that of KCC2. This was indeed what occurred in the mouse brain in vivo when we analyzed the time course of the expression of both NL2 and KCC2 from postnatal day 1 to day 20 (Figure 5A-B). In neonatal mouse brain (P1-P4), NL2 was already expressed in a significant amount whereas the expression of KCC2 was minimal, consistent with a delayed KCC2 expression that correlates with GABA functional switch. Quantitatively, NL2 expression at P4 reached about 50% of P20 level, whereas KCC2 did not reach 50% of P20 expression level even at P11 (Figure 5B). Therefore, the in vivo sequential expression of NL2 and KCC2 makes it possible for NL2 to regulate KCC2.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Neuroligin-2 expression precedes that of KCC2. A, Representative immunoblot showing the time-course of NL2 and KCC2 expression in mouse cortex at postnatal day 1 (P1), P4, P8, P11, P15, and P20. Actin was used as loading control. B, Quantification of the NL2 and KCC2 expression level as measured by immunoblot (n = 3). Protein expression was normalized to the expression level at P20. C, Representative images showing the expression of HA-NL2 (green) and KCC2 (red) in mouse cortical neurons (14 DIV) transfected with HA-NL2. Empty arrowhead indicates KCC2 puncta without NL2 colocalization; Filled arrowhead indicates NL2 puncta without KCC2 colocalization; Arrow indicates colocalized NL2 and KCC2 puncta. Scale bar, 10 μm. D, Representative images showing the expression pattern of NL2 (green), KCC2 (red), and NL2 + KCC2 in HEK cells. Scale bar, 10 μm. E, Lack of co-immunoprecipitation revealed by HEK cells co-transfected with HA-NL2 and KCC2. Total protein lysate was immunoprecipitated with normal mouse IgG or mouse anti-HA antibody and immunoblotted with rabbit anti-KCC2 and anti-NL2 antibodies. F, No co-immunoprecipitation detected between NL2 and KCC2 in mouse brain cortical tissue. Total mouse cerebral cortex protein lysate was immunoprecipitated with normal rabbit IgG, rabbit anti-NL2, or rabbit anti-KCC2 antibodies. The precipitated proteins were then probed with rabbit anti-KCC2 or anti-NL2 antibodies to test whether co-immunoprecipitation occurred.
We next investigated whether NL2 directly interacts with KCC2. Because both available NL2 (129203, Synaptic Systems) and KCC2 (07–432, Millipore) antibodies are raised in rabbit, it is not feasible to detect endogenous NL2 and KCC2 simultaneously. Instead, we overexpressed HA-NL2 in cortical neurons and found partial co-localization between exogenous HA-NL2 and endogenous KCC2 (Figure 5C). However, because HA-NL2 was overexpressed in neurons, it raised a concern whether HA immunostaining truly represents the endogenous NL2 localization. We therefore further investigated possible interaction between NL2 and KCC2 in HEK 293T cells (Figure 5D).
As expected, expression of NL2 alone showed mainly membrane localization, and expression of KCC2 alone showed both membrane and intracellular localization. Coexpression of NL2 and KCC2 showed comparable subcellular localization to that of NL2 or KCC2 expression alone (Figure 5D). Next, we employed co-immunoprecipitation to examine whether NL2 and KCC2 interact with each other. Protein lysate prepared from HEK cells coexpressing HA-NL2 and KCC2 was immuno-precipitated with HA antibody. The HA-NL2 was clearly precipitated but KCC2 was not co-immunoprecipitated (Figure 5E).
We further performed co-immunoprecipitation experiment using mouse cerebral cortex protein lysate and obtained similar result to HEK cells that no co-immunoprecipitation could be detected between NL2 and KCC2, regardless which protein was immuno-precipitated first (Figure 5F). These results suggest that NL2 and KCC2 may not directly interact, or their interaction is too weak to be detected with current method.
## Neuroligin-2 regulates glutamatergic transmission and dendritic spines through KCC2
*In the **Neuroligin-2 regulates glutamatergic transmission and dendritic spines through KCC2** section:*
Previous study reported that knockdown of NL2 decreased not only GABAergic but also glutamatergic synapse numbers, but the underlying mechanism was not well understood. We transfected cortical neurons at 2 DIV with NL2shRNA and employed patch clamp recordings to analyze synaptic events at 10–12 DIV. As expected, the miniature inhibitory postsynaptic currents (mIPSCs) were largely abolished in neurons transfected with NL2shRNA (Figure 6A and C; Control, 0.67 ± 0.25 Hz, n = 10; NL2shRNA, 0.002 ± 0.002 Hz, n = 10; p < 0.05, unpaired Student’s t-test). Interestingly, the frequency of miniature excitatory postsynaptic currents (mEPSCs) was also significantly decreased (Figure 6B-C; Control, 1.2 ± 0.3 Hz, n = 10; NL2shRNA, 0.01 ± 0.01 Hz, n = 10; p < 0.01), confirming that NL2 not only regulates GABAergic synapses, but also affects glutamatergic synapses. We initially wondered whether the NL2 effect on glutamatergic synapses might be due to any alterations of neuronal intrinsic excitability.
To test this idea, we recorded voltage-dependent sodium and potassium currents but found no difference between control and NL2shRNA-transfected neurons (Figure 6D-E). Quantitatively, the I-V curves of both Na+ and K+ currents showed no significant difference between control and NL2shRNA-transfected neurons (Figure 6F; INa+, p > 0.9; IK+, p > 0.2, Two-way ANOVA).[](https://www.ncbi.nlm.nih.gov/mesh/D012964)
Neuroligin-2 regulates both GABAergic and glutamatergic synapse formation. A-B, Sample traces showing mIPSCs (A) and mEPSCs (B) from 10–12 DIV control and NL2shRNA-transfected neurons. C, Bar graphs showing the frequency of mIPSCs (left Y-axis) and mEPSCs (right Y-axis). n = 10 neurons per group, *p < 0.05, **p < 0.01, unpaired Student’s t-test. D-E, Sample traces showing whole-cell sodium currents (D) and potassium currents (E) in response to depolarizing voltage steps. F, I-V plot of the mean amplitude of sodium and potassium currents in control (n = 7) and NL2shRNA-transfected neurons (n = 9).[](https://www.ncbi.nlm.nih.gov/mesh/D012964)
Considering the specific targeting of NL2 to GABAergic synapses, the reduction in mEPSC frequency after NL2 knockdown is not easy to interpret. Since NL2 is not directly localized at glutamatergic synapses, it is possible that a different protein might mediate the effect of NL2 on glutamatergic synapses. Interestingly, KCC2 has recently been shown to modulate dendritic spines and AMPA receptor diffusion through interaction with cytoskeleton proteins. Our above results suggest that KCC2 may be a novel downstream effector of NL2. We therefore hypothesized that NL2 may indirectly regulate glutamatergic synapses through the mediation of KCC2.
To test this hypothesis, we examined the effect of NL2 and KCC2 on dendritic spine morphogenesis in mature neurons (15 DIV) (Figure 7). Consistent with the electrophysiology data, knockdown of NL2 significantly reduced dendritic spines compared to the GFP controls (Figure 7A-B). Interestingly, coexpression of KCC2 with NL2shRNA significantly rescued dendritic spines (Figure 7C). The dendritic spines were confirmed with overlaying glutamatergic presynaptic marker VGlut1 (Figure 7D).
Quantitatively, the spine density in control neurons transfected with GFP alone was 10 ± 1 per 20 μm, and decreased to 4 ± 1 per 20 μm after transfection with GFP-NL2shRNA, but rescued to 9 ± 1 per 20 μm when KCC2 was coexpressed with GFP-NL2shRNA (Figure 7E). Knockdown of NL2 in mature neurons also significantly decreased KCC2 expression, and KCC2 co-transfection with NL2shRNA effectively rescued the KCC2 level (Figure 7F). Moreover, when KCC2 was coexpressed with GFP-NL2shRNA, it partially rescued the decrease of mEPSC frequency, but mIPSCs were not rescued (Figure 7G). These data suggest that NL2 directly regulates GABAergic synapse formation, while indirectly regulate glutamatergic synapse formation through KCC2.
Neuroligin-2 regulates dendritic spines through KCC2. A-C, Representative images showing dendritic spines in neurons (15 DIV) transfected with GFP (A), GFP-NL2shRNA (B), or GFP-NL2shRNA + KCC2 (C). Scale bar, 20 μm. D, Enlarged inlets in A-C showing dendritic spines (green) overlaid with glutamatergic presynaptic terminals labeled by vGlut1 (red). Scale bar, 5 μm. E, Quantification of dendritic spine numbers per 20 μm in neurons expressing GFP, GFP-NL2shRNA, or GFP-NL2shRNA + KCC2. n = 11 per group. F, Bar graphs showing somatic KCC2 immunofluorescence intensity: control neurons, 153 ± 9, n = 17; GFP-NL2shRNA transfected neurons, 88 ± 14, n = 11; GFP-NL2shRNA + KCC2 transfected neurons, 168 ± 13, n = 9. a.u., arbitrary unit. G, Bar graphs showing the frequency of mIPSCs (n = 9 per group; left Y-axis) and mEPSCs (n = 11 per group; right Y-axis). mIPSCs: Control, 0.58 ± 0.14 Hz; GFP-NL2shRNA, 0.02 ± 0.01 Hz; GFP-NL2shRNA + KCC2, 0.03 ± 0.02 Hz. mEPSCs: Control, 2.55 ± 0.42 Hz; GFP-NL2shRNA, 0.12 ± 0.03 Hz; GFP-NL2shRNA + KCC2, 1.07 ± 0.22 Hz. *p < 0.05, ***p < 0.001, n.s., not significant, one-way ANOVA.
## Discussion
*In the **Discussion** section:*
In this study, we report a novel function of neuroligin-2 in regulating KCC2 and GABA functional switch in cortical neurons. This is supported by several lines of evidences: 1) Knockdown of NL2 induced a significant decrease of the expression of KCC2, but not NKCC1; 2) Overexpressing and knockdown of NL2 caused negative and positive shift of EGABA, respectively; 3) After knockdown of NL2, GABA application induced large Ca2+ influx and even action potentials, indicating an excitatory action; 4) The decrease of KCC2 expression and GABA functional changes after knockdown of NL1-3 were rescued selectively by shRNA-proof NL2*, but not NL1*. Therefore, KCC2 is specifically regulated by NL2.
We further demonstrate that NL2 may also modulate glutamatergic transmission and dendritic spines through the regulation of KCC2. Therefore, knockdown of NL2 will decrease GABAergic synapses, reduce glutamatergic synapses, and make GABA function more excitatory (summarized in Figure 8A). These novel findings support a central role of NL2 in governing the delicate balance between GABAergic and glutamatergic functions (Figure 8B).[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Graphic summary of major findings and significance. A, Working model illustrating multiple changes induced by knockdown of NL2, including reduced GABAergic synapses, reduced glutamatergic synapses, and reduced KCC2 expression, which results in increased intracellular Cl- concentration and more excitatory GABA function. B, Schematic drawing illustrating NL2 as a central player in regulating the excitation and inhibition balance in the brain.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
## Classical function of neuroligin-2 in GABAergic synaptogenesis
*In the **Classical function of neuroligin-2 in GABAergic synaptogenesis** section:*
Neuroligins and their presynaptic binding partner neurexins are important trans-synaptic cell adhesion molecules that play a critical role in synaptogenesis. The synaptogenic effect of neuroligins was demonstrated by their potent induction of presynaptic differentiation when expressed in non-neuronal cells. The NL1-3 triple knockout in vivo and acute NL1-3 knockdown by shRNAs in vitro both showed significant deficits in synaptic transmission. It was later found that neuroligins might also play a role in synaptic validation and plasticity, possibly through a trans-synaptic feedback. Most recent studies have revealed that NL1 may be cleaved by matrix metalloproteinases in an activity-dependent manner, which regulates glutamatergic synaptic transmission.
It will be interesting to investigate whether other NLs could also be cleaved and how the cleavage process regulates synaptic functions.
We have previously demonstrated that overexpression of NL2 with GABAA receptors in HEK cells can induce fully functional GABAergic innervations from surrounding neurons. We have also identified a novel loss-of-function mutation of NL2 (R215H) from human schizophrenic patient, which is incapable of inducing GABAergic innervations. In the current study, we further demonstrate that knockdown of NL2 significantly reduced GABAergic synaptogenesis, consistent with previous studies. The function of NL2 in regulating GABAergic synapse formation is likely mediated by interactions with scaffolding proteins like gephyrin and collybistin. All these studies are consistent with the role of neuroligin-2 as a cell adhesion molecule to regulate GABAergic synapse formation and plasticity.
## Novel function of neuroligin-2 in regulating KCC2 and GABA functional switch
*In the **Novel function of neuroligin-2 in regulating KCC2 and GABA functional switch** section:*
The most surprising finding of this work is the regulation of KCC2 by NL2. After knockdown of NL2, the KCC2 expression level was significantly decreased. We tested whether this might be caused by an off-target effect of the shRNAs, but coexpression of NL2shRNA with KCC2 in HEK cells showed no effect on the expression of KCC2 at all. Moreover, the shRNA-proof NL2*, but not NL1*, rescued the KCC2 expression, suggesting a NL2-specific regulation of KCC2. Similar to previously reported knockdown of KCC2, we discovered that the GABA reversal potential was shifted to more depolarized level after NL2 knockdown.
Moreover, after NL2 knockdown, bath application of GABA induced large Ca2+ influx and even triggered action potentials. These results discover a novel function of NL2 in regulating KCC2 and GABA functional switch. The regulation of KCC2 by NL2 suggests that the time-course of GABAergic synaptogenesis and GABA functional switch are tightly coordinated by NL2.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
KCC2 is a Cl- transporter with a major function in controlling intracellular Cl- concentration and therefore playing an important role in determining whether GABA function is excitatory or inhibitory. The regulation of KCC2 by NL2 greatly expands the function of NL2 beyond its classical role as a cell adhesion molecule. Since both NL2 and KCC2 are transmembrane proteins, we investigated whether NL2 directly interacts with KCC2. Our co-immunoprecipitation experiments in HEK cells and mouse brain lysate suggest a lack of direct interaction between NL2 and KCC2. However, we cannot exclude the possibility that NL2 may indirectly interact with KCC2 through other mediating proteins, such as gephyrin or GABAA receptors, since overexpressed HA-NL2 showed partial colocalization with KCC2.
NL2 may also regulate KCC2 activity through a variety of pathways, such as transcriptional regulation, phosphorylation, membrane trafficking, and oligomerization.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
Since both NL2 and KCC2 regulate GABA functions, we tested whether the regulation of KCC2 by NL2 may be dependent upon GABAA receptor activation. Our pharmacological experiments with BIC and TTX demonstrated that developmental KCC2 up-regulation and GABA functional switch are independent of GABAA receptor activation, which is in agreement with previous studies. What may be responsible for KCC2 up-regulation during neuronal development? Our new findings suggest that KCC2 may be regulated by NL2, a cell adhesion molecule that expresses early during development. Indeed, we found that NL2 expression precedes that of KCC2 in vivo in early postnatal period, consistent with previous studies reporting that NL2 expression was first detected at embryonic day 16 while KCC2 was first detected at postnatal day 1.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Our discovery of the NL2 regulation of KCC2 makes it easy to connect some interesting findings previously reported by different labs that are seemingly unrelated. For example, overexpression of NL2 in cerebellar granule cells has been shown to accelerate GABAergic synapse maturation by promoting the switch of GABAA receptor α3 subunit to α1 subunit during early development. Interestingly enough, overexpression of KCC2 was also found to accelerate this switch in cerebellar granule cells. Our finding that NL2 regulates KCC2 may explain why NL2 and KCC2 both showed similar functions in promoting α subunit switch, although other links may also be possible. Therefore, NL2 is a chief conductor in orchestrating a variety of GABA functions including GABAergic synapse formation, GABA functional switch, and GABAA receptor maturation.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Neuroligin-2 regulates glutamatergic synapses through KCC2
*In the **Neuroligin-2 regulates glutamatergic synapses through KCC2** section:*
Given the fact that NL2 is mainly localized at GABAergic synapses, it is initially puzzling that knockdown of NL2 not only reduced GABAergic synapses, but also reduced glutamatergic synapses. The original report using the same NL2shRNA also showed reductions in both glutamatergic and GABAergic synapses. Interestingly, NL2 overexpression increased both glutamatergic and GABAergic synapse formation. On the other hand, it has been shown that GABAergic, but not glutamatergic, transmission was decreased in NL2 knockout mice. Such discrepancy between in vitro and in vivo data regarding the role of NL2 at glutamatergic synapses could be due to the difference between global knockout and shRNA-mediated knockdown.
A recent study demonstrated that the transcellular differences in the relative amounts of NL1, rather than the absolute NL1 amount, governs the number of glutamatergic synapses in vivo. Nevertheless, how NL2 might regulate glutamatergic synapses is not very clear.
KCC2 has been found to regulate dendritic spines. Specifically, Rivera and colleagues first reported that neurons from KCC2 knockout mice showed abnormally long dendritic protrusions and low frequency of mEPSCs. Similarly, knockdown of KCC2 in developing neurons decreased the frequency of glutamatergic events. KCC2 has also been shown to regulate the diffusion of AMPA receptors in dendritic spines. Here, we demonstrated that NL2 regulates KCC2, raising a potential link between NL2 and glutamatergic synapses through KCC2.
Indeed, overexpression of KCC2 together with NL2shRNA was able to rescue decreased glutamatergic synapses induced by NL2 knockdown, suggesting that KCC2 is likely the mediator of NL2 regulation of glutamatergic synapses. Therefore, NL2 not only regulates GABAergic synapses, but also regulates glutamatergic synapses through regulating KCC2, making NL2 a central player in balancing GABAergic and glutamatergic functions in the brain.
## Conclusion
*In the **Conclusion** section:*
To conclude, our data discovered a novel function of NL2 in regulating KCC2 and consequently affecting GABA functional switch as well as glutamatergic synapses. This finding extends the function of NL2 beyond its classical role in cell adhesion, and put NL2 in a central position in coordinating GABAergic synaptogenesis with GABA functional switch, and in balancing GABAergic and glutamatergic functions. We propose that NL2 may function as a master regulator of the delicate excitation-inhibition balance in the brain, and NL2 may be a novel drug target for developing the next generation of antipsychotic drugs.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Materials and methods
*In the **Materials and methods** section:*
## Cell culture
*In the **Cell culture** section:*
Primary mouse cortical neurons were cultured as previously described. Briefly, cerebral cortices of newborn C57BL/6 mice of either sex were dissociated and plated on a monolayer of cortical astrocytes at a density of 8,000-12,000 cells/cm2 in 24-well plates. Culture medium contained MEM (500 ml, Invitrogen), 5% FBS (HyClone), 10 ml B-27 supplement (Invitrogen), 100 mg NaHCO3, 20 mM D-glucose, 2 mM Glutamax (Invitrogen), and 25 units/ml penicillin/streptomycin. Neurons were maintained at 37°C in a 5% CO2-humidified incubator. All experiments were repeated in at least three independent cultures.[](https://www.ncbi.nlm.nih.gov/mesh/D017693)
## Transfection
*In the **Transfection** section:*
Calcium-phosphate transfection in cultured neurons was performed similar to a protocol developed in our laboratory. Plasmid at 1 μg each was used for transfection per well in a 24-well plate. HEK 293T cells were transfected using polyethylenimine (Polysciences). GFP or mCherry was coexpressed to identify transfected cells.[](https://www.ncbi.nlm.nih.gov/mesh/C020243)
## Plasmids
*In the **Plasmids** section:*
NLmiR with IRES GFP or mCherry, HA-tagged shRNA-proof mouse NL1* and rat NL2* with IRES mCherry were generously provided by Dr. Roger Nicoll (University of California at San Francisco, San Francisco, CA). HA-tagged WT mouse NL2 and EGFP-NL2shRNA were provided by Dr. Peter Scheiffele (University of Basel, Basel, Switzerland). A non-tagged NL2shRNA was generated by cutting off EGFP with restriction enzymes. NLmiR and NL2shRNA have been characterized previously. The NL2 target sequence of NL2shRNA and NLmiR is identical (ATGGAGCAAGTTCAACAGCAA) and conserved in mouse and rat.
Rat KCC2 (pIRES2-EGFP) was provided by Dr. Yun Wang (Fudan University, Shanghai, China). mCherry (pEGFP-C1) was provided by Dr. Yingwei Mao (Pennsylvania State University, University Park, PA).
## Immunostaining and imaging analysis
*In the **Immunostaining and imaging analysis** section:*
Neurons were fixed in 4% paraformaldehyde for 8 min, permeabilized with 0.2% Triton X-100 for 5 min, and blocked with 5% normal donkey/goat serum for 30 min. Primary antibodies in blocking solution were incubated overnight at 4°C. Dylight-conjugated secondary antibodies (Jackson ImmunoResearch) were incubated at room temperature for 45 min. Following antibodies were used: KCC2 (07–432, Millipore), HA (sc-7392, Santa Cruz Biotechnology), MAP2 (ab5392, Abcam), NKCC1 (T4, Developmental Studies Hybridoma Bank), vGlut1 (135302, Synaptic Systems), GFP (ab13970, Abcam). Confocal images were collected on an Olympus FV1000 confocal microscope. For the quantification of HA, KCC2, and NKCC1 immunostaining, neuronal soma was selected and the mean intensity (0–255) was analyzed by ImageJ software.
For spine density analysis, two secondary dendritic segments of 20 μm each were analyzed per neuron.[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
## Electrophysiology
*In the **Electrophysiology** section:*
The electrophysiological experiments were performed as previously described. Briefly, Multiclamp 700A amplifier and pClamp software (Molecular Devices) were used for acquiring data (sampling at 10 kHz and filtered at 1 kHz). Neurons were continually perfused with bath solution (in mM): 128 NaCl, 30 glucose, 25 HEPES, 5 KCl, 2 CaCl2 and 1 MgCl2 (320 mOsm, adjusted to pH 7.3 with NaOH). Pipette solution contained (in mM): 147 KCl, 5 Na2-phosphocreatine, 2 EGTA, 10 HEPES, 2 MgATP, 0.3 Na2GTP (300 mOsm adjusted to pH 7.3 with KOH). Gramicidin (40 μg/ml, Sigma) was included in the pipette solution for perforated patch recording.
A Picospritzer (Parker Instrumentation) was used to eject GABA directly to neuronal soma through a fine pipette (~2 μm tip). In whole-cell patch clamp mode (holding at -70 mV), mEPSCs were recorded in the presence of TTX (0.5 μM) and BIC (20 μM); mIPSCs were recorded in the presence of TTX (0.5 μM) and DNQX (10 μM).[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## Immunoblot
*In the **Immunoblot** section:*
HEK 293T cells in 6-well plates were transfected using polyethylenimine and total protein lysate was harvested after 2–3 days in lysis buffer (20 mM HEPES, 1% Triton X-100, 0.1 mM EDTA, 2 mM CaCl2, 1mM MgCl2 and 50 mM NaCl with PMSF, protease and phosphatase inhibitors, pH 7.3 with NaOH). For cultured neurons, 2 wells from a 24-well plate were lysised at each time point. For mouse cortical proteins, cortices were dissected out and homogenized. After 2 hr rotation at 4°C, supernatant was harvested by centrifugation (12,000 g, 30 min). Protein concentration was measured by Bradford Protein Assay Kit (Thermo Scientific Pierce Protein Biology Products).
Samples were incubated with NuPAGE LDS sample buffer and 1% β-mercaptoethanol at 95°C (for NL2) or 50°C (for KCC2) for 15 min before resolved in 10% SDS-PAGE and transferred to PVDF membrane. Primary antibodies including rabbit anti-KCC2 (07–432, Millipore), rabbit anti-NL2 (129202, Synaptic Systems), mouse anti-HA (sc-7392, Santa Cruz Biotechnology) and mouse anti-actin (612656, BD Transduction Laboratories) together with HRP-conjugated secondary antibodies (Abcam) were used. Immunoblot band intensities were measured with ImageJ software.[](https://www.ncbi.nlm.nih.gov/mesh/D011094)
## Co-immunoprecipitation
*In the **Co-immunoprecipitation** section:*
HEK 293T cells in 10-cm dish were transfected and total protein lysate was harvested after 2 days in Pierece IP buffer (Thermo Scientific) with protease and phosphatase inhibitors. Protein lysate was pre-cleaned by incubation with Dynabeads M-280 sheep anti-mouse IgG (11201D, Invitrogen) for 2 hr at 4°C. About 2 mg protein lysate was then incubated with Dynabeads and 2 μg normal mouse IgG (PP54, Millipore) or mouse anti-HA antibodies overnight at 4°C. After washing with IP buffer and PBS, the immunoprecipitated proteins were eluted by NuPAGE LDS sample buffer (Invitrogen). Protein lysate from 2–3 month old mouse brain was processed following the same protocol with Dynabeads M-280 sheep anti-rabbit IgG (11203D, Invitrogen).
Normal rabbit IgG (PP64, Millipore) was used as control.[](https://www.ncbi.nlm.nih.gov/mesh/C010615)
## Calcium imaging
*In the **Calcium imaging** section:*
Cells were incubated in 2.5 μM Fura-2 AM (Invitrogen) for 45 min at 37°C and washed for 15 min in bath solution. Coverslips were transferred to a perfusion chamber mounted on a Nikon TE-2000-S inverted microscope with a 20× objective and imaged with a 340/380 nm transmittance filter set (Chroma Technology). SimplePCI (HCImage, Hamamatsu) was used to measure the ratio of 340/380 fluorescence signal in neuronal soma. Sister coverslips from 3–4 independent cultures were taken for imaging at 3–4 time points. mCherry was coexpressed to identify transfected cells in all calcium imaging experiments.
All recordings were done in the presence of DNQX (10 μM) to block AMPA receptor activations. The threshold of a significant Ca2+ response was set as 10 times of baseline standard deviation.[](https://www.ncbi.nlm.nih.gov/mesh/D016257)
## Statistical analysis
*In the **Statistical analysis** section:*
Unpaired Student’s t-test was used for comparisons between two groups. One-way ANOVA with Bonferroni multiple comparisons was used for comparisons between multiple groups. Two-way ANOVA with Bonferroni multiple comparisons was used for comparisons between multiple time points and groups. GraphPad Prism (GraphPad Software) was used for all statistical analysis. Data were shown as mean ± standard error in all bar graphs.
## Abbreviations
*In the **Abbreviations** section:*
NL2: Neuroligin-2; KCC2: Potassium-chloride transporter 2; NL2shRNA: Short hairpin RNA targeting NL2; NLmiR: Short hairpin RNA targeting NL1-3; BIC: Bicuculline; TTX: Tetrodotoxin; DIV: Days in vitro; EGABA: GABA equilibrium potential.[](https://www.ncbi.nlm.nih.gov/mesh/D001640)
## Competing interest
*In the **Competing interest** section:*
The authors declare that they have no competing interests.
## Authors’ contributions
*In the **Authors’ contributions** section:*
CS and LZ performed the experiments; CS, LZ, and GC analyzed the data; CS and GC designed the experiments and wrote the paper. All authors read and approved the final manuscript.
## Acknowledgments
*In the **Acknowledgments** section:*
We thank Drs. Roger Nicoll, Peter Scheiffele, Yun Wang, and Yingwei Mao for generously sharing plasmid constructs. We thank Yuting Bai for the preparation of neuronal culture. This work was supported by grants from National Institutes of Health (MH092740 and MH083911) to GC.
|
# Introduction
A Pivotal Role for [Tryptophan](https://www.ncbi.nlm.nih.gov/mesh/D014364) 447 in Enzymatic Coupling of Human Endothelial Nitric Oxide Synthase (eNOS)
# Abstract
*In the **Abstract** section:*
Background: Interaction of tetrahydrobiopterin (BH4) with a key tryptophan residue in the NOS active site is critical for activity.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
Results: Mutation of tryptophan 447 causes eNOS uncoupling and monomerization.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Conclusion: Tryptophan 447 determines enzymatic coupling of human eNOS.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Significance: The development of BH4-based strategies to restore NOS function must consider the structural effects of BH4 binding and their role in NOS coupling.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
Tetrahydrobiopterin (BH4) is a required cofactor for the synthesis of NO by NOS. Bioavailability of BH4 is a critical factor in regulating the balance between NO and superoxide production by endothelial NOS (eNOS coupling). Crystal structures of the mouse inducible NOS oxygenase domain reveal a homologous BH4-binding site located in the dimer interface and a conserved tryptophan residue that engages in hydrogen bonding or aromatic stacking interactions with the BH4 ring. The role of this residue in eNOS coupling remains unexplored. We overexpres[sed human eNOS W447](https://www.ncbi.nlm.nih.gov/mesh/C003402)A [and](https://www.ncbi.nlm.nih.gov/mesh/C003402) W447F mutants in novel cell lines with tetrac[yc](https://www.ncbi.nlm.nih.gov/mesh/D009569)line-regulated expression of[ hu](https://www.ncbi.nlm.nih.gov/mesh/C003402)man GTP cyclohydrolase I, the rate-limiting enzyme in BH[4 ](https://www.ncbi.nlm.nih.gov/mesh/D009569)synth[esis, to d](https://www.ncbi.nlm.nih.gov/mesh/D013481)etermine the importance of BH4 and Trp-447 in eNOS uncoupling. NO production was abolished in eNOS-W447A cells and diminished in ce[lls](https://www.ncbi.nlm.nih.gov/mesh/C003402) expressing W447F, despite high BH4 levels. eNOS-derived supe[roxide pro](https://www.ncbi.nlm.nih.gov/mesh/D014364)duction was significantly[ elevate](https://www.ncbi.nlm.nih.gov/mesh/D006859)d in W447A and W447F versus wild-type eNOS, and this[ wa](https://www.ncbi.nlm.nih.gov/mesh/C003402)s sufficient to oxidize BH4 to 7,8-dihydrobiopterin. In uncoupled, BH4-deficient cells, the deleterious effects of W447A mutation were greatly exa[cerbated, re](https://www.ncbi.nlm.nih.gov/mesh/D013752)sulting in further attenuation of NO and greatly increased superoxide production.[ eN](https://www.ncbi.nlm.nih.gov/mesh/C003402)OS dimerization was attenuated in W447A eNO[S c](https://www.ncbi.nlm.nih.gov/mesh/C003402)ells and further reduced in BH4-d[ef](https://www.ncbi.nlm.nih.gov/mesh/D009569)icient cells, as demonstrated using a novel split Renilla luciferase biosensor. Reduction of cellular[ BH](https://www.ncbi.nlm.nih.gov/mesh/C003402)4 levels resulted in a[ switch fr](https://www.ncbi.nlm.nih.gov/mesh/D013481)om an eNOS dimer to an eNOS monomer. These data reveal a key role for Trp-447 in determining NO versus superoxide pr[odu](https://www.ncbi.nlm.nih.gov/mesh/C003402)ctio[n by eNOS, by effect](https://www.ncbi.nlm.nih.gov/mesh/C017226)s on BH4-depende[nt ](https://www.ncbi.nlm.nih.gov/mesh/C003402)catalysis, and by modulating eNOS dimer formation.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
NOSs catalyze the five-electron oxidation of l-arginine to produce l-citrulline and NO, an important signaling molecule in immunological, neuronal, and cardiovascular homeostasis and disease. In mammals, NO is generated by three distinct NOS isoforms, referred to as inducible (iNOS),2 neuronal, and endothelial (eNOS). All NOS isoforms are obligate homodimers with similar structures. The N-terminal catalytic oxygenase domain (NOSox) binds heme, the substrate l-arginine, and the essential pterin cofactor tetrahydrobiopterin (BH4), whereas the C-terminal reductase domain (NOSred) supplies electrons through bound flavin mononucleotide, flavin adenine dinucleotide, and NADPH. A central linker binds calmodulin and modulates the transfer of electrons from NOSred of one subunit to the NOSox of the other dimer subunit.[](https://www.ncbi.nlm.nih.gov/mesh/D001120)
BH4 is required for NO synthesis by all NOS isoforms. Although fully reduced tetrahydropterins support catalysis by NOSs, oxidized pterin species such as 7,8-dihydrobiopterin (BH2) and biopterin are catalytically incompetent, having the same allosteric effects without the ability to catalyze NO production. EPR studies showed that in the absence of BH4 (or presence of excess BH2), superoxide is the sole in vitro product of recombinant eNOS. In the absence of BH4, electron transfer from NOS flavins becomes “uncoupled” from l-arginine oxidation, the ferrous-dioxygen complex dissociates, and superoxide is released from the oxygenase domain. This eNOS-derived superoxide production has been implicated in a wide variety of molecular, animal, and clinical models of vascular disease, including diabetes, cigarette smoking, hypertension, and atherosclerosis.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
As well as being pivotal in the transfer of electrons to the Fe(II)O2 complex and radical formation, the binding of BH4 to iNOS also has positive effects on dimerization and increases the binding affinity of the enzymatic substrate arginine. Previous studies looking at the binding of BH4 to iNOS have demonstrated that BH4 makes hydrogen bonds with the heme propionate and exhibits extensive interactions with the residues Trp-455, Trp-457, Phe-470, Arg-375, and Arg-193, as shown in Fig. 1. In particular, mutation of Trp-457 to Phe and Ala residues in iNOS greatly disrupts the interaction of Trp-457 with BH4 and decreases NO synthesis activity by 3.3-fold and 8-fold, respectively. However, the role of BH4 binding in eNOS, the specific interaction of BH4 with this residue, and the role of eNOS Trp-447 (Trp-457 in iNOS) in enzymatic uncoupling remain unexplored.[](https://www.ncbi.nlm.nih.gov/mesh/D005296)
Human eNOS and murine iNOS have closely related structures.
A, the active site of iNOS (yellow), showing the heme group (green) and arginine substrate (pale blue). B, the active site of eNOS (gray), demonstrating its close structural similarity with iNOS. Structural superposition of human eNOS and murine iNOS (PDB codes 3NOS and 1NOD, respectively) was performed using PDBe Fold. The top-ranked root mean square deviation between 365 matched residues in a single protein chain was 0.81Å. C, the active site of iNOS showing the W457F mutation. D, the active site of iNOS showing the W457A mutation. Figures were produced using PyMOL on the basis of Fig. 1 by Wang et al. using PDB codes 3NOS (eNOS), 1NOD (iNOS), 1JWJ (iNOS W457F), and 1JWK (iNOS W457A).[](https://www.ncbi.nlm.nih.gov/mesh/D006418)
eNOS uncoupling is often thought to occur in parallel with eNOS monomerization, and confusion exists as to whether changes in the dimer/monomer ratio are directly related to the functional uncoupling of eNOS because current literature suggests that only the dimeric form of eNOS is biochemically active and able to generate either NO or superoxide. Questions also remain as to whether eNOS in the uncoupled state exists as a monomer, therefore suggesting that the influences of BH4 on “dimer stabilization” and the coupling of eNOS are not necessarily one and the same effect.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Accordingly, we sought to elucidate a mechanistic role for the interaction of BH4 with eNOS Trp-447 in the regulation of eNOS uncoupling and monomerization. To address these questions, we expressed eNOS W447F and eNOS W447A mutants in HEK293 cells and in cells that stably express doxycycline-regulatable GTPCH protein to determine the effects of high or low intracellular BH4. We also developed a novel biosensor of eNOS dimerization on the basis of the reconstitution of split Renilla luciferase, revealing for the first time that the Trp-447 residue within the BH4 binding site of eNOS is required for efficient NO production by the enzyme, is critical for the coupling of eNOS, and also plays a role in dimerization.
These findings have significant consequences for the therapeutic potential of BH4-dependent eNOS catalysis and dimerization.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## EXPERIMENTAL PROCEDURES
*In the **EXPERIMENTAL PROCEDURES** section:*
## Molecular Modeling of the eNOS Active Site
*In the **Molecular Modeling of the eNOS Active Site** section:*
Figures were produced using PyMOL, on the basis of Fig. 1 by Wang et al. using PDB codes 3NOS (eNOS), 1NOD (iNOS), 1JWJ (iNOS W457F), and 1JWK (iNOS W457A).
## Generation of Tet-regulatable Cells
*In the **Generation of Tet-regulatable Cells** section:*
We used National Institutes of Health 3T3 murine fibroblasts stably transfected with a Tet-Off transactivator construct as described previously. In the presence of doxycycline, binding of the transactivator is blocked, and gene expression is prevented. These 3T3-Tet-Off cells, previously shown not to express GTPCH and also confirmed to be devoid of eNOS protein, were stably transfected with a plasmid encoding hemagglutinin antigen-tagged human GTPCH under the control of a tetracycline-responsive element. Individual colonies were isolated and analyzed for GTPCH expression, and a cell line termed “GCH cells” was established from expansion of a single colony. GCH/eNOS cells were produced by stable transfection of GCH cells with a plasmid encoding a human eNOS-eGFP fusion protein as described.
In this model, addition of doxycycline to GCH/eNOS cells results in diminished GCH mRNA, GTPCH protein, and BH4 levels, leading to uncoupling of eNOS.[](https://www.ncbi.nlm.nih.gov/mesh/D004318)
## Measurement of eNOS Protein Levels by eGFP Fluorescence
*In the **Measurement of eNOS Protein Levels by eGFP Fluorescence** section:*
Cell pellets were lysed in phosphate-buffered saline containing 1 mmol/liter dithioerythritol and 100 μmol/liter EDTA as for BH4 analysis. Sample fluorescence was quantified using a TEKAN fluorescence plate reader and a standard curve generated using recombinant eGFP. Because recombinant eNOS was expressed as an eNOS-eGFP fusion protein, eGFP fluorescence and eNOS levels were directly proportional.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## Generation of Tryptophan eNOS Mutants
*In the **Generation of Tryptophan eNOS Mutants** section:*
QuikChange site-directed mutagenesis (Stratagene) was used to create W447A, W447F, and C99A human eNOS mutants. Mutant primers were designed and used as follows: W447A, 5′-TGCAGACTGGGCCGCGATCGTGCCCCCC-3′ (sense) and 5′-GGGGGGCACGATCGCGGCCCAGTCTGCA-3′ (antisense); W447F, 5′-GCAGACTGGGCCTTCATCGTGCCCCCCA-3′ (sense) and 5′-TGGGGGGCACGATGAAGGCCCAGTCTGC-3′ (antisense); and C99A, 5′-CACCCCAAGACGCGCCCTGGGCTCCCTG-3′ (sense) and 5′-CAGGGAGCCCAGGGCGCGTCTTGGGGTG-3′ (antisense). The sequence of each eNOS single and double mutant was confirmed by DNA sequencing (SourceBioscience, UK). Mutant constructs were transiently expressed in GCH cells using FuGENE 6 (Roche).[](https://www.ncbi.nlm.nih.gov/mesh/D011119)
## Cell Culture
*In the **Cell Culture** section:*
Cells were cultured in DMEM (Invitrogen) supplemented with 2 mm glutamine, 100 units/ml penicillin, and 0.1 mg/ml streptomycin. Additionally, GCH cells were maintained using media containing the antibiotics hygromycin (200 μg/ml) and genetecin (200 μg/ml), whereas GCH/eNOS-eGFP cell medium also included 2 μg/ml puromycin. Where appropriate, 1 μg/ml doxycycline was added to cell culture medium to abolish transcription of GCH1 mRNA.[](https://www.ncbi.nlm.nih.gov/mesh/D005973)
## Monomer and Dimer Western Blotting
*In the **Monomer and Dimer Western Blotting** section:*
Low-temperature SDS-PAGE was performed for detection of the eNOS monomer and dimer. Briefly, cells lysates were prepared by homogenization in ice-cold CelLytic M buffer (Sigma) containing protease inhibitor (Roche Applied Science) and subjected to three freeze-thaw cycles in liquid nitrogen. Lysates were centrifuged at 13,200 rpm for 10 min at 4 °C, and samples were prepared using Laemmli sample buffer (Sigma). Protein lysates were resolved using a 6% Tris-glycine gel (Invitrogen) under reducing conditions. All gels and buffers were pre-equilibrated to 4 °C before electrophoresis, and the buffer tank was placed in an ice bath during electrophoresis to maintain the gel temperature below 15 °C.
Standard blotting techniques were used, and membranes were incubated with mouse anti-eNOS polyclonal antibody (BD Transduction Laboratories) as described previously.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## Creation of eNOS Dimerization Biosensors
*In the **Creation of eNOS Dimerization Biosensors** section:*
Rluc8.1 and 8.2 fragments were PCR-amplified using the rluc8 plasmid as a template (a gift from Prof. Sanjiv Gambhir, Stanford University). The rluc8 PCR fragments were cloned into pcDNA3. Gateway destination plasmids were then created by cloning the vector conversion kit (Invitrogen) into the rluc8.1 and rluc8.2 vectors. For the creation of stable cell lines, this destination cassette was cloned into the pIRES-puro and Neo plasmids (Clontech). pENTR clones encoding our genes of interest were either created by PCR or purchased from Geneservice or Thermo Fisher.
To create expression plasmids, pENTR clones were recombined with the destination vectors using LR clonase according to the instructions of the manufacturer. Recombinant Rluc8 expression plasmids and a LacZ-encoding control plasmid were then transfected into HEK293T or GCH cells and grown on white 96-well plates using FuGENE HD (Roche). We placed our rluc8.1 and rluc8.2 halves onto either end of eNOS. rluc8.1 was added to the C terminus (reductase domain), and rluc8.2 was added to the N terminus (oxygenase domain).
## Detection of Biosensor Luminescence by Protein Fragment Complementation Assay
*In the **Detection of Biosensor Luminescence by Protein Fragment Complementation Assay** section:*
24 h following transfection, cells were treated with drugs as indicated. Cells were then washed with PBS, and reconstituted Rluc8 activity was measured using benzyl-coelenterazine (Nanolight) on a BMG Polarstar plate reader (5-s read time). Cells were lysed by adding LacZ lysis buffer (120 μm TrisPO4 (pH 7.8), 10 mm 1,2-cyclohexylenedinitrilotetraacetic acid, 30% glycerol, and 1% Triton X-100) to each well, and LacZ activity was measured using ortho-nitrophenyl-β-galactopyranoside. Rluc8 readings were normalized using LacZ activity.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## Analysis of NO Synthesis by eNOS
*In the **Analysis of NO Synthesis by eNOS** section:*
Cellular NO synthesis by eNOS was assessed by measuring the conversion of 14C l-arginine to citrulline with HPLC detection, in the presence and absence of NG-monomethyl-l-arginine, as described previously.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## Quantification of Superoxide Production by HPLC
*In the **Quantification of Superoxide Production by HPLC** section:*
Measurement of 2-hydroxyethidium formation by HPLC was used to quantify superoxide production, as described previously. Cells were washed three times in PBS and incubated in Krebs-Hepes buffer in the presence or absence of 100 μm L-NAME. After 30 min, 25 μm dihydroethidium was added, and cells were incubated for an additional 20 min at 37 °C. Cells were then harvested by scraping, centrifuged, and lysed in ice-cold methanol. 100 mm hydrochloric acid was added (1:1 v/v) prior to loading into the autosampler for analysis.
All samples were stored in darkened tubes and protected from light at all times. Separation of ethidium, oxyethidium, and dihydroethidium was performed using a gradient HPLC system (Jasco) with an ODS3 reverse phase column (250 mm, 4.5 mm; Hichrom) and quantified using a fluorescence detector set at 510 nm (excitation) and 595 nm (emission). A linear gradient was applied from mobile phase A (0.1% TFA) to mobile phase B (0.1% TFA in acetonitrile) over 23 min (30–50% acetonitrile).[](https://www.ncbi.nlm.nih.gov/mesh/C506392)
## Biopterin Quantification by HPLC with Electrochemical Detection
*In the **Biopterin Quantification by HPLC with Electrochemical Detection** section:*
BH4, BH2, and biopterin levels in cell lysates were determined by HPLC, followed by electrochemical and fluorescent detection, as described previously. Briefly, cells were grown to confluency and harvested by trypsinization. Sample pellets were resuspended in 50 mm phosphate-buffered saline (pH 7.4) containing 1 mm dithioerythritol and 100 μm EDTA and subjected to three freeze-thaw cycles. Following centrifugation (15 min at 13,000 rpm and 4 °C), samples were transferred to new, cooled microtubes and precipitated with 1 m cold phosphoric acid, 2 m TCA, and 1 mm dithioerythritol. Samples were mixed vigorously and then centrifuged for 15 min at 13,000 rpm and 4 °C.
Samples were injected onto an isocratic HPLC system and quantified using sequential electrochemical (Coulochem III, ESA Inc., UK) and fluorescence (Jasco) detection. HPLC separation was performed using a 250-mm ACE C-18 column (Hichrom) and a mobile phase comprising 50 mm sodium acetate, 5 mm citric acid, 48 μm EDTA, and 160 μm dithioerythritol (pH = 5.2) (all ultrapure electrochemical HPLC grade) at a flow rate of 1.3 ml/min. Background currents of +500 μA and −50 μA were used for the detection of BH4 on electrochemical cells E1 and E2, respectively. 7,8-BH2 and biopterin were measured using a Jasco FP2020 fluorescence detector. Quantification of BH4, BH2, and biopterin was done by comparison with authentic external standards and normalized to sample protein content.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## Statistical Analysis
*In the **Statistical Analysis** section:*
Data are presented as mean ± S.E. Data were subjected to the Kolmogorov-Smirnov test to determine distribution. Groups were compared using Mann-Whitney U test for non-parametric data or Student's t test for parametric data. When comparing multiple groups, data were analyzed by analysis of variance with Newman-Keuls post-test for parametric data or Kruskal-Wallis test with Dunn's post-test for non-parametric data. A value of p < 0.05 was considered statistically significant.
## RESULTS
*In the **RESULTS** section:*
## Characterization of Mutant eNOS Expression
*In the **Characterization of Mutant eNOS Expression** section:*
We first investigated the effect of Trp-447 mutation on eNOS expression in murine fibroblasts. Western blotting confirmed that transient expression of wild-type and mutant eNOS was equal (Fig. 2A), as quantified by GFP fluorescence (B). The effect of Trp-447 mutation on eNOS dimerization was assessed by immunoblotting, and the ratio of eNOS monomer/dimer was quantified in GCH cells in the presence and absence of doxycycline, generating low or high levels of BH4. As shown in Fig.
2C, in conditions of high BH4 levels, we observed that human eNOS and the respective Trp-447 mutants existed almost entirely as dimers, as did a murine eNOS-positive control. As a positive control for monomerization, expression of C99A resulted in complete loss of the eNOS dimer, perhaps because of disruption of Zn coordination, as described previously. In doxycycline-treated, BH4-deficient cells, W447A eNOS was predominantly monomeric, and the amount of dimer in the W447F mutant was also decreased significantly (Fig. 2C).[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Overexpression of eNOS Trp-447 mutants results in monomerization and altered localization of eNOS protein. GCH cells were cultured in the presence or absence of DOX for 7 days and then transiently transfected with wild-type eNOS, W447A or W447F mutant eNOS DNA. A, Western blot analysis showed equal expression of our wild-type and mutant eNOS constructs using both anti-eNOS and anti-GFP antibodies. sEnd.1 murine endothelial cells and cells stably expressing eNOS-GFP were used as controls. B, equivalent expression levels were demonstrated by fluorescence quantification. C, eNOS dimerization was compared by low temperature Western blotting in cells in the absence (high BH4) or presence (low BH4) of DOX. Only the C99A mutation of eNOS resulted in significant monomerization compared with endothelial cell (EC) cold and boiled sample controls. In contrast, in BH4-deficient cells, following DOX treatment, the eNOS dimer was totally abolished in W447A and significantly attenuated in W447F mutant-expressing cells. These changes were linked with increases in detectable levels of eNOS monomer. Western blot analyses are representative of three separate experiments. N.D. = not detectable.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## Mutation of Trp-447 Uncouples eNOS
*In the **Mutation of Trp-447 Uncouples eNOS** section:*
We next sought to establish the effect of Trp-447 mutation on eNOS uncoupling and subsequent NO and superoxide production. eNOS activity, as assessed by measuring the conversion of radiolabeled arginine to citrulline, was significantly attenuated in BH4-deficient, DOX-treated cells (Fig. 3; †, p < 0.05), as described previously. Expression of C99A or W447A in cells containing either high or low levels of BH4 totally abolished eNOS activity (p < 0.001) with no further effect of doxycycline. In contrast, W447F mutation resulted in an attenuation of eNOS activity in cells containing high levels of BH4, with a further significant decrease observed in BH4-deficient, DOX-treated cells (p < 0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Trp-447 mutation decreases eNOS activity. GCH cells were cultured in the presence or absence of DOX for 7 days and then transiently transfected with wild-type eNOS, W447A, or W447F mutant eNOS DNA. eNOS activity, as assessed by measuring the conversion of radiolabeled arginine to citrulline, was significantly attenuated in BH4-deficient, DOX-treated cells (†, p < 0.05). Expression of C99A or W447A in cells containing either high, or low levels of BH4 totally abolished eNOS activity (*, p < 0.001) with no further effect of doxycycline. W447F mutation resulted in an attenuation of eNOS activity in cells containing high levels of BH4 with a further significant decrease observed in BH4-deficient, DOX-treated cells (†, p < 0.05). n = 6.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
To further investigate the enzymatic uncoupling of eNOS induced by mutation at Trp-447, we next investigated the ability of eNOS to produce superoxide using dihydroethidium fluorescence. eNOS-derived superoxide was distinguished using the arginine analog L-NAME. As described previously, prevention of BH4 production upon doxycycline exposure of untreated cells caused a striking elevation of superoxide production in an eNOS-independent manner (p < 0.001). We hypothesize that this demonstrates the general antioxidant role of BH4 within the cell. Furthermore, although having no effect in BH4-replete cells, eNOS overexpression in BH4-deficient cells increased the production of superoxide in an L-NAME-inhibitable manner, indicating eNOS uncoupling (p < 0.05).
In contrast, W447A mutation significantly elevated eNOS-derived superoxide production in cells with high BH4 levels and in BH4-deficient cells (p < 0.05). The levels of superoxide produced by the W447A mutant eNOS were significantly greater than those produced by wild-type eNOS (p < 0.05). The increased production of superoxide observed in W447F-expressing cells was not inhibitable with L-NAME in BH4-replete cells but totally abolished by L-NAME in cells containing low levels of BH4. In contrast to the differences in 2-hydroxyethidium, specific for superoxide production, the accumulation of ethidium (an indicator of production of other reactive oxygen species) remained unchanged in eNOS wild-type or mutant-expressing cells (Fig.
4).[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
eNOS-derived superoxide production is exacerbated by Trp-447 mutation in a BH4-dependent manner. GCH cells were cultured in the presence or absence of DOX for 7 days and then transiently transfected with wild-type eNOS, W447A, or W447F mutant eNOS DNA. Accumulation of 2-hydroxyethidium following exposure of cells to dihydroethidium was used as an indirect measure of superoxide production and was quantified by HPLC. A, in high BH4 conditions (no DOX), eNOS overexpression had no effect on superoxide production. C99A, W447A, and W447F mutation increased the detectable levels of 2-hydroxyethidium ∼2.5-fold ($, p < 0.05), which was only inhibitable with L-NAME in W447A but not C99A or W447F mutant cells (†, p < 0.05). DOX exposure elevated superoxide levels in untreated and mock control cells, as described previously. However, in these low-BH4 cells, levels of 2-hydroxyethidium were elevated in wild-type and mutant cells (wild type < Trp-eNOS-Phe < W447A) (*, p < 0.05; **, p < 0.01). This elevation in 2-hydroxyethidium accumulation was totally abolished following L-NAME treatment in Trp-447 but not C99A mutant cells. n = 6. B, ethidium accumulation was used as an indicator of overall oxidative stress. DOX treatment strikingly elevated the levels of ethidium within cells with no additional effect of wild-type or mutant eNOS transfection. *, p < 0.05, n = 6.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
BH4 was only oxidized to BH2 by superoxide produced from wild-type eNOS when the total levels of BH4 were low and insufficient to efficiently couple the enzyme. eNOS wild-type overexpression had no effect on BH4 homeostasis when total levels of BH4 were saturating. Intracellular BH4 was markedly oxidized to BH2 by superoxide produced from W447A and W447F eNOS mutants under conditions of both high and low levels of BH4. This oxidation of BH4 in W447A eNOS-expressing cells with high BH4 levels was comparable with that observed in WT eNOS in low BH4 conditions. In BH4-deficient cells, oxidation of BH4 and the accumulation of BH2 were further exacerbated in W447A eNOS
but not either W447F eNOS or wild-type eNOS, as revealed by the diminished ratio of BH4/BH2 in cells expressing W447A eNOS following doxycycline exposure (Fig. 5).[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
Mutant eNOS-derived superoxide exacerbates BH4 oxidation. GCH cells were cultured in the presence or absence of DOX for 7 days and then transiently transfected with wild-type eNOS, W447A, or W447F mutant eNOS DNA. Intracellular biopterins were measured using HPLC as described under “Experimental Procedures.” BH4 (A) was oxidized to BH2 (B) by mutant but not wild-type eNOS-derived superoxide in high BH4-containing cells. Following induction of BH4 deficiency with DOX, eNOS wild-type expression also triggers BH4 oxidation to BH2. C, total biopterin levels remains unaffected by eNOS overexpression. The BH4:BH2 ratio is attenuated by uncoupled eNOS (†, p < 0.05) because of DOX treatment and greatly exacerbated in W447A versus W447F eNOS overexpression (**, p < 0.01), suggesting that W447A mutation worsens enzymatic uncoupling compared with wild-type eNOS (n = 4).[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
Overall, the diminished NO production, elevated superoxide production, and the accumulation of BH2 suggest that the W447A and W447F mutants uncouple eNOS independently of total BH4 levels. However, the effects of W447A mutation are exacerbated when levels of BH4 are very low, resulting in a worsening of eNOS uncoupling.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## The eNOS Dimerization Biosensor Reveals That Trp-447 Mutation Causes Monomerization in BH4 Deficiency
*In the **The eNOS Dimerization Biosensor Reveals That Trp-447 Mutation Causes Monomerization in BH4 Deficiency** section:*
Having established that Trp-447 is critical for efficient NO production from eNOS and that its mutation initiated superoxide production rather than NO, was important to determine whether these changes in signaling of eNOS were because of changes in dimerization. To this end, we developed a novel biosensor on the basis of the reconstitution of Renilla luciferase as depicted in Fig. 6. The eNOS dimerization biosensor was created from the rluc8 construct as described under “Experimental Procedures.”
We first characterized the biosensor by Western blotting and luminometry. Following transient transfection with either the experimental eNOS-Rluc fusion constructs or a GST control biosensor, HEK293 cells were shown to express either eNOS-Rluc8.1, eNOS-Rluc8.2, GST-Rluc8.1, or GST-Rluc8.2, where appropriate, using Rluc8.1- or Rluc8.2-specific antibodies (Fig. 7A). We placed our rluc8.1 and rluc8.2 halves onto either end of eNOS. rluc8.1 was added to the C terminus (reductase domain), and rluc8.2 was added to the N terminus (oxygenase domain). This was done using two individual constructs, with eNOS tagged only at one end at a time.
Cells were then transfected with both constructs, either tagged at the N or C terminus, and luminescence was measured by protein fragment complementation assay. During the setup and initial characterization of our system, we tested the ability of luciferase to generate a detectable signal when eNOS-rluc8.1N and eNOS-rluc8.2N or eNOS-rluc8.1C and eNOS-rluc8.2C were co-overexpressed. A detectable signal was generated from both “N-” or “C-overexpressed” constructs that was significantly smaller than that obtained from eNOS-rluc8.1C and eNOS-rluc8.2N overexpression (Fig. 7B). In cells expressing both the Rluc8.1 and Rluc8.2 fusion proteins, experiments were conducted to assess the effect of BH4-treatment on eNOS dimer formation.
These data were then compared with that from the removal of intracellular zinc using N,N,N,N-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN) as a positive control for monomerization. In HEK293 cells, control GST-Rluc and eNOS-Rluc, biosensor expression, and subsequent reconstitution were detected. As an important control to rule out nonspecific reconstitution of Renilla luciferase, no luminescence was detected when eNOS-Rluc8.1 and GST-Rluc8.2 were coexpressed (or vice versa, data not shown). Following exposure to BH4 (10 μm, 4 h), a significant increase in luminescence and, hence, dimer formation was observed in eNOS-Rluc but not control or GST-Rluc-expressing cells. Removal of intracellular Zinc, required for the dimerization of both eNOS and GST, abolished the luminescent signal (Fig. 7C).
The next objective was to specifically investigate whether the Trp-447 residue within eNOS has any role in determining eNOS dimerization. The robust signal generated by eNOS-Rluc was dramatically attenuated following expression of C99A-mutant eNOS, described previously to be required for eNOS dimerization. The W447A and W447F eNOS mutants both exhibited a marked attenuation of the luminescence signal and, therefore, eNOS dimerization in the order C99A < W447A < W447F < WT-eNOS (Fig.
7D, *, p < 0.05). The ability of W447A and W447F mutation of eNOS to determine eNOS coupling and the effect of total intracellular BH4 on eNOS dimerization was next determined in GCH-Tet cells where intracellular BH4 deficiency was induced by doxycycline exposure. Mutation of eNOS-W447 in cells with saturating levels of BH4 had no effect on eNOS dimerization. In contrast, W447A and W447F overexpression in BH4-deficient cells markedly decreased eNOS dimerization in a graded manner, with W447A having a more significant effect than W447F (Fig. 8, **, p < 0.001 and *, p < 0.05, respectively).[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Novel Renilla luciferase biosensor of eNOS dimerization. We developed a novel protein-protein interaction assay on the basis of the reconstitution of split Renilla luciferase (rluc8). eNOS-separate constructs were generated to express eNOS tagged with Rluc-8.1on the N terminus and Rluc-8.2 on the C terminus. Cotransfection of these two constructs allows dimerization. Following the addition of coelenterazine substrate, the magnitude of dimerization was quantified by protein fragment complementation assay and luminescence as described under “Experimental Procedures.”[](https://www.ncbi.nlm.nih.gov/mesh/C017144)
Overexpression of W447A and W447F results in dramatic monomerization of eNOS. HEK cells were transfected with both eNOS-Rluc constructs as well as GST-Rluc control plasmids. A, characterization of plasmid overexpression was done by Western blotting using specific eNOS and Rluc antibodies. B, different rluc biosensor constructs were tested where rluc8.1 and rluc 8.2 were placed on the N- and C terminus (N-C), the C- and C terminus (C-C), or the N- and N terminus (N-N) on each eNOS construct, respectively. Un = untransfected; *, p < 0.05 versus C-C and N-N. C, transfected cells were then subjected to treatment with either BH4 or N,N,N,N-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN) and eNOS dimerization compared with control GST-overexpressing cells. Overexpression of both eNOS-Rluc8.1 and eNOS-Rluc8.2 revealed a robust luminescence signal. eNOS dimerization only, not GST dimerization, was significantly increased by BH4 treatment (†, p < 0.05, *, p < 0.01 versus control, $, p < 0.05 versus GST expressing cells). Sequestration of intracellular zinc with TPEN abolished all dimerization in eNOS- and GST-expressing cells (n = 3). D, eNOS dimerization is prevented in C99A mutants and significantly attenuated in HEK cells expressing W447A and W447F mutant eNOS. *, p < 0.01; n = 3.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
BH4 deficiency exacerbates monomerization of eNOS. GCH cells were cultured in the presence or absence of DOX for 7 days and then transiently transfected with either wild-type, W447A, or W447F mutant eNOS in the Rluc constructs, as described previously. Dimerization was again assessed by protein fragment complementation luminescence assay. Modulation of BH4 levels by DOX had no effect on eNOS dimerization. In contrast, W447A and W447F dramatically inhibited the ability of eNOS to dimerize (*, p < 0.05; **, p < 0.001). W447A eNOS dimerization was significantly lower than W447F (†, p < 0.05). C99A mutation resulted in significantly attenuated dimerization of eNOS (**, p < 0.001; N.D., none detectable with a limit of detection of 2 × 105 RLU, equal to 0.02 on the y axis. n = 4.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## DISCUSSION
*In the **DISCUSSION** section:*
In this study, we investigated the role of Trp-447 in human eNOS uncoupling and reveal for the first time that Trp-447, situated within the BH4 binding site, is critical for the enzymatic coupling of eNOS and, therefore, efficient NO production by the enzyme. Moreover, mutation of this key residue results in an attenuated interaction of BH4 with eNOS and substantial superoxide production. We also use novel biosensors to demonstrate the effect of mutation of this residue on eNOS dimerization. This study identifies a pivotal role for Trp-447 and reveals new aspects of the relationship between eNOS uncoupling and dimerization, two independent mechanisms of eNOS activity and regulation.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
The major findings of this study are as follows. First, mutation of W447A prevents eNOS dimer formation in low but not high BH4-containing cells. Second, W447A expression effectively abolishes, whereas W447F significantly attenuates, NO production from eNOS independently of BH4 levels.
These effects are exacerbated when BH4 levels are diminished. Third, these changes in eNOS activity are associated with changes in superoxide production and BH4 oxidation. Finally, discordance exists between superoxide production and the monomerization of eNOS, thus suggesting that both the eNOS monomer and dimer are significant sources of superoxide following mutation of Trp-447. Interestingly, this source of superoxide production appears to switch from eNOS dimers in replete BH4 conditions to eNOS monomers when BH4 levels are deficient. Taken together, these findings provide clear evidence to support an important catalytic role for Trp-447 in the regulation of eNOS coupling and eNOS dimerization.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
These findings provide important insights into the role of BH4 in regulating eNOS activity and eNOS coupling, which is well established to occur in a variety of in vitro, cell culture, and animal models of vascular disease. Previous studies have used models of BH4 deficiency to investigate vascular diseases such as hypercholesterolemia, hypertension, and diabetes and showed that many pathophysiological effects can be effectively rescued using BH4 supplementation either by pharmacological or genetic intervention. This is the first study to specifically investigate the impact of altering the interaction of BH4 with the eNOS active site on eNOS coupling rather than modulation of BH4 levels per se, therefore eliminating nonspecific effects of reduced BH4 levels on systems where BH4 also plays important regulatory roles in a NOS-independent manner, such as other intracellular antioxidant or redox effects.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
We show a critical role for Trp-447 in eNOS function and further advance previous work on the basis of the equivalent Trp-457 mutation in mouse iNOS where possible enzymatic uncoupling was not addressed. Structural studies of the iNOSoxy domain distinguish functional roles of BH4 that are modulated by Trp-457 (the Trp-447 equivalent in eNOS). We hypothesized that induction of eNOS uncoupling because of the mutation of W447A is due to altered BH4 binding to the enzyme. However, although W457A and W457F mutation in iNOS abolishes or inhibits NO synthesis, crystallographic structures demonstrate how W457A and W447F iNOS mutation does not result in dramatic changes in wild type position, orientation, and hydrogen bonding of BH4. However, the structural basis for changes in dimer formation and iNOS function observed result from the rearrangement of Arg-193 following W457A mutation and the ability of Arg-193 to fill the open space left by replacement of the large tryptophan with the small arginine residue.
In the W457A mutant iNOSox structure, the rearrangement of positively charged Arg-193 to form a T-shaped π-cation interaction with BH4 further stabilizes protein binding to BH4 while further destabilizing protein binding to the pterin radical that forms during catalysis. When BH4 forms the cation radical, as suggested by the eNOSox crystallographic structure and by EPR experiments, the protein binding of BH4 would be further destabilized by repulsive positive charge distribution between Arg-193 and the cofactor. Consistent with these predictions on the basis of the crystallographic structures, the rate of radical formation is further decreased or remains unchanged, whereas the rate of radical decay is further increased in W457A iNOSox relative to the W457F mutant, revealing that Trp-457 is implicated in the regulation of electron transfer during NO synthesis. Trp-457 mutants of iNOS in the presence of BH4 and l-Arg lead to reduced BH4 and l-Arg affinity and slower synthesis of NO.
Full-length W457A iNOS, as well as the W457A and W457F iNOSox mutants, all exhibit diminished rates of NO synthesis and are uncoupled with respect to enzyme NADPH oxidation. NO synthesis activities of full-length and corresponding neuronal NOS mutants W678L and W678H were less than the wild-type ones, and iNOS W457 mutations primarily slowed the rate of BH4 radical formation and sped up radical decay. Slower electron transfer from BH4 to Fe(II)O2 could uncouple NOS oxygen activation from NO synthesis if Fe(II)O2 decay is sufficiently fast in full- length NOS. In such a circumstance, superoxide release from the Fe(II)O2 intermediate would occur at the expense of NADPH oxidation and would uncouple NADPH oxidation from NO synthesis in NOS.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Because of the lack of eNOS mutant crystal structures, we assume the same electronic and structural effect of the Trp-447 mutant to be true in eNOS. For the first time, we present that Trp-447 plays an important role in dimer assembly in low BH4 conditions, more similar to neuronal NOS W678L mutation in the presence of both BH4 and l-Arg, where only 15% of the purified enzyme is dimeric and a drastically diminished production of NO is observed. As with neuronal NOS, mutation of Trp-447 did not result in totally abolished dimer formation, therefore suggesting that BH4 is not an absolute requirement for dimerization.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
The BH4 dependence of dimer formation and superoxide production by W447A and W447F mutants differs compared with wild-type eNOS. The W447A mutation is more sensitive to intracellular BH4 concentration than either W447F or wild-type eNOS, respectively. Evidence for this comes from the different magnitude of superoxide production in altered BH4 conditions. In cells containing high levels of BH4, an elevation in eNOS mutant-derived superoxide production occurs with no concurrent change in eNOS dimer formation. In contrast, DOX-induced BH4 deficiency increases superoxide to a greater extent but with a corresponding decrease (WT > W447F > W447A) in dimerization, suggesting substantial monomer-derived superoxide.
This supports previous reports that indicate that BH4 deficiency-induced eNOS uncoupling occurs simultaneously with monomerization. Studies of high glucose-treated endothelial cells and diabetic mice reveal BH4 oxidation, superoxide production with a simultaneous increase in eNOS monomer. This is in discordance with other publications that state that the eNOS dimer is required for superoxide production. The inhibition of superoxide production from the eNOS monomer by L-NAME is somewhat confusing because all superoxide would be derived from the reductase domain and would, therefore, not be expected to be inhibited with L-NAME.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
We hypothesize that there is a small amount of electron flow in the eNOS monomer that enables some degree of superoxide production from its reduced flavins. However, there is no precedent to expect that this monomer would be able to transfer electrons to its heme or would generate more superoxide per mole than a BH4-free eNOS dimer. As we report here, the superoxide produced by both W447F and W447A when BH4 levels are limiting occur along with a decrease in dimer formation, and the superoxide production became mostly inhibitable by L-NAME. We hypothesize that, in Trp-447 mutant cells in this circumstance, the bulk of the superoxide is produced from the remaining dimer (as evidenced by the background dimerization detected with our biosensor and by Western blotting in W447F mutant cells), which is likely BH4-deficient but still capable of reducing its heme and generating superoxide. This path would have significantly greater activity than flavin auto-oxidation in the monomer and, therefore, would overwhelm the small amount of superoxide produced by the monomer, despite the monomer being the major form of the eNOS
enzyme present under this condition in the cell. This explains both the increased superoxide and the ability of L-NAME to inhibit the signal. These data are supported by our findings in C99A mutants. We now provide evidence that monomeric eNOS, produced by the mutation of C99A, is capable of producing superoxide in a BH4-independent manner. The lack of an effect of L-NAME is also striking.
Production of superoxide from C99A eNOS appears to be different from that of WT or the W447 mutants.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
The studies presented here are of great significance to advance our understanding of BH4 as a regulator of eNOS coupling, affecting signaling cascades dependent on NO and reactive oxygen species, and, importantly, as a therapeutic to restore endothelial function in vascular disease. Indeed, several studies have already explored the effect of BH4 administration, either intravascularly or orally, on endothelial functions. In clinical studies, pharmacological supplementation of BH4 improves endothelium-dependent relaxations and augments NO-mediated effects on forearm blood flow in smokers and those with diabetes and elevated cholesterol. These studies have been limited to acute or short-term administration, used very high doses, and only determined the effects on endothelial-dependent relaxation rather than other variables related to vascular disease progression or risk. Indeed, numerous studies have found that pharmacologic supplementation of BH4 augments NO-mediated effects in either cell culture or in vitro vessel rings, animal models, or patients with vascular disease risk factors.
Specifically, increasing BH4 biosynthesis in cultured endothelial cells, which are relatively BH4-deficient, restores eNOS activity and increases the proportion of eNOS protein present as the homodimeric form. Gene transfer of GTPCH in carotid arteries of DOCA-salt hypertensive rats restores BH4 levels and improves endothelial function, and when GTPCH is constitutively overexpressed specifically within endothelial cells in transgenic mice, tissue BH4 levels were increased and eNOS activity was restored. In GTPCH transgenic mice rendered diabetic with streptozotocin, the loss of vascular BH4 was prevented, leading to reduced evidence of eNOS uncoupling, and restored endothelial function. When GTPCH transgenic mice were crossed with ApoE KO mice, endothelial function was improved, and atherosclerotic plaque progression was reduced. Previous clinical studies from our group show that oral BH4 treatment in patients with coronary artery disease significantly elevates BH4 levels in blood, but this effect is significantly limited by systemic oxidation of exogenous BH4 to BH2, which lacks eNOS cofactor activity.
Accordingly, the ratio of reduced to oxidized biopterins in blood and vascular tissue is unchanged by exogenous BH4 treatment, resulting in no net effect on eNOS coupling, endothelial function, or vascular superoxide production. Targeting BH4 is a rational therapeutic strategy in cardiovascular disease, but future studies should be directed toward interventions that can favorably alter the endogenous BH4/BH2 ratio and augment BH4 binding to eNOS in human vascular endothelium via a selective increase in absolute BH4 levels, prevention of BH4 oxidation, or increased BH4 recycling.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
Because of the sensitivity of BH4 to oxidation, there is therapeutic potential for non-BH4, alternative pteridine analogues that possess more stable oxidative states. These compounds may further improve the impaired relaxation associated with endothelial dysfunction and substitute for BH4 within the vasculature. Two analogues of BH4 that can act as oxidatively stable alternatives to BH4, causing NO-mediated vasorelaxation, have been developed. Treatment with 6-hydroxymethyl pterin and 6-acetyl-7,7- dimethyl-7,8-dihydropterin improved endothelium-dependent vasorelaxation in isolated perfused lungs from both normoxic and hypoxic rats and increased eNOS expression in these rats. The development of future compounds to restore NOS function in vascular disease states must consider the structural effects of BH4 binding to eNOS W447 and the role of this residues in eNOS coupling for BH4-based strategies to be successful.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
We propose that, rather than characterizing uncoupled eNOS as “dysfunctional,” uncoupling by mechanisms such as impaired BH4 binding, BH4 deficiency, or posttranslantionally by glutathione, are in fact a tightly regulated mechanism that renders eNOS as a redox “hub.” This would mean that Trp-447 is a critical residue that ultimately determines BH4 binding and BH4-dependent uncoupling, linking BH4 binding with of a plethora of targets and pathways that lie downstream of eNOS that have been demonstrated to be modulated by cellular redox state. Further studies using purified protein in vitro will give further insights into the role of Trp-447 on eNOS enzyme kinetics, electron flow, and triihydrobiopterin radical formation. Observations suggest that changes in redox status can exert a powerful influence on cellular homeostasis via eNOS, and further studies are required to elucidate the mechanisms for this redox-sensitive, downstream signaling.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
This research was supported, in whole or in part, by National Institute for Health Research (NIHR). This work was also supported by British Heart Foundation Centre of Research Excellence Intermediate Fellowship Award RE/08/004 (to M. J. C.), by RE/08/004 Programme Grant RG/12/5/29756, by The Wellcome Trust Grant 090532/Z/09/Z, and by the Oxford Biomedical Research Centre.
iNOS
inducible NOS
eNOS
endothelial NOS
BH4[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
tetrahydrobiopterin[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
BH2[](https://www.ncbi.nlm.nih.gov/mesh/C017226)
7,8-dihydrobiopterin[](https://www.ncbi.nlm.nih.gov/mesh/C017226)
eGFP
enhanced GFP
DOX[](https://www.ncbi.nlm.nih.gov/mesh/D004318)
doxycycline[](https://www.ncbi.nlm.nih.gov/mesh/D004318)
GTPCH
GTP cyclohydrolase I
l-NAME[](https://www.ncbi.nlm.nih.gov/mesh/D019331)
NG-nitro-l-arginine methyl ester.[](https://www.ncbi.nlm.nih.gov/mesh/D019331)
The abbreviations used are:
# REFERENCES
*In the **REFERENCES** section:*
|
# Introduction
Monitoring the efficacy and safety of three [artemisinin](https://www.ncbi.nlm.nih.gov/mesh/D037621) based-combinations therapies in Senegal: results from two years surveillance
# Abstract
*In the **Abstract** section:*
Background
Malaria remains a major public health problem in developing countries. Then in these countries prompt access to effective antimalarial treatment such as Artemisinin based-Combination Therapies (ACT) proves to be an essential tool for controlling the disease. In Senegal, since 2006 a nationwide sc[aling up pr](https://www.ncbi.nlm.nih.gov/mesh/D037621)ogram of ACT is being implemented. In this context it has become relevant to monitor ACT efficacy and provide recommendations for the Senegalese national malaria control program.
Methods
An open randomized trial was conducted during two malaria transmission seasons (2011 and 2012) to assess the efficacy and safety of three combinations: dihydro-artemisinin-piperaquine (DHAPQ), artemether-lumefantrine (AL) and artesunate-amodiaquine (ASAQ). The primary end point of the study was represented by a PCR adjusted adequate clinical and parasitological response (ACPR) at day 28. Secondary end points included: (i) a ACPR at days 35 and 42, (ii) a parasite and fever clearance time, (iii) ACTs safety and tolerability. The 2003 WHO’s protocol for antimalarial drug evaluation was used to assess each outcome.[](https://www.ncbi.nlm.nih.gov/mesh/C034759)
Results
Overall, 534 patients were randomized selected to receive, either ASAQ (n = 180), AL (n = 178) or DHAPQ (n = 176). The PCR adjusted ACPR at day 28 was 99.41% for the group ASAQ, while that was 100% in the AL and DHAPQ groups (p = 0.37). The therapeutic efficacy was evaluated at 99.37% in the ASAQ arm versus 100% in AL and DHAPQ arm at day 35 (p = 0.37). At day 42, the ACPR was 99.27% in the ASAQ group versus 100% for both AL and DHAPQ groups, (p = 0.36). No serious adverse event was noted during the study period. Also a similar safety profile was noted in the 3 study groups.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Conclusion
In the context of scaling up of ACTs in Senegal, ASAQ, AL and DHAPQ are highly effective and safe antimalarial drugs. However, it’s remains important to continue to monitor their efficacy.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Trial registration
PACTR 201305000552290.
## Background
*In the **Background** section:*
Despite increasing efforts to control malaria, the disease is still a public health problem. According to the World Health Organization (WHO), there are 216 million new malaria infection cases per year in the world. Also, the disease causes 655 000 deaths per year in the world. More than 81% of the new infection cases and 90% of the deaths occur in Africa mainly in children under five years [1]. The emergence and spread of the P. falciparum resistance to the monotherapies such as chloroquine, sulfadoxine, pyrimethamine, has been a major obstacle for malaria control in sub-saharan countries.
Then to deal with these resistance issues, the WHO recommended Artemisinin based-Combination Therapies (ACT) for the management of uncomplicated malaria cases [2]. ACTs reduce malaria related morbidity and mortality and the transmission of Plasmodium falciparum by acting on gametocytes and reducing the chances of development of drug resistance [3-5].[](https://www.ncbi.nlm.nih.gov/mesh/D002738)
In Senegal, the National Malaria Control Program (NMCP) initiated in 2006 a nationwide scaling up program of ACT [5]. Several ACTs are currently being used in Senegal including Artemether-lumefantrine (AL) and Artesunate-Amodiaquine (ASAQ) as first line treatment and dihydro-artemisinin-piperaquine DHAPQ (Duocotexcin*) as second line treatment [6].[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
Recent studies demonstrated a decline in ACTs efficacy as well as artesunate monotherapy in the Asian region [7,8]. This raised to some concerns related to ACT efficacy in the context of scaling up antimalarial intervention in West African countries particularly in Senegal. Then it becomes relevant to monitor ACT efficacy in Senegal.[](https://www.ncbi.nlm.nih.gov/mesh/D000077332)
In Senegal, notification of adverse events has been a great challenge for the National Malaria Control Programme although a pharmacovigilance system for monitoring ACT drug related adverse events has been established for several years [9]. Thus there is a need to document the safety profile of commonly used ACTs in Senegal when scaling up these antimalarial drugs. This study was undertaken to assess the efficacy and safety of three artemisinin combinations therapies for the treatment of uncomplicated Plasmodium falciparum malaria in Senegal.[](https://www.ncbi.nlm.nih.gov/mesh/D037621)
## Methods
*In the **Methods** section:*
## Study period and area
*In the **Study period and area** section:*
The study was carried out during two malaria transmission seasons in two health centers: (i) Deggo which is located 20 km form Dakar, the capital city and (ii) Keur Soce located 200 km of South from Dakar. In the areas around the health posts, malaria is highly seasonal during the rainy season (July to October) with a peak of transmission from September to December. Plasmodium falciparum is the predominant species and transmission is mainly due to Anopheles gambiae s.l.
The enrollment of patients started in 2011 and was completed in 2012.
## Study design
*In the **Study design** section:*
The study was designed as an open randomized trial comparing three ACTs for the treatment of uncomplicated Plasmodium falciparum malaria: Artesunate-Amodiaquine (ASAQ), Artemether-Lumefantrine (AL) and Dihydroartemisinin-Piperaquine (DHAPQ). Randomization was permuted in blocks of 9. The primary endpoint was the PCR adjusted adequate clinical and parasitological response (ACPR) at day 28. Secondary end points included: (i) PCR adjusted ACPR at day-35 and day-42, (ii) the parasite clearance time, (iii) the fever clearance time and (iv) ACT tolerability and safety.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The study was conducted as part of a national surveillance program aimed at monitoring ACTs efficacy under routine conditions.
## Study population
*In the **Study population** section:*
Subjects were enrolled if their age was above 6 months and they presented with uncomplicated Plasmodium falciparum malaria with parasite density ranged from 1000 to 100,000 trophozoites/μl. Ability to take oral medication and written informed consent were required as part of the inclusion criteria. Patients presenting with mono-infection by another species or mixed infectation, severe vomiting, severe malnutrition, severe signs of malaria (such as severe anemia, convulsion, respiratory distress), a positive pregnancy test and patients who had a history of allergy to study drugs or did not given informed consent were excluded from the study.
## Antimalarial treatment
*In the **Antimalarial treatment** section:*
After inclusion, all patients were weighed and randomized to receive one of three study drugs for three days. The drugs were administered under the direct supervision of the medical staff. In case of vomiting within the 30 minutes following the first administration, the same dose was administrated again.
Participants who vomited a second time were excluded from the study and received intravenous quinine treatment in accordance with the national malaria control program guidelines (25 mg/kg/day for seven days 3 times daily). The dosages of the study drugs were as follows:[](https://www.ncbi.nlm.nih.gov/mesh/D011803)
Artesunate-Amodiaquine (ASAQ): the tablet contains 4 mg/kg/day Artesunate (AS) plus 10 mg/kg/day Amodiaquine (AQ). The drug was given once a day. The dosage was adjusted according to the weight: one tablet per day containing 25 mg/67.5 mg (4.5 - 9 kg), one tablet per day containing 50 mg/135 mg (9 - 18 kg), one tablet per day containing 100 mg/270 mg (18 - 36 kg) and two tablets per day containing 100 mg/270 mg if weight was more than 36 kg.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Artemether-Lumefantrine (AL): the drug was not given with additional fat. The tablet contains 20 mg of Artemether plus 120 mg of Lumefantrine. The drug was given 2 times a day. The dosage was adjusted according to the weight: two tablets per day (5 – 14 kg); four tablets per day (15 – 24 kg); six tablets per day (25 – 45 kg) and eight tablets per day if weight was more than 45 kg.[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
Dihydroartemisinin-Piperaquine (DHAPQ): the tablet contains 40 mg of Dihydroartemisinin (DHA) plus 320 mg of Piperaquine (PQ). The drug was given once a day. The dosage was adjusted according to the age: 3 tablets per day if age was more than 16 years; 2 tablets per day if age was between 11 and 16 years and 1.5 tablets if the age was between 6 and 11 years.[](https://www.ncbi.nlm.nih.gov/mesh/C034759)
## Data collection
*In the **Data collection** section:*
After inclusion, patients were followed at day 0 (day of inclusion), 1, 2, 3, 7, 14, 21 and 28. A random sub-sample of study participants was followed up to day 35 and day 42 to assess the long term protective effect of each drug after curative doses. A clinical and biological assessment was performed for all patients. The 2003 WHO’s protocol for antimalarial drug efficacy evaluation was used [10].
## Clinical assessment
*In the **Clinical assessment** section:*
A clinical examination and an interview were performed before the inclusion. After inclusion and first dose administration, all patients were examined during the first 4 days. At each follow up visit, a clinical examination and interview to evaluate the patient’s clinical conditions as well as the occurrence of adverse events were done. Patients were seen by the medical team at any time if they did not feel well.
## Biological assessment
*In the **Biological assessment** section:*
A blood sample was collected for thick and thin smears for all study patients. Both tests were used to determine the parasite density and the plasmodium species at the day 0, 1, 2, 3, 7, 14, 21, and 28. Both tests were repeated at day 35, 42 and other days of follow up to evaluate parasite clearance times.
To distinguish recrudescence form new infection, blood was collected on filter paper at day 0 and at day of parasite reappearance.
The haemoglobin level was determined at day 0 and day 7 using Sysmex XS 1000i automate. Creatinine, urea, bilirubin, aspartate amino-transferase (ASAT) and alanine-aminotransferase (ALAT) were also examined at day 0 and day 7 using automate A15 of Biosystems laboratories.[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
## Laboratory methods
*In the **Laboratory methods** section:*
## Thick and thin smear
*In the **Thick and thin smear** section:*
Finger prick blood was used to collect blood samples. Thick and thin smears were stained with Giemsa. The parasite density was evaluated by counting the number of asexual parasites per 200 white blood cells and calculated per μl: number of parasites × 8000/200 assuming a white blood cell count of 8000 cells per μl. Thick and thin smears were negative after 100 field microscopics reading.
## Haematological and biochemical assessment
*In the **Haematological and biochemical assessment** section:*
Haematological and biochemical parameters were performed at enrolment and day 7 to determine the haemoglobin level, the concentration of urea, creatinine, bilirubine, asparta-amino-transferase (AST) and alanine-amino-transferase (ALAT).[](https://www.ncbi.nlm.nih.gov/mesh/D014508)
## Polymerase chain reaction (PCR)
*In the **Polymerase chain reaction (PCR)** section:*
The PCR was used to distinguish recrudescence from new infection in case of treatment failure. Nested PCR was conducted to compare the genetic polymorphism of P falciparum genes (Merozoïte Surface Protein): MSP1 and MSP2[11]. Recrudescence was defined as at least one identical allele for each of the two markers in the pre-treatment and post-treatment samples. New infections were diagnosed when all alleles for at least one of the markers differed between the two samples.
## Definition of early and late parasitological failure
*In the **Definition of early and late parasitological failure** section:*
## Early treatment failure
*In the **Early treatment failure** section:*
This was defined as a development of danger signs or severe malaria on days 1–3 in the presence of parasitaemia, a patient with parasitaemia on day 2 higher than the day 0 count irrespective of axillary temperature; parasitaemia on day 3 with axillary temperature ≥ 37.5°C and parasitaemia on day 3 that is ≥ 25% of count on day 0.
## Late parasitological failure
*In the **Late parasitological failure** section:*
This was defined as a presence of parasitaemia on any day from day 7 to day 28 and axillary temperature < 37.5°C, without previously meeting any of the criteria of early treatment failure or late clinical failure.
## Adequate clinical and parasitological response (ACPR)
*In the **Adequate clinical and parasitological response (ACPR)** section:*
The ACPR was defined as an absence of parasitaemia on day 28 irrespective of axillary temperature without previously meeting any of the criteria of early treatment failure, late clinical failure or late parasitological failure [2].
## Statistical methods
*In the **Statistical methods** section:*
Based on an expected therapeutic effect of AL not low than 95% [12] assuming a non-inferiority margin of 7% (two side) and power at 80%, using a 95% confidence level and accounting for 10% of lost to follow up, sample size for each study arm was evaluated at 155 participants.[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
Data collected were entered into Excel software and the analysis was done with Stata IC 12 software. Intention to treat and per protocol analysis were performed.
The intention to treat included all randomized subjects who took at last one full dose and had one post-baseline efficacy without major protocol deviation. The per protocol analysis include all subjects who received the three dose and had no major protocol deviation.
Data were analysed by estimation of difference in proportion according to a 95% confidence interval. Groups were compared using Chi Square test or Fisher exact test for categorical variables and Student’s t-test for continuous variables when these tests were applicable. Otherwise, non- parametric tests (Mann–Whitney, Kruskall-Wallis) were used.
The cumulative incidence of failure rate was calculated in each group and compared using Kaplan Meier method. Changes in biological parameters from day 0 to day 7 were calculated and compared between treatment arms using Bonferroni test.
Statistical significance for all tests was set at 0.05 (p value < 0.05 two side).
## Ethical considerations
*In the **Ethical considerations** section:*
This study was conducted according to the Declaration of Helsinki and existing national legal and regulatory requirements. The protocol was reviewed and approved by the Senegalese National Ethical Committee (Conseil National d’Ethique et de Recherche en Santé). Written informed consent was obtained from each participant or their parent or guardian for particiapants with an age below 18 years. The study was registered at the Pan African Clinical Trial Registry: registration number: PACTR201305000552290.
## Results
*In the **Results** section:*
## Trial profile
*In the **Trial profile** section:*
Overall, 1495 febrile patients were screened for malaria. 947 of them were positive for Plasmodium falciparum malaria. For positives subjects, 534 of them meet the inclusion criteria: 180 in the ASAQ arm, 178 in the AL arm and 176 in the DHAPQ arm. Withdrawal of consent was noted in 2 patients in the ASAQ arm, 1 patient in the AL arm and 3 patients in the DHAPQ arm. During the follow up, 7 patients in ASAQ group, 1 patient in AL group and 8 patients in DHAPQ group were lost to follow up.
Protocol violations were observed in 2 subjects in ASAQ group, 4 patients in AL group and 3 subjects in DHAPQ group. At the end of the study, 534 patients were included in ITT analysis and 503 patients in PP analysis at day 28. The PP analysis at day 35 and day 42 concerned respectively 470 and 411 patients (Figure 1).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Trial profile.
## Baseline characteristics of subjects at inclusion in the three treatment groups
*In the **Baseline characteristics of subjects at inclusion in the three treatment groups** section:*
At inclusion, the three groups were comparable in term of age, weight, sex ratio, temperature and parasitemia. The median age was 14 years in the ASAQ and DHAPQ groups and 13 years in the AL group. The sex ratio was 1.85, 1.17 and 1.41 respectively in the ASAQ, AL and DHAPQ arm. The mean weight in each group was 38.9 ± 18 kg, 37.5 ± 20 kg and 42.4 ± 19 kg for ASAQ, AL and DHAPQ, significantly higher in the last group. The mean temperature was almost similar in the three treatment groups.
The proportion of subjects with fever on admission was 75% (135/180), 75.8% (135/178) and 76.14% (134/176) respectively in ASAQ, AL and DHAPQ group.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The median parasitemia was 13132.5 trophozoites/μl in ASAQ arm, 26192.5 trophozoites/μl in AL arm and 21347.5 trophozoites/μl in DHAPQ arm (Table 1).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Baseline characteristics of subjects at the inclusion in the three treatment groups
## Biological characteristics of patients at inclusion
*In the **Biological characteristics of patients at inclusion** section:*
No significant difference was observed in the mean level of haemoglobin, creatinine, bilirubin and liver enzymes such as ASAT and ALAT in the three groups treatment (Table 1).[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
## Therapeutic efficacy
*In the **Therapeutic efficacy** section:*
There were no early treatment failures and cure rates, both PCR uncorrected and corrected, for all three treatment groups were higher than 95% by ITT and PP analysis, with no significant differences observed between the groups (Table 2).
Treatment outcomes of ASAQ, AL and DHAPQ at day 28[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The Kaplan Meier survival analysis resulted in a very similar cumulative incidence failure rate at day 28 in all three groups (log rank test, p = 0.83) (Figure 2).
Kaplan Meier survival estimates of efficacy in three treatment groups in ITT analysis at day 28.
A very low rate of late parasitological failures were detected (Table 2).
There were 88% (n = 470) and 77% (n = 411) of all patients seen at day 35 and day 42 respectively. Again very good cure rates were observed with no significant difference detected between the groups (Table 3).
Treatment outcomes of ASAQ, AL and DHAPQ at day 35 and day 42[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Fever and parasite clearance
*In the **Fever and parasite clearance** section:*
Fever clearance was similar in the three treatment groups. At inclusion, the proportion of subjects with fever was 75% (135/180), 75.8% (135/178) and 76.14% (134/176) respectively in ASAQ, AL and DHAPQ group. The difference was not significative at inclusion between the groups (p = 0.96). After first dose administration, 3.8% (7/180) patients in ASAQ group and 3.4% (6/176) patients in DHAPQ group were found with fever. The proportion of patients with fever was more important in the AL group 8.4% (15/178) with no significant difference (p = 0.06).
The fever clearance was noted at day 2 after the first dose treatment.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
The three treatments showed rapid parasite clearance time. The parasitemia at inclusion was 13132.5 trophozoites/μl in ASAQ group, 26192.5 trophozoites/μl in AL group and 21347.5 trophozoites/μl in DHAPQ group. At day 1 after first dose administration, the parasitemia decreased to 1139.8 trophozoites/μl in ASAQ group, 583.03 trophozoites/μl in AL group and 277.73 trophozoites/μl in DHAPQ group. The proportion of patients remaining parasitemic at day 1 was 75% (135/180), 66.8% (119/178) and 48.8% (86/176) respectively in ASAQ, AL and DHAPQ group (p < 10-3). At day 2, this proportion was 3.3% (6/180), 6.7% (12/178) and 2.3% (4/176).
The difference was not significative at day 2 after first dose administration (p = 0.08) between the three treatments groups. Complete parasite clearance was obtained at day 3 after inclusion.[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
## Clinical and biological tolerance
*In the **Clinical and biological tolerance** section:*
The three treatments were well tolerated during the study period. No serious adverse event was observed. No patient died during the study period and no signs of neurotoxicity were observed. The main adverse events were minor and they were represented by vomiting, abdominal pain, herpes labialis, dizziness and diarrhea.
Abdominal pains were more frequent in ASAQ group 16.66% (30/180) versus 10.79% (19/176) in DHAPQ group and 7.86% (14/178) in AL group (p = 0.31). Vomiting was more frequent in DHAPQ group 5.68% (10/176) compared to AL group 2.45% (4/178) and ASAQ group 1.66% (3/180) (p = 0.06). Labial herpes was more frequent in AL group 3.37% (6/178).[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
Biological tolerance was good during the three treatments. No major biological disorder was observed.
The mean of haemoglobin was lower at day 7 in the three treatments groups compared to hemoglobin mean at day 0. The difference between the three treatment groups at day 7 was not significative (p = 0.19). Anemia was most frequent at day 7 in three groups compared to inclusion. This was more important in AL group (70.79%). The difference was not significative between the three groups (p = 0.31).[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
An improvement of liver function was observed between day 0 and day 7. The number of patients with normal transaminase was higher at day 7 compared to the enrollment.
Regarding the mean of creatinine, no significant variation was observed between day 0 and day 7 in the three treatment arms. A significant decrease of bilirubin level was noted at day 7 in the three treatment arms (Table 4).[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
Patients’ biological profile at day 7 in the three treatment groups
At the pair analysis, the haemoglobin mean decreased from day 0 to day 7 with no statistical difference between AL and DHAPQ treatment groups (p = 1). From day 0 to day 7, this mean decreased to 1.1 g/dl versus 0.52 g/dl respectively in AL and ASAQ groups. Haemoglobin level decreased to 1.4 g/dl from day 0 to day 7 in the DHAPQ group versus 0.52 g/dl in the ASAQ group (p < 0.001). Overall the ALAT mean decreased from day 0 to day 7 in the three treatments groups. No statistical difference was noted between ASAQ (2.57 UI/L) and AL (3.65 UI/L) groups (p = 1) and between AL (16.2 UI/L) and DHAPQ (2.58 UI/L) groups (p = 1).
From day 0 to day 7, the mean ALAT decreased to 3.65 UI/L in AL group versus 2.58 UI/L in DHAPQ group (p = 0.48). The mean ASAT decreased to 6.4 UI/L in ASAQ group versus 9.5 UI/L in AL group from day 0 to day 7 (p = 0.54). Same tendency was noted in DHAPQ (8.7 UI/L) and AL (9.5 UI/L) groups from day 0 to day 7 (p = 1). There was no significant difference in ASAT level from day 0 to day 7 in ASAQ and DHAPQ arms (p = 0.94). From day 0 to day 7, the mean production of creatinin decreased to 0.18 mg/l in ASAQ group versus 0.41 mg/l in DHAPQ group (p = 1).
In AL group this mean decreased to 1.39 mg/l from day 0 to day 7 versus 0.18 mg/l in ASAQ group (p = 0.77). Regarding AL group versus DHAPQ group, the mean creatinin decreased respectively to 1.39 mg/L and 0.41 mg/l from day 0 to day 7 (p = 1).[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
Overall, the mean bilirubin decreased from day 0 to day respectively to 0.17 mg/dl, 0.91 mg/dl and 0.78 mg/dl in ASAQ, AL and DHAPQ group. The difference was significative between ASAQ and AL groups and between ASAQ and DHAPQ groups (p < 10-3). There was not significative between AL and DHAPQ groups (p = 0.46) (Table 5).[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
Biological parameter changes from day 0 to day 7 by treatment arms
Overall, the pair analysis showed an improvement of liver and kidney function from day 0 and day 7.
## Discussion
*In the **Discussion** section:*
This study was undertaken to assess the efficacy and the safety of the three artemisinin combination therapies commonly used in Senegal. The study showed that AL, DHAPQ and ASAQ are highly effective for the treatment of uncomplicated Plasmodium falciparum malaria. These findings are consistent with results reported from other trials. In Senegal, Tine et al. in 2010 reported a cure rate of 96.7% at day 28 for AL in a clinical trial assessing the efficacy and tolerability of new formulation of Artesunate-Mefloquine [13]. Faye et al. obtained a cure rate of 97% for AL in a multicentric study (Senegal and Ivory Coast) from September 2007 to November 2008
[14].[](https://www.ncbi.nlm.nih.gov/mesh/D037621)
Menan et al. from, in a multicentric study including Cameroon, Ivory Coast and Senegal September 2008 to February 2009, obtained a PCR adjusted ACPR at day 28 of 99% in AL group [15]. Makanga et al. obtained a therapeutic efficacy for AL above 95% for the treatment of uncomplicated malaria [16]. Regarding ASAQ combination, Faye et al. in 2008 in a multicentric study (Senegal, Cameroon and Ivory Coast) evaluating the non-inferiority of the new paediatric formulation of Artesunate/Amodiaquine versus Artemether/Lumefantrine for malaria treatment reported a therapeutic efficacy at day 28 at 98.7% when ASAQ was given to children under 5 years of age [17].[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
Ndiaye et al., in randomized trial assessing the efficacy of a fixed dose of ASAQ combination from March to December 2006, showed an adjusted ACPR at day 28 more than 98% [18].[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
NDounga et al. in 2005 in Brazaville, obtained 94.4% of cure rate at day 28 when ASAQ combination was given to children [19]. Zwang et al. in a multi-centre analysis of the efficacy of ASAQ combination showed good ACPR after PCR correction. For DHAPQ combination similar results were obtained [20]. Yavo et al. from November 2006 to May 2008 in multi-centre study assessing the efficacy of DHAPQ obtained 99.5% of ACPR after PCR correction
[21].[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
In Thaïland, Ashley et al. between July 2002 and April 2003 showed a cure rate of 98.3% in the DHAPQ group [22]. Many studies have reported good efficacy of DHAPQ combination for malaria treatment [23-25]. The three antimalarial drugs remained highly effective at days 35 and 42. Compared to others drugs, DHAPQ has good cure rate more than 98%.
This result was demonstrated by Grande et al. and by Zwang et al[26,27].[](https://www.ncbi.nlm.nih.gov/mesh/C034759)
After the administration of the three ACT used in this study, it resulted in a rapid decrease of fever and parasite. The fever and parasite clearance was obtained respectively at day 2 and day 3 after the initial dose. Similar results were obtained by Tine et al., Faye et al. in Senegal and Gbotosho et al. in Nigeria [13,14,28].
The three antimalarial treatments were well tolerated with a similar safety profile. Indeed, no serious adverse event was noted. The main adverse events were minor and they were represented by vomiting, abdominal pain, herpes labialis, dizziness and diarrhea. However abdominal pains were more frequent in patients treated with ASAQ, vomiting was more frequent in patients treated with DHAPQ and labial herpes was more frequent in AL group. Previous study demonstrated that these ACTs are well tolerated [13-15].[](https://www.ncbi.nlm.nih.gov/mesh/C515299)
No major biological disorder was observed in our study. Anemia was most frequent at day 7 in three groups compared at day 0. Similar results were noted by others study. Olliaro et al. observed a decrease of haemoglobin level at day 7 after malaria treatment [29].
Same trends were noted by Price et al. and by Zwang et al.[30,31].
In Senegal, notification of adverse events has been a great challenge for the National Malaria Control Programme although a pharmacovigilance system to monitor ACT drug related adverse events has been established since 2006 [9]. This study provided scientific evidence that can contribute to supplement existing data regarding on ACT safety in Senegal.
## Study limitation
*In the **Study limitation** section:*
AL was not given with fat diet in our study. This could explain the decrease of the efficacy. However many studies have reported good efficacy when AL was given with fat. Thus Ashley et al. in pharmacokinetic study of AL in 2002 showed that the high-fat allowed having a good absorption of the drug. This had resulted in an increase of therapeutic efficacy
[32].[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
Mayxay M et al., in the efficacy study of AL in Southern Laos between June and November from 2008–2010 showed good efficacy of AL with a cure rate more than 95% when AL was given with fatty food [33].[](https://www.ncbi.nlm.nih.gov/mesh/D000077611)
Many studies reported an increase of haemoglobin level from day 7 to 28 after treatment with ACT [20,31]. In this study haemoglobin level was assessed at day 0 and day 7. Additional haemoglobin dosages at day 14, 21 and 28 would provide more information on haemoglobine changes after treatment with ACTs; theses assessments were not done in the study.
## Conclusions
*In the **Conclusions** section:*
DHAPQ, AL and ASAQ are highly effective and safe antimalarial drugs. These ACTs remain useful antimalarial interventions for effective malaria control. It is however important to continue to monitor their efficacy in the context of scaling up of ACTs in Senegal.[](https://www.ncbi.nlm.nih.gov/mesh/C034759)
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interest.
## Authors’ contributions
*In the **Authors’ contributions** section:*
KS, RCT, BF, DS, JLN, OG conceived and designed the study. KS and MN monitored the data collection. KF and LAN collected data in the site. KS and RCT analysed the data. MN, ACL, AA were responsible for the PCR analysis.
KS and AA wrote the first draft of the manuscript. All authors read and approved the final manuscript.
## Pre-publication history
*In the **Pre-publication history** section:*
The pre-publication history for this paper can be accessed here:
http://www.biomedcentral.com/1471-2334/13/598/prepub
## Acknowledgment
*In the **Acknowledgment** section:*
We thank all patients who agreed to participate in the study. We also acknowledge support from the national malaria control program, which provided us funding for the study and gave us all the drugs used in this study.
|
# Introduction
Marker-Assisted Selection for Recognizing Wheat Mutant Genotypes Carrying HMW Glutenin Alleles Related to Baking Quality
# Abstract
*In the **Abstract** section:*
Allelic diversity of HMW glutenin loci in several studies revealed that allelic combinations affect dough quality. Dx5 + Dy10 subunits are related to good baking quality and Dx2 + Dy12 are related to undesirable baking quality. One of the most regular methods to evaluate the baking quality is SDS-PAGE which is used to improve baking quality labs. Marker-assisted selection is the method which can recognize the al[lel](https://www.ncbi.nlm.nih.gov/mesh/D012967)es related to baking quality and this method is based on polymerase chain reaction. 10 pairs of specific primers related to Dx2, Dx2.1, Dx5, Dy10, and Dy12 subunits were used for recognizing baking quality of some wheat varieties and some mutant genotypes. Only 5 pairs of them could show the specific bands. All subunits were recognized by the primers except Dx2.1. Some of the primers were extracted from previous studies and the others were designed based on D genome subunits of wheat. SDS-PAGE method accomplished having confidence in these marker's results. To realize the effect of mutation, seed storage[ pr](https://www.ncbi.nlm.nih.gov/mesh/D012967)oteins were measured. It showed that mutation had effect on the amount of seed storage protein on the mutant seeds (which showed polymorphism).
## 1. Introduction
*In the **1. Introduction** section:*
Bread wheat (Triticum aestivum) is a kind of allohexaploid which has A, B, and D genomes. High molecular weight glutenin (HMW-Gs) subunits are important combinations of wheat glutenin proteins which play an important role in viscosity and elasticity of wheat dough [1]. Nullisomic, tetrasomic, nulli-tetrasomic, and ditelocentric series lines showed that high molecular weight glutenin are controlled by the loci on long arm of 1A, 1B, and 1D chromosomes near centrometer (recombination index = 9%) [2]. Each locus has two linked genes which are named x and y type with different molecular weight [3].
Bread wheat contains six HMW glutenin subunit-coding genes (two x type and two y type), but silencing of specific genes causes the creation of only three to five HMW protein subunits [4]. The y genes in hexaploid wheat A genome are silent but they are active in several diploid and tetraploid wheat [5].
Glu-A1 locus is distinguished by three main x type alleles which are coded 1, 2*, and null subunits [6]. In several bread wheat and durum wheat, there is Bx7 allele in Glu-B1 locus commonly [7]. The subunits Ax1 and Ax2* HMW glutenin alleles are in Glu-A1, Bx17 + By18, Bx7 + By8, or By9 are in Glu-1B locus, and Dx5 + Dy10 are in Glu-1D locus, related to baking quality, on the other hand, AxNull, Bx6 + By8, and Dx2 + Dy12 are related to undesirable baking quality [5]. SDS-PAGE is one of the most regular techniques which is used in wheat improvement labs to recognize different allelic forms related to baking quality [8].
Wheat breeders selected different variety for baking quality by using SDS tests, Zeleny sedimentation, micrograph assessment, and other methods. These procedures are not useful in initial hybrids because these methods need a large amount of seed, also these methods destroy experimental samples, but improved PCR based on HMW glutenin subunit markers showed baking quality in initial segregation hybrids [9].[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
Glu-1 loci sequences are recognized by Gu et al. [10], Kong et al. [11], Anderson and Greene [12], Halford et al. [4], and Wan et al. [13].
Many researchers studied different HMW glutenin subunits in wheat cultivars by analysis based on PCR [8, 14–19]. Moczulski and Salmanowicz [16] reported that rapid recognition of Glu-1 genes molecular markers by using multiplex PCR could be standard and effective to select initial useful wheat genotypes.
DNA markers are easy and rapid way for selection. These markers could be replaced as standard and creditable techniques to select genotypes which have HMW glutenin subunits related to baking quality [20].
The goal of this study was recognition of wheat genotypes carrying glutenin allele combinations by using specific primers and PCR methods, studying about the effect of mutation on seed storage proteins, and selecting these mutant genotypes to continue breeding programs.
## 2. Materials and Methods
*In the **2. Materials and Methods** section:*
## 2.1. Plant Material
*In the **2.1. Plant Material** section:*
48 genotypes (mutant lines: Roshan, Tabasi, Azar, Azadi, and Omid with their parents, also some other cultivars like Chamran, Chinese spring and so on) were used. These mutant lines only were supplied.
Tabasi and the mutants of Tabasi: T-66-58-6, T-65-9-IP, T-65-5-1, T-65-6, T-66-58-9, T-66-58-12, T-66-58-10, T-65-9-1, T-65-4, T-58-8, T-67-7-1, T-66-I-II, T-58-7, T-67-60, T-58-14, T-65-9-II-4, and T-66-58-6; Omid and the mutants of Omid: O-64-4, O-64-1-1, and O-6-1-1; Roshan and the mutants of Roshan: Ro-1, Ro-2, Ro-3, Ro-4, Ro-5, Ro-6, Ro-7, Ro-8, Ro-9, Ro-10, Ro-11, and Ro-12; Azadi and the mutants of Azadi: As-48, Azar, Azar mutant, Chamran, Tajan, Tajan-e-garm, Atrak, Navid, Inia, Bezostaya, Pishtaz, and Chinese spring.
## 2.2. DNA Extraction and Polymerase Chain Reaction (PCR)
*In the **2.2. DNA Extraction and Polymerase Chain Reaction (PCR)** section:*
Genomic DNA of each genotype was extracted from young leaf by CTAB procedure [21]. Quantity and quality of genomic DNA were measured by spectrophotometer in 260 nm and 1% Agarose gel. Polymerase chain reaction was accomplished based on sequence tagged sites (STS) primers which were specified for alleles related to HMW glutenin alleles. The sequences of primers exist in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/C004322)
Reaction combinations in 25 μL were 1X PCR buffer, 2 mM MgCl2, 5 mM dNTPs, 10 mM of each primer, 1.5 U Taq DNA polymerase, 250 ng of genomic DNA, and ddH2O (up to 25 μL), also the PCR program was 94°C for 2′ (initial denaturing) and 30 cycles: 94°C for 1′ (denaturing), 58–66°C for 1′ (annealing), and 70–72°C for 1′ (extension) and the final extension was 72°C for 6′ (Table 2). Amplified products separated by agarose gel 1.7% and staining were by Ethidium Bromide.[](https://www.ncbi.nlm.nih.gov/mesh/D015636)
## 2.3. SDS-PAGE Method
*In the **2.3. SDS-PAGE Method** section:*
The HMW glutenin subunits were extracted from the seeds of each accession without embryo, based on modified Laemmli [23], then separated by electrophoresis in polyacrylamide gel (main gel was 10% and stacking gel was 4%). After electrophoresis, the gels were stained by Coomassie Brilliant Blue R250 for 24 hours; the gels were left in TCA 10% for 3-4 hours, then in distilled water for 12 hours. Finally, the gels scanned and analyzed.[](https://www.ncbi.nlm.nih.gov/mesh/C016679)
## 3. Results and Discussion
*In the **3. Results and Discussion** section:*
## 3.1. PCR and SDS-PAGE
*In the **3.1. PCR and SDS-PAGE** section:*
By P1P2, the genotypes, which have Dy10 and Dy12, showed the band 650 bp. All of the genotypes revealed this band. In addition, two other bands were observed. The first one was 350 bp, only Chamran showed this band, and the other one was 600 bp in Tabasi, Omid, Roshan, and some of the other mutants. The 600 bp band was reported by Smith et al.
[22], but 350 bp band was not reported by any researchers (Figure 1). T-66-58-12 genotype is similar to Azadi (both of them have only 650 bp band), but Tabasi and Omid have both 650 bp and 600 bp bands. These results confirm the SDS-PAGE result. For example, Tabasi and Omid have similar banding pattern in STS and also in SDS-PAGE method.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
The P3P4 primers amplified four kinds of bands in which three of them were reported by Smith et al. [22]. The first one was 576 bp for Dy10 and the second one was 612 bp that belongs to Dy12. These bands according to the sequences of wheat are specific. In addition, Ahmad [8] used these primers and reported two bands which belong to Dy10 and Dy12.
Like P1P2, all of the genotypes had one of these alleles (Dy10 or Dy12) also, SDS-PAGE results confirm the result of P1P2 and P3P4 (Figure 2). The third band was 675 bp which belongs to By9; all of the genotypes showed this allele except three mutants (Ro-9, T-65-9-1P, and T-65-9-1). The last band was only in Azar mutant, this band was not reported by Smith et al. [7] and this has occurred because of the mutation.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
P5P6 primers were for identifying Dx5 [8]. Only the genotypes that have Dx5 showed the band 450 bp. In this study, 11 genotypes showed this band. Five genotypes were Navid, Atrak, Tajan, Bezostaya, and Chamran, and the other genotypes were the mutants of Tabasi, Omid, and Roshan (Figure 3).
Because nonmutant genotypes did not show this band, maybe mutation was the cause of this event. Ahmad [8] used these primers to separate the genotypes that had Dx5 allele from the genotypes which had Dx2 allele. In SDS-PAGE method, because of closeness of Dx2 and Dx5, the recognition is very hard. By PCR method and using these specific primers, the recognition of these alleles is so easy.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
P7P8 primers belong to Dx5 like P5P6. According to De Bustos et al. [15], these primers amplify a band in 2775 bp, but in this study there were two bands (550 and 1350 bp) for all genotypes. Therefore these primers cannot be used for marker-assisted selection, because of the absence of specific and polymorphic band.
By using P9P10, primers that amplify Dy10 allele, one single band occurred in eleven genotypes. Five of these genotypes were the controller cultivars (Navid, Atrak, Tajan, Bezostaya, and Chamran) and the others were the mutants of Tabasi, Omid, and Roshan. Therefore, by using these primers, the Dy10 alleles can be recognized.
P11P12 primers were for recognition of Dx2. The genotypes, which had this allele, showed single 2799 bp band. All of the genotypes except eleven (which had Dy10 allele) showed this band.
P13P14 primers were the last primers, which were extracted from the study of De Bustos et al. [15]. They reported a 2190 bp band that belongs to Dy12. However, this band did not appear in spite of changes like combination of PCR materials, PCR cycle, and so forth.
P15P16, P17P18, and P19P20 primers which were for amplifying Dx5 (2520 bp), Dx2.1 (651 bp), and Dy10 (1947 bp) respectively, like the previous primers did not show any specific band appearing in spite of all the changes. Therefore, these primers cannot be used for selection by using marker.
## 3.2. Protein Percentage
*In the **3.2. Protein Percentage** section:*
Comparing of protein mean with Duncan's multiple-range test showed that there is a significant difference between some mutants and their origin genotypes (Table 3). The Roshan mutant genotypes, which are named Ro-1 and Ro-5, had significantly lower protein percentage in comparison with Roshan, on the other hand, Ro-3 had higher protein percentage than Roshan. There was not any significant difference of protein percentage between Tabasi and Omid with their mutants.
Usually the movement of HMW glutenin subunits in SDS-PAGE method does not have a complete relationship with its molecular weight and this event may cause trouble for breeders to select the lines, as Shewry et al. [7] reported. Goldsbrough et al. [24] reported that a significant difference exists between molecular weight and movement of HMW subunits in SDS-PAGE method. These alleles can be recognized by comparing these HMW glutenin subunits in SDS-PAGE method.
Marker-assisted selection is an effective and reliable method which is based on the gene sequence and polymerase chain reaction.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
Glutenins are combinations of high molecular weight subunits (HMW) and low molecular weight (LMW) which are named according to the migration ratio in gel [25]. Allelic diversity in each glutenin locus and several studies showed allelic combinations affected dough quality [5]. HMW glutenin subunits 5 + 10 were related to high elasticity and good baking quality; however, 2 + 12 subunits were related to low baking quality [26]. 31% of smooth wheat cultivars have 5 + 10 subunits and many of breeders reported the existence of 5 + 10 in smooth wheat lines
[27].
In this study, ten pairs of specific primers were used for recognition of Dx2, Dx2.1, Dx5, Dy10, and Dy12 but only five pair of them showed specific bands. All of the alleles were recognized except Dx2.1, Dy10, and Dy12 were recognized by P1P2 and P3P4 primers. By using P1P2 and P3P4 primers, Smith et al. [22] and Ahmad [8] identified the genotypes that carried Dy10 and Dy12. The results showed all the genotypes had Dy10 or Dy12 allele.
P1P2 primers could not separate Dy10. Ahmad [8] reported the coding genes of Dy10 and Dy12 had 98.9% similarity, he found three regions for designing of primers and these three regions were Dy10, Dy12, and Dy9.
The results from amplification of fragments showed a high linkage between Dy12 and Dy10.
Liu et al. [19] amplified the genes related to Glu D1 (x2, x5, y10, and y12) alleles by codominant markers. Also Schwarz et al. [28] used SNP markers for recognition of Dx2 and Dx5 subunits.
HMW glutenin subunits were related to elasticity and baking quality and a large amount of diversity between cultivars is because of HMW glutenin subunits [25]. The sequence of Glu-A1x2* and Glu-D1x5 genes was analyzed and compared by Anderson and Greene [12]. These results showed a high similarity in their structures, therefore separating of the alleles by SDS-PAGE method is hard. In addition, there was a complete conformity between the results of HMW glutenin allelic combination in this study, also D'Ovidio and Anderson
[18] reported a complete conformity between PCR and allelic combination of wheat samples in SDS-PAGE method.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
High genetic linkage between x and y subunits causes trouble to recognize of subunits that are related to baking quality [18].
## 4. Conclusion
*In the **4. Conclusion** section:*
All of the studies in this field emphasize on the importance of this kind of marker to recognize the gene which affects dough characteristics. The method needs only a few amount of leaf DNA and to prevent the injury (by separating of a piece of endosperm) to embryo [14]. By the method (marker-assisted selection), selecting and studying about the plants can be possible in all of the life cycle of plant.
By marker-assisted selection, the recognition between DNA samples which carried alleles related to baking quality is simple and the results could be achieved in about three hours. The method omits using dangerous material like Acrylamide and this method is more reliable. Because of the closeness of x2 and x5 protein coding molecule weight and movement of HMW glutenin subunits in gel some miss scoring occurred. On the other hand, some mutants in x5 cannot be recognized by SDS-PAGE method, therefore marker-assisted selection is a reliable way to select genotypes. Another benefit to this method is that by using any part of the plant (leaf, root, seed, and stem) this selection can be done.
Using multiplex, PCR (amplifying some alleles at the same time) can decrease the payment [17]. By marker-assisted selection, more then 1000 plants can be selected and evaluated at the first stage of plant growth by breeders [8]. Using of diploid part of a plant in comparison to endosperm (triploid) is easier and faster in commenting of the results in the heterozygous plants [22]. Finally, the last benefit of this marker is that the marker is not affected by internal or external environment.[](https://www.ncbi.nlm.nih.gov/mesh/D020106)
## Conflict of Interests
*In the **Conflict of Interests** section:*
The authors declare that there is no conflict of interests regarding the publication of this paper.
|
# Introduction
Gla-Rich Protein Is a Potential New [Vitamin K](https://www.ncbi.nlm.nih.gov/mesh/D014812) Target in Cancer: Evidences for a Direct GRP-Mineral Interaction
# Abstract
*In the **Abstract** section:*
Gla-rich protein (GRP) was described in sturgeon as a new vitamin-K-dependent protein (VKDP) with a high density of Gla residues and associated with ectopic calcifications in humans. Although VKDPs function has been related wit[h γ](https://www.ncbi.nlm.nih.gov/mesh/D015055)-carboxylation, the Gla status of GRP in humans is still unknown. Here, we investigated the expression of recently identified GRP s[pli](https://www.ncbi.nlm.nih.gov/mesh/D015055)ced transcripts, the γ-carboxylation status, and its association with ectopic calcifications, in skin basal cell and breast carcinomas. GRP-F1 was identified as the predominant splice variant expressed in healthy and cancer tissues. Patterns of γ-carboxylated GRP (cGRP)/undercarboxylated GRP (ucGRP) accumulation in healthy and cancer tissues were determined by immunohistochemistry, using newly developed conformation-specific antibodies. Both GRP protein forms were found colocalized in healthy tissues, while ucGRP was the predominant form associated with tumor cells. Both cGRP and ucGRP found at sites of microcalcifications were shown to have in vitro calcium mineral-binding capacity. The decreased levels of cGRP and predominance of ucGRP in tumor cells suggest[ that G](https://www.ncbi.nlm.nih.gov/mesh/D002118)RP may represent a new target for the anticancer potential of vitamin K. Also, the direct interaction of cGRP and ucGRP with BCP crystals provides a possible mechanism expla[ining GRP](https://www.ncbi.nlm.nih.gov/mesh/D014812) association with pathological mineralization.[](https://www.ncbi.nlm.nih.gov/mesh/C020243)
## 1. Introduction
*In the **1. Introduction** section:*
Gla-rich protein (GRP), also known as cartilage matrix associated protein or upper zone of growth plate and cartilage matrix associated protein (UCMA), was identified in sturgeon as a new vitamin-K-dependent protein (VKDP), exhibiting an unprecedented high density of Gla residues (16 Gla residues among 74 amino acids) and a high affinity for calcium mineral. Highly conserved GRP orthologs presented conserved features specific to all VKDPs, in particular a remarkably well conserved Gla domain, thus suggesting GRP to be a new vertebrate-specific γ-carboxylated protein. While in sturgeon GRP was predominantly found in cartilaginous tissues, in mammals it was shown to have a wider tissue distribution and to accumulate both in skeletal and connective tissues including bone, cartilage, skin, and vasculature. GRP was found to be a circulating protein and to be associated with calcifying pathologies affecting skin and arteries, where it accumulates at sites of ectopic calcifications and colocalizes with calcium mineral deposits. Although the function of GRP is still unknown, it has been suggested to act as a negative regulator of osteogenic differentiation, a modulator of calcium availability in the extracellular matrix, and as a potential inhibitor of soft tissue calcification in connective tissues.
In concordance, recent functional studies pointed to an essential role of GRP in zebrafish skeletal development and calcification, albeit GRP-deficient mice did not reveal evident phenotypic alterations in skeletal architecture, development, or calcification. While four alternatively spliced transcripts of the GRP gene (GRP-F1, -F2, -F3, and F4) were described in mouse chondrocytes and zebrafish, new alternatively spliced transcripts were recently [ide](https://www.ncbi.nlm.nih.gov/mesh/D015055)ntified in hum[ans](https://www.ncbi.nlm.nih.gov/mesh/D015055). Besides GRP-F1, t[he new vari](https://www.ncbi.nlm.nih.gov/mesh/D000596)ants GRP-F5 and GRP-F6 wer[e chara](https://www.ncbi.nlm.nih.gov/mesh/D002118)cterized by the loss of full γ-carboxylation and partial secretion functional motifs, due to deletion of exon 3 in F5 and exons 2 and 3[ in](https://www.ncbi.nlm.nih.gov/mesh/D015055) F6.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Although the precise function of the novel human GRP variants and the importance of their γ-carboxylation need to be further addressed, the perception that crucial differences exist between mouse and human GRP highlights the need for additional characterization of GRP expression/accumulation patterns in health and disease. Considering VKDPs for which the function is known, γ-carboxylation was shown to be required for biological activity, and under- or uncarboxylated species are generally regarded as proteins with low or no functional activity. Several factors, such as insufficient dietary intake of vitamin K, mutations in the γ-glutamyl carboxylase enzyme (GGCX), and warfarin treatment, result in decreased γ-carboxylation of VKDPs, which has been associated with an increased risk for osteoporotic bone loss as well as for arterial and skin calcifications. Microcalcifications are also often associated with different types of cancer and are considered as a hallmark for early detection of breast cancer, with recognized prognostic relevance. Accumulating evidence suggests that mineralization occurring in breast cancer is a cell-specific regulated process sharing molecular mechanisms involved in arterial pathological mineralization and physiological mineralization in bone.
Several reports have described an association of VKDPs with different types of cancer, namely, matrix Gla protein (MGP) and uncarboxylated prothrombin (des-γ-carboxy prothrombin, DCP). Although a relation between MGP γ-carboxylation status and neoplasias is still unknown, for prothrombin it was shown that increased circulating DCP can be used as a diagnostic and prognostic marker for hepatocellular carcinoma. It is remarkable that, despite the widely reported vitamin K anticancer potential [, and references therein], its mechanism of action in cancerous processes remains elusive. Therefore, the identification of new vitamin K targets in cancer, such as GRP, may contribute to unveil the role and functional mechanism of vitamin K in cancer development.[](https://www.ncbi.nlm.nih.gov/mesh/D014812)
Here, we report on the association of GRP in human skin and breast cancers with the specific accumulation pattern of carboxylated (cGRP) and undercarboxylated (ucGRP) protein forms. In these studies we used newly developed conformation-specific antibodies against ucGRP and cGRP. We investigated the association and γ-carboxylation status of GRP with microcalcifications occurring in skin and breast cancers and tested the mineral binding capacity of both protein forms, which may help understanding the mechanism behind the previously reported association of GRP with pathological mineralization.
## 2. Materials and Methods
*In the **2. Materials and Methods** section:*
## 2.1. Ethics Statement
*In the **2.1. Ethics Statement** section:*
This study complies with the guidelines for good clinical practice and was performed in accordance with the Declaration of Helsinki and approved by the ethics committees of all the hospitals and institutions involved, namely Algarve Medical Centre, Lisbon Central Hospital, HPP-Santa Maria Hospital, and National Institute of Legal Medicine and Forensic Sciences, Public Institute. Written informed consent was obtained from all participants.
## 2.2. Biological Material
*In the **2.2. Biological Material** section:*
Control breast (mammary gland, MG) and skin (Sk) tissues were obtained from three healthy volunteers at the time of esthetic surgeries. Nontumorous areas, adjacent to tumors, present in four skin and four breast cancer patient samples, were also used as controls in immunohistochemical (IHC) staining experiments. Samples of malignant breast lesions were retrieved from patients who had undergone breast surgery at the Algarve Medical Centre. Skin biopsies of malignant lesions were taken under local anesthesia at Lisbon Central Hospital. A total of eleven tissue samples from patients diagnosed with basal cell carcinoma (BCC, Table 1) and eleven from patients diagnosed with invasive ductal carcinoma (IDC, Table 2) were studied.
IHC was performed in eight BCC and seven IDC samples, and gene expression analysis was performed in five BCC and four IDC samples.
## 2.3. Sample Processing
*In the **2.3. Sample Processing** section:*
Tissue samples were embedded in paraffin at the Pathology Departments of Algarve Medical Centre and Lisbon Central Hospital and histologically classified by pathologists. Physiological structures were identified by regular haematoxylin-eosin staining and mineral deposits were detected with silver nitrate (Sigma-Aldrich) by the von Kossa method. Samples used in gene expression studies were collected into RNAlater (Sigma-Aldrich) immediately after surgery.[](https://www.ncbi.nlm.nih.gov/mesh/D010232)
## 2.4. RNA Extraction
*In the **2.4. RNA Extraction** section:*
Total RNA was extracted from Sk, MG, BCC, and IDC tissues as described by Chomczynski and Sacchi. RNA integrity was evaluated by agarose-formaldehyde gel electrophoresis and concentration determined by spectrophotometric analysis at 260 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D012685)
## 2.5. Gene Expression
*In the **2.5. Gene Expression** section:*
One microgram of total RNA was treated with RQ1 RNase-free DNase (Promega) and reverse-transcribed at 37°C with MMLV-RT (Invitrogen) using a dT adapter. PCR amplifications for GRP-F1, -F5, and -F6 splice variants were performed with SsoFast EvaGreen Supermix (BioRad) for 50 cycles and specific primer sets A/B, C/D, and C/E, respectively. Ribosomal 18S was used as loading control. A list of all PCR primer sequences is presented in Table 3.
## 2.6. Quantitative Real-Time PCR (qPCR)
*In the **2.6. Quantitative Real-Time PCR (qPCR)** section:*
Quantitative PCR was performed with an iCycler iQ apparatus (Bio-Rad) using 25 ng cDNA and the conditions described above. In addition to GRP-F1, -F5, -F6 and 18S, MGP, GGCX, VKOR (vitamin K epoxide reductase), OPN (osteopontin), TNFα (tumor necrosis factor alpha), and GAPDH were amplified using primer sets as described in Table 3. Fluorescence was measured at the end of each extension cycle in the FAM-490 channel and melting profiles of each reaction were performed to check for unspecific product amplification. Levels of gene expression were calculated using the comparative method (ddCt) and normalized using gene expression levels of both GAPDH and 18S housekeeping genes, with the iQ5 software (BioRad); qPCR was performed in duplicates and a normalized SD was calculated.
## 2.7. Conformation-Specific Antibodies against Carboxylated (cGRP) and Undercarboxylated (ucGRP) GRP Protein Forms
*In the **2.7. Conformation-Specific Antibodies against Carboxylated (cGRP) and Undercarboxylated (ucGRP) GRP Protein Forms** section:*
Affinity-purified chicken polyclonal antibody against cGRP (cGRP pAb) (GenoGla Diagnostics, Faro, Portugal) was produced by immunizing chickens with a synthetic peptide corresponding to a γ-carboxylated region of the human GRP Gla-domain located within exon 4 (aa 29-42: QRNEFENFVEEQND, in which all E are Gla-residues and termed cGRP29-42, Figure 1). An equivalent, but noncarboxylated peptide (aa 29-42, where all E are Glu residues), was termed ucGRP29-42 (Figure 1). The conformation-specific affinity-purified antibody was obtained by passing the chicken serum through an ucGRP29-42 affinity column followed by immunopurification of the flow-through on a cGRP29-42 column.[](https://www.ncbi.nlm.nih.gov/mesh/D015055)
Monoclonal antibody against ucGRP (ucGRP mAb) (VitaK BV, Maastricht, The Netherlands) was raised against an epitope of human GRP located within exons 4 and 5 (aa 31-54: NEFENFVEEQNDEQEERSREAVEQ), in which all E are Glu residues and termed ucGRP31-54 (Figure 1). An equivalent, but carboxylated peptide (aa 31-54, where all E are Gla residues), was termed cGRP31-54 (Figure 1). The conformation-specific ucGRP mAb was raised in BALB/c mice and postimmune sera were screened for their conformational affinity toward synthetic c and ucGRP31-54 peptides. After a satisfying antibody titer was reached, splenocytes were fused with a mouse myeloma cell line (Sp 2/01-Ag, CRL 8006, ATCC). Clones strongly reacting with ucGRP31-54 and nonreactive with cGRP31-54 were selected, and the antibodies produced were purified by protein G affinity chromatography.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
CTerm-GRP polyclonal antibody (GenoGla Diagnostics) detecting total GRP was produced against a synthetic peptide corresponding to the C-terminus of rat GRP, following a previously described procedure.
## 2.8. Immunohistochemistry
*In the **2.8. Immunohistochemistry** section:*
Immunohistochemical staining was performed on paraffin-embedded tissue sections as described elsewhere. Briefly, endogenous peroxidase activity was blocked with 3% (v/v) H2O2 in TBST buffer (TBST: 0.1 mol/L Tris, 0.15 mol/L NaCl, 0.1% (v/v) Triton X-100) for 15 min. Nonspecific antibody binding was blocked with TBT (0.5% (w/v) BSA in TBST) for 1 h at 37°C. Incubation with CTerm-GRP, cGRP pAb, and ucGRP mAb (5, 1, and, 1 μg/mL diluted in TBT, resp.) was performed overnight (O/N) in a humidified chamber at room temperature (RT). Blocking assays were performed by incubating cGRP pAb (1 μg/mL) with 10-6 M of cGRP29-42 or ucGRP29-42 peptides and ucGRP mAb (1 μg/mL) with 10−6 M of ucGRP31-54 or cGRP31-54 peptides, for 2 h at RT prior to tissue incubation. Primary antibodies were detected using species specific HRP-conjugated secondary antibodies (Sigma-Aldrich) and 0.025% (w/v) 3,3-diaminobenzidine (Sigma-Aldrich).
Negative controls consisted in the substitution of primary antibody with TBT. Counterstaining was performed with haematoxylin. Microscopic images were acquired in a Zeiss AXIOIMAGER Z2 microscope, with an AxioCam ICc3 camera and AxioVision software version 4.8 (Carl Zeiss), at the light microscopy facility, Department of Biomedical Sciences and Medicine, University of Algarve (Portugal).[](https://www.ncbi.nlm.nih.gov/mesh/D010232)
## 2.9. Cloning of hGRP-F1 into pET151 Expression Vector
*In the **2.9. Cloning of hGRP-F1 into pET151 Expression Vector** section:*
The complete open reading frame (ORF) of the human GRP-F1 isoform (hGRP) was amplified by nested PCR from reverse transcribed Sk total RNA, using HsGRPCDS1_Fw and HsGRPCDS3_Rv specific primers in the first PCR, followed by nested amplification with HsGRPCDS2_Fw/HsGRPEx5R2 primers. PCR products were cloned into pCRIITOPO (Invitrogen) and sequenced (CCMAR sequencing facilities, Faro, Portugal). Human GRP cDNA coding for the secreted GRP-F1 protein was amplified using an Sk-derived positive clone and specific primers ReHsGRP_CFw/ReHsGRP_Rv designed to allow directional cloning into a pET151/D-TOPO vector (Champion pET Directional TOPO Expression kit, Invitrogen). A His6 tag, a V5 epitope, and a tobacco etch virus (TEV) protease cleavage site were added to the N-terminus of the expressed protein. Correct cloning was verified by DNA sequencing (CCMAR).
A list of all PCR primer sequences is presented in Table 3.
## 2.10. Recombinant Protein Expression and Purification
*In the **2.10. Recombinant Protein Expression and Purification** section:*
Escherichia coli BL21star (DE3) cells (Champion pET Directional TOPO Expression kit) were transformed according to manufacturer's instructions and induction was performed with 1 mM IPTG for 4 h. Cells were pelleted by centrifugation, resuspended in binding buffer (20 mM sodium phosphate, 0.5 M NaCl, 20 mM imidazole, pH 7.4), and sonicated for 3 min in 10 sec pulses series at 60 V. The resulting cleared supernatant was loaded onto a 1 mL HisTrap HP column (GE Healthcare) according to manufacturer's instructions, and recombinant protein was eluted with 20 mM sodium phosphate, 0.5 M NaCl, 500 mM imidazole, pH 7.4. Recombinant human GRP (rhGRP) protein purity was assessed by SDS-PAGE.[](https://www.ncbi.nlm.nih.gov/mesh/D007544)
## 2.11. Extraction and Purification of GRP and MGP from Calcified Tissues
*In the **2.11. Extraction and Purification of GRP and MGP from Calcified Tissues** section:*
Sturgeon GRP (sGRP) was extracted and purified as previously described. Identification of purified protein, obtained after RP-HPLC purification, was confirmed by N-terminal amino acid sequence. Bovine MGP (bMGP) was extracted from bovine calcified costal cartilage, obtained from local slaughterhouse, as described. Briefly, the formic acid demineralized fraction containing mineral-binding proteins was dialyzed against 50 mM HCl using 3,500 molecular weight tubing (Spectra/Por 3, Spectrum) over two days and then freeze-dried and dissolved in 6 M guanidine-HCl, 0.1 M Tris, pH 9.0. Subsequent partial purification was achieved by precipitation of insoluble proteins (mainly MGP) through dialysis against 5 mM ammonium bicarbonate.
Precipitated MGP was dissolved in 6 M guanidine-HCl, 0.1 M Tris, pH 9.0. HisTrap rhGRP was further purified through RP-HPLC as described above for sGRP, and recombinant Thermus thermophilus S6 ribosomal protein (S6) was a kind gift from Professor Eduardo Melo (CBME, University of Algarve, Portugal).[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## 2.12. Protein Mineral Complex (PMC) In Vitro Assay
*In the **2.12. Protein Mineral Complex (PMC) In Vitro Assay** section:*
Basic calcium phosphate (BCP) crystals were produced as previously described by incubating 2 mM CaCl2 and 10 mM sodium phosphate buffer pH 7.0 for 2 h at 37°C and then centrifuged at 20 000 ×g for 20 min at RT. BCP crystals were incubated for 30 min at 37°C, with approximately 5 μg of each protein (rhGRP, sGRP, bMGP, and S6) in 25 mM boric acid, pH 7.4. After centrifugation at 20 000 ×g for 20 min at RT, supernatants containing non-bound mineral proteins were collected, lyophilized, and analyzed by SDS-PAGE. Pellets containing PMCs were first washed with 25 mM boric acid, pH 7.4 and then demineralized with 30% (v/v) formic acid for 2 h at 4°C with agitation. After centrifugation at 20 000 ×g for 20 min at 4°C, the supernatant containing the BCP-binding proteins was collected, lyophilized, and analyzed by SDS-PAGE.[](https://www.ncbi.nlm.nih.gov/mesh/C020243)
## 2.13. Electrophoresis and Dot-Blot Analysis
*In the **2.13. Electrophoresis and Dot-Blot Analysis** section:*
Aliquots of protein were separated on a 4 to 12% gradient SDS-PAGE (NuPage, Invitrogen) gel and proteins were visualized by Coomassie Brilliant Blue (CBB, USB) staining as described. For dot-blot immunodetection, 100, 50, and 25 ng of synthetic peptides were applied onto a nitrocellulose membrane (BioRad), as described, and incubated O/N with a 1 : 1000 dilution of cGRP pAb and ucGRP mAb, respectively. Immunodetection was achieved using species-specific secondary horseradish-conjugated antibodies and Western Lightning Plus-ECL (PerkinElmer).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## 3. Results
*In the **3. Results** section:*
## 3.1. GRP-F1 as Main GRP Splice Variant Expressed in Human Skin and Mammary Gland
*In the **3.1. GRP-F1 as Main GRP Splice Variant Expressed in Human Skin and Mammary Gland** section:*
In order to study the expression pattern of GRP splice variants in skin (Sk) and mammary gland (MG) control tissues, specific primers were designed (Figure 2(a)) and tested for the unique amplification of each splice variant using GRP-F1, -F5, and -F6 clones. Primer sets A-B, C-D, and C-E were shown to specifically amplify GRP-F1, -F5, and -F6, respectively (results not shown). GRP-F1 was consistently amplified in all control samples of Sk and MG analyzed, while GRP-F5 and -F6 expressions were shown to be heterogeneous: barely detectable in some Sk samples and mostly undetectable in the MG tissues analyzed (Figure 2(b)). Overall, the expression pattern of GRP splice variants shows that the GRP-F1, coding for the full protein, is the main transcript present in control skin and mammary gland.
## 3.2. GRP Accumulates in Both Skin and Mammary Gland Control Tissues
*In the **3.2. GRP Accumulates in Both Skin and Mammary Gland Control Tissues** section:*
The pattern of total GRP accumulation was determined in control human skin (Figures 3(a)–3(c)) and mammary gland (Figure 3(d)). Strong positive staining for GRP was observed in all strata of the epidermis (Ep), in small blood vessels (BV) at the dermis level (Figure 3(a)), in hair follicles (results not shown) and in sweat (SwG; Figure 3(b)) and sebaceous (SG; Figure 3(c)) glands; this is consistent with the previously described pattern of GRP accumulation in human skin. In normal mammary tissue, GRP was mainly detected in the cytoplasm of ductal cells (DC) forming the lobules (Figure 3(d)) and in small arterioles (results not shown). Negative controls (NC) showed absence of signal.
## 3.3. GRP-F1 and Genes Involved in γ-Carboxylation Share Gene Expression Pattern in Skin and Breast Cancers
*In the **3.3. GRP-F1 and Genes Involved in γ-Carboxylation Share Gene Expression Pattern in Skin and Breast Cancers** section:*
Expression levels of GRP splice transcripts were determined in control and cancerous tissues and correlated with gene expression of MGP, GGCX, VKOR, and the tumor markers OPN and TNFα (Figures 4 and 5). Both in skin cancer (SC) and in the control samples (Sk), the levels of GRP-F1 were found to be heterogeneous without a clear tendency for up- or downregulation in cancer cases (Figure 4). Interestingly, the same heterogeneous pattern was found for MGP, GGCX, and VKOR, while OPN and TNFα were found clearly upregulated in tumor samples (Figure 4). These results suggest a concerted expression of the VKDPs, GRP and MGP, and the genes involved in the γ-carboxylation process, which cannot be associated with growth, progression, or metastasis of cancer processes at this time. Of notice, skin cancer samples analyzed were devoid of microcalcifications, as determined by von Kossa staining and confirmed by histological evaluation by pathologists (Table 1).
Similar gene expression results were found in control MG and breast cancer (BC) samples (Figure 5), with heterogeneous levels of GRP-F1, MGP, GGCX, and VKOR and increased expression of OPN and TNFα in cancer cases (Figure 5). However, higher levels of GRP-F1, MGP, GGCX, and VKOR were found in BC samples that include microcalcifications (Table 1), suggesting an upregulation associated with calcification, but not necessarily with tumor development. Gene expression of GRP-F5 and -F6 transcripts was found to be nearly undetectable in the majority of samples from both skin and breast cancers (results not shown), highlighting the predominance of the GRP-F1 transcript in all tissues and conditions analyzed.
## 3.4. Validation of Novel Conformation-Specific Antibodies against Human Carboxylated (cGRP) and Undercarboxylated (ucGRP) GRP Protein Forms
*In the **3.4. Validation of Novel Conformation-Specific Antibodies against Human Carboxylated (cGRP) and Undercarboxylated (ucGRP) GRP Protein Forms** section:*
Conformational specificity of cGRP and ucGRP antibodies was initially screened by a cross-reactivity test of immune sera with each peptide in both γ-carboxylated and non-γ-carboxylated forms. Cross-reactivity of purified cGRP pAb and ucGRP mAb with all available synthetic peptides was further tested by dot-blot, confirming specificity of cGRP pAb for cGRPpep29-42 and of ucGRP mAb for ucGRPpep31-54 (Figure 6). In addition, quenching assays were performed for both antibodies to validate their use in IHC. Blocking recognition of cGRP pAb and ucGRP mAb epitopes was performed using the corresponding synthetic peptides in both forms; incubations of cGRP pAb with cGRPpep29-42 and ucGRP mAb with ucGRPpep31-54 resulted in a decreased signal intensity, while incubations of cGRP pAb with ucGRPpep29-42 and ucGRP mAb with cGRPpep31-54 showed similar signal staining as nonblocked antibody assays (results not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D010455)
## 3.5. Differential Accumulation Pattern of Human cGRP and ucGRP in Skin and Breast Cancers
*In the **3.5. Differential Accumulation Pattern of Human cGRP and ucGRP in Skin and Breast Cancers** section:*
Fifteen individual cases (eight skin and seven breast cancers, see Tables 1 and 2) were analyzed by IHC using cGRP and ucGRP antibodies and compared with control tissues (Figure 7). In control skin, cGRP and ucGRP (Figures 7(a) and 7(e), resp.) colocalized with total GRP (Figure 3(a)), although most of the fibroblasts (Fb) of the upper dermis were only stained for cGRP (Figure 7(a)). In BCC tumors, a clear difference is observed between cGRP (Figures 7(b) and 7(c)) and ucGRP (Figures 7(f) and 7(g)) accumulation associated with tumor cells (TC): staining for cGRP was decreased compared to ucGRP, while in healthy skin areas adjacent to tumors, both GRP forms are similarly accumulated (Figures 7(d) and 7(h)) with a pattern comparable to control skin (Figures 7(a) and 7(e)). These results indicate that although both GRP forms are present in control skin, cGRP is mainly present in the healthy tissue while ucGRP is the protein form predominantly occurring in association with tumor cells in BCC.
In healthy mammary glands cGRP was consistently detected in the cytoplasm of ductal cells (DC) forming the lobules (Figures 7(i) and 7(j)) and colocalized with total GRP (Figure 3(d)), while ucGRP was found to be either colocalized (Figure 7(m)) or undetectable (Figure 7(n)), depending on samples analyzed. In contrast, in IDC samples ucGRP was consistently detected throughout the cytoplasm of tumor cells (TC; Figures 7(o) and 7(p)), while cGRP was found to be highly localized with a pointed spot pattern to certain tumor cells (Figures 7(k) and 7(l)). Overall, ucGRP is the predominant form accumulating in IDC-tumor cells, while cGRP preferentially accumulates in healthy mammary gland. Of notice, not all areas of IDC analyzed were found positive for GRP, but in areas with a positive signal, the described patterns were always observed. Negative controls were performed for each antibody and each sample analyzed and showed absence of signal (results not shown).
## 3.6. Both cGRP and ucGRP Protein Forms Are Accumulated at Sites of Microcalcifications in BCC and IDC
*In the **3.6. Both cGRP and ucGRP Protein Forms Are Accumulated at Sites of Microcalcifications in BCC and IDC** section:*
From all samples analyzed by von Kossa staining, two BCC and four IDC samples were found to contain microcalcifications (results not shown) and classified as light, moderate, or massive, according to the quantity and size of the mineral present (Tables 1 and 2). In all samples containing microcalcifications, both cGRP (Figures 8(a)–8(c)) and ucGRP (Figures 8(d)–8(f)) were detected colocalizing with mineral deposits in BCC (Figures 8(a) and 8(d)) and IDC (Figures 8(b), 8(c), 8(e), and 8(f)). These results strongly suggest that both cGRP and ucGRP have a high affinity for calcium mineral deposits.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## 3.7. In Vitro Association of Both c and ucGRP Protein Forms with Basic Calcium Phosphate (BCP) Crystals
*In the **3.7. In Vitro Association of Both c and ucGRP Protein Forms with Basic Calcium Phosphate (BCP) Crystals** section:*
Calcium/phosphate (Ca/P) mineral-binding capacity of carboxylated and noncarboxylated GRP protein forms was evaluated using protein mineral complex (PMC) assays.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Human recombinant GRP-F1 protein (rhGRP) was expressed as a noncarboxylated secretory 107-aa protein, comprising the 74-aa GRP-F1 protein, with 33 aa of His6, epitope V (V5), and the TEV recognition site (TEV_RS) at its N-terminus. Purified rhGRP with an apparent molecular weight of 14 kDa on SDS-PAGE (Figure 9(a)) was further identified through LC-MS/MS analysis (Mass Spectroscopy facilities, ITQB-Lisbon, results not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
Protein fractions obtained in the PMC assays using noncarboxylated rhGRP, carboxylated sturgeon GRP (sGRP), S6, and bovine MGP (bMGP) were analyzed by SDS-PAGE. Results demonstrate that most of rhGRP, sGRP, and bMGP (used as positive control) were present in the demineralized fraction corresponding to the mineral-bound proteins, while S6, used as negative control, was predominantly found in the supernatant containing the non-mineral bound proteins (Figure 9(b)). These results confirm the BCP-binding capacity of both carboxylated and noncarboxylated GRP protein forms.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## 4. Discussion
*In the **4. Discussion** section:*
γ-carboxylation of VKDPs is widely accepted to be determinant for their proper function, highlighting the importance of investigating γ-carboxylation status of GRP in humans. Human GRP carboxylation has been hypothesized on the basis of its high sequence similarity with sturgeon protein, previously shown to be γ-carboxylated, and the identification of specific domains and motifs conserved in other VKDPs. In this work, we first investigated GRP γ-carboxylation status in human healthy tissues and further determined its association with ectopic calcification in cancers. Recently we have found additional GRP alternatively splice transcripts in human tissues (GRP-F5 and -F6,) which were different from those previously described for mouse and zebrafish. However, the GRP-F1 transcript was clearly shown to be the predominantly expressed variant in the control and cancer tissues analyzed.
The newly developed conformation-specific antibodies were designed to detect the complete form of human secreted GRP (i.e., the GRP-F1 isoform), containing 15 Glu residues potentially γ-carboxylated. Since both GRP-F5 and -F6 contain exons 4 and 5 it would be possible that in certain conditions both cGRP pAb and ucGRP mAb colocalize different GRP isoforms. However, in this study the expression of GRP-F5 and -F6 was undetectable in the majority of the samples analyzed and their contribution for the GRP accumulation pattern was considered to be negligible. By using the new conformation-specific antibodies we were able to demonstrate the differential accumulation patterns of cGRP and ucGRP species in healthy skin and mammary gland tissues, their relation with neoplasias, and particular association with microcalcifications in skin and breast cancers. In healthy tissues, cGRP and ucGRP were found to be colocalized, suggesting an incomplete GRP γ-carboxylation status under normal physiological conditions.
This result is consistent with the knowledge that all extra hepatic Gla proteins presently investigated are undercarboxylated in non-vitamin-K supplemented healthy individuals. Moreover, in tumor cells (both in BCC and IDC) cGRP was clearly lower than in non-affected areas, whereas ucGRP preferentially associated with tumor cells; high amounts of ucGRP were also found at sites of microcalcifications. Since conformation-specific GRP antibodies were produced against synthetic peptides covering small regions of GRP and possible γ-carboxylated Glu residues are present throughout the entire mature protein, the possibility of simultaneous detection of c/ucGRP protein forms cannot be completely discarded yet. Further characterization of monospecificity against native GRP species is currently under investigation. Nevertheless, these antibodies were found to have high specificity towards the respective synthetic GRP-related peptides used as antigens and clearly demonstrate different patterns of cGRP and ucGRP protein accumulation in the human tissues analyzed.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
We have previously suggested that GRP may be a physiological inhibitor of soft tissue calcification accumulating at sites of mineral deposition, and the clear association of GRP with microcalcifications present in BCC and IDC further supports a global association of GRP with ectopic calcifications, independent of disease etiology. The presence of high amounts of ucGRP at sites of calcification, together with (i) the knowledge that Gla residues increase calcium binding capacity of VKDPs and (ii) that calcification inhibitors are known to accumulate at sites of mineral deposition, suggests a pivotal role for GRP in the regulation of mineralization that can be compromised in situations of low γ-carboxylase activity (e.g., by poor vitamin K status). In analogy, impaired carboxylation of MGP, leading to the accumulation of substantial amounts of ucMGP at sites of calcification, was previously suggested to be associated with a suboptimal capacity of arterial and skin calcification inhibition. In concordance, our results show higher levels of GRP-F1 expression in IDC cases where ectopic calcifications were present. Although increasing sample sets, not yet available in our laboratory, would be required to clearly establish a relation between GRP-F1 expression levels and cancer development, it is interesting to note that expression pattern of GGCX, VKOR, and MGP was found highly similar to that of GRP-F1.
This suggests that genes required for γ-carboxylation respond in a concerted manner according to demands of substrate and should not be limiting factors for carboxylation of VKDPs, such as GRP and MGP. However, increased gene expression might not reflect protein functionality, and impaired GGCX activity has been associated with insufficient γ-carboxylation of prothrombin in cancers, while levels of GGCX mRNA have been shown either increased or heterogeneous among hepatocellular carcinomas. Although prolonged subclinical vitamin K deficiency has been demonstrated as a risk factor for cancer development, a relation between vitamin K status or intake and decreased carboxylation of VKDPs is still controversial. However, the increased ucGRP accumulation in BCC and IDC and concomitant decreased of cGRP in relation to healthy tissues could be explained by decreased levels of vitamin K in tumor areas in contrast to non-tumorous, as previously reported. Although a number of studies have shown that different forms of vitamin K (notably vitamin K2) exert antitumor activity on various rodent- and human-derived neoplastic cell lines, most of these effects have been correlated with increased γ-carboxylation of prothrombin leading to decreased levels of DCP.
Moreover, although levels of MGP mRNA have been suggested as a molecular marker for breast cancer prognosis, with overexpression and downregulation of MGP gene reported in different types of cancer and cell lines [reviewed in ], its γ-carboxylation status in neoplasias remains unknown. Special attention should be given to the suggested therapeutic effect of vitamin K on cancer progression and to the potential detrimental effects of vitamin K antagonists widely used in therapy of patients with cancer, on the functionality of VKDPs present in tumor tissues such as GRP and MGP.[](https://www.ncbi.nlm.nih.gov/mesh/D015055)
Our protein-mineral complex in vitro studies provide insights into a possible mechanism explaining the accumulation of GRP at sites of pathological calcifications, since we demonstrated that both cGRP and ucGRP have calcium mineral-binding capacity and can directly bind BCP crystals. Similarly, MGP was shown to directly interact with HA crystals involving both phosphoserine and Gla residues; also for MGP the direct protein-HA interaction was suggested to be the mechanism underlying MGP arterial calcification inhibition. Interactions between proteins and biological calcium crystals are believed to play a central role in preventing or limiting mineral formation in soft tissues and biological fluids, being determinant in several physiological processes and associated with pathological conditions. Additional studies are required to further clarify the role of cGRP and ucGRP species in calcium crystal nucleation and growth and to determine their precise mechanism of action at the molecular level.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Although further efforts will be made to highlight the relevance of GRP in cancer processes, we showed that GRP is associated with pathological mineralization in cancer and has the in vitro capacity to directly interact with calcium crystals. Our results emphasize that the involvement of this protein should be considered whenever conditions for correct carboxylation of VKDPs are affected. Furthermore, the measurement of carboxylation degree of Gla proteins, such as MGP, osteocalcin, and prothrombin, has been proposed as a marker for certain pathological conditions and vitamin K status. Further investigation aiming to correlate circulating levels and γ-carboxylation status of GRP with the degree of calcification and disease progression are currently in progress in our labs. We expect that our work will contribute to the evaluation of GRP potential usage as an additional marker for ectopic calcification.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## 5. Conclusions
*In the **5. Conclusions** section:*
Here we report the γ-carboxylation status of GRP-F1 in human healthy tissues and its association with skin and breast cancers. The new conformation-specific GRP antibodies enable us to demonstrate that in healthy tissues, cGRP and ucGRP were found to be colocalized, suggesting an incomplete GRP γ-carboxylation status under normal physiological conditions, while ucGRP was the predominant form associated with tumor cells.
Our results strengthen the previously reported association of GRP with ectopic calcifications, which are particularly relevant in the diagnosis of breast tumors. Our findings suggest that GRP may represent a new target for the anticancer potential of vitamin K, while the degree of GRP γ-carboxylation might be useful as a potential marker for vitamin K status and ectopic calcification occurrence.[](https://www.ncbi.nlm.nih.gov/mesh/D014812)
## Conflict of Interests
*In the **Conflict of Interests** section:*
The tools and methods described in this manuscript are included in a PCT patent application PCT/PT2009000046. Dina C. Simes and Carla S. B. Viegas are cofounders of GenoGla Diagnostics; Cees Vermeer is founder of VitaK. The authors declare that there is no conflict of interests regarding the publication of this paper.
Conformation-specific antibodies for γ-carboxylated (cGRP) and undercarboxylated (ucGRP) GRP forms. Amino acid sequence alignments of mature hGRP-F1 (hGRP) and synthetic peptides used as antigens for development of conformational c/ucGRP antibodies: cGRP29-42 and ucGRP31-54, respectively, and ucGRP29-42 and cGRP31-54 peptides used to test respective conformational specificity. GRP exons (Ex3, Ex4, and Ex5) are denoted in the corresponding amino acid sequence in gray scale; bold and underlined E in cGRP29-42 and cGRP31-54 indicate Gla residues.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
GRP-F1 is the predominant splice variant expressed in healthy human skin and mammary gland. (a) Schematic representation of GRP-F1, -F5, and -F6 splice transcripts and strategy to obtain corresponding transcript-specific primer sets. Structural organization of full length GRP-F1 is represented in white boxes above corresponding coding exons and limited to the open reading frame (between ATG and Stop codons): SP, signal peptide; PP, propeptide; MP, mature protein. (b) Gene expression analysis of GRP-F1, -F5, and -F6 splice transcripts in three control samples of both skin (Sk 1–3) and mammary gland (MG 1–3) tissues, using the above-described primer sets. 18S was used as loading control for sample integrity; transcript sizes are 118, 172, and 102 bp for F1, F5, and F6, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D010455)
GRP accumulates in healthy skin and mammary gland tissues. Immunolocalization of total GRP was performed with the CTerm-GRP antibody in healthy skin (a–c) and mammary gland (d) tissue sections. In skin, GRP accumulates at the epidermis (Ep) and blood vessels (BV) (a), sweat (SwG, b) and sebaceous (SG, c) glands. In mammary gland, GRP accumulates in the cytoplasm of ductal cells (DC) forming the lobules (d). NC represents negative controls performed by omitting the CTerm-GRP antibody in consecutive tissue sections. Samples were counterstained with haematoxylin. Scale bar represents 100 μm.[](https://www.ncbi.nlm.nih.gov/mesh/D006416)
GRP-F1, MGP, and γ-carboxylation related genes exhibit similar expression patterns in control and skin cancer. Levels of GRP-F1, MGP, GGCX, VKOR, OPN, and TNFα gene expression were determined by qPCR in three control skin (Sk 1–3) and five skin cancer (SC 1–5) samples and normalized using 18S and GAPDH as housekeeping genes. Expression values are relative to zero and represent the mean of duplicates; standard deviations are indicated.
GRP-F1 is upregulated in mineralization-containing breast cancer samples. Levels of GRP-F1, MGP, GGCX, VKOR, OPN, and TNFα gene expression were determined by qPCR in three control mammary gland (MG 1–3) and four breast cancer (BC 1–4) samples and normalized using 18S and GAPDH as housekeeping genes. Expression values are relative to zero and represent the mean of duplicates; standard deviations are indicated.
Validation of conformation-specific antibodies for cGRP and ucGRP forms. Conformation specificity of cGRP and ucGRP antibodies was tested by dot blot using synthetic peptides (100, 50, and 25 ng peptide each lane): cGRP29-42; ucGRP31-54; ucGRP29-42; and cGRP31-54, and shows specific detection of cGRP and ucGRP antibodies to cGRP29-42 and ucGRP31-54 peptides, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D010455)
cGRP preferentially accumulates in healthy tissues while ucGRP is the predominantly form associated with tumor cells. Immunolocalization of cGRP and ucGRP in control skin (Sk; (a, e), resp.), basal cell carcinoma (BCC; (b–d), (f–h), resp.), control mammary gland (MG; (i–j), (m–n), resp.), and invasive ductal carcinoma (IDC; (k–l), (o–p), resp.) tissue sections was performed with cGRP and ucGRP antibodies, respectively. In control skin, cGRP and ucGRP are similarly accumulated in the epidermis, although cGRP is the predominant form in blood vessels (BV) and fibroblasts (Fb) ((a, e), resp.). In BCC tumor cells (TC), cGRP levels are significantly decreased (b, c), while ucGRP is the predominant form (f, g) compared to both healthy skin (a, e) and non-affected areas adjacent to tumor cells (d, h). In control mammary gland (MG), cGRP is accumulated in the ductal cells (DC; (i, j)), while ucGRP accumulation is either similar (m) or decreased (n) compared to cGRP. In IDC tumor cells, the amount of cGRP is significantly decreased (k, l) in relation to ucGRP (o, p). Sections were counterstained with haematoxylin. Scale bar represents 100 μm.[](https://www.ncbi.nlm.nih.gov/mesh/D006416)
cGRP and ucGRP are highly accumulated at sites of microcalcifications in BCC and IDC. Sites of cGRP (a, b, c) and ucGRP (d, e, f) accumulation were determined by IHC in BCC (a, d) and IDC (b, c and e, f) tissue sections, using the c/ucGRP antibodies. Both GRP protein forms are highly accumulated at sites of mineral deposits in all BCC and IDC calcification-containing samples analyzed. White arrows show examples of microcalcifications; sections were counterstained with haematoxylin. Scale bar represents 100 μm.[](https://www.ncbi.nlm.nih.gov/mesh/D006416)
Both cGRP and ucGRP forms show in vitro mineral-binding affinity. (a) SDS-PAGE analysis of purified noncarboxylated recombinant human GRP-F1 (rhGRP) protein expressed in E. coli, detected with CBB. (b) SDS-PAGE analysis of non-mineral bound and mineral-bound proteins obtained from in vitro basic calcium-phosphate (BCP) protein mineral-binding assays, performed with purified rhGRP, sturgeon GRP (sGRP), bovine MGP (bMGP), and S6 ribosomal protein (S6). Relevant molecular weight markers (MW, kDa) are indicated on the left side of panels (a) and (b).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
∗1Spinocellular carcinoma; ∗2basal cell carcinoma.
Histopathological features of skin cancer samples.
∗1IDC: invasive ductal carcinoma.
Histopathological features of breast cancer samples.
*Overhang sequence to pET151/D-TOPO for directional cloning.
Primers for PCR amplifications.
|
# Introduction
Activation of JNK Contributes to [Evodiamine](https://www.ncbi.nlm.nih.gov/mesh/C049639)-Induced Apoptosis and G2/M Arrest in Human Colorectal Carcinoma Cells: A Structure-Activity Study of Evodiamine
# Abstract
*In the **Abstract** section:*
Evodiamine (EVO; 8,13,13b,14-tetrahydro-14-methylindolo[2′3′-3,4]pyrido[2,1-b]quinazolin-5-[7H]-one derived from the traditional herbal medicine Evodia ruta[ecarpa was](https://www.ncbi.nlm.nih.gov/mesh/C049639) r[epo](https://www.ncbi.nlm.nih.gov/mesh/C049639)rt[ed to possess anticancer activity; however, the anticancer mechanism is still uncl](https://www.ncbi.nlm.nih.gov/mesh/C049639)ear. In this study, we investigated the anticancer effects of EVO on human colon COLO205 and HT-29 cells and their potential mechanisms. MTT and lactate dehydrogenase (LDH) release assays showed that the viability of C[OLO](https://www.ncbi.nlm.nih.gov/mesh/C049639)L205 and HT-29 cells was inhibited by EVO at various concentrations in a[cco](https://www.ncbi.nlm.nih.gov/mesh/C022616)rdance with increases in the percentage of apoptotic cells and cleavage of caspase-3 and poly(ADP ribose) polymerase (P[ARP](https://www.ncbi.nlm.nih.gov/mesh/C049639)) proteins. Disruption of the mitochondrial membrane potential by EVO was accompanied by increased Bax, caspase-9 protein cleavage, and cytochrome (Cyt) c protein translocation in COLO205 and HT-29 cells. Application of th[e a](https://www.ncbi.nlm.nih.gov/mesh/C049639)ntioxidant N-acetyl-L-cysteine (NAC) inhibited H2O2-induced reactive oxygen species (ROS) production and apoptosis, but did not affect EVO-induced apoptosis of COLO205[ or HT-29 cells. Si](https://www.ncbi.nlm.nih.gov/mesh/D000111)gn[ifi](https://www.ncbi.nlm.nih.gov/mesh/D000111)cant increas[es i](https://www.ncbi.nlm.nih.gov/mesh/D006861)n the G2/[M ratio and cyclinB1/cd](https://www.ncbi.nlm.nih.gov/mesh/D017382)c2[5c ](https://www.ncbi.nlm.nih.gov/mesh/D017382)protein expression by EVO were respectively ide[nti](https://www.ncbi.nlm.nih.gov/mesh/C049639)fied in colon carcinoma cells via a flow cytometric analysis and Western blotting. Induction of extracellular signal-regulated k[ina](https://www.ncbi.nlm.nih.gov/mesh/C049639)se (ERK) and c-Jun N-terminal kinase (JNK) protein phosphorylation was detected in EVO-treated cells, and the JNK inhibitor, SP600125, but not the ERK inhibitor, U0126, inhibited EVO-induced phosphorylated JNK protein expression, apoptosis[, a](https://www.ncbi.nlm.nih.gov/mesh/C049639)nd G2/M arrest of colon carcinoma cells[. Data o](https://www.ncbi.nlm.nih.gov/mesh/C432165)f the structure-activity anal[ysis ](https://www.ncbi.nlm.nih.gov/mesh/C113580)showed that [EVO](https://www.ncbi.nlm.nih.gov/mesh/C049639)-related chemicals containing an alkyl group at position 14 were able to induce apoptosis, G2/M arrest associated with increased DNA ladder formation, cl[eav](https://www.ncbi.nlm.nih.gov/mesh/C049639)age of caspase-3 and PARP, and elevated cycB1 and cdc25c protein expressions in COLO205 and HT-29 cells. Evidence supporting JNK activation leading to EVO-induced apoptosis and G2/M arrest in colon carcinoma cells is provided, and alkylation at position 14 of EVO is a critical substitution for treatment of[ co](https://www.ncbi.nlm.nih.gov/mesh/C049639)lonic cancer.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Colorectal cancer (CRC) is the second leading diagnosed cancer with high mortality, and remains a significant global health problem , . Many therapeutic strategies such as surgery and chemotherapy are used to treat CRC; however, there are troublesome side effects with chemotherapy, and surgical treatment is associated with high mortality and local recurrence , . Natural products have served as a leading source of drug development for centuries, and many of the new antitumor drugs such as taxol and cisplatin are natural products or derived from natural products , . Evodiamine (EVO) is a natural chemical isolated from Evodia rutaecarpa, and several biological effects of EVO including antitumor, antinociceptive, and vasorelaxant properties were reported , . EVO showed an inhibitory effect on tumor cell migration in vitro, and induced cell death in several cell types, but had little effect on normal human peripheral blood mononuclear cells .
Ogasawara et al. (2004) indicated the inhibitory effects of EVO against the invasion and lung metastasis of colon carcinoma cells . In addition to anti-tumor effect, EVO may inhibit insulin-Stimulated mTOR-S6K activation in adipocytes and improves glucose tolerance in Obese/Diabetic Mice . These results reveals the beneficial effects of EVO, however the mechanism underlying its antitumor activities and the structure-activity relationship of EVO are still poorly defined.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
Recent studies suggested that eradication of cancer cells by anticancer agents was mediated by induction of apoptosis of those cells –. There are several apoptotic pathways in cells in response to apoptotic stimuli, and induction of apoptosis by chemotherapeutic agents mostly occurs through mitochondrial apoptotic pathways , . The release of mitochondrial apoptotic proteins such as cytochrome (Cyt) c initiates caspase activation, and Cyt c release leads to activation of caspase-9, which in turn activates effector caspases such as caspase-3 causing caspase-dependent DNA fragmentation, a characteristic of apoptosis. Members of the Bcl-2 family proteins with either proapoptotic (e.g., Bax, and Bak) or antiapoptotic (e.g., Bcl-2, and Bcl-xL) functions regulate the mitochondrial membrane permeability (MMP) in apoptosis, and decreases in antiapoptotic and increases in proapoptotic Bcl-2 family proteins were observed during apoptosis of cancer cells under chemical stimulation. Previous papers indicated that the subtle balance of the Bcl-2/Bax complex led to an anti- or proapoptotic effect, and the overexpression of Bax may induce loss of the MMP that initiates apoptosis progression , .
It was indicated that disruption of the MMP via disturbing the Bcl-2/Bax balance leading to activation of caspases-9 and -3 plays an important role in apoptosis induced by chemotherapeutic agents. Reactive oxygen species (ROS) are mediators of apoptosis induction, and a number of studies showed that increased ROS production can cause cellular apoptosis via a mitochondrion-dependent pathway . EVO was shown to induce apoptosis in various cancer cells; however, the mechanisms and roles of ROS in EVO-induced apoptosis are still unclear.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
Current drug development in cancer therapy is to induce mitogenic arrest via blocking diverse signal transduction pathways in cancer cells, and several chemotherapeutic agents such as paclitaxel and nocodazol that act against cancer cell cycle progression have been explored , . It was indicated that mitotic arrest is a fundamental cause of cytotoxicity by these chemotherapeutic agents. Alternative expressions of cyclin-dependent kinases (CDKs) and cyclines drive progression of the cell cycle, and cyclinE/CDK2 for G1/S and cyclinB/CDK2 regulated by cdc25 for the G2/M transition were reported . Clinical chemotherapeutic agents mainly cause cell cycle arrest at the G2/M phase and induce apoptosis in cancer cells. Activation of intracellular kinase cascades contributes to the proliferation and survival of cancer cells, and previous studies showed that activation of mitogen-activated protein kinases (MAPK), including extracellular signal-regulated kinase (ERK) and c-Jun N-terminal kinase (JNK) participates in apoptosis and cell cycle progression of cancer cells.
Although induction of mitogenic arrest by EVO was reported, the role of MAPK activation in EVO-induced cell cycle arrest remains undefined.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
In this study, we examined the mechanisms of EVO-inhibited viability and cell cycle progression of COLO205 and HT-29 colorectal carcinoma cells, and the structure-activity relationship (SAR) of EVO was analyzed. We found that EVO was able to reduce the viability of colorectal carcinoma cells via apoptosis induction, and G2/M arrest, which were independent of ROS production. Increased caspase-9 and -3 protein cleavage, and cyclin B1 and cdc25c proteins through induction of JNK protein phosphorylation by EVO were observed in colorectal carcinoma cells. Additionally, substitution at N14 of EVO is critical for apoptosis and G2/M arrest of colorectal carcinoma cells, and the intracellular pathway of apoptosis and G2/M arrest elicited by EVO was also investigated.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Results
*In the **Results** section:*
## EVO reduced the viability of colorectal carcinoma cells via apoptosis induction
*In the **EVO reduced the viability of colorectal carcinoma cells via apoptosis induction** section:*
In order to examine the effect of EVO on the viability of two colon carcinoma cells COLO205 and HT-29, MTT and LDH release assays were applied in the present study. In the study, NIH3T3 and WI-38 cells were used to test if the limited cytotoxicity of EVO against the viability of colon carcinoma cells. NIH3T3 cells were established cells from murine embryo with no tumor formation in mice, and WI-38 cells were isolated from normal embryonic lung tissue with a finite lifetime. As shown in Fig. 1A, concentration-dependent reductions in the viability of COLO205 and HT-29 cells were detected by the MTT assay, and EVO exhibits the more potent cytotoxicity again the viability of COLO205/HT-29 than NIH3T3/WI-38 cells.
Data of the LDH release assay showed that EVO concentration-dependently increased LDH in the medium of COLO205 and HT29 cells (Fig. 1B). The ratio of apoptotic bodies indicated that increased apoptotic bodies were detected in EVO-treated COLO205 and HT-29 cells (Fig. 1C). Loss of DNA integrity with the appearance of DNA ladders was observed in EVO-treated COLO205 and HT-29 cells via DNA electrophoresis (Fig. 1D). Examination of apoptotic proteins including caspase-3 and PARP protein expressions showed that increased cleavage of caspase-3 and PARP proteins was detected in COLO205 and HT-29 cells under EVO stimulation (Fig. 1E).
Additionally, caspase-3 activity induced by EVO was identified in COLO205 and HT-29 cells using the colorimetric peptidyl caspase-3 substrate, Ac-DEVD-pNA (Fig. 1F). These results supported the reduction in viability of colorectal carcinoma cells by EVO being mediated by induction of apoptosis.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
EVO reduction of viability of colorectal carcinoma COLO205 and HT-29 cells via apoptosis induction.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
(A) EVO reduction of cell viability of COLO205, HT-29, NIH3T3, and WI-38 cells by an MTT assay. These cells were treated with indicated concentrations (0.5, 1, 2, 4, and 8 µM) of EVO for 24 h, and cell viability was examined by an MTT assay. (B) EVO induction of lactate dehydrogenase (LDH) release by COLO205 and HT-29 cells according to an LDH release assay. As described in (A), the amount of LDH in the medium was examined by LDH kits. (C) Increased percentages of hypodiploid cells in EVO-treated COLO205 and HT-29 cell lines. Cells were treated with EVO (2 µM) for 24 h, and the percentage of hypodiploid cells was measured by flow cytometric analysis using PI staining. (D) EVO-induced loss of DNA integrity through increased DNA ladder formation. As described in (C), DNA integrity was analyzed by agarose electrophoresis. (E) Induction of caspase-3 (Casp 3) and poly(ADP ribose) polymerase (PARP) protein cleavage by EVO was detected in COLO205 and HT-29 cells by Western blotting using specific antibodies. (F) A significant increase in Casp 3 enzyme activity in EVO-treated colorectal carcinoma cells. As described in (C), activity of Casp 3 was measured by adding the Casp 3-specific colorimetric peptidyl substrate, Ac-DEVD-pNA. Each data point was calculated from three triplicate groups, and data are displayed as the mean ± S.D. ** p<0.01 denotes a significant difference compared to the control (C or CON) group.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Mitochondrion-mediated apoptosis by EVO in colorectal carcinoma cells
*In the **Mitochondrion-mediated apoptosis by EVO in colorectal carcinoma cells** section:*
We further examined the role of mitochondria in apoptosis induction by EVO in COLOL205 and HT-29 colorectal carcinoma cells. Data from the MMP analysis using a fluorescent mitochondria-binding dye (DiOC6) showed that EVO addition significantly reduced the MMPs in both cell lines. Reduction of the MMP by H2O2 was described as a positive control (Fig. 2A). Alternative expressions of pro- and antiapoptotic Bcl-2 family proteins appeared, in that an increase in the proapoptotic Bax protein and a decrease in the antiapoptotic protein Bcl-XL were observed by Western blotting using specific antibodies (Fig. 2B).
Induction of cleavage of caspase-9 and Cyt c in cytosol was detected in EVO-treated COLO205 and HT-29 cells (Fig. 2C). Incubation of both cell lines with the peptidyl caspase-9 inhibitor, Ac-YVAD-FMK, inhibited EVO-induced DNA ladder formation (Fig. 2D). An increase in caspase-9, but not caspase-8, activity in EVO-treated COLO205 and HT-29 cells was observed using a specific peptidyl colorimetric substrate (Fig. 2E). These results indicated that disruption of the MMP contributed to EVO-induced apoptosis in colorectal carcinoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
Disruption of the mitochondrial membrane potential (MMP) with increased Bax protein and cytosolic cytochrome (Cyt) c protein expressions, and caspase-9 (Casp 9) protein cleavage in EVO-treated COLO205 and HT-29 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
(A) Loss of the MMP by EVO and H2O2 in COLO205 and HT-29 cells. Cells were treated with EVO (2 µM) or H2O2 (100 µM) for 12 h, and the MMP was detected by a flow cytometric analysis using DiOC6 as a fluorescent dye. (upper) A representative example of flow cytometric data is shown; (lower) quantification of the M1 ratio from three independent experiments is shown. (B) Alternative Bcl-2 family protein expression by EVO was detected by Western blotting using specific antibodies. Cells were treated with different concentrations of EVO for 24 h, and expressions of indicated proteins were detected by Western blotting. (C) EVO induction of Casp 9 protein cleavage and cytosolic Cyt c protein in COLO205 and HT-29 cells. As described in (C), expressions of Casp 9, cytosolic Cyt C, and mitochondrial Cyt c proteins were examined by Western blotting using specific antibodies. (D) The peptidyl Casp 9 inhibitor, Ac-YVAD-FMK (YVAD; 100 µM), inhibited EVO-induced DNA ladder formation by COLO205 and HT-29 cells. Cells were incubated with Ac-YVAD-FMK (100 µM) for 2 h followed by EVO (2 µM) treatment for 24 h, and DNA integrity was examined by agarose electrophoresis. (E) A significant increase in Casp 9, but not Casp 8, enzyme activity in EVO-treated colorectal carcinoma cells. As described in (C), activities of Casp 9 and 8 were respectively measured by adding the Casp 9-specific colorimetric peptidyl substrate, Ac-DEVD-pNA, or the Casp 8-specific colorimetric peptidyl substrate, Ac-IETD-pNA. Each data point was calculated from three triplicate groups, and data are displayed as the mean ± S.D. ** p<0.01 denotes a significant difference compared to the control (C or CON) group. The intensity of each band was examined by a densitometric analysis (Imag J), and expressed as multiples of the control.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## ROS-independent apoptosis by EVO in colorectal carcinoma cells
*In the **ROS-independent apoptosis by EVO in colorectal carcinoma cells** section:*
We further examined the role of ROS in EVO-induced apoptosis in COLO205 and HT-29 cells. Intracellular peroxide levels were detected by a flow cytometric analysis using DCHF-DA as a fluorescent dye. As illustrated in Fig. 3A, the addition of H2O2 induced intracellular peroxide production in COLO205 and HT-29 cells, but no effect on peroxide production in either cell line by EVO was observed. Additionally, H2O2-induced DNA ladders and cytotoxicity were detected in COLO205 and HT-29 cells, and those were abolished by adding the antioxidant, NAC, via agarose electrophoresis and an MTT assay, respectively.
However, NAC addition was unable to reduce EVO-induced cell death via MTT assay and DNA ladders via agarose electrophoresis in either cell line (Fig. 3B). Increases in cleaved caspase-3 and PARP proteins with H2O2 treatment were detected by Western blotting using specific antibodies, and they were inhibited by the addition of NAC (Fig. 3C). No alteration in the expressions of cleaved caspase-3 or PARP protein was observed in EVO- and EVO+NAC-treated cells. This indicates that EVO-induced apoptosis might not be mediated in an ROS-dependent manner in colorectal carcinoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
Role of reactive oxygen species (ROS) in EVO-induced apoptosis of colorectal carcinoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
(A) EVO shows no effect on intracellular peroxide production in COLO205 and HT-29 cells. Both cells were treated with EVO (2 µM) or H2O2 (100 µM) for 3 h followed by adding a fluorescent dye (DCHF-DA) to examine intracellular peroxide levels via a flow cytometric analysis. (upper) A representative example of flow cytometric data is shown; (lower) quantification of the M1 ratio from three independent experiments is shown. (B) N-Acetyl-L-cysteine (NAC; N; 20 mM) protected cells from H2O2-induced cell death and DNA ladder formation, but had no effect on EVO-induced apoptosis. Both cells were treated with NAC (20 mM) for 30 min followed by EVO (2 µM) or H2O2 (100 µM) treatment for 24 h. DNA integrity (upper panel) and the viability (lower panel) of cells under different treatments are examined by agarose electrophoresis and MTT assay, respectively. (C) NAC inhibited H2O2-induced caspase (Casp) 3 and poly(ADP ribose) polymerase (PARP) protein cleavage, but did not affect EVO-induced events in COLO205 and HT-29 cells. As described in (B), the indicated protein expression was examined by Western blotting using specific antibodies. Each data point was calculated from three triplicate groups, and data are displayed as the mean ± S.D. **p<0.01 denotes a significant difference compared to the control (C) in (A) or between indicated groups (B).[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Activation of JNK is involved in EVO-induced apoptosis of colorectal carcinoma cells
*In the **Activation of JNK is involved in EVO-induced apoptosis of colorectal carcinoma cells** section:*
The roles of MAPK activation were reported, and we investigated if EVO-induced apoptosis occurred through altered MAPK activation in colorectal carcinoma cells. As shown in Fig. 4A, increased ERK and JNK protein phosphorylation was detected in COLO205 and HT-29 cells under EVO stimulation. Incubation of both cell lines with the ERK inhibitor, U0126, or the JNK inhibitor, SP600125, followed by EVO stimulation was applied to examine the roles of ERK and JNK activation in EVO-induced apoptosis. Data of the MTT assay showed that SP600125, but not U0126, addition protected both colorectal carcinoma cell lines from EVO-induced cell death (Fig. 4B).
Examination of DNA integrity showed that SP600125 addition attenuated EVO-induced DNA ladder formation in COLO205 and HT-29 cells (Fig. 4C). Data of Western blotting showed that U0126 inhibited EVO-induced pERK protein expression, but did not affect EVO-induced cleavages of caspase-3 protein in COLO205 and HT-29 cells. In the same part of the experiment, SP600125 exhibited an inhibitory effect on EVO-induced pJNK protein in accordance with the decrease in cleavage of the caspase-3 protein in both cell lines. No changes in the expressions of α-tubulin, tERK, or tJNK were described as internal controls (Fig. 4D).[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
c-Jun N-terminal kinase (JNK) activation participates in EVO-induced apoptosis of COLO205 and HT-29 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
(A) Induction of extracellular signal-regulated kinase (ERK) and JNK protein phosphorylation by EVO in colorectal carcinoma cells. Both cell lines were treated with EVO (2 µM) for different times, and expressions of phosphorylated (p)ERK/(p)JNK and total (t)ERK/(t)JNK were detected by Western blotting using specific antibodies. (B) The JNK inhibitor, SP600125 (SP; 20 µM), but not the ERK inhibitor, U0126 (U0; 20 µM), protected COLO205 and HT-29 cells from EVO-induced cytotoxicity according to an MTT assay. (C) SP600125 attenuates EVO-induced DNA ladder formation in colorectal carcinoma cells. Cells were treated with SP600125 (10 µM) for 30 min followed by EVO stimulation for an additional 24 h, and DNA integrity was examined by agarose electrophoresis. (D) SP600125 inhibited EVO-induced JNK protein phosphorylation and caspase (Casp) 3/poly(ADP ribose) polymerase (PARP) protein cleavage; however, U0126 inhibited EVO-induced ERK protein phosphorylation without affecting EVO-induced Casp 3/PARP protein cleavage in both cell lines. Both cell lines were treated with different concentrations of SP600125 or U0126 for 30 min followed by EVO stimulation for 30 min (for ERK and JNK protein expressions) or 24 h (for Casp 3 and PARP protein expressions) via Western blotting. Each data point was calculated from three triplicate groups, and data are displayed as the mean ± S.D. **p<0.01 denotes a significant difference compared between indicated groups.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Induction of G2/M arrest and cyclin B1/cdc25c protein expression by EVO were identified in COLO205 and HT-29 cells, and these were blocked by adding the JNK inhibitor, SP600125
*In the **Induction of G2/M arrest and cyclin B1/cdc25c protein expression by EVO were identified in COLO205 and HT-29 cells, and these were blocked by adding the JNK inhibitor, SP600125** section:*
Cell cycle progression was analyzed by flow cytometry using PI as a fluorescent dye. As illustrated in Fig. 5A, an increase in the G2/M ratio and a decrease in the G1 ratio were detected in EVO-treated COLO205 and HT-29 cells in concentration-dependent manners. Similarly, the EVO-induced G2/M ratio and EVO-inhibited G1 ratio were observed in both COLO205 and HT-29 cells in time-dependent manners (Fig. 5B).
Altered expressions of cell cycle-regulatory proteins including cyc B1, cdc 2, cyc E, cdc 25c, and p27 were detected by Western blotting using specific antibodies, and data shown in Fig. 5C reveal that time-dependent increases in cyc B1 and cdc 25c, but not the others, were detected in both EVO-treated cell lines. Increases in cyc B1 and cdc 25c proteins by EVO were reduced with a decrease in EVO-induced G2/M by adding the JNK inhibitor, SP (Fig. 5D). These results indicated that JNK activation contributes to EVO-induced apoptosis and G2/M arrest of colorectal carcinoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/D011419)
Induction of G2/M arrest and cyclin B1/cdc 25c protein expressions by EVO in COLO205 and HT-29 cells was significantly inhibited by adding the c-Jun N-terminal kinase (JNK) inhibitor, SP600125.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
(A) Concentration-dependent increases in the G2/M ratio and decreases in the G1 ratio in EVO-treated COLO205 and HT-29 cells. Both cell lines were treated with different concentrations (1, 2, and 4 µM) of EVO for 24 h, and ratios of cells in the G1 and G2/M phases were measured by a flow cytometric analysis via PI staining. (B) Time-dependent increases in the G2/M ratio and decreases in the G1 ratio were detected in EVO-treated colorectal carcinoma cells. Cells were treated with EVO (2 µM) for different times (6, 12, and 24 h), and ratios of cells in the G1 and G2/M phases were measured by a flow cytometric analysis via PI staining. (Left panel) A representative example of flow cytometric data is shown. (Right panel) Data of the G1 and G2/M ratios from three independent experiments are presented. (C) Alternative expressions of cell cycle regulatory proteins including cyc B1, cdc 2, cyc E, cdc 25c, p27, and α-tubulin in colorectal carcinoma cells under EVO stimulation. Cells were treated with EVO (2 µM) for different times (4, 8, 12, and 24 h), and expressions of the indicated proteins were examined by Western blotting using specific antibodies. (D) The JNK inhibitor, SP600125 (SP), inhibited EVO-induced cdc25c and cyc B1 protein expressions accompanied by decreases in the G2/M ratio in COLO205 and HT-29 cells. (Upper panel) Cells were treated with different concentrations of SP600125 for 30 min followed by EVO (2 µM) stimulation for 24 h, and expression of the indicated protein was examined by Western blotting. (Lower panel) Cells were treated with SP600125 (10 µM) for 30 min followed by EVO (2 µM) treatment for 24 h, and cell cycle progression was analyzed by flow cytometry via PI staining. The intensity of each band was examined by a densitometric analysis (Imag J), and expressed as multiples of the control. Each data point was calculated from three triplicate groups, and data are displayed as the mean ± S.D. **p<0.01 denotes a significant difference compared to the control (CON).[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Structure-activity analysis of effects of EVO-related compounds on the viability of colorectal carcinoma cells
*In the **Structure-activity analysis of effects of EVO-related compounds on the viability of colorectal carcinoma cells** section:*
EVO and 12 EVO-related compounds were applied to examine their effects on apoptosis induction of COLO205 and HT-29 cells. Structures of these chemicals are depicted in Fig. 6A, and different substitutions at position 14 of EVO are marked. Analysis of the DNA integrity in COLO205 and HT-29 cells under treatment with the indicated EVO-related chemicals showed that EVO, EVO-2, -4, -7, -8, and -12 possessed the ability to induce DNA ladder formation in both cell lines, whereas EVO-1, -3, -5, -6, -9, -10, and -11 did not (Fig. 6B). This implies that adding an alkyl group, such as a methyl or butyl, at position 14 of quinazolin is critical to apoptosis induction by EVO.
Furthermore, four compounds (i.e., EVO, -4, -5, and -8) were selected for a mechanism study, and these EVOs contained the same structure except for a methyl of EVO, an ethyl of EVO-4, a hydrogen of EVO-5, and a butyl of EVO-8 at position 14. Western blotting data showed that EVO, EVO-4, and EVO-8 induced cleavage of caspase-3 and PARP proteins with increased cyc B1 and cdc25c protein levels in COLO205 and HT-29 cells; however, EVO-5 did not (Fig. 6C). Analysis of the G2/M ratio in EVO-treated COLO205 and HT-29 cells showed that significant increases in the G2/M ratio were detected in EVO, EVO-4, and EVO-8-treated cells, but those were not observed in EVO-5-treated cells (Fig. 6D).[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
Structure-activity relationship of EVO and related chemicals on apoptosis and G2/M arrest elicited by EVO in colorectal carcinoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
(A) The chemical structures of EVO and structurally related chemicals (EVO-1∼12) are depicted. (B) Differential apoptotic effects elicited by EVOs in colorectal carcinoma cells. Cells were treated with the indicated EVOs (2 µM) for 24 h, and DNA integrity was analyzed by agarose electrophoresis. (C) Four EVOs with different substitutions at the position 14 of quinazolin showed differential effects on caspase (Casp) 3/poly(ADP ribose) polymerase (PARP) protein cleavage and cycB1/cdc 25c protein expressions in colorectal carcinoma cells. Cells were treated with the indicated chemicals (2 µM) for 24 h, and expressions of Casp 3/PARP, cycB1/cdc 25c, and α-tubulin (TUB) were detected by Western blotting using specific antibodies. (D) EVO, EVO4 (4), and EVO-8 (8), but not EVO-5 (5), increased the G2/M ratio of COLO205 and HT-29 cells. As described in (C), the G2/M ratio of COLO205 and HT-29 cells under different treatments was examined by flow cytometric analysis via PI staining. Each data point was calculated from three triplicate groups, and data are displayed as the mean ± S.D. **p<0.01 denotes a significant difference compared to the control (CON).[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Discussion
*In the **Discussion** section:*
Cell cycle checkpoints play important roles in coordination of cell cycle transitions in eukaryotic cells, and abnormal regulation of cell cycle checkpoints are frequently occurred in tumor cells. The function of arrest at cell cycle checkpoints is for DNA repair when cellular damage occurred, and the cell cycle arrest signaling might activate apoptotic pathways, leading to cell death when cellular damage is irreparable. According to flow cytometric analysis, EVO-induced G2/M arrest in COLO205 and HT-29 cells occurred at an early time point (Fig. 5B; 6 h) with a subsequent increase in apoptotic ratio at later times (12 and 24 h). It indicates G2/M arrest resulted in the inhibition of cell proliferation leading to apoptosis by EVO in COLO205 and HT-29 cells.
There is growing evidence indicating that inappropriate activation of cdc25c and cyclin B1 has an important role in antitubulin agent induced mitotic arrest and apoptosis. Increases in the G2/M ratio and cyclin B1/cdc 25c protein expression by EVO were observed in COLO205 and HT-29 cells, and the JNK inhibitor, SP600125, but not the ERK inhibitor, U0126, inhibited EVO-induced JNK protein phosphorylation with suppression of EVO-induced apoptotic events and G2/M arrest in COLO205 and HT-29 cells. Taken together, the molecular mechanism for EVO's induction of apoptosis and G2/M arrest in a JNK-mediated manner in colorectal carcinoma cells was demonstrated in the present study.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
Caspase-3 is a critical executioner of apoptosis through cleaving several essential cellular proteins such as PARP and D4-GDI. Induction of apoptosis through activation of caspase activity by EVO was reported in several previous studies. Huang et al. (2004) and Zhang et al. (2010) reported that EVO induced apoptosis and cleavage of caspases in human leukemic T lymphocytes and colon LOVO cells , . Zhang et al. (2013) reported that EVO induced caspase-dependent and -independent apoptosis in human U937 leukemia cells . Wang et al. (2013) reported that EVO inhibited the proliferation and induced cleavage of caspase-7 and PARP in breast carcinoma cells .
Our investigations revealed that EVO has the ability to increase caspase-3 activity and expressions of cleaved caspase-3 and PARP proteins, accompanied by apoptosis induction in COLO205 and HT-29 cells. These findings show that activation of the caspase cascade contributes to EVO-induced apoptosis in colon carcinoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
A low level of reactive oxygen species (ROS) is important for cellular function and survival signaling, while excessive ROS-elicited oxidative stress leads to cell death via apoptosis induction. Accumulated evidence indicates that chemotherapeutic agents can induce apoptosis through ROS production in various cancer cells. Yang et al. (2011) reported that increased ROS production by gelomulide K potentiates the lethality of breast carcinoma cells . Our previous publications supported an involvement of ROS production in apoptosis of cancer cells , , . Imatinib mesylate, gossypol, vitamin K3, and flavonoids induced apoptosis in several cell lines including melanoma, leukemia, glioma, and colorectal carcinoma cells through elevating ROS production.
However, ROS-independent apoptosis by chemical stimulation was also reported , . Although apoptosis induced by EVO was reported, the roles of ROS are still undefined. In the present study, NAC inhibited H2O2-induced DNA ladder formation and caspase-3/PARP protein cleavage, but was unable to block EVO-induced apoptosis. Data of DCHF-DA staining indicated that no alteration in intracellular peroxide levels by EVO was observed in COLO205 or HT-29 cells. These results suggested that ROS might not be involved in EVO-induced apoptosis of colorectal carcinoma cells.
In contrast to those results, EVO elevation of ROS and NAC inhibition of EVO-induced apoptosis in human cervical carcinoma HeLa cells were observed. Lower concentrations (1∼4 µM) of EVO in colon carcinoma cells in the present study and a higher concentration (21 µM) in cervical carcinoma cells in the previous study are possibly why ROS played differential roles in EVO-induced apoptosis. Additionally, Bcl-2 family proteins participate in maintenance of MMP regulation of the release of mitochondrial Cyt c to the cytosol and activation of caspase-9 activity which contribute to apoptosis of cancer cells. A significant increase in the proapoptotic Bax protein with decreases in antiapoptotic Bcl-2/Bcl-xL proteins was identified in both COLO205 and HT-29 cells under EVO stimulation. Accordingly, loss of the MMP with the occurrence of caspase-9 protein cleavage and release of Cyt C from mitochondria to the cytosol was observed in EVO-treated cells.
Mitochondrion-dependent apoptosis by EVO was indicated to occur in colorectal carcinoma cells.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
MAPK is implicated in regulating survival and cell death responses of tumor cells, and several studies reported the involvement of MAPK in cancer deregulation; however the precise mechanisms of MAPK in apoptosis and cell cycle progression of cancer cells remain elusive. Du et al. (2013) reported that EVO-induced apoptosis was enhanced by its combination with the ERK inhibitor, PD98059, or the p38 MAPK inhibitor, SB203580 . The relationship of MAPK to EVO-induced apoptosis and cell cycle arrest is still unclear. Data of the present study indicated that induction of ERK and JNK protein phosphorylation by EVO was detected in COLO205 and HT-29 cells, and EVO-induced apoptotic events, including DNA ladder formation and caspase-3 protein cleavage, were inhibited by adding the JNK inhibitor, SP600125, but not the ERK inhibitor, U0126. Additionally, control of cell cycle progression in cancer cells is regarded as an effective strategy for inhibiting tumor cell proliferation.
Previous studies reported that EVO inhibited the proliferation of various cancer cells that were arrested at the G2/M or S phase , , but the mechanism for mitogenic arrest by EVO is still poorly understood. In the present study, an increased G2/M ratio by EVO with induction of cyclinB1 and cdc25c protein expressions was detected in COLO205 and HT-29 cells. Addition of the JNK inhibitor, SP600125, decreased EVO-induced G2/M arrest and cyclinB1/cdc25c protein expression in both colon carcinoma cell lines. The promoters of cyclin B and CDC25C conserved cell cycle-dependent element (CDE), cell cycle genes homology region (CHR) sites, and CCAAT-boxes. Several factors such as E2F, CDF-1, and CBP have been reported to bind with CHR/CDE in cyclin B and CDC25C promoters .
Muller et al (2012) found that CHR is a central element in transcriptional regulation of cyclin B by the DREAM and MMB complexes . Chae et al (2011) found a transcriptional factor NF-Y binds to CCAAT in the promoters of cell cycle G2 regulators such as cyclin B and CDC25C . Seo et al (2008) indicated that phosphorylated c-Myc bound to the promoter of cyclin B1, resulting in increased cyclin B1 promoter activity . Inhibition of JNK protein phosphorylation reduces cdc25c/cyclinB1 protein expression in EVO-treated COLO 205 and HT-29 cells, however the mechanism of JNK inhibition leading to reduce EVO-induced cdc25c/cyclinB1 protein expression is still unclear. Contribution of JNK to transcriptional regulations of cyclin B1 and CDC25C gene via modulating the binding of transcriptional factors to their promoters needs to be further investigated.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
In order to estimate the structures that contribute to the apoptosis and G2/M arrest induced by EVO in colorectal carcinoma cells, the effects of compounds (EVO-1∼12) possessing structures similar to that of EVO on apoptosis and cell cycle progression of both colon cancer COLO205 and HT-29 cell lines were examined. As shown in Fig. 6, EVO-2, -4, -7, -8, and -12 containing an alkyl group such as ethyl or butyl at position 14 compared to the methyl group of EVO induced significant apoptosis in COLO205 and HT-29 cells. Furthermore, EVO and its structurally related compounds including EVO-4, -5, and -8 were used to study the effects on caspase-3, PARP, cyclinB1, and cdc25c protein expressions with cell cycle progression in both colorectal carcinoma cell lines. EVO, EVO-4, -5, and -8 share the same chemical structure except for different substitutions including a methyl of EVO, an ethyl of EVO-4, a hydrogen of EVO-5, and a butyl of EVO-8 at position 14.
Our results showed that EVO, EVO-5, and EVO-8, but not EVO-4, significantly induced G2/M arrest with increased cyclin B1/cad25c protein expressions and caspase-3/PARP protein cleavage in both colon carcinoma cell lines. Ogasawara et al. (2002) also indicated the role of a methyl group at position 14 for EVO in inhibiting invasion by Lewis lung cancer and melanoma cells . The critical roles of alkyl substitutions such as methyl and butyl at position 14 for apoptosis and G2/M arrest by EVO against colorectal carcinoma cells were demonstrated.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
In conclusion, we showed in the present study that EVO possesses antitumor activities including apoptosis and G2/M arrest against the viability of colorectal carcinoma cells. EVO induced disruption of the MMP, which was accompanied by activation of caspases-3/9, and increases in cyclin B1/cdc25c protein expressions in COLO205 and HT-29 cells. Activation of JNK by EVO was detected, and EVO-induced apoptotic and G2/M arrest were blocked by the JNK inhibitor, SP600125, indicating the critical role of JNK activation in the anti-colorectal carcinoma activity of EVO. Furthermore, a structure-activity study showed that methyl at position 14 is important for EVO's action against the viability of colon cancer cells. Further studies will investigate whether these effects of EVO can be extended to colon cancer cells in vivo, especially chemotherapy-resistant colon cancer cells.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Methods
*In the **Methods** section:*
## Cell culture
*In the **Cell culture** section:*
COLO205, HT-29, NIH3T3, and WI-38 cells were obtained from the American Type Culture Collection (Manassas, VA, USA). COLO205/HT-29 colon carcinoma cells in RPMI 1640, WI-38 in MEM containing 10% heat-inactivated fetal bovine serum (FBS; Gibco/BRL, Grand Island, NY, USA), and NIH3T3 in DMEM containing 10% heat-inactivated calf serum (CS; Gibco/BRL, Grand Island, NY, USA), supplemented with antibiotics (100 U/mL penicillin A and 100 U/mL streptomycin) were maintained in a 37 °C humidified incubator containing 5% CO2.[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Agents
*In the **Agents** section:*
The chemical reagents of EVO, N-acetyl-l-cysteine (NAC), SP600125, U0126, BCIP, 3-(4,5,-dimethylthiazol)-2-yl-2,5-diphenyltetrazolium bromide (MTT), and NBT were obtained from Sigma Chemical (St. Louis, MO, USA). Antibodies of α-tubulin, poly(ADP ribose) polymerase (PARP), caspase-3, caspase-9, Bcl-2, and Bax were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies of total (t) and phosphorylated (p) MAPKs (tERK/pERK and tJNK/pJNK), and cyclinB1/cdc25c proteins were obtained from Cell Signaling Technology (Danvers, MA, USA). The colorigenic synthetic peptide substrates, Ac-DEVD-pNA (a caspase-3 substrate), Ac-YVAD-pNA (a caspase-9 substrate), and Ac-IETD-pNA (a caspase-8 substrate) were purchased from Calbiochem. Other chemicals not mentioned above were obtained from Sigma Chemical.[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## Synthesis of structure-related chemicals of EVO
*In the **Synthesis of structure-related chemicals of EVO** section:*
The synthesis of EVO-related compounds were based on the coupling of 3,4-dihydro-β-carboline with substituted N-alkyl isatoic anhydride in pyridine. 3,4-dihydro-β-carboline was prepared by reacting tryptamine with ethyl formate and followed by intramolecular ring closure in the presence of POCl3. In the presence of NaH and DMF, Isatoic anhydride was alkylated with alkyl halide such as iodomethane, iodoethane, iodoprpopane, 2-methoxy ethyl chloride to afford N-alkyl isatoic anhydride analogues. The purities of them were more than 95% when analyzed by HPLC (Fig. S1).[](https://www.ncbi.nlm.nih.gov/mesh/C049639)
## MTT (3-(4,5,-dimethylthiazol)-2-yl-2,5-diphenyltetrazolium bromide) assay
*In the **MTT (3-(4,5,-dimethylthiazol)-2-yl-2,5-diphenyltetrazolium bromide) assay** section:*
Cell viability was assessed by MTT staining as described previously . Briefly, cells were plated at a density of 105 cells/well into 24-well plates. At the end of treatment, the supernatant was removed, and 30 µl of the tetrazolium compound, MTT, and 270 ml of fresh RPMI medium were added. After incubation for 4 h at 37°C, 200 µl of 0.1 N HCl in 2-propanol was placed in each well to dissolve the tetrazolium crystals. Finally, the absorbance at a wavelength of 600 nm was recorded using an enzyme-linked immunosorbent assay (ELISA) plate reader.[](https://www.ncbi.nlm.nih.gov/mesh/C022616)
## Lactate dehydrogenase (LDH) release assay
*In the **Lactate dehydrogenase (LDH) release assay** section:*
The percentage of LDH release was expressed as the production of LDH released into the medium compared to the total amount of LDH present in cells treated with 2% Triton X-100. The activity was monitored by the oxidation of the reduced form of NADH at 530 nm by an LDH assay kit (Roche, Indianapolis, IN, USA). Cytotoxicity was determined by the equation: [(OD530 of the treated group – OD530 of the control group)/(OD530 of the Triton X-100-treated group – OD530 of the control group)] × 100%.[](https://www.ncbi.nlm.nih.gov/mesh/D017830)
## Western blotting
*In the **Western blotting** section:*
Total cellular extracts (30 µg) were prepared and separated on 8% sodium dodecylsulfate (SDS)-polyacrylamide mini gels for PARP detection and 12% SDS-polyacrylamide minigels for caspase-3, caspase-9, the Bcl-2 family, tERK, pERK and α-tubulin detection, and transferred to Immobilon polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA). Membranes were incubated at 4 °C with 1% bovine serum albumin (BSA) and then incubated with the indicated antibodies for a further 3 h at room temperature followed by incubation with an alkaline phosphatase-conjugated immunoglobulin G (IgG) antibody for 1 h. Proteins were visualized by incubating with the colorimetric substrates, nitroblue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## Measurement of reactive oxygen species (ROS) generation by colonic carcinoma cells
*In the **Measurement of reactive oxygen species (ROS) generation by colonic carcinoma cells** section:*
Intracellular production of ROS by colonic carcinoma cells under different treatments was measured by oxidation of DCFH-DA to DCF. DCFH-DA is a non-polar compound that readily diffuses into cells, where it is hydrolyzed to the non-fluorescent polar derivative, DCFH, and is thereby trapped within cells. When DCFH-DA is oxidized, it turns into the highly fluorescent DCF. After treatment, cells were incubated in the dark for 10 min at 37 °C with 50 µM DCFH-DA, then 104 cells were acquired per sample, the fluorescence of cells was analyzed using a FACScan (Becton Dickinson, Sunnyvale, CA, USA) flow cytometer with excitation at 488 nm and emission at 530 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
## DNA fragmentation assay
*In the **DNA fragmentation assay** section:*
Trypsinized cells were washed with ice-cold phosphate-buffered saline (PBS) and fixed in 70% ethanol at −20 °C for at least 1 h. After fixation, cells were washed twice with PBS and incubated in 1 ml of 0.5% Triton X-100/PBS at 37 °C for 30 min containing 1 mg/ml of RNase A, followed by staining with 1 ml of 50 µg/ml propidium iodide (PI) for 10 min. Fluorescence emitted from the PI-DNA complex was quantitated after excitation of the fluorescent dye by FACScan flow cytometry (Becton Dickenson, San Jose, CA, USA). Ratios of cells at the G2/M and sub-G1 phases were measured, and expressed as percentages (%) of total counts.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Measurement of the mitochondrial membrane potential (MMP)
*In the **Measurement of the mitochondrial membrane potential (MMP)** section:*
After different treatments, cells were incubated with 40 nM DiOC6(3) for 15 min at 37 °C, then washed with ice-cold PBS, and collected by centrifugation at 500×g for 10 min. Collected cells were resuspended in 500 ml of PBS containing 40 nM DiOC6(3). Fluorescence intensities of DiOC6(3) were analyzed on a flow cytometer (FACScan, Becton Dickinson) with excitation and emission settings of 484 and 500 nm, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/C007392)
## Analysis of caspase-3, -8, and -9 activities
*In the **Analysis of caspase-3, -8, and -9 activities** section:*
Ac-DEVD-pNA was used as a colorimetric protease substrate to detect caspase-3 activity. After different treatments, cells were collected and washed three times with PBS and resuspended in 50 mM Tris–HCl (pH 7.4), 1 mM EDTA, and 10 mM EGTA. Cell lysates were clarified by centrifugation at 15,000 rpm for 3 min, and clear lysates containing 200 µg of protein were incubated with 100 µM of the indicated specific colorimetric substrates at 37 °C for 1 h. Alternative activities of indicated caspase-3, -8, and -9 enzymes were described as the cleavage of colorimetric substrate by measuring the absorbance at 405 nm.[](https://www.ncbi.nlm.nih.gov/mesh/C430230)
## Detection of cell cycle progression and hypodiploid cells by EVO in colon carcinoma cells
*In the **Detection of cell cycle progression and hypodiploid cells by EVO in colon carcinoma cells** section:*
Cells were plated in 24 well plates in duplicate, then incubated for 24 h. Media were removed and different treatments were added to each well. Cells were treated for 12 h and the supernatant and cells were then harvested by exposing the cells to 0.25%, Trypsin-EDTA solution for 10 min, then centrifuged and washed in phosphate buffered saline (PBS), fixed in 3 mL ice-cold 100% ethanol. All samples were incubated for 30 min at room temperature in the dark. Cell cycle distribution and hypodiploid cells were determined using FACSan Flow Cytometer (FACScan, Becton Dickinson).[](https://www.ncbi.nlm.nih.gov/mesh/D004492)
## Statistical analysis
*In the **Statistical analysis** section:*
Values are expressed as the mean±standard error (SE) of triplicate experiments. The significance of the difference from the respective controls for each experimental was assayed using a one-way analysis of variance (ANOVA) with a post-hoc Bonferroni analysis when applicable, and p values of <0.05 or <0.01 were considered statistically significant.
## Supporting Information
*In the **Supporting Information** section:*
# References
*In the **References** section:*
|
# Introduction
Biantennary [oligoglycines](https://www.ncbi.nlm.nih.gov/mesh/D009842) and [glyco-oligoglycines](https://www.ncbi.nlm.nih.gov/mesh/D006020) self-associating in aqueous medium
# Abstract
*In the **Abstract** section:*
Summary
Oligoglycines designed in a star-like fashion, so-called tri- and tetraantennary molecules, w[ere found to ](https://www.ncbi.nlm.nih.gov/mesh/D009842)form highly ordered supramers in aqueous medium. The formation of these supramers occurred either spontaneously or due to the assistance of a mica surface. The driving force of the supramer formation is hydrogen bonding, the polypeptid[e ch](https://www.ncbi.nlm.nih.gov/mesh/C011934)ain conformation is related to the folding of helical pol[yglycine](https://www.ncbi.nlm.nih.gov/mesh/D006859) II (PG II). Tri- and tetraantennary molecules are capable of association if the a[ntenna length ](https://www.ncbi.nlm.nih.gov/mesh/C011080)re[ach 7](https://www.ncbi.nlm.nih.gov/mesh/C011080) glycine (Gly) residues. Properties of similar biantennary molecules have not been investigate[d yet, ](https://www.ncbi.nlm.nih.gov/mesh/D005998)an[d w](https://www.ncbi.nlm.nih.gov/mesh/D005998)e compared their self-aggregating potency with similar tri- and tetraantennary analogs. Here, we synthesized oligoglycines of the general formula R-Glyn-Х-Glyn-R (X = -HN-(СН2)m-NH-, m = 2, 4, 10; n = 1[–7) without p](https://www.ncbi.nlm.nih.gov/mesh/D009842)endant ligands (R = H) a[nd with two pen](https://www.ncbi.nlm.nih.gov/mesh/D009842)dant sialoligands (R = sialic acid or sialooligosaccharide). Biantennary o[l](https://www.ncbi.nlm.nih.gov/mesh/D006859)igoglycines formed PG II aggregates, thei[r propertie](https://www.ncbi.nlm.nih.gov/mesh/D012794)s, h[owever, differ from ](https://www.ncbi.nlm.nih.gov/mesh/C016735)those of the co[rresponding t](https://www.ncbi.nlm.nih.gov/mesh/D009842)ri- and [tetra](https://www.ncbi.nlm.nih.gov/mesh/C011080)antennary oligoglycines. In particular, the tendency to aggregate starts from Gly4 motifs instead of Gl[y7. The antiv](https://www.ncbi.nlm.nih.gov/mesh/D009842)iral activity of end-glycosylated peptides was studied,[ an](https://www.ncbi.nlm.nih.gov/mesh/D005998)d all capable of ass[emb](https://www.ncbi.nlm.nih.gov/mesh/D005998)ling glycopeptides demonstrated an antiviral potency which was up to 50 times higher than the acti[vity of pepti](https://www.ncbi.nlm.nih.gov/mesh/D006020)de-free glycans.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Recently, we have synthesized and described tetraantennary [2] and triantennary [3] oligoglycines capable of spontaneous or surface-promoted formation of flat layers in aqueous medium. These layers are one or two molecules thick. The stability was attributed to the formation of a network of hydrogen bonds. This class of supramers has been called tectomers. Tectomers in a layer are packed by polyglycine II type (PG II)
[4–5]. Their helical polypeptide chain fundamentally differs from the canonical α-helix. The association of symmetrical tetra- and triantennary (star-like) oligoglycines spontaneously proceeds only when the number of glycine (Gly) residues in a chain (n) is equal or greater to seven. Oligoglycines with an antennae size less than seven do either not associate at all or require extremely favorable conditions, in particular surface promotion. Yet, the properties of similar biantennary molecules were not investigated.
Here, we synthesized biantennary oligoglycines and studied them in order to determine the necessary and sufficient conditions for self-association. More specifically, we investigated the combination of structure elements, such as the n value, the type of terminal substituents, and the type of structural motifs (core), where the antennae are connected to each other. The knowledge of the rules found for the unsubstituted assembly of oligoglycines may be suitable for us for the design of corresponding sialo derivartives, which are candidate therapeutics for the blocking of the influenza virus [6].[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Results and Discussion
*In the **Results and Discussion** section:*
## Synthesis of biantennary oligoglycines and their glyco derivatives
*In the **Synthesis of biantennary oligoglycines and their glyco derivatives** section:*
The synthesized biantennary oligoglycines and their glyco derivatives are presented in Fig. 1. Analogously to tri- and tetraantennary molecules, oligoglycine antennas are connected according to the ‘head-to-head’ principle, i.e., by their C-termini, so that the two amino groups are terminal. The obtained compounds differ threefold. Firstly, they differ by core X nature: hydrophilic oligoethylene glycol (OEG), hydrophobic flexible decamethylene (C10), or short ethylene (C2). Secondly, the length of oligoglycine antennas, i.e., the number of glycine residues in a chain (n = 1–7) is different.
Thirdly, the substances differ by the presence or absence of carbohydrate fragment (Sug), containing α-N-acetylneuraminic moiety (Neu5Acα).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
The structure of biantennary oligoglycines and their glyco derivatives (sp = spacer group).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Diamines NH2–X–NH2 were the starting substances for the synthesis, oligoethylene glycol diamine was obtained from ditosylate as described in [7–8]. The synthesis of biantennary oligoglycines was carried out by means of the activated esters method (Scheme 1) [9]. The glycine chains were elongated stepwise by their N-terminus by using N-oxysuccinimidyl esters (BocGlyONSu or BocGly2ONSu, Boc = tert-butyloxycarbonyl). Boc-peptides were isolated from the reaction mixture by the removal of the solvent and the re-crystallization of the reaction product from aqueous methanol (yields 75–95%).
In the case of poorly soluble products the impurities and starting materials were washed off with methanol (yields 60–90%). In the case of an oil-like substance (Х = ОEG, n = 2) chromatography on silica gel was performed. The quantitative removal of Boc groups was achieved by the treatment of the obtained peptides with trifluoroacetic acid. Salt forms (trifluoroacetates or hydrochlorides) of diamino derivatives were obtained by sedimentation from an aqueous solution by methanol (yield ≥95%). At later stages of elongation the salts were converted to the respective free bases by treatment with a slight excess of triethylamine.
The preparation of oligoglycines with a chain length exceeding five glycine residues for the derivatives with core C10 and six residues for core C2 failed due to their low solubility and, consequently, the impossibility of separating them from the intermediates of the synthesis.[](https://www.ncbi.nlm.nih.gov/mesh/D003959)
Synthesis of biantennary oligoglycines and their glycoderivatives.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Sialo conjugates of biantennary oligoglycines were obtained from the corresponding diamines and derivatives of α-N-acetylneuraminic acid (Sug-ONp), where the carboxyl group of the spacer was activated with 4-nitrophenol (Np) (Scheme 1). The synthesis of these compounds was described in [10–11]. Owing to the poor solubility of the diamine form of oligoglycines with cores C2 and C10 in DMSO, the reaction was carried out in a saturated aqueous solution of lithium bromide, which prevented the formation of hydrogen bonds and thus increased the solubility of oligoglycines. Glycopeptides were isolated from the reaction mixture by gel-permeation chromatography (yields 70–75%). The peptide modification by the amino group with mono- or oligosaccharides dramatically increased their solubility in water.
This may support their antiviral action (see below), because glycopeptides act topically, in the respiratory tract, and are administered as a spray.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
We then investigated the ability of synthesized biantennary oligoglycines to assemble in aqueous media as well as the antiviral activity of glycoderivatives.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Study of biantennary oligoglycines association in solution by dynamic light scattering
*In the **Study of biantennary oligoglycines association in solution by dynamic light scattering** section:*
The size of the particles formed by the biantennary oligoglycines in solution was measured with the dynamic light scattering method (DLS). We found that the ability of association depends on the number of the glycine units in the antennae, the nature of the core, the pH, and the peptide concentration.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
It is known that the charge of terminal amino groups of the protonated form of oligoglycines hinders association. To overcome this obstacle an equimolar quantity of NaHCO3 or Nа2CO3 was added to aqueous solutions of oligoglycine salts. In the absence of the base, pH values of oligoglycine salt solutions varied from 3.5 to 4.5 (hereinafter denoted as рН < 5). In the case of the addition of one base equivalent per one amino group the solution becomes neutral (рН 6.5), in the case of two Nа2CO3 equivalents the pH value is more than 8.5 (basic solution, denoted as pH > 8.5).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
At n < 4 peptides with an oligoethylene glycol core and the cores С2 and С10 did not form associates in aqueous medium in all the studied ranges of pH and concentration (0.1–1.0 mg/mL).[](https://www.ncbi.nlm.nih.gov/mesh/D011092)
Biantennary oligoglycines, cores С2 and С10, n ≥ 4, are capable of forming associates (700–900 nm) in acidic solutions in the studied concentration range, except for Н-Gly4NH(СН2)10NHGly4-Н·2HCl, which associates in concentrations ≥0.5 mg/mL.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Molecules with the core С2 (n = 4–6) and C10 (n = 5) associate so rapidly in neutral and basic media that a precipitate is formed (data for peptide with n = 5 are given in Fig. 2). Only the peptide Н-Gly4NH(СН2)10NHGly4-Н in concentration ≤0.1 mg/mL is capable of forming associates (800–1200 nm) stable in aqueous media (Fig. 2,b).[](https://www.ncbi.nlm.nih.gov/mesh/C036216)
Dynamics of associate formation by biantennary oligoglycines Н-Glyn-NH(СН2)10NH-Glyn-Н. a) n = 4–5, in aqueous solution at рН 6.5 and a concentration of 0.1 mg/mL at t = 0 and рН < 5 before the addition of base and at t > 0 and рН 6.5 after the addition of base. The region of sedimentation is marked with a dotted line. b) n = 4, in aqueous solution at рН 6.5 and 8.5.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Study of biantennary peptides association using scanning force microscopy
*In the **Study of biantennary peptides association using scanning force microscopy** section:*
Scanning force microscopy (SFM) elucidates information not only about the association process both in solution and on a surface, but also about fine details of the formed architectures. Of particular interest are cases characterized by the active participation of the surface in accelerating the self-assembly. To discriminate the processes taking place on the surface from similar processes in liquid volume, measurements were carried out immediately after the deprotonation of oligoglycine salts at incubation times insufficient for a spontaneous association in solution (found out as ≤1 min). The solution was placed on a freshly cleaved surface of mica or graphite, exposed for fixed time intervals (denoted as t exp), followed by the removal of the liquid phase from the surface and the scanning of the sample in tapping mode in air. The contact mode of scanning was used for experiments in a liquid cell.
Experiments in a liquid cell allowed us to study the kinetics of the process without a possible distortion of the nanostructures resulting from the drying of the sample.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
The Raman spectra (Fig. 3) of biantennary oligoglycines capable of association as well as the spectra of tri- and tetraantennary peptides described earlier display bands at 884, 1261, 1382, 1424 and 1654 cm−1, which are characteristic and specific for crystalline PG II. Based on the presence of these bands we conclude that the structure organization of associates formed in solution corresponds to PG II. The sensitivity of routine Raman scattering method is insufficient for the work with oligoglycine monolayers, so indirect methods were used in order to attribute formed material to a PG II structure. More specifically, geometrical parameters of the layers were determined by using SFM and compared with: 1) those for tectomers (attributed to PG II, see above) formed in solution and 2) calculated values for different, not only PG II, models. As shown below for particular examples, in most cases spontaneous and surface-mediated assembly led to associates of PG II structure, i.e., tectomers.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Raman spectra of a) [H-Gly7-NHCH2]4C; b) H-Gly4-NH(CH2)2NH-Gly4-H; c) H-Gly4-NH(CH2)10NH-Gly4-H. The spectra contain characteristic bands at 884, 1261, 1382, 1424 and 1654 cm−1, corresponding to the structure PG II [12]. Spectra were recorded for the samples in solid phase.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
The formation of the PG II structure for oligoglycines with a short rigid spacer С2 is only possible if the molecule is extended, i.e., antennas are pointing in opposite directions (conformation “1 + 1”, Fig. 4а). The presence of a flexible core (C10, OEG) allows the molecule to adopt the conformation ”2 + 0”, which is characterized by unidirectional oligoglycine antennas. The hydrophobic side of the tectomer should initiate the formation of the second layer with the opposite orientation of the monomer in aqueous solutions (Fig. 4) in order to minimize the thermodynamically unfavorable contact with water.[](https://www.ncbi.nlm.nih.gov/mesh/C011080)
Model of the formation of tectomer layers by biantennary oligoglycines on a mica surface. The heights are given for Н-Gly4NH(СН2)10NHGly4-Н.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
It was demonstrated for biantennary oligoglycines that a concentration of 0.1 mg/mL is optimal for the study of the dynamics of tectomer growth on a mica surface. At higher concentrations the growth both in solution and on the surface proceeded so rapidly that the dynamics study was considered impossible. No tectomer structure was observed under acidic conditions (рН < 5), whereas under neutral and basic conditions the reaction proceeded similarly in terms of both the velocity and the morphology of formed tectomers. The oligoethylene glycol derivatives Н-Glyn-NH(СН2СН2О)3СН2СН2NH-Glyn-Н (n = 2–7), non-associating in aqueous solutions as well as oligoglycines with cores С2 and С10 (n < 4) did not form associates on a mica surface under all studied ranges of pH (from 4.5 to 8.5).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
According to dynamic light scattering data (see above) only peptide Н-Gly4NH(СН2)10NHGly4-Н was capable of forming tectomers in neutral and basic solutions which were unchanged in an aqueous phase for a long time. Fig. 5 demonstrates the dynamics of layer growth on mica with the characteristic formation of islet structures (t exp = 0.5 min, Fig. 5а), growing laterally (t exp = 1 min, Fig. 5), and covering the whole surface with an even layer (t exp = 2 min, Fig. 5). Presumably, longer times (t exp > 2 min) are characterized by the appearance of multilayer tectomers resulting from the sorption of associates formed in solution. The multilayer tectomers can be readily removed by washing with buffer solution (рН 6.5 or 9.0).
The morphology of the first layer remains unchanged and the available defects (‘holes’) are preserved. The layer height is 3.7–4.0 nm, which may correspond to both mono- and bilayer (conformations “1 + 1” and “2 + 0”, respectively, see Fig. 4).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Growth of the tectomer formed by the peptide Н-Gly4-NH(СН2)10-NHGly4-Н (concentration 0.1 mg/mL) on a mica surface at рН 6.5, with tapping mode, SFM on air, and a t еxp of а) 0.5 min, b) 1 min, c) 2 min. Here, the shown field is completely covered with tectomer layer; its roughness found to be ± 0.1 nm. Т indicates the tectomer layer, M the uncovered mica regions.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
The dynamics of Н-Gly4NH(СН2)10NHGly4-Н association was studied in more detail in a liquid cell (Fig. 6). After 3 min the surface was virtually completely covered with a uniform defect-free layer. It should be noted that the stepwise surface profile (typical for bilayer structures) was not observed in a liquid cell. The layer morphology was identical to the one observed in experiments in air (Fig. 5).[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Growth of the tectomer formed by peptide Н-Gly4-NH(СН2)10NH-Gly4-Н (concentration 0.1 mg/mL) on a mica surface in a liquid cell at рН 6.5. Phase SFM images were taken with a) t exp = 1 min, b) t exp = 2 min and c) t exp = 3 min where t exp is the time after the experiment was started. Т indicates the tectomer layer, M the uncovered mica regions.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Dynamic light scattering data give evidence that in neutral aqueous solutions the association of compounds Н-Gly5-NH-X-NH-Gly5-Н (cores С2 and С10) leads to the formation of large aggregates. By means of SFM it was demonstrated that the peptide with core C2 formed islet-like tectomers on mica surface (t exp = 10 min) with a height of 3.3 nm and planar dimensions of 500–700 nm (Fig. 7). The compound with core C10 associated more rapidly (Fig. 7), though the surface was not completely covered (t exp = 10 min). This is in contrast to the structure analog with four glycines in the antenna, where a time period of only two minutes was sufficient for complete covering. The measured heights fit the model “1 + 1”.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
SFM images of associates formed by peptides а) Н-Gly5-NH(СН2)2NH-Gly5-Н and b) Н-Gly5-NH(СН2)10NH-Gly5-Н (concentration 0.1 mg/mL) on a mica surface at рН 6.5 with t еxp = 10 min SFM tapping mode in air, and an incubation time in solution of 1 min. Surface profiles, schematic layer models, and their calculated heights are given on the right.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
The obtained data give evidence that mica promotes the formation of tectomers from biantennary oligoglycines in neutral and basic solutions. Layer growth proceeds due to the surface co-participation. In the case of the molecule Н-Gly4-NH(СН2)10NH-Gly4-Н growth continues until the surface is completely covered, whereas in the bulk of the liquid phase dimensions remain unchanged over time according to dynamic light scattering data.[](https://www.ncbi.nlm.nih.gov/mesh/C011934)
There are no direct data pointing at the particular conformation (“1 + 1” or ”2 + 0”) the peptide monomer has in the layer (Fig. 4). The height value does not allow for the unambiguous assignment of one of the proposed models. Nevertheless, all intrinsic data supports the model ”1 + 1”: 1) the steps typical for a bilayer profile are not present in the SFM images, 2) intense washing does not lead to the formation of half-height structures, 3) the lack of any association on a graphite surface where the formation of ”2 + 0” is expected to be preferable (see below).[](https://www.ncbi.nlm.nih.gov/mesh/D006108)
Mica promotes the assembling of amino-terminated chains due to the negative charge of the surface. In contrast, graphite did not participate in the association of biantennary oligoglycines, we observed only irregular associates formed in solution. This experimental result is unexpected, because the formation of a monolayer with the monomer conformation ”2 + 0” is favorable according to molecular dynamics simulations [13].[](https://www.ncbi.nlm.nih.gov/mesh/C011934)
## Minimal size of Glyn fragment providing association
*In the **Minimal size of Glyn fragment providing association** section:*
In the case of biantennary molecules association formally starts from the value n = 4, but, in fact, this value is supposedly equal to 8 because biantennary peptides form a polyglycine structure in the extended conformation ”1 + 1”. It is noteworthy that related polymers, nylons with the formula -NH(CH2)xCO- are known to form a PG II structure [14], i.e., additional methylene groups (-(CH2)x- instead of -CH2-) have no significant influence on its ability to form PG II. The association of tri- and tetraantennary peptides leads to structures with the monomer conformations ”2 + 2” and ”3 + 0”, respectively. The association starts from n = 7.
Presumably, the first and closest to forking Gly residue takes a distorted conformation and does not take part in the formation of hydrogen bonds with neighboring residues. In the case of tetraantennary oligoglycines the plane of one pair of antennas is rotated by 90° [8] with respect to another pair, so that the glycines cannot form a continuous chain. On the other hand, in the biantennary analog the chain Gly4-Х-Gly4 has the ability to form a PG II structure despite the core fragment -(CH2)n-.[](https://www.ncbi.nlm.nih.gov/mesh/C011080)
## The nature of the core fragment Х in H-Glyn-Х-Glyn-H
*In the **The nature of the core fragment Х in H-Glyn-Х-Glyn-H** section:*
Biantennary molecules with core С10 form tectomers on a mica surface more readily than the molecules with core C2 and an equal number of glycines. The more flexible core C10 should lead to a entropy driven destabilization. The opposite effect observed in reality is most probably caused by van der Waals interactions of hydrophobic fragments C10 closely situated in the PG II structure. The oligoethylene glycol core abolishes the formation of a PG II structure, presumably due to competitive hydrogen bonding with spatially close oligoglycine fragments.[](https://www.ncbi.nlm.nih.gov/mesh/D000473)
## Spontaneous and surface-promoted association
*In the **Spontaneous and surface-promoted association** section:*
The formation of tectomers on mica proceeds considerably more rapidly than association in solution (the formation of associates in solutions just starts when assembling on the surface is already finished), i.e., the mica surface obviously plays an active role in the process. Tectomer growth starts from the formation of islet structures that increase in lateral direction and cover the whole surface in an even layer. This growth is limited only by the dimensions of the support itself. It should be noted that graphite, in contrast to mica, does not promote association.[](https://www.ncbi.nlm.nih.gov/mesh/C011934)
## Effect of рН value
*In the **Effect of рН value** section:*
Depending on the pH value, the free terminal group of the oligoglycine chain can be heavily charged, weakly charged, or neutral. In acidic solutions antennas are repulsed due to the positive charge, which hinders tectomer assembly or even abolishes it. The pH value effects not only the ability to assemble but also the morphology of forming supramers. Thus, in neutral solutions biantennary oligoglycines form multilayer tectomers. The process is unstoppable at the stage of the monolayer formation.
At the same time, monolayer tectomers are exclusively formed in basic solutions.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Concentration range
*In the **Concentration range** section:*
Most parts of the experiments were carried out in the concentration range of 0.1–1.0 mg/mL. A concentration of 0.1 mg/mL was used for the adequate comparison of association of all investigated peptides. The association in the liquid phase proceeds slower at low concentrations leading to an increased size of the formed supramers.
In summary, based on our investigations related to unglycosylated molecules we can conclude that the association of biantennary oligoglycines is affected by several factors. 1) Mica but not graphite promotes the formation of tectomers. 2) The spatial organization of oligoglycine molecules in supramers corresponds to PG II conformation. 3) Not less than four glycine residues in each of two antennas are required for the assembling of monomer layers into surface tectomer layers or into long-living associates in solution.
4) Oligoethylene glycol core ‘inhibits’ the association both in the liquid phase and on a mica surface.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Antiviral activity of glycoderivatives
*In the **Antiviral activity of glycoderivatives** section:*
The idea of antiadhesion influenza virus therapy is based on the inhibition or the blocking of the binding of the influenza virus with target cells [15]. Monovalent oligosaccharides are incapable of an efficient competition for analogous glycans on the cell surface due to the low binding constant with viral hemagglutinin. An attractive way of increasing the affinity of a blocker (inhibitor) is the design of multivalent receptor analogs such as the oligoglycine-based tectomers described above. A first success for an application in this regard was achieved by inhibiting the influenza virus by sialo derivatives of the associating tetraantennary peptides, which demonstrated an antiviral activity three orders of magnitude higher than the activity of non-associating analogs [2]. Similar triantennary molecules with sialo-glycan located in the molecule “head”, however, appear to display a low activity [16].
Thus, it was interesting to study the antiviral activity of sialo derivatives of biantennary oligoglycines in relation to their propensity to associate in aqueous solutions.[](https://www.ncbi.nlm.nih.gov/mesh/D009844)
The fact that a sialylated biantennary peptide is capable of association in an aqueous solution similarly to glycan-free peptides is confirmed by DLC data. The average size of the sialoglycopeptide aggregates in aqueous solution was about 1 μm (data not shown).
The antiviral activity of biantennary glycopeptides was studied by means of a fetuin binding inhibition test (FBI-test) [17] (the glycoprotein fetuin contains several sialylated carbohydrate chains). In this test, the glycopeptides inhibited the binding of a fetuin peroxidase conjugate to a virus immobilized on a plastic (related to the corresponding monomer). Results are given in Table 1. The activity of associating sialooligoglycines with core C2 was only 3–6 times higher than the activities of their non-associated counterparts (n = 2–4) and the monomeric reference sialoside, Neu5AcαBn.
The compound with core С10 and n = 4 demonstrated the highest activity from the studied biantennary glycopeptides, which was 50 times higher than the activity of the monomer. The activity of bivalent derivatives with core OEG and n = 2–5 did not exceed that of monovalent sialoside. However, the activity increased dramatically when n = 6 (up to 50 times), although it was still orders of magnitude smaller compared to the high activity of polymeric inhibitors [18]. As sialooligoglycines of the OEG series did not associate in aqueous solution, we suppose that the reason for the increased activity is related to a critical distance, which facilitates the realization of a divalent interaction of this bivalent molecule with a viral hemagglutinin. Indeed, a simple calculation demonstrates that this distance in a maximally extended molecule with n = 6 is about 100 Å. This distance value corresponds to the distance between the carbohydrate binding sites in one molecule of a hemagglutinin homotrimer and slightly exceeds the distance between a couple of hemagglutinin trimmers, which are closely situated on the virion surface.[](https://www.ncbi.nlm.nih.gov/mesh/D006020)
Relative activity of biantennary glycopeptides in the influenza virus receptor binding inhibition assay [17].[](https://www.ncbi.nlm.nih.gov/mesh/D006020)
aAbbreviations: sp1 = -OCH2(p-C6H4)NHCOCH2NH-CO(CH2)4CO; Bn – benzyl; sp2 = -O(CH2)3NHCO(CH2)4CO; 6’SLN = Neu5Acα2-6Galβ1-4GlcNAcβ; sp3 = -NHCOCH2NHCO(CH2)4CO; 3’SL = Neu5Acα2-3Galβ1-4Glcβ. bValues of IC50, μM, for monomeric Neu5AcαOBn, 6'SLN and 3’SL are given in parentheses.
## Experimental
*In the **Experimental** section:*
Reagents and solvents were bought from Merck and Sigma–Aldrich. Activated esters BocGlyONSu and BocGly2ONSu were prepared as described earlier [9] from glycine or glycylglycine (Acros). Ethylenediamine and 1,10-diaminodecane were supplied from Sigma-Aldrich, and diamine NH2(CH2CH2O)3CH2CH2NH2 (1) was synthesized from ditosylate TosO(CH2CH2O)3CH2CH2OTos (Sigma–Aldrich) according to the described methods [7–8].[](https://www.ncbi.nlm.nih.gov/mesh/D004952)
Silica gel (Kieselgel 60, Merck, Germany) was used for low-pressure column chromatography. Sephadex LH-20 (Pharmacia Biotech, Austria) was employed for gel chromatography. Thin-layer chromatography (TLC) was performed on foil plates covered with silica gel (Kieselgel 60, Merck, Germany).[](https://www.ncbi.nlm.nih.gov/mesh/D058428)
1Н NMR spectra were recorded on a Bruker spectrometer (600, 700, 800 MHz) at 303 K. Chemical shifts (δ) for characteristic signals in 1Н NMR spectra are given in ppm and spin–spin coupling constants (J) in Hz. The scale of the chemical shifts was calibrated against the signals of residual protons of solvents (CDCl3: δ 7.26 ppm; DMSO-d 6: δ 2.50 ppm; D2O: δ 4.75 ppm). Mass-spectra were recorded on the time-of-flight spectrometer Vision-2000 (Thermo Bioanalysis, UK) with MALDI with 2,6-dihydroxybenzoic acid as reference. Raman spectra were recorded on a spectrometer Ramanor HG-2S (Jobin Yvon) with the monochromator Anaspec 300S and argon (λ = 514.5 nm, Spectra Physics, model 164-03).[](https://www.ncbi.nlm.nih.gov/mesh/D002725)
## Synthesis of biantennary oligoglycines
*In the **Synthesis of biantennary oligoglycines** section:*
Protocol 1: Elongation of the oligoglycine chain (Boc-GlynNH-X-NHGlyn-Boc; n = 1–7, X = С2, С10 and OEG). Et3N (8 mmol) followed by BocGlyONSu or BocGly2ONSu (3 mmol) were added to a solution of diamine (1 mmol) in dimethyl sulfoxide (DMSO; 5 mL). The reaction mixture was stirred until the disappearance of the starting diamine (1–24 h, TLC control) and the solvent was removed under vacuum. The dry residue was suspended in methanol, filtered, dissolved in water, sedimentated with methanol, and dried in vacuo.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
Protocol 2: Preparation of oligoglycines (HCl·GlynNH-X-NHGlyn·HCl; n = 1–7, X = С2, С10 and OEG). The Boc-derivative (0.5 mmol) was dissolved in trifluoroacetic acid (5 mL), the reaction mixture was kept for 2 h at room temperature, co-evaporated with toluene (2 × 10 mL) and 1 M HCl aqueous solution (1–2 mL), and finally with a mixture iPrOH/methanol 1:1 (2 × 10 mL). The obtained product was sedimentated from the aqueous solution by the addition of methanol and dried in vacuo.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
## Synthesis of associating glycopeptides
*In the **Synthesis of associating glycopeptides** section:*
Protocol 3: Neu5Acα-sp1-ONp, 3'SL-sp3-ONp or 6'SLN-sp2-ONp (4 μmol) were added to a solution of diamine (1 μmol) in DMSO or saturated aqueous solution of LiBr (200 μL). NEt3 (4 μmol) was added until a рН of 8 was reached, and the mixture was stirred for 24 h at room temperature. Exclusion chromatography on Sephadex LH-20 (eluent: 0.1 М solution of NH3 in the mixture acetonitrile/water, 1:1). Fractions containing pure product were combined and evaporated. Dry residue was dissolved in water and freeze-dryed.[](https://www.ncbi.nlm.nih.gov/mesh/D012794)
## Dynamic light scattering experiments
*In the **Dynamic light scattering experiments** section:*
The light scattering of aqueous solutions of biantennary oligoglycines was studied with an analyzer of submicron particle size “Malvern HPPS” (UK). After the preparation of aqueous (Milli-Q) solutions of oigoglycine salts in a concentration of 0.01–0.1 mg/mL, the instrument reading was recorded (t = 0, pH < 5). Then, 1–2 equiv of base (0.1 М aqueous solution of NaHCO3 or Na2CO3) per amino group was added to the solution of the analyzed oligoglycine salt (t > 0, pH 6–8), and instrument readings were recorded in fixed periods of time. In the case of the formation of large associates (intense opalescence, sedimentation), whose dimensions exceeded the working limit of the instrument, the experiment was stopped.[](https://www.ncbi.nlm.nih.gov/mesh/D009842)
For experiments with biantennary sialooligoglycines their aqueous (Milli-Q) solutions with a concentration of 0.1 mg/mL were used.[](https://www.ncbi.nlm.nih.gov/mesh/D006020)
## Scanning force microscopy (SFM)
*In the **Scanning force microscopy (SFM)** section:*
The samples were imaged with a Nanoscope IIIa instrument (Digital Instruments, USA). Commercial silicon nitride cantilevers with force constants of 0.06, 0.12, and 0.32 Nm−1 were used for the measurements in contact mode in liquid cell. Cantilevers with a resonance frequency of about 300 kHz and a force constant of 42 Nm−1 were used for the SFM tapping mode in air. Software WSxM (Nanotec Electronica, Spain) was employed for the image treatment. Pure water (Fluka) was used for the preparation of solutions.[](https://www.ncbi.nlm.nih.gov/mesh/C032734)
Scanning in air. 1–2 equiv of 0.1 M of aqueous solution of NaHCO3 or Na2CO3 per amino group (pH ~ 6–8) was added for the deprotonation to a freshly prepared solution of oligoglycine salt (0.1–1.0 mg/mL; pH < 5), and incubated for a specified time period in the range of 0 to 90 min. Then the solution was applied on the freshly cleaved mica or graphite, and kept for a specified period of time within the range of 0 to 10 min. Liquid was removed from the surface by spin coating or in nitrogen flow. Structures formed on the surface were visualized in tapping mode SFM.[](https://www.ncbi.nlm.nih.gov/mesh/D017693)
Scanning in liquid cell . A plate of freshly cleaved mica (1 × 1 cm2) was placed in a liquid cell. The cell was filled with water (25 μL) and the instrument was set up. Then, water was changed with a freshly prepared solution of deprotonated peptide (see scanning in air above) and the surface was scanned in contact mode SFM in fixed time periods.[](https://www.ncbi.nlm.nih.gov/mesh/C011934)
The influenza virus receptor-binding inhibition assay was carried out as described in [17].
## Supporting Information
*In the **Supporting Information** section:*
|
# Introduction
[Unsaturated Fatty Acid](https://www.ncbi.nlm.nih.gov/mesh/D005231), [cis-2-Decenoic Acid](https://www.ncbi.nlm.nih.gov/mesh/C052476), in Combination with Disinfectants or Antibiotics Removes Pre-Established Biofilms Formed by Food-Related Bacteria
# Abstract
*In the **Abstract** section:*
Biofilm formation by food-related bacteria and food-related pathogenesis are significant problems in the food industry. Even though much disinfection and mechanical procedure exist for removal of biofilms, they may fail to eliminate pre-established biofilms. cis-2 decenoic acid (CDA), an unsaturated fatty acid messenger produced by Pseudomonas aeruginosa, is reportedly capable of inducing the dispersion of establis[hed biofilms by mul](https://www.ncbi.nlm.nih.gov/mesh/C052476)ti[ple](https://www.ncbi.nlm.nih.gov/mesh/C052476) types[ of microorganisms. Ho](https://www.ncbi.nlm.nih.gov/mesh/D005231)wever, whether CDA has potential to boost the actions of certain antimicrobials is unknown. Here, the activity of CDA as an inducer of pre-established biofilms dispersal, for[med](https://www.ncbi.nlm.nih.gov/mesh/C052476) by four main food pathogens; Staphylococcus aureus, Bacillus cereus, Salmonella enterica and E.[ co](https://www.ncbi.nlm.nih.gov/mesh/C052476)li, was measured using both semi-batch and continuous cultures bioassays. To assess the ability of CDA combined biocides treatments to remove pre-established biofilms formed on stainless steel discs, CFU counts were performed for both treated and untreated c[ult](https://www.ncbi.nlm.nih.gov/mesh/C052476)ures. Eradication of the biofilms by CDA combined antibiotics was evaluated[ using crystal ](https://www.ncbi.nlm.nih.gov/mesh/D013193)violet staining. The effect of CDA combined treatments (antibiotics and disinfectants) on biofilm surface [are](https://www.ncbi.nlm.nih.gov/mesh/C052476)a and bacteria viability was evaluated usi[ng fluorescenc](https://www.ncbi.nlm.nih.gov/mesh/D005840)e microscopy, digital ima[ge ](https://www.ncbi.nlm.nih.gov/mesh/C052476)analysis and LIVE/DEAD staining. MICs were also determined to assess the probable inhibitory effects of CDA combined treatments on the growth of tested microorganisms' planktonic cells. Treatment of pre-established biofilms with only 310 nM CDA resulted in at lea[st ](https://www.ncbi.nlm.nih.gov/mesh/C052476)two-fold increase in the number of planktonic cells in all cultures. While antibiotics or disinfectants alone exerted a trivial effect[ on](https://www.ncbi.nlm.nih.gov/mesh/C052476) CFU counts and percentage of surface area covered by the biofilms, combinational treatments with both 310 nM CDA and antibiotics or disinfectants led to approximate 80% reduction in biofilm biomass. These data suggests that combined treatments with CDA would pave the [way](https://www.ncbi.nlm.nih.gov/mesh/C052476) toward developing new strategies to control biofilms with widespread applications in industry as well as medicine.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
The biofilm mode of growth is a basic survival strategy deployed by microorganisms in a wide range of environmental, industrial and clinical settings . Biofilms are defined as sessile communities of cells attached to each other and/or to surfaces or interfaces which are embedded in a self-produced matrix of extracellular polymeric substances (EPS) , . A function frequently attributed to EPS is their general protective effect on sessile microorganisms against adverse conditions including presence of most antimicrobial agents . This is supposed to be due mainly to physiological characteristics of biofilm bacteria, but also to a barrier function of EPS . According to Körstgens et al. the EPS matrix also provides biofilm mechanical stability by filling and forming the space between the bacterial cells, keeping them together.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
Biofilm formation by food-related bacteria and food-related pathogenesis are significant problems in the food industry. The attachment of the bacteria to the food product or the product contact surfaces leads to serious hygienic problems and economic losses due to food spoilage , .
For the sanitation and removal of biofilms in food industry, chemical agents and mechanical forces (sonication, flushing, etc.) are parameters often involved simultaneously. Mechanical actions only allow the removal of the biofilms from the surfaces and once established, biofilms are harder to be removed completely . They also cannot kill biofilms and biofilm cells might later re-attach to other surfaces and form a biofilm , . Thus, disinfection procedure is indispensible with the intention of killing them. However, it is important to note that most of the disinfection processes that are implemented are based upon the results of planktonic tests .
Therefore, such tests do not mimic the behavior of sessile cells and can be highly ineffective when applied to control biofilms. Biofilms have been reported as possessing susceptibilities towards antimicrobials that are 100–1000 times less than equivalent populations of planktonic counterparts . If a microbial population faces high concentrations of an antimicrobial product, susceptible cells will be inactivated. Although some cells may possess a degree of natural resistance and physiological plasticity or they may acquire it later through mutation or genetic exchange. These processes allow the microorganisms to survive and grow .
To address the need for novel and improved measures against biofilms especially pre-established biofilms, a clear strategy is to study the biofilm life cycle and identify key trigger points that regulate biofilm development. To control biofilm, the last stage of biofilm development presents several advantages, where a coordinated dispersal of biofilm cells is possible. Induction of biofilm dispersal could potentially use the microorganisms' own energy to remove established biofilms, revert cells to a planktonic phenotype and restore their susceptibility to disinfectants and antibiotics.
It has been recently reported that P. aeruginosa produces an un-saturated fatty acid, cis-2-decenoic acid (C10: Δ2, CDA), which is capable of inducing the dispersion of pre-established biofilms by multiple types of bacteria . Furthermore, CDA is also capable of inducing dispersion in biofilms of Candida albicans, indicating that this signalling molecule is involved in inter-species and inter-kingdom signalling where it can modulate the behavior of other microorganisms that do not produce the signal . CDA is a promising candidate for control of biofilms in different industrial and clinical settings as it has a broad-spectrum of activity in addition to the fact that it has no cytotoxic effects to human cells at nano-molar ranges . However, whether CDA has potential to boost the actions of certain disinfectants and antibiotics is unknown.[](https://www.ncbi.nlm.nih.gov/mesh/D005231)
Therefore, in the current work, the ability of nano-molar concentrations of CDA to induce dispersal in pre-established biofilms, formed by four main food-borne biofilm producer bacteria (Bacillus cereus, Staphylococcus aureus, Salmonella enterica and E. coli) as well as to remove and kill their biofilms when combined with biocides or antibiotics were studied? Besides, the ability of CDA to increase the inhibitory effects of antimicrobials on the growth of tested microorganisms' planktonic cells was investigated.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Materials and Methods
*In the **Materials and Methods** section:*
## Bacterial strains, media and growth conditions
*In the **Bacterial strains, media and growth conditions** section:*
The microorganisms used in the present study included E. coli (ATCC 25922), Staphylococcus aureus (ATCC 25923), Bacillus cereus (ATCC 11778) and Salmonella enterica (ATCC 14028). Overnight cultures were grown at optimum temperature for each microorganism in Luria Bertani (LB) medium (Merck, Germany) for E. coli, B. cereus and S. enterica and in Tryptic Soy Broth (TSB) medium (Merck, Germany) for S. aureus. Biofilm experiments were performed in 1/5 strength LB for E. coli, B. cereus and S. enterica, and in 1/5 strength TSB for S. aureus.[](https://www.ncbi.nlm.nih.gov/mesh/C026375)
## Chemicals and antimicrobial compounds
*In the **Chemicals and antimicrobial compounds** section:*
Three different concentrations of CDA (U-Chemo, China) (100, 310 or 620 nM) were used. These concentrations were previously observed to have the most effect on inducing the dispersion of pre-established biofilms with no cytotoxic effects on human cells . Ethanol (10%) (Merck, Germany) was used as a carrier for CDA. Two commercial disinfectants, Epimax S (Epimax, Iran) and Percidine (Behban chemistry, Iran), were used for their widespread applications in food industry in Iran. Their active ingredients were hydrogen peroxide (45–50%) and peracetic acid (15%), respectively.
Final concentration of 120 ppm hydrogen peroxide for Epimax S and 70 ppm peracetic acid for Percidine was used. These concentrations were respectively 3 and 4 times lower than the manufacturer's recommended concentration for disinfection purposes. This study also examined three antibiotics commonly used in medical and veterinary practice; ciprofloxacin (Sigma, USA) for both gram positive and gram negative tested microorganisms, vancomycin (Sigma, USA) for only gram positive bacteria, and ampicillin (Sigma, USA) for gram negative strains. Ciprofloxacin (Sigma) was used at a final concentration of 1 µg.ml−1, vancomycin at (4 µg.ml−1 and 256 µg.ml−1 for S. aureus and B. cereus, respectively) and ampicillin at 256 µg.ml−1. The concentrations of antibiotic selected for use were established in our laboratory to be effective against planktonic cells but have no inhibitory effect on the tested pathogens' biofilm cells.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Biofilm dispersal bioassays in petri dishes
*In the **Biofilm dispersal bioassays in petri dishes** section:*
Biofilms were grown on the inside surface of petri dishes by using a semi-batch culture method in which the medium was replaced every 24 h. This was done to reduce the accumulation of native dispersion inducing factors and to allow mature biofilms form. Biofilms grown in this manner were then treated with three different concentrations of CDA (100, 310 or 620 nM) as dispersion inducer or just the carrier (10% ethanol) as a control to release cells into the bulk liquid and evaluate dispersed cell number by measuring the optical density (OD). To cultivate biofilms, overnight cultures of tested microorganisms were diluted 1∶1,000 into fifteen ml of growth medium, (except for B. cereus that was diluted 200 times), inoculated in sterile petri dishes and incubated at room temperature with 30 rpm shaking. Medium in the plates was replaced every 24 h for 5 days.
After the last exchange of medium, the cells were allowed to grow for about 1 h and then dispersion induction was tested by replacing the growth medium with fresh medium containing one of the indicated concentrations of CDA or just the carrier as a control and the cells were incubated for a further 1 h. Afterward, Medium containing dispersed cells was transferred by pipette to a 50 ml Erlenmeyer and was homogenized for 30 s at 5,000 rpm with a WiseTis-Homogenizer model HG-150 (Daihan Scientific Co., Ltd., Korea) to ensure the separation of cells. The cell density was then determined based on the OD600 with an UV/VIS spectrophotometer model T80+ (PG Instruments, Ltd., China). Biofilm dispersal bioassays were performed in triplicates in at least three individual experiments for each concentration.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Dispersion bioassays of biofilms in biofilm tube reactors
*In the **Dispersion bioassays of biofilms in biofilm tube reactors** section:*
Biofilms were also grown on the interior surfaces of tubing reactors. A continuous once-through tube reactor system was configured by using eight silicone reactor tubes (40-cm length by 3-mm inner diameter), connected to an eight-roller head peristaltic pump (Baoding Longer Precision Pump Co., Ltd., China) and medium reservoir, via an additional silicone tubing. Medium was pumped through the tubing to a closed effluent medium reservoir. The entire system was closed to the outside environment but maintained in equilibrium with atmospheric pressure by a 0.2-µm-pore-size gas-permeable filter fitted to medium reservoir. The assembled system was sterilized by autoclaving prior to inoculation.
The silicone tubes were inoculated by syringe injection through a septum 1 cm upstream from each reactor tube, with 3 ml of overnight cultures of each microorganism. Bacteria cells were allowed to attach (static incubation) to the tubing for 1 h, after which the flow was started at an elution rate of 280 µl.min−1. After 5 days of biofilm cultures, the influent medium was switched from fresh medium in the test lines to one of the three concentrations of CDA. Control lines were switched to new lines containing just the carrier (ethanol 10%). Samples were collected in test tubes on ice and were subsequently homogenized and cell density was determined as mentioned above.
All experiments were repeated three times.[](https://www.ncbi.nlm.nih.gov/mesh/D012828)
The concentration of CDA that induced the most dispersal in the examined biofilms in both petri dish and tube reactor cultures was used for further studies.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Combined CDA and biocide treatment of pre-established biofilms, formed on stainless steel discs
*In the **Combined CDA and biocide treatment of pre-established biofilms, formed on stainless steel discs** section:*
For disinfectants alone and combined CDA susceptibility testing, biofilms were formed on stainless steel (SS) type 316 discs with a surface area of 2.7 cm2, placed at the bottom of wells in 24-well plates. To grow biofilms, 2.5 ml of overnight cultures of each microorganism, previously diluted 1∶1,000 in biofilm medium (except for B. cereus as indicated above), was added to each well and incubated at room temperature with gentle shaking. Medium in the wells was replaced every 24 h for 5 days to allow mature biofilms form. Biofilms were then treated for 1 h with indicated concentrations of disinfectants alone or combined with 310 nM CDA as CDA at this concentration induced the most dispersal in the tested biofilms in both petri dish and tube reactor cultures. At the end of the experimental period, the SS discs were washed with PBS to remove non-adherent bacteria, carefully transferred to sterile glass tubes containing 1 ml of sterile 0.89% NaCl and washed with another 1 ml of 0.89% NaCl.
To remove the biofilm from the SS discs, the glass tubes with the biofilms were placed in an ultrasonic bath for 10 min at room temperature. CFU were enumerated after plating on LB agar to assess bacterial viability. All experiments were repeated three times.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Antibiotics combined CDA biofilm microtiter plate assays
*In the **Antibiotics combined CDA biofilm microtiter plate assays** section:*
To assess the effect of antibiotics alone and in combination with CDA, biofilms were grown on the inside surface of sterile polystyrene 96-well plates. For biofilm cultures, plates were inoculated with 150 µl/well of overnight culture containing the tested organism, previously diluted in growth medium (as indicated above) and incubated at 37 °C with shaking at 120 rpm. Medium within each well was replaced every 24 h for 5 days. Biofilms were then treated for 1 h with indicated concentrations of antibiotics alone or combined with 310 nM CDA. The plates were gently rinsed twice with PBS to remove planktonic and loosely adherent organisms.
After rinsing, the plates were shaken dry and each well of each plate stained with 160 µl of an aqueous 0.1% crystal violet solution in distilled water. After allowing the stain to adhere to the biofilms for 15 min, each plate was again rinsed with PBS until no more stain could be rinsed from the plate. Each plate was again shaken dry, inverted and allowed to dry thoroughly for 30 min. Finally, 170 µl of a 30% acetic acid solution was pipetted into each well to desorb the adhered stain back into solution. After allowing 30 min for the adhered stain to dissolve into the destaining solution, the biofilm in each well was quantified via absorbance at OD590 using a ELx808 Absorbance Microplate Reader (BioTek Instruments, Inc., Winooski, VT) .
All experiments were repeated at least three times.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Combined CDA and antimicrobial treatment of planktonic cells
*In the **Combined CDA and antimicrobial treatment of planktonic cells** section:*
We have evaluated the probable inhibitory effects on the growth of tested microorganisms' planktonic cells by biocides or antibiotics alone and in combination with three different concentrations of CDA (100, 310 or 620 nM). The MICs were determined in triplicate in Mueller-Hinton broth by using microdilution assay with bacteria at a density of 105 CFU/ml. Plates were incubated for 24 h at optimum temperature for each bacterium. The lowest concentration of antibiotics or biocides where there was no growth after 24 h was taken as the MIC , .[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Flow cell (continuous-culture) biofilm experiments; disinfectants and antibiotics sensitivity assays and surface area coverage
*In the **Flow cell (continuous-culture) biofilm experiments; disinfectants and antibiotics sensitivity assays and surface area coverage** section:*
To observe the effect of CDA combined antimicrobial treatments on biofilm surface area and bacteria viability, biofilms were also grown in continuous culture flow cells (channel dimensions, 1×4×40 mm). Appropriate sterile biofilm medium was pumped from a 5-Liter vessel through silicone tubing to the flow cell using an eight-roller-head peristaltic pump (Baoding Longer Precision Pump Co., Ltd., China) at a flow rate of 280 µl.min−1. Medium leaving the flow cell was discharged to an effluent reservoir via silicone tubing. The entire system was closed to the outside environment but maintained in equilibrium with atmospheric pressure by a 0.2-µm-pore-size gas-permeable filter fitted to each vessel. Channels were inoculated with overnight cultures of tested organism and incubated without flow for 1 h, at room temperature.
After 48 h of biofilm cultures, the influent medium was switched from fresh medium in the test lines to the antimicrobials in combination with 310 nM CDA. Control lines were switched to new lines containing only examined antimicrobial agents. After 1 h treatment, biofilms were stained with a LIVE/DEAD BacLight bacterial viability kit (Molecular Probes). The two stock solutions of the stain (SYTO 9 and propidium iodide) were diluted to 3 µl.ml−1 in biofilm medium and injected into the flow channels. Live SYTO 9-stained cells and dead propidium iodide-stained cells were visualized using epifluorescence microscopy (CETI, Belgium).
15 selected fields of view per flow cell were imaged in the XY plane, at regular intervals and across the entire channels. Image analysis (ImageJ Software, NIH) was performed and results were presented as the percentage of total biofilm surface reduction in cultures treated with combined CDA and antimicrobial treatments relative to the total biofilm surface in control cultures that were not exposed to CDA. Three replicates per experiment were used and at least 2 independent repetitions of experiments were performed.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Statistical Analysis
*In the **Statistical Analysis** section:*
All data were analyzed using analysis of variance (ANOVA) by the general linear model procedure of Minitab data analysis software (release 16, Minitab Inc., PA. USA). Pairwise comparisons were then made between all of the groups using Tukey's method. P values <0.05 were regarded as significant. All measurements were carried out in triplicate.
## Results
*In the **Results** section:*
## Very low concentrations of CDA induce biofilm dispersal
*In the **Very low concentrations of CDA induce biofilm dispersal** section:*
We investigated the effect of exposure to nano-molar concentrations of CDA on pre-established biofilms in the petri dish cultures. In all cultures tested, CDA treatments resulted in a significant increase in the populations of planktonic cells released into the bulk liquid compared to untreated control samples (Figure 1A). The greatest effect was repeatedly observed with 310 nM CDA with at least two-fold increase in the number of planktonic cells. However, no significant differences were detected in the number of planktonic cells after exposure of S. enterica biofilms to 310 and 620 nM CDA (P-value <0.05) (Figure 1A). Following exposure to 310 nM CDA, the most significant increase in planktonic population was observed in the case of E. coli biofilms (OD600 = 0.9±0.02, SE, P-value <0.05) versus untreated controls (OD600 = 0.66±0.01, SE, P-value <0.05) (Figure 1A).
The results from these experiments are summarized in Figure 1A. We also examined the effect of exposure to very low concentrations of CDA on pre-established biofilms grown in continuous cultures on the inner surface of silicone tubing. We again observed an increase in population of planktonic cells after treatment with CDA, indicating the release of biofilm bacteria into the effluent of cultures treated with CDA. As for semi-batch biofilm cultures, the most increase in population of planktonic cells in the effluents, with more than two-fold increase in the number of planktonic cells in comparison with control biofilms were observed when cultures were treated with 310 nM CDA (Figure 1B).[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
Induction of planktonic mode of growth in pre-established biofilms formed by food pathogens using CDA.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
(A) Biofilms were grown for 5 days in petri dishes in which the medium was replaced every 24 h. Dispersion induction was tested by replacing the growth medium with fresh medium containing three different concentrations of CDA (100, 310 or 620 nM) or just the carrier as a control and the cells were incubated for a further 1 h. Medium containing dispersed cells was then homogenized and cell density was determined by measuring the optical density. (B) After 5 days of biofilm growth in flow cell continuous cultures, the influent medium was switched from fresh medium in the test lines to three indicated concentrations of CDA and control lines were switched to new lines containing just the carrier. Effluent runoffs were then collected and cell density was determined by measuring the OD. Error bars indicate standard errors (n = 3) and mean values sharing at least one common lowercase letter shown above the bars are not significantly different (P-value <0.05).[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
At this concentration, the most significant increase in population of planktonic cells was observed in S. enterica biofilms (OD600 = 0.37±0.01, SE, P-value <0.05) compared to results for untreated controls (OD600 = 0.19±0.005, SE, P-value <0.05) and no significant differences were detected between B. cereus and E. coli biofilms.
The results from these two different dispersal bioassays demonstrated the ability of nano-molar ranges of CDA to stimulate the release of cells from pre-established biofilms formed by different species of food related- bacteria.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Antimicrobial combined CDA survival assays of pre-established biofilms on stainless steel and polystyrene surfaces
*In the **Antimicrobial combined CDA survival assays of pre-established biofilms on stainless steel and polystyrene surfaces** section:*
To examine the effect of CDA combined antimicrobial agents on removal of biofilms; we tested Epimax S (hydrogen peroxide) and Percidine (peracetic acid) against pre-established biofilms grown on the surface of SS discs, in the presence and absence of 310 nM CDA. When 120 h biofilms were treated in the absence of CDA, both disinfectants caused approximate two-fold decrease in CFU counts compared to the untreated controls, while combined exposure of cultures to 310 nM CDA and 70 ppm Percidine or 120 ppm Epimax S, resulted in approximate five-fold decrease in CFU counts. No significant differences were observed between these two different combinational treatments in reduction of CFU counts (P-value <0.05). The results from these experiments are illustrated in Figure 2A.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
Effect of CDA combined antimicrobial treatments on eradication and killing of pre-established biofilms.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
(A) After 120 h of growth on the surface of SS discs, biofilms were treated for 1 h with biocides alone or combined with 310 nM CDA; CFU plate counts were then performed to assess the viability of the bacteria. (B) The amount of biofilm remaining was determined by the absorbance at 590 nm of crystal violet after staining the 120 h different biofilms in a microtiter plate assay after treatment with tested concentrations of antibiotics alone (- CDA) or in combination with 310 nM CDA (+CDA) for 1 h. All readings are corrected to reflect 0% and 100% controls (blank well, 0%; biofilms without any treatments, 100%). Error bars indicate standard errors (n = 3) and mean values sharing at least one common lowercase letter shown above the bars are not significantly different (P-value <0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D013193)
We have also tested effectiveness of CDA combined with three antibiotics (ciprofloxacin, vancomycin and ampicillin). We observed that combined treatments with both CDA and antibiotics had a significant effect on removing pre-established biofilms formed by examined microorganisms on polystyrene surfaces. For example, ciprofloxacin treatment of biofilms formed by S. aureus and B. cereus caused approximately 11% and 13% reductions in their biofilms, respectively (compared to biofilms without any treatments) while combined treatment of their biofilms with 1 µg of ciprofloxacin and 310 nM CDA resulted in 87% and 89% removal of their biofilms, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
Significant differences were detected between two different combinational treatments applied for gram positive and gram negative bacteria; since the combination of CDA and ciprofloxacin was more effective than CDA combined ampicillin to eradicate biofilms formed by gram negative organisms. Similarly, combined treatments with both CDA and vancomycin were more effective to eliminate biofilms formed by gram positive bacteria. Results from these experiments are summarized in Figure 2B.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
Thus, combined treatments using only low concentrations of CDA together with biocides or antibiotics were highly effective in removal and killing of pre-established biofilms formed by food pathogens.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Combined CDA and antimicrobial treatment of planktonic cells (cont.)
*In the **Combined CDA and antimicrobial treatment of planktonic cells (cont.)** section:*
To further investigate the effect of CDA on the sensitivity of tested microorganisms towards antimicrobial agents, we also evaluated very low concentrations of CDA for any inhibitory effects on growth of their planktonic cells. Compared to antibiotics or biocides alone, combination of antimicrobial treatment with nano-molar concentrations of CDA had no additional inhibitory effects on the growth of planktonic cells; for that reason only Minimum Inhibitory Concentrations (MICs) for antibiotics and disinfectants alone are presented in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
MICs of tested microorganisms' planktonic cells to examined disinfectants and antibiotics.
## Biofilm surface coverage reduction by CDA combined biocides or antibiotics
*In the **Biofilm surface coverage reduction by CDA combined biocides or antibiotics** section:*
To further examine the effect of CDA on biofilm surface area and bacteria viability, we also tested various disinfectants and antibiotics alone or combined with CDA against pre-established biofilms grown in continuous culture flow cells. When 48-h biofilms were treated in the absence of CDA, none of the disinfectants or antibiotics reduced biofilm biomass effectively (Figure 3). In contrast, after combined treatment, the biofilm cells remaining on the surface were easily removed and killed by antimicrobial compounds when examined by using the LIVE/DEAD staining kit (Figure 4). The most significant reduction in biofilm surface area (P-value <0.05) was observed when biofilms were treated with combination of Epimax S and 310 nM CDA. For example, this combination resulted in eradication of more than 90% of the E. coli biofilms from the surface (Figure 3).[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
Effect of CDA combined disinfectant or antibiotic treatments on biofilms surface area.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
Following dispersion of biofilms by CDA, cells remaining on the surface are easily killed and removed by various disinfectants (Epimax S and Percidine) or antibiotics (vancomycin; Van, ampicillin; Amp, ciprofloxacin; Cip) in biofilms grown in continuous culture flow cells. Pre-established biofilms were grown for 48 h without any treatment and then were treated with indicated concentrations of antimicrobials alone (- CDA) or combined with 310 nM CDA (+ CDA) for 1 h, stained with LIVE/DEAD staining and quantified (percent surface coverage) using digital image analysis. The bars show the levels of biofilm biomass after treatment with antimicrobials alone or combined with 310 nM CDA. Error bars indicate standard errors (n = 3) and mean values sharing at least one common lowercase letter shown above the bars are not significantly different (P-value <0.05).[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
Effect of CDA combined antimicrobial treatments on killing of pre-established biofilms.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
CDA treatment reverses biofilm formation in pre-established biofilms and cells remaining on the surface are easily removed and killed various disinfectants (Epimax S and Percidine) or antibiotics (vancomycin; Van, ampicillin; Amp, ciprofloxacin; Cip) in biofilms grown in continuous culture flow cells. Pre-established biofilms were grown for 48 h without any treatment, then were treated with indicated concentrations of antimicrobials alone (- CDA) or combined with 310 nM CDA (+CDA) for 1 h and stained with LIVE/DEAD staining to allow analysis using fluorescence microscopy. The images show microscopic pictures of the biofilms on the surface of cover slip after combinatorial treatments. Images are top-down views (x-y plane); scale bars: 50 µm. Results are representative of 3 separate experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
Combined treatments with both CDA and antibiotics or biocides caused almost-complete eradication of pre-established biofilms.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
## Discussion
*In the **Discussion** section:*
The EPS matrix acts as a barrier in which diffusive transport prevails over convective transport . EPS delay or prevent antimicrobials from reaching target microorganisms within the biofilm by diffusion limitation (like ciprofloxacin and ampicillin) , and/or chemical interaction with the matrix material (like peroxides such as peracetic acid and hydrogen peroxide) . Against such a drawback still oxidizing compounds (like peroxides) for their nonspecific mode of actions and because of variation in the chemical composition of biofilms are among widely used disinfectants in food industry in most countries including Iran. Several studies have shown that strategies to induce biofilm dispersal could potentially use the microorganisms' own energy to disrupt EPS and remove pre-established biofilms . In a previous study Davies and Marques showed that a synthesized signalling molecule by P. aeruginosa induces dispersion of pre-established biofilms in P. aeruginosa as well as many other strains of microorganisms.
They concluded that CDA most likely induce the production of degradative enzymes of EPS by these microorganisms. Differential microarray analysis, by Rahmani et al. (under preparation) indicated that 100 nM CDA (added exogenously to P. aeruginosa pre-established biofilms) significantly up regulates the expression of P. aeruginosa genes including EPS, alginate, degradative enzyme (alginate lyiase; algL) and negative regulator for this EPS biosynthesis (mucB). Their results also showed that CDA down regulates the expression of genes involved in P. aeruginosa attachment to the surfaces (cupA and cupB), which results in reversion of biofilms to a population of planktonic cells with increased susceptibility to antimicrobial agents compared to their sessile counterparts (Rahmani et al., under preparation). Therefore, in this investigation we first examined the action of nano-molar concentrations of CDA (as an inducer of biofilm dispersal) on dispersion of pre-established biofilms, formed by four main food-borne pathogenic or spoilage microorganisms.
Our results interestingly showed that only 310 nM of the signal was enough to reverse pre-established biofilms, formed by distant genera of bacteria, to their planktonic mode of growths. Since disinfectants and antibiotics have greater bactericidal efficacy against planktonic bacteria than their sessile counterparts, the combination of CDA with common antimicrobial agents could have improved bactericidal efficacy. Thus, we then tried to remove and kill pre-established biofilms by using the combination of CDA and traditional disinfectants or antibiotics which are broadly used in food processing environments and their related medical issues, at concentrations that had no significant effects against biofilms, to reach a novel mechanism for enhancing the activity of these treatments through the disruption of biofilms. The results presented here demonstrated that following exposure to low concentrations of CDA, biofilm cells on the surface were easily detached and then killed by antimicrobial agents where the combination of 310 nM CDA with examined disinfectants (Percidine and Epimax S) or antibiotics (ciprofloxacin, vancomycin and ampicillin), when added to their solutions, resulted in approximate 80% reduction in biofilm biomass in all cultures.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
Numerous strategies to control microbial biofilms have been proposed, with different degrees of success. In various industrial settings, a range of biocides and toxic metals (e.g., tin and copper) has been used for antifouling coatings and sanitizing purposes , ; however, these substances are not appropriate for use in food industries and clinical settings. In this work, we showed that CDA-based strategies to induce biofilm dispersal involve only nano-molar concentrations of CDA that should be safe to humans and to the environment. Besides, previous findings showed that CDA has no cytotoxic or stimulatory effect on human cells even at high concentrations (up to 250 µg.ml−1) . Because CDA mediates the transition from a biofilm to a planktonic phenotype via a signalling mechanism (because acts at nano-molar concentrations which are consistent with all known cell-to-cell signalling molecules) rather than toxic effect, CDA-based biofilm control strategies would not be expected to select for resistant strains as seen with antibiotics.
Therefore, in this study we examined two different combination of CDA; CDA combined disinfectants and CDA combined antibiotics, to introduce a promising strategy which is appropriate to control biofilms both in food industry and clinical settings.[](https://www.ncbi.nlm.nih.gov/mesh/D014001)
While some free fatty acids have antimicrobial properties , and play a vital role in maintaining the microbial flora of the skin , , we demonstrated that CDA does not inhibit bacterial growth at nano-molar ranges that induce biofilm dispersal. These results were highly in consistent with the results from Jennings et al. study where they showed that CDA inhibited bacterial growth only at high (micro-molar to milli-molar) concentrations. This lack of growth inhibition at lower concentrations was not surprising since bacteria produce this unsaturated fatty acid and use it as a signalling molecule .[](https://www.ncbi.nlm.nih.gov /mesh/D005227)
## Conclusions
*In the **Conclusions** section:*
Data from this study suggest that application of CDA prior to or in combination with disinfectants or antibiotics may allow for novel and improved strategies to control biofilms in industrial as well as clinical settings, with clear benefits such as reduced ecological impact and reduced treatment costs.[](https://www.ncbi.nlm.nih.gov/mesh/C052476)
# References
*In the **References** section:*
|
# Introduction
Antidiabetic Activity of [Acacia tortilis (Forsk.) Hayne ssp. raddiana Polysaccharide](https://www.ncbi.nlm.nih.gov/mesh/D011134) on [Streptozotocin](https://www.ncbi.nlm.nih.gov/mesh/D013311)-[Nicotinamide](https://www.ncbi.nlm.nih.gov/mesh/D009536) Induced Diabetic Rats
# Abstract
*In the **Abstract** section:*
The present study was designed to investigate the antidiabetic activity of aqueous extract of Acacia tortilis polysaccharide (AEATP) from gum exudates and its role in comorbidities associated with diabetes in STZ-nicotinamide induce[d diabetic rats. Male albino W](https://www.ncbi.nlm.nih.gov/mesh/D011134)is[tar r](https://www.ncbi.nlm.nih.gov/mesh/D011134)ats were divided into control, diabetic control, glimepiride treated (10 mg/kg[), ](https://www.ncbi.nlm.nih.gov/mesh/D013311)a[nd diabetic ](https://www.ncbi.nlm.nih.gov/mesh/D009536)rats treated with 250, 500, and 1000 mg/kg dose of AEATP groups and fasting blood glucose, gl[ycated hemo](https://www.ncbi.nlm.nih.gov/mesh/C057619)globin, total cholesterol, triglyceride, LDL, VLDL, HDL, SGOT, and SGPT levels were m[easur](https://www.ncbi.nlm.nih.gov/mesh/D011134)ed. STZ significantly incr[eased f](https://www.ncbi.nlm.nih.gov/mesh/D005947)asting blood glucose level, g[lycated hem](https://www.ncbi.nlm.nih.gov/mesh/D002784)og[lobin, total](https://www.ncbi.nlm.nih.gov/mesh/D014280) cholesterol, triglyceride, LDL, VLDL, SGOT, and SGPT l[eve](https://www.ncbi.nlm.nih.gov/mesh/D013311)ls, whereas HDL level was reduced as co[mpared ](https://www.ncbi.nlm.nih.gov/mesh/D005947)to control group. After 7 days of a[dministrati](https://www.ncbi.nlm.nih.gov/mesh/D002784)on[, 500 and 10](https://www.ncbi.nlm.nih.gov/mesh/D014280)00 mg/kg dose of AEATP showed significant reduction (P < 0.05) in fasting blood glucose level compared to diabetic control. AEATP has also reduced total ch[olest](https://www.ncbi.nlm.nih.gov/mesh/D011134)erol, triglyceride, LDL, VLDL, SGOT, and SGPT levels and i[mproved](https://www.ncbi.nlm.nih.gov/mesh/D005947) HDL level as compared to diabetic co[ntrol](https://www.ncbi.nlm.nih.gov/mesh/D011134) group. Our study is the[ first to r](https://www.ncbi.nlm.nih.gov/mesh/D002784)ep[ort the norm](https://www.ncbi.nlm.nih.gov/mesh/D014280)alization of fasting blood glucose level, lipid profile, and liver enzyme in AEATP treated diabetic rats. Thus, it can be concluded that AEATP may have potentials fo[r the t](https://www.ncbi.nlm.nih.gov/mesh/D005947)reatment[ of T](https://www.ncbi.nlm.nih.gov/mesh/D008055)2DM and its comorbidities.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 1. Introduction
*In the **1. Introduction** section:*
Diabetes mellitus (DM) is characterized by hyperglycemia and is associated with a group of metabolic disorders, that is, abnormalities in carbohydrate, fat and protein metabolism which further result in chronic complications including microvascular, macrovascular, and neuropathic disorder [1]. It is dispersed worldwide with prevalence from 171 million in 2000 to 366 million in 2030 [2].[](https://www.ncbi.nlm.nih.gov/mesh/D002241)
The currently available oral hypoglycemic and antihyperglycemic drugs for type-II diabetes have their own limitations, adverse effects, and secondary failures. Therefore, to reduce their cost, limitation, and adverse effects, focus has been shifted towards the medicinal herbs for safe and effective use. Recently a lot of medicinal herbs are being investigated for their role in pharmacotherapy of diabetes [3].
Israeli babool and umbrella thorn are the other names of Acacia tortilis and are widespread distributed around the globe (Africa, Algeria, Egypt, Asia, Israel, Somalia, Pakistan, and India). In India, this tree was introduced in 1958 from Israel [4]. Various extracts of this plant have the following actions like smooth muscle relaxing activity [5], effective in treatment of α 2-adrenoceptor related diseases [6], antimicrobial activity against Bacillus subtilis, Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa, and Candida albicans [7], in-vitro antiplasmodial and antileishmanial activity [8], antiviral effect against human immunodeficiency virus-1 [9], antiasthmatic [10], and hypotensive and diuretic property [11].
Moreover, various species of Acacia are reported to have antidiabetic activity like Acacia arabica, Acacia catechu, Acacia mollissima, Acacia polyacantha, and so forth [12–15]. Previously numerous studies on polysaccharides from Phellinus linteus [16], Ascophyllum [17], Taxus [18], Acanthopanax [19], and Andrographis [20] demonstrated the antidiabetic activity. Further, the seed extract of Acacia tortilis has been also found to have an antihyperglycemic activity [21].
Thus, with the same line of research, the present study was designed to explore the antidiabetic activity of aqueous extract of Acacia tortilis (Forsk.) Hayne ssp. raddiana polysaccharide from gum exudates.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 2. Material and Methods
*In the **2. Material and Methods** section:*
## 2.1. Chemicals
*In the **2.1. Chemicals** section:*
Streptozotocin, Glimepiride, and Nicotinamide were procured form Sigma-Aldrich, Milwaukee, USA, and all the other chemicals were of analytical grade.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## 2.2. Plant Material
*In the **2.2. Plant Material** section:*
Gum exudates from the stem and branches of Acacia tortilis was collected from Central Arid Zone Research Institute Campus, Jodhpur, India.
## 2.3. Animals
*In the **2.3. Animals** section:*
Male albino Wistar rats (150–200 gm) were used in this study and experimental protocol was approved by Institutional Animal Ethics Committee. Animals were kept as per the guidelines of the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA), Ministry of Environment and Forest, Government of India (Chitkara College of Pharmacy Animal Facility Registration number: 1181/ab/08/CPCSEA). Animals were fed normal chow diet and ad libitum under controlled environmental condition of temperature (24–28°C), relative humidity 60–70%, and natural light/dark cycle (12 : 12) and maintained on standard food pellets and tap water ad libitum.
## 2.4. Acute Oral Toxicity Studies
*In the **2.4. Acute Oral Toxicity Studies** section:*
Acute oral toxicity studies were performed according to OECD (Organization for Economic Cooperation and Development) 423 guidelines at the dose of 300, 2,000, and 5,000 mg/kg. The general behavior was observed continuously for 48 hr, 3 days and mortality was observed for 14 days [22].
## 2.5. Isolation of Polysaccharide
*In the **2.5. Isolation of Polysaccharide** section:*
Gum exudate was crushed into fine particles using laboratory grinder. Fine powder of gum exudates (100 gm) was stirred vigorously in distilled water (200 mL) for 6 hours at room temperature and centrifuged to remove water-insoluble part. The supernatant solution was decanted off. The concentrated aqueous solution was poured into 3 times its volume of ethanol with constant stirring. The polysaccharide was precipitated out in the form of a fluffy precipitate.
The precipitate was again dissolved in water and added to ethanol. Precipitate was treated successively with dry solvent ether and acetone. It was filtered under vacuum and dried in vacuum desiccators at room temperature [23].[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## 2.6. Complete Hydrolysis
*In the **2.6. Complete Hydrolysis** section:*
The pure polysaccharide was subjected to hydrolysis with sulfuric acid (2N) for 18 hr on steam bath. The hydrolyzate was cooled, neutralized with saturated solution of barium carbonate by dropwise addition till the pH of the solution reached at 7, filtered, and the residue washed with water. The combined filtrate was concentrated at or below 40°C in rotary evaporator under reduced pressure. This hydrolyzed mass was used for paper chromatography.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 2.7. Experimental Protocol
*In the **2.7. Experimental Protocol** section:*
In experimental animals, diabetes was induced by intraperitoneal injection of Nicotinamide (230 mg/kg) 15 min before streptozotocin (65 mg/kg, i.p) administration [24]. The STZ was freshly prepared by dissolving in 0.1 M citrate buffer, pH 4.5, and nicotinamide was prepared in normal saline. Diabetes mellitus was confirmed after 14 days of STZ administration when fasting blood glucose level had become constant above 250 mg/dL. As STZ is capable of inducing fatal hypoglycemia as a result of massive pancreatic insulin release, STZ-treated rats were provided with 10% glucose solution after 6 hr for the next 24 hr to prevent fatal hypoglycemia [25]. Different doses of AEATP (250, 500, and 1000 mg/kg) were administered to the animals and doses were selected on the basis of acute toxicity studies.
[26].[](https://www.ncbi.nlm.nih.gov/mesh/D009536)
Group 1: control: no intervention.
Group 2: diabetic control + vehicle distilled water.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
Group 3: diabetic rats + glimepiride (10 mg/kg body weight).[](https://www.ncbi.nlm.nih.gov/mesh/C057619)
Group 4: diabetic rats + 250 mg/kg AEATP.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Group 5: diabetic rats + 500 mg/kg AEATP.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Group 6: diabetic rats + 1000 mg/kg AEATP.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Further, animals with blood glucose level above 250 mg/dL were selected and divided into six groups comprising ten animals in each group. After 29 days of continuous treatment, blood sample was collected from retroorbital plexus under anesthesia for biochemical estimation of total cholesterol, triglycerides, LDL, HDL, VLDL, HbA1c, SGOT, and SGPT by commercially available kits of Reckon Diagnostics Pvt., Ltd. Body weight was measured before induction of diabetes and during treatment period. Fasting blood glucose level of overnight fasted rats was measured using glucometer on the 1st, 7th, 14th, 21st, and 28th days of pharmacological interventions. Serum insulin was measured by using ELISA kits (EMD Millipore-EZRMI-13 K).
The assay is based on Sandwich ELISA technique and enzyme activity was measured spectrophotometrically by the increased absorbency at 450 nm, corrected from the absorbency at 590 nm, after acidification of formed products. Pancreatic insulin content was extracted by taking 0.2 g pancreas portion with 5.0 mL of ice-cold acid-alcohol in a centrifuge tube, homogenized, followed by sonication, stored at −20°C overnight, and finally centrifuged at 3000 rpm at 4°C for 15 minutes. The supernatant was transferred into a new centrifuge tube and stored at −20°C, while the pellet was subjected to extraction again. After extraction insulin level was measured at room temperature by ELISA assay kit [27].[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
## 2.8. Statistical Analysis
*In the **2.8. Statistical Analysis** section:*
Statistical analysis was performed using Graph pad Prism 6. Values are expressed as mean ± SEM and statistical analysis was carried out by using by ANOVA followed by Tukey's as post hoc multiple comparison test.
## 3. Results
*In the **3. Results** section:*
The pure polysaccharide was extracted as amorphous white powder with a percentage yield of 24.5% from the gum exudates. Complete hydrolysis of the polysaccharide followed by paper chromatography revealed the presence of four spots, corresponding to D-galactose, D-glucose, L-rhamnose, and D-glucuronic acids, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 3.1. Acute Oral Toxicity Studies
*In the **3.1. Acute Oral Toxicity Studies** section:*
In the present study, oral toxicity was carried out according to OECD guidelines, up to an elevated concentration of 5,000 mg/kg. However, at this dose Acacia tortilis polysaccharide did not exhibit any sign of toxicity, behavioral changes, and mortality. Thus Acacia tortilis polysaccharide was found to be nontoxic and therefore can be safely used.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 3.2. Effect of AEATP on Body Weight
*In the **3.2. Effect of AEATP on Body Weight** section:*
Animals of the same weight range were used in experimental protocol. During the study, the body weight of control group was increased naturally, whereas body weight was found to be significantly attenuated in STZ-induced diabetic group as compared to control group (Figure 1). After 28 days of continuous administration of glimepiride, a significant increase in body weight was observed as compared to diabetic control group. Oral administration of AEATP to diabetic rats also significantly increased the body weight at the 14th day of intervention and reversed the effect of STZ comparable to glimepiride treated group. Furthermore, on 21st and 28th days of treatment, no significant difference in body weight was observed between glimepiride, 250, 500, and 1000 mg/kg of AEATP treated groups and the effect of AEATP administration produced dose-independent effect on body weight in diabetic rats.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## 3.3. Effect of AEATP on Fasting Blood Glucose Level
*In the **3.3. Effect of AEATP on Fasting Blood Glucose Level** section:*
The basal values of fasting blood glucose level were almost the same and statistically no significant difference was observed while including the animals for experimentation. Fasting blood glucose level of control group ranged from 106.5 ± 6.64 to 110.16 ± 3.33 mg/dL in 28 days of study, while there was a significant increase in fasting blood glucose level in STZ + nicotinamide treated rat 400.5 ± 7.8 mg/dL as compared to control group (Figure 2). Fasting blood glucose level glucose level was measured on the 7th, 14th, 21st, and 28th days. Different doses of AEATP (250, 500, and 1000 mg/kg) were administered continuously for 28 days and significant reduction in blood glucose level was observed in the STZ + NAD treated diabetic rats. Similarly, glimepiride significantly reduced fasting blood glucose level measured on the 7th, 14th, 21st, and 28th days.
However, hypoglycemia was not observed even on the 28th day of continuous administration of AEATP since as per CPCSEA normal blood glucose range of Wistar albino rat is 50–135 mg/dL. Interestingly, on the 7th and 14th days, both 500 and 1000 mg/kg of AEATP and the 21st day only 1000 mg/kg of AEATP also showed significant reduction in blood glucose level compared to glimepiride and 250 mg/kg of AEATP, whereas, on the 28th day, only 1000 but not 500 mg/kg of AEATP showed significant reduction in blood glucose level compared to 250 mg/kg of AEATP.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
## 3.4. Effect of AEATP on Glycated Hemoglobin (HbA1c)
*In the **3.4. Effect of AEATP on Glycated Hemoglobin (HbA1c)** section:*
Glycated hemoglobin was significantly increased in STZ-induced diabetic group 9.79 ± 0.2% as compared to control group 3.92 ± 0.19% (Figure 3). After 28 days of treatment, with 250, 500, and 1000 mg/kg of AEATP and glimepiride shown significant attenuation in elevated glycated hemoglobin level as compared to diabetic control group, that is, 7.01 ± 0.12%, 6.43 ± 0.09%, 6.28 ± 0.13%, and 6.76 ± 0.11, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## 3.5. Effect of AEATP on Total Cholesterol Level
*In the **3.5. Effect of AEATP on Total Cholesterol Level** section:*
The present result showed that total cholesterol level in serum was significantly elevated in diabetic control group 206.1 ± 3.7 mg/dL as compared to control group 115.1 ± 3.1 mg/dL (Figure 4). Different doses of (250, 500, and 1000 mg/kg) of AEATP produced statistically significant reduction in cholesterol level 148.1 ± 4.58 mg/dL, 129.6 ± 2.55 mg/dL, and 126.0 ± 6.51 mg/dL, respectively, as compared to diabetic control group after 28 days of treatment. 500 and 1000 mg/kg of AEATP except 250 mg/kg brought down the elevated cholesterol level to normal and also significantly reduced the total cholesterol level comparable to glimepiride treated group, that is, 134.2 ± 3.52 mg/dL.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
## 3.6. Effect of AEATP on Total Triglycerides, LDL, and VLDL Levels
*In the **3.6. Effect of AEATP on Total Triglycerides, LDL, and VLDL Levels** section:*
There was significant increase in total triglyceride (TG), low density lipoprotein (LDL), and very low density lipoprotein (VLDL) levels in diabetic control group when compared to control group. The administration of AEATP and glimepiride for 28 days significantly attenuated diabetes induced high level of TG, LDL, and VLDL. 250, 500, and 1000 mg/kg AEATP and glimepiride significantly reduced the TG level by 106.73 ± 2.6, 104.87 ± 2.0, 102.3 ± 3.7, and 123.14 ± 4.8 mg/dL, respectively, as compared to diabetic control group 194.67 ± 3.41 mg/dL (Figure 5). Similarly, significant attenuation was observed in LDL level with 250, 500, and 1000 mg/kg AEATP and glimepiride treatment, that is, 59.62 ± 5.24, 54.62 ± 5.9, 53.76 ± 3.77, and 81.79 ± 4.17 mg/dL, respectively as compared to diabetic control group 160.97 ± 6.92 mg/dL (Figure 6). Moreover, 250, 500, and 1000 mg/kg AEATP and glimepiride significantly reversed the VLDL level, that is, 21.01 ± 0.28, 20.98 ± 0.4, 22.09 ± 0.7, and 24.63 ± 0.96 mg/dL, respectively as compared to diabetic control 38.93 ± 0.68 mg/dL (Figure 7).[](https://www.ncbi.nlm.nih.gov/mesh/D014280)
## 3.7. Effect of AEATP on High Density Lipoprotein Level (HDL)
*In the **3.7. Effect of AEATP on High Density Lipoprotein Level (HDL)** section:*
HDL level was found to be significantly reduced in diabetic control group 17.75 ± 2.73 mg/dL compared to control group 51.43 ± 2.95 mg/dL. After 28 days of intervention with 250, 500, and 1000 mg/kg of AEATP, a significant increase was observed in HDL level 28.35 ± 1.28 mg/dL, 31.55 ± 2.79 mg/dL, and 32.87 ± 3.3 mg/dL, respectively, as compared to diabetic control group 17.75 ± 2.73 mg/dL. Administration of glimepiride also increased HDL level of diabetic rats 24.44 ± 1.3 (Figure 8).[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 3.8. Effect of AEATP on Liver Enzymes (unit/lit.)
*In the **3.8. Effect of AEATP on Liver Enzymes (unit/lit.)** section:*
A significant elevation in SGOT and SGPT was observed in serum of STZ-induced diabetic rats 284.5 ± 10.21 and 161.9 ± 5.21 unit/lit. as compared to control group 177.0 ± 2.4 and 55.76 ± 3.09 unit/lit., respectively. A statistically significant attenuation was observed in SGOT enzyme level after 28 days of administration of 250, 500, and 1000 mg/kg of AEATP (71.16 ± 6.79 and 65.76 ± 4.31 unit/lit. 69.87 ± 3.61, resp.) and glimepiride (88.50 ± 5.22 unit/lit.) as compared to diabetic control 284.5 ± 10.21 unit/lit.
(Figure 9). Similar results were observed in SGPT enzyme level with 250, 500, and 1000 mg/kg of AEATP and glimepiride, that is, 70.11 ± 6.75, 67.53 ± 5.70, 68.87 ± 4.10, and 83.63 ± 4.92 unit/lit., respectively, when compared to diabetic control group 161.9 ± 5.21 unit/lit. (Figure 10).[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## 3.9. Effect of AEATP on Fasting Insulin Level and Pancreatic Insulin Content
*In the **3.9. Effect of AEATP on Fasting Insulin Level and Pancreatic Insulin Content** section:*
A significant decrease was observed in fasting insulin level (0.373 ± 0.026 ng/mL) and pancreatic insulin content (56.0 ± 2.81 ng/mg pancreas) in STZ-induced diabetic rats as compared to control group 0.820 ± 0.024 ng/mL and 103.9 ± 5.24 ng/mg pancreas, respectively. Fasting insulin level was statistically increased after 28 days administration of 250, 500, and 1000 mg/kg of AEATP (0.550 ± 0.024, 0.613 ± 0.020, and 0.683 ± 0.024 ng/mL, resp.) and glimepiride (0.723 ± 0.024 ng/mL) as compared to diabetic control (0.373 ± 0.026 ng/mL) (Table 1). Similar results were observed in pancreatic insulin content with 250, 500, and 1000 mg/kg of AEATP and glimepiride (67.0 ± 2.98, 77.33 ± 5.68, 92.46 ± 3.14, and 99.9 ± 2.22 ng/mg pancreas, resp.) when compared to diabetic control group (56.0 ± 2.81 ng/mg pancreas) (Table 2).[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## 4. Discussion
*In the **4. Discussion** section:*
The present study was designed to investigate antidiabetic activity of AEATP along with its effect on lipid profile and liver enzymes. STZ (N-nitro derivative of glucosamine) has potent alkylating property [28] and is specifically cytotoxic to the pancreatic beta cells in mammals. The pancreatic beta cell preferentially uptakes STZ resulting in formation of superoxide radicals. Moreover, NO moiety is liberated from STZ leading to the destruction of β-cells by necrosis [25].
Recently, a new animal model of type 2 diabetes has been introduced in which a combination of STZ and nicotinamide administration is able to induce DM in adult rats. The rats administered nicotinamide (230 mg/kg, ip) 15 min before STZ (65 mg/kg, ip) was found to develop moderate and stable nonfasting hyperglycaemia without any significant change in plasma insulin level. Nicotinamide is an antioxidant which exerts protective effect on the cytotoxic action of STZ by scavenging free radicals and causes only minor damage to pancreatic β-cell mass producing type 2 diabetes [24]. Like previous reports, significant increase in fasting blood glucose level was observed in STZ induced diabetic rats compared to control group [29, 30].
Administration of AEATP for 28 days resulted in significant reduction in the fasting blood glucose level as compared to diabetic rats. It is evident from this investigation that the aqueous extract was effective in maintaining the blood glucose levels in STZ and nicotinamide induced diabetic rats. Interestingly, on the 7th and 14th days, both 500 and 1000 mg/kg of AEATP and the 21st day only 1000 mg/kg of AEATP also showed significant reduction in blood glucose level as compared to diabetic control and 250 mg/kg of AEATP, whereas on the 28th day, only 1000 but not 500 mg/kg of AEATP showed significant reduction in blood glucose level as compared to 250 mg/kg of AEATP.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Induction of diabetes is associated with the characteristic loss of body weight [31, 32], which is due to muscle wasting and catabolism of tissue proteins leading to significant reduction in the body weight in diabetic rats. Similar effect was also observed in the present study. Diabetic rats treated with AEATP, independent of dose, showed an increase in the body weight as compared to the diabetic control which might be due to its protective effect in controlling muscle wasting, that is, reversal of gluconeogenesis, and might also be due to the improvement in insulin secretion and glycemic control.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Glycated hemoglobin (HbA1c) is the standard biochemical marker in assessment of diabetes. In our study, diabetic rats showed higher level of glycated hemoglobin indicating their poor glycemic control which is also supported by other previously reported studies [33, 34]. Oral administration of AEATP at all the doses significantly reduced HbA1c to near normalcy by 28 days of intervention as compared to diabetic control group. Several studies have demonstrated that flavonoids attenuate hyperglycemia [35] and reduced nonenzymatic glycation of proteins in animals [36] as Acacia tortilis was also reported to have flavonoids content [37] and might show the similar activity.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Lipid plays an important role in the pathogenesis of complications associated with diabetes mellitus. The elevated level of serum cholesterol and reduced level of serum HDL cholesterol in diabetic condition, poses to be a risk factor for developing microvascular complication leading to atherosclerosis and cardiovascular diseases like coronary heart disease. The abnormal high concentration of serum lipid in diabetic mainly due to increased mobilization of free fatty acids from peripheral fat depots, since insulin inhibits the hormone sensitive lipase, insulin deficiency, or insulin resistance may be responsible for dislipidemia [38]. The present study showed that diabetic rats has abnormal lipid profile as earlier reports [39, 40], whereas the AEATP treated group showed significant improvement in the lipid profile comparable to glimepiride treated group. Interestingly, 500 and 1000 mg/kg of AEATP have shown significant reduction of TG and LDL-C as compared to glimepiride treated group indicating additional hypolipidemic activity over available standard drugs.
Hypolipidemic effect could represent a protective mechanism against the development of atherosclerosis. It is well known that hyperlipidemia has an association with atherosclerosis and the incidence of atherosclerosis is increased in diabetics [41].[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
There was increase in liver enzymes in serum of diabetic rats like earlier reports [42] which might be primarily due to leakage of these enzymes from liver cytosol into bloodstream as a consequence of hepatotoxic effect of STZ. AEATP, at all the doses, lowered serum SGPT and SGOT levels which showed the protective effect and normal functioning of liver in reversing the organ damage due to diabetes which was clearly observed by high levels of SGOT and SGPT in diabetic control.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
The possible mechanism of action of AEATP could be correlated with promoting insulin secretion by closure of K+-ATP channels, membrane depolarization, and stimulation of Ca2+ influx, an initial key step in insulin secretion. As studies have shown that the gut incretin content is altered in animal models of type 2 diabetes [43], therefore AEATP may act by increasing incretin, that is, increasing glucagon-like-peptide-1 (GLP-1) or inhibiting dipeptidyl peptidase-4 (DPP-4). Furthermore, glucagon-like-peptide-1 reportedly promotes islet cell growth and inhibits apoptosis in animal models; an increase in GLP-1 secretion might also be beneficial for islet cell function and mass in humans [44]. AEATP exerts protective effects in experimental diabetes, possibly by reducing oxidative stress, and hence protects rats from oxidative damage and dyslipidemia due to STZ treatment.
However, further studies are necessary to confirm these effects. The results of our study confirm that the AEATP has a potent antidiabetic activity reversing the disturbances in lipid profile and liver toxicity.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
This study revealed first time the antidiabetic activity of AEATP and its effective role in maintaining lipid profile and liver toxicity in diabetic rats. This new insight allows an understanding of the use of Acacia tortilis polysaccharide in prevention and treatment of diabetes, hyperlipidemia, and its associated complications. However, the precise mechanism by which AEATP reduced fasting blood glucose level in diabetic rats will require further detailed study. Therefore, future research and clinical trials in this area may lead to the use of Acacia tortilis polysaccharide as a new type of therapeutic agent in treatment of type 2 diabetes.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
# Abbreviations
*In the **Abbreviations** section:*
T2DM:
Type 2 diabetes mellitus
HbA1c:
Glycated hemoglobin
LDL:
Low density lipoprotein
HDL:
High density lipoprotein
VLDL:
Very low density lipoprotein
S.E.M:
Standard error of means
STZ:[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
Streptozotocin[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
ANOVA:
Analysis of variance
AEATP:[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Aqueous extract of Acacia tortilis polysaccharide[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
DM:
Diabetes mellitus.
## Conflict of Interests
*In the **Conflict of Interests** section:*
The authors declare that there is no conflict of interests regarding the publication of this paper.
Effect of AEATP on body weight (gms) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on fasting blood glucose level (mg/dL) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on glycated hemoglobin (HbA1c) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on total cholesterol (mg/dL) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on total triglyceride level (mg/dL) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on low density lipoprotein (LDL) level (mg/dL) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on very low density lipoprotein (VLDL) level (mg/dL) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on high density lipoprotein (HDL) level (mg/dL) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on SGOT level (unit/lit.) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on SGPT level (unit/lit.) in type-2 diabetic Wistar rats. Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using Two way ANOVA followed by Tukey's multiple test; a versus control, b versus Diabetic control, c versus Glimepiride treated, d versus 250 mg/kg of AEATP, e versus 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on fasting insulin level (ng/mL) in type-2 diabetic Wistar rats.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using one-way ANOVA followed by Tukey's multiple test; a control, b diabetic control, c 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Effect of AEATP on pancreatic insulin content (ng/mg pancreas) in type-2 diabetic Wistar rats.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Each group (n = 6) represents mean ± standard error of means. Data was analyzed by using one-way ANOVA followed by Tukey's multiple test; a control, b diabetic control, c 250 mg/kg of AEATP, d 500 mg/kg of AEATP. *P < 0.0001, # P < 0.001, † P < 0.01, ‡ P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
|
# Introduction
[Cyclic phosphatidic acid](https://www.ncbi.nlm.nih.gov/mesh/D010712) relieves osteoarthritis symptoms
# Abstract
*In the **Abstract** section:*
Background
Cyclic phosphatidic acid (cPA) is a naturally occurring phospholipid [mediator with a unique c](https://www.ncbi.nlm.nih.gov/mesh/D010712)yc[lic](https://www.ncbi.nlm.nih.gov/mesh/D010712) phosphate ring at the sn-2[ and sn-3 po](https://www.ncbi.nlm.nih.gov/mesh/D010743)sitions of its glycerol [backbone. Natura](https://www.ncbi.nlm.nih.gov/mesh/D010710)l cPA and its chemically stabilized cPA deri[vative, ](https://www.ncbi.nlm.nih.gov/mesh/D005990)2-carba-cPA (2ccPA)[, i](https://www.ncbi.nlm.nih.gov/mesh/D010712)nhibit chronic and acute inflam[mat](https://www.ncbi.nlm.nih.gov/mesh/D010712)ion, and 2ccP[A attenuate](https://www.ncbi.nlm.nih.gov/mesh/C558317)s [neuro](https://www.ncbi.nlm.nih.gov/mesh/C558317)pathic pain. Osteoarthritis (OA) is a degenerat[ive d](https://www.ncbi.nlm.nih.gov/mesh/C558317)isease frequently associated with symptoms such as inflammation and joint pain. Because 2ccPA has obvious antinociceptive activity, we hypothesized that 2ccP[A mig](https://www.ncbi.nlm.nih.gov/mesh/C558317)ht relieve the pain caused by OA. We aimed to characterize t[he ef](https://www.ncbi.nlm.nih.gov/mesh/C558317)fects of 2ccPA on the pathogenesis of OA induced by total meniscectomy in the [rabbi](https://www.ncbi.nlm.nih.gov/mesh/C558317)t knee joint.
Results
Intra-articular injection of 2ccPA (twice a week for 42 days) significantly reduced pain and articular swelling. Histopathology showed that 2ccPA suppressed cartilage degeneration in OA. We also examined the effects of 2ccPA on the inflammatory and catabolic responses of human OA synoviocytes and chondrosarcoma SW1353 cells in vitro. 2ccPA stimulated synthesis of hyaluronic acid and suppressed production of the metalloproteinases MMP-1, -3, and -13. However, it had no effect on the production of interleukin (IL)-6, an inflammatory cytokine. The suppressive effect of 2ccPA on MMP-1 and -3 production in synoviocytes and on MMP-13 production in SW1353 cells was not mediated by the lysophosphatidic acid receptor, LPA1 receptor (LPA1R).[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
Conclusions
Our results suggest that 2ccPA significantly reduces the pain response to OA by inducing hyaluronic acid production and suppressing MMP-1, -3, and -13 production in synoviocytes and chondrocytes.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Background
*In the **Background** section:*
Cyclic phosphatidic acid (cPA) is a naturally occurring phospholipid mediator that was originally isolated from the myxoamoebae of a true slime mold, Physarum polycephalum, in 1992 . Later, cPA was found in mammalian tissues . cPA has distinct biological activities; it inhibits autotaxin , suppresses cancer cell invasion and metastasis , attenuates ischemia-induced delayed neuronal cell death in rat hippocampal CA1 regions , and inhibits chronic and acute inflammation-induced C-fiber stimulation, and attenuates neuropathic pain . Therefore, cPA is a promising candidate of therapeutic agent for pain.[](https://www.ncbi.nlm.nih.gov/mesh/D010712)
To develop this idea, we synthesized several chemically stabilized derivatives of cPA. cPA has a unique structure comprising of a cyclic phosphate ring at the sn-2 and the sn-3 positions of the glycerol backbone (Figure 1A), which is required for the biological activities of cPA . In 2-carba-cPA (2ccPA), one of the phosphate oxygens is replaced with a methylene group at the sn-2 position (Figure 1B) . In vivo, tritium labeled 2ccPA ([3H]2ccPA) was intravenously injected to rats (30 mg/kg). Then, it has been revealed the half-life of the tritium labeled compound(s) was 81.7 h (unpublished data obtained by Mitsubishi Chemical Medience Corporation).
In vitro, we have previously shown that cPA 18:1 is stable in neutral-buffered aqueous medium for up to 24 h using liquid chromatography-mass spectroscopy (LC-MS) . The modest drop in cPA concentration over this time frame was not accompanied by a significant increase in LPA levels , suggesting that cPA may not be converted into LPA. Furthermore, we investigated the stability of 2ccPA 18:1 in 1 mM phosphate buffered saline using LC-MS/MS, and we revealed that 2ccPA 18:1 was stable for more than 2 weeks at 37°C and it was not converted into LPA (unpublished data). We previously showed that 2ccPA retains many of the biological functions of cPA and that it is a much more potent inhibitor of cancer cell invasion and metastasis and a stronger suppressor of the nociceptive reflex than natural cPA .[](https://www.ncbi.nlm.nih.gov/mesh/D010712)
Structure of cPA and 2ccPA. (A) Structure of natural occurring cPA 18:1, and (B) chemically synthesized its derivative, 2ccPA 18:1, used for the present experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D010712)
Osteoarthritis (OA) is a degenerative disease frequently associated with inflammation, joint pain, swelling, and stiffness, leading to significant functional impairment and disability at articular joints . OA is caused by characteristic structural alterations of the joint, including focal degradation of articular cartilage, subchondral bone alterations, and synovitis . OA joints are the biological site of inflammation and catabolism. Synovial inflammation likely contributes to the dysregulation of cartilage homeostasis, favoring an imbalance between the catabolic and anabolic activities of chondrocytes in remodeling the cartilage extracellular matrix (ECM) .
2ccPA inhibites chronic and acute inflammation-induced C-fiber stimulation and attenuates neuropathic pain ; we then assessed the ability of 2ccPA to relieve OA-related pain in a rabbit model in vivo. In addition, to investigate the molecular mechanisms of 2ccPA in OA-related cells, we examined the effects of 2ccPA on the inflammatory and catabolic responses of human OA synoviocytes and chondrosarcoma SW1353 cells in vitro.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Results and discussion
*In the **Results and discussion** section:*
## Pain assessment
*In the **Pain assessment** section:*
Figure 2B shows the change in right hind paw weight distribution (%) from Day 0 (surgery) to Day 42 (sacrifice) for the vehicle- and 2ccPA-treated groups. From Days 7 to 14, both the vehicle- and 2ccPA-treated groups recovered from the surgical stress observed between Days 0 and 7. After Day 14, the weight distribution (%) of the right hind paw of the vehicle-treated group gradually decreased with time, indicating that OA symptoms were induced by meniscectomy. In the 2ccPA-treated group beginning on Day 21, the weight distribution (%) was higher than in the vehicle-treated group. On Day 42, the weight distribution (%) of the 2ccPA-treated group was 1.7-fold higher than that of the vehicle-treated group.
These results suggest 2ccPA reduces OA pain in the rabbit meniscectomy model.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
Effect of 2ccPA on hind-paw weight distribution and articular swelling in a rabbit model of OA. (A) Experimental schedule. OA was induced by total meniscectomy in rabbits. Vehicle or 2ccPA was intra-articularly injected starting 7 days after surgery. (B) The change in weight distribution (%) of the right hind paw from Day 0 to Day 42 in vehicle- (open-circles) and 2ccPA-(closed circles) treated groups. All data are expressed as means ± standard error (SE). (C) The percent swelling score was also calculated and expressed as means ± SE. (*P < 0.05 and **P < 0.01 vs. vehicle-treated controls).[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Inflammatory swelling assessment
*In the **Inflammatory swelling assessment** section:*
Figure 2C shows the swelling ratio on Day 42 for the vehicle- and 2ccPA-treated groups. The swelling ratio was 11.87 ± 3.54% in the vehicle-treated group and 0.45 ± 1.78% in the 2ccPA-treated group. These results suggest 2ccPA exerts a constraining influence on the swelling resulting from OA inflammation in the rabbit meniscectomy model. Therefore, we believe 2ccPA is effective for relieving pain and reducing inflammation caused by OA.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Histopathology
*In the **Histopathology** section:*
We observed remarkable attenuation of pain and articular swelling in 2ccPA-treated animals 6 weeks after surgery (Figure 2B and C). Therefore, we chose Day 42 to evaluate the effects of 2ccPA on OA.Figure 3 shows a representative section of hematoxylin and eosin (HE)- and Safranine-O (Saf-O)-stained sections of cartilage from the right medial condyle of the femur and tibia in the vehicle- and 2ccPA-treated groups on Day 42. Stained-sections revealed typical changes of OA such as disorcanization of chondrocytes (black circles), exposure of subchondral bone (black arrowheads), cluster formation (blue arrowheads), loss of chondrocytes (green arrowheads), loss of the superficial layer (yellow arrowheads), and fissure (red arrowhead).[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
Histopathology of the medial condyle of the femur and tibia with or without 2ccPA treatment. Hematoxylin and eosin (HE)- and Safranine (Saf)-O–stained sections of the cartilage of the right medial condyle of the femur and tibia on Day 42 after surgery. Original magnification × 10 (A), ×200 (B), and × 600 (C1 and C2). Disorganization of chondrocytes (black circles), exposure of subchondral bone (black arrowheads), cluster formation (blue arrowheads), loss of chondrocytes (green arrowheads), loss of the superficial layer (yellow arrowheads), and fissure (red arrowhead) are shown.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
HE-staining revealed that chondrocyte disorganization in the medial condyle of the femur was significant in the vehicle-treated group. However, chondrocytes were much more ordered in the 2ccPA-treated group than in the vehicle-treated group. Cluster formation was also attenuated in the 2ccPA-treated versus the vehicle-treated group. Saf-O staining showed a significant loss of stainable proteoglycan in the vehicle-treated group. However, proteoglycan loss was imperceptible in the 2ccPA-treated group.
These results suggest 2ccPA relieved chondropathy and cartilage degeneration; in addition, the loss of stainable proteoglycan, chondrocyte disorganization, and cluster formation were considerably lower in the 2ccPA-treated group.[](https://www.ncbi.nlm.nih.gov/mesh/D004801)
Like the medial condyle of the femur, chondrocyte disorganization and cluster formation in the medial condyle of the tibia were significant in the vehicle-treated group; however, these morphologic changes were reduced in the 2ccPA-treated group. Saf-O staining showed that the loss of stainable proteoglycan in the vehicle-treated group was attenuated by 2ccPA.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
Other morphologic changes in the medial condyle of the femur and tibia, including loss of the superficial layer, cartilage erosion, and fibrillation and/or fissures were less substantial in the 2ccPA-treated group than in the vehicle-treated group. Due to mechanical friction during meniscectomy in the rabbit OA model, serious cartilage degeneration occurs within a week after surgery and progresses gradually . In this study, cartilage degradation induced by meniscectomy was suppressed in the 2ccPA-treated group, suggesting that in the rabbit OA model, 2ccPA may influence pain and catabolic regulation; therefore, 2ccPA provided chondroprotective effects during OA progression.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Measurement of hyaluronic acid in synoviocyte and chondrosarcoma SW1353 cells in vitro
*In the **Measurement of hyaluronic acid in synoviocyte and chondrosarcoma SW1353 cells in vitro** section:*
Synovial inflammation likely contributes to dysregulation of cartilage homeostasis, favoring an imbalance between the catabolic and anabolic activities of the chondrocyte in remodeling cartilage extra cellular matrix (ECM). Cartilage tissue is altered and damaged by inflammatory mediators and degradative enzymes in the synovial fluid of OA and the proportion of high-molecular-weight hyaluronic acid decrease while the inflammatory cytokines and ECM-degrading enzyme MMPs increase . The pain and swelling scores in this study demonstrated the antinociceptive and anti-inflammatory effects of 2ccPA administration. Therefore, we investigated the chondroprotective effects of 2ccPA in vitro by using synoviocytes and chondrosarcoma cell line SW1353, an in vitro model for primary chondrocytes in OA.[](https://www.ncbi.nlm.nih.gov/mesh/D006820)
We initially studied the effects of 2ccPA on the production of hyaluronic acid by enzyme-linked immunosorbent assay, ELISA. As shown in Figure 4, the amount of hyaluronic acid increased with incubation time in both cell types and 2ccPA significantly increased hyaluronic acid secretion in a dose-dependent manner in synoviocytes (Figure 4A). Compared with vehicle, 10 μM 2ccPA enhanced hyaluronic acid secretion by 3.2-fold in synoviocytes. On the other hand, SW1353 cells increased production of hyaluronic acid over time, but were unaffected by 2ccPA (Figure 4B). Hyaluronic acid synthetic capacity is much lower in SW1353 cells than in synoviocytes; therefore 2ccPA might not stimulate hyaluronic acid synthesis in SW1353 cells.
We suggest 2ccPA stimulates synoviocytes to synthesize hyaluronic acid, thus providing an apparent chondroprotective effect. High-molecular-weight hyaluronic acid inhibits IL-1β-stimulated production of MMP-1, -3, and -13 in chondrocytes . Therefore, hyaluronic acid induced by 2ccPA may inhibit the production of inflammatory cytokines and MMP-1, -3, and -13 in synoviocytes and SW1353 cells. We investigated the effects of 2ccPA on the inflammatory cytokines and MMPs production of both cells.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
Effect of 2ccPA on production of hyaluronic acid by synoviocytes and SW1353 cells. Synoviocytes (A) and SW1353 cells (B) were cultured with 1, 3, or 10 μM 2ccPA for the time indicated. The concentrations of hyaluronic acid in culture media were determined with ELISA. Data represent the mean ± SE of triplicate independent experiments (*P < 0.05 and **P < 0.01 vs. vehicle-treated controls).[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Inflammatory cytokine expression and secretion from IL-1β-stimulated synoviocytes and chondrosarcoma SW1353 cells
*In the **Inflammatory cytokine expression and secretion from IL-1β-stimulated synoviocytes and chondrosarcoma SW1353 cells** section:*
By using quantitative real-time PCR, we investigated transcript expression of inflammatory cytokines IL-1β, IL-6, IL-8, and TNF-α in synoviocytes and SW1353 cells. As shown in Figure 5A, IL-1β treatment for 24 h increased transcript expression of IL-6 and -8, but not TNF-α and IL-1β (data not shown) in synoviocytes. Primary rheumatoid synovial fibroblasts stimulated by IL-1β exhibit increased IL-6 and IL-8 secretion, but TNF-α is not affected . Although the synoviocytes used in this study were obtained from OA patients, the induction of inflammatory cytokines upon stimulation with IL-1β was similar to that of rheumatoid synovial fibroblasts. IL-1β treatment of SW1353 cells induced transcript expression of IL-6, -8, and TNF-α with a maximum at 3 h (data not shown) that was maintained for 24 h (Figure 5B).
Expression of IL-1β mRNA did not change over the 24 h assay period (data not shown). The induction of inflammatory cytokines by IL-1β was slightly affected by 2ccPA, and these results suggest that 2ccPA did not have a dramatic effect on inflammatory cytokine production.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
Effect of 2ccPA on mRNA expression and production of inflammatory cytokines. To measure mRNA expression, synoviocytes (A) and SW1353 cells (B) were cultured with 1 or 10 μM 2ccPA for 24 h in the presence or absence of 10 ng/mL IL-1β. mRNA levels of each gene were determined by quantitative real-time PCR. Data represent the mean ± SE of triplicate independent experiments (*P < 0.05 and **P < 0.01 vs. vehicle-treated controls). To measure IL-6 concentrations in the culture media of synoviocytes (A) and SW1353 cells (B), cells were cultured with 10 μM 2ccPA, IBF or DCF in the presence of 10 ng/mL IL-1β. After 24 h incubation, culture media were collected, and IL-6 concentrations were determined by ELISA. Data represent the mean ± SE of triplicate independent experiments (*P < 0.05 and **P < 0.01 vs. vehicle-treated controls).[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
It has been reported that the production of inflammatory cytokine IL-6 increases in OA joints and triggers catabolic reactions . Then, we performed ELISA to measure the production of IL-6 by both cells in culture media. The amount of IL-6 produced by IL-1β–stimulated synoviocytes and SW1353 cells increased to 12.9 ± 0.2 ng/mL and 0.254 ± 0.011 ng/mL, respectively (Figure 5). On synoviocytes, NSAIDs treatment reduced the inductive effect while 10 μM 2ccPA had no effect on IL-6 production (Figure 5A). On the other hand, NSAIDs and 2ccPA showed negligible effects on IL-6 production in SW1353 cells (Figure 5B).[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## MMP-1, -3, and -13 expression and secretion from IL-1β-stimulated synoviocytes and chondrosarcoma SW1353 cells
*In the **MMP-1, -3, and -13 expression and secretion from IL-1β-stimulated synoviocytes and chondrosarcoma SW1353 cells** section:*
To investigate the effects of 2ccPA on production of ECM-degrading enzymes by synoviocytes and SW1353 cells, we measured MMP-1, -3, and -13, which increase in OA joints and play a role in OA progression . Transcript expression of MMP-1, -3, and -13 in IL-1β-stimulated synoviocytes and SW1353 cells was assessed by quantitative real-time PCR. IL-1β induced expression of MMP-1, -3, and -13 in both synoviocytes and SW1353 cells after 24 h treatment as shown in Figure 6. The induction of MMPs in synoviocytes and SW1353 cells were significantly reduced by 1, and 10 μM 2ccPA (Figure 6). To investigate the effect of 2ccPA on IL-1β-stimulated MMP-1, -3, and -13 protein expression in synoviocytes and SW1353 cells, we assessed the levels of MMPs secreted into the culture medium.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
Effect of 2ccPA on mRNA expression and production of MMP-1, -3, and -13. To measure mRNA expression, synoviocytes (A) and SW1353 cells (B) were cultured with 1 or 10 μM 2ccPA for 24 h in the presence or absence of 10 ng/mL IL-1β. mRNA levels were determined by quantitative real-time PCR. Data represent the mean ± SE of triplicate independent experiments (*P < 0.05 and **P < 0.01 vs. vehicle-treated controls). To measure MMPs in the culture media of synoviocytes (A) and SW1353 cells (B), cells were cultured with 1, 3, or 10 μM 2ccPA in the presence of 10 ng/mL IL-1β. After 24 h incubation, culture media were collected and the concentration of each MMP was measured by ELISA. Data represent the mean ± SE of triplicate independent experiments (*P < 0.05 and **P < 0.01 vs. vehicle-treated controls). N.D. stands for not detected.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
MMP-1 production increased to 4.9 ± 0.2 ng/mL and 0.60 ± 0.06 ng/mL with IL-1β stimulation in synoviocytes and SW1353 cells, respectively. In synoviocytes, MMP-1 secretion was suppressed dose-dependently by 3 and 10 μM 2ccPA. In SW1353 cells, MMP-1 secretion was suppressed by 1 and 3 μM 2ccPA but dose-dependency was poor. These results suggest that to elicit suppressive function of 2ccPA, it is necessary to choose a certain dose depending on cell types.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
MMP-3 production increased to 68.7 ± 4.8 ng/mL and 20.3 ± 1.3 ng/mL with IL-1β stimulation in synoviocytes and SW1353 cells, respectively. MMP-3 production was much higher than that of MMP-1 and -13. In synoviocytes, MMP-3 secretion was suppressed dose-dependently by 2ccPA. In contrast, MMP-3 secretion in SW1353 cells was not affected by 2ccPA treatment. MMP-13 production increased to 0.81 ± 0.18 ng/mL with IL-1β stimulation in SW1353 cells, and the MMP-13 secretion was suppressed by 2ccPA.
Although expression of MMP-13 mRNA was observed in synoviocytes (Figure 6A), the amount of protein was not detected.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
There are some reports that MMP-1 and -13 degrade collagen and are expressed in synoviocytes and chondrocytes, respectively . MMP-3 degrades non-collagen matrix components of the joint and contributes to proteoglycan loss ; its expression is high in comparison to other MMPs . Our results are consistent with these reports. We suggest that the obvious suppressive effects of 2ccPA on MMP-1 and -3 production in synoviocytes, and on MMP-13 in chondrosarcoma SW1353 cells may offer the appropriate evidence to explain the chondroprotective effect we observed in vivo (Figure 3).[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## LPA1 receptor function and the suppressive effect of 2ccPA in OA
*In the **LPA1 receptor function and the suppressive effect of 2ccPA in OA** section:*
The pathophysiology of OA appears to be mediated by imbalances between the anabolic and catabolic activity of articular chondrocytes and other joint tissue cells such as synoviocytes . Inhibition of the LPA1R has been reported to have therapeutic benefits in Japanese OA , although these results were not replicated in a larger sample . Although 2ccPA is an agonist of LPA1R , it suppressed OA pathogenesis in vivo. We investigated the association of these properties of 2ccPA with the LPA1R signaling pathway and found that synoviocytes and SW1353 cells expressed high levels of LPA1R (Figure 7). In order to examine the involvement of LPA1R in IL-1β-stimulated MMPs production, we tested Ki16425, a selective antagonist for LPA1R and LPA3R. In synoviocytes, Ki16425 did not influence on MMP-1 and -3 production, and it showed no influence on the suppressive effects of 2ccPA on MMP-1 and -3 production (Figure 8A).
These results suggest the major receptor for 2ccPA suppression of MMPs in synoviocytes may not be LPA1R. On the other hand, in SW1353 cells, Ki16425 attenuated MMP-1 inhibition by 1 and 3 μM 2ccPA (Figure 8B), but had no effect on MMP-13 production (Figure 8B). Thus, we suggest the major receptor for MMP-13 suppression by 2ccPA in SW1353 cells may not be LPA1R, although this receptor may be involved in MMP-1 suppression. We need to identify the receptor involved in 2ccPA-mediated suppression of MMP-1, -3, and -13 expression hereafter.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
Expression of LPA receptors in human synoviocytes and SW1353 cells. Total RNA was extracted and expression of each LPA receptor in synoviocytes (A) and SW1353 cells (B) was determined by quantitative real-time PCR.
Effect of Ki16425 on the suppressive function of 2ccPA on MMP-1, -3, and -13 production in synoviocytes and SW1353 cells. Synoviocytes (A) or SW1353 cells (B) were pre-incubated with 10 μM Ki16425 for 30 min, then treated with 1, 3, or 10 μM 2ccPA for 24 h in the presence of 10 ng/mL IL-1β. The culture media were collected and the concentration of each MMP was measured by ELISA. Data represent the mean ± SE of triplicate independent experiments (*P < 0.05 and **P < 0.01 vs. vehicle-treated controls).[](https://www.ncbi.nlm.nih.gov/mesh/C477898)
Previous studies have suggested that hyaluronic acid reduces MMPs expression in synovial fluid . Therefore, MMP-1, -3, and -13 suppression by 2ccPA might be due to a 2ccPA-mediated increase in hyaluronic acid synthesis. Further studies are expected to clarify how 2ccPA modulates MMPs expression.[](https://www.ncbi.nlm.nih.gov/mesh/D006820)
OA causes morbidity, activity limitation, physical disability, excess health care utilization, and reduces health-related quality of life (QOL), especially in people over 60 years old. However, OA management is now limited to the symptomatic treatment of pain and inflammation without reducing joint destruction, which leads to inevitable referral for total joint replacement. Given this unresolved therapeutic need, many challenges remain in the discovery and development of disease-modifying OA drugs (DMOADs) aimed at slowing, halting, or reversing the progression of structural damage of the articular cartilage . We believe 2ccPA is a promising DMOAD candidate. The pain-relieving mechanisms of 2ccPA in the pathogenesis of OA are now under investigation.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Conclusions
*In the **Conclusions** section:*
Our results suggest 2ccPA significantly reduces the pain response to OA by inducing hyaluronic acid production and suppressing MMP-1, -3, and -13 production in synoviocytes and chondrocytes. These activities might protect chondrocytes from destruction. As a result, pain and inflammatory swelling are relieved. It is strongly suggested that 2ccPA is a promising candidate of DMOAD.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Materials and methods
*In the **Materials and methods** section:*
## Drug
*In the **Drug** section:*
We used chemically synthesized 2-carba-cPA 18:1 (2ccPA) . In the in vivo experiments, 2ccPA was dissolved in saline, and saline was used as vehicle. In the in vitro experiments, 2ccPA was dissolved in phosphate-buffered saline (PBS) containing 0.1% fatty acid-free bovine serum albumin (BSA), and PBS containing 0.1% BSA solution was used as vehicle.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## OA model
*In the **OA model** section:*
A rabbit model was used to investigate the effects of 2ccPA on the pathogenesis of OA. The design of this animal study was approved by the ethics committee of KAC Corporation (Ethics approval number: 12–0218), a contract research organization (Shiga Japan). All animals were purchased from KITAYAMA LABES Co., Ltd. (Japan), and all animal experiments were performed by KAC Corporation using 11- or 12-week-old male SPF New Zealand white rabbits (n = 12, body weight 2.1–2.3 kg). Animals were anesthetized with intravenous (i.v.) pentobarbital (32.4 mg/kg) prior to 1–4% isoflurane followed by subcutaneous infusion of lidocaine (approx. 3 mL) during surgery.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
The meniscus of the right leg was totally removed. Briefly, the boundary between the patellar ligament and articular capsule of the right hind leg and the lateral-collateral ligament were dissected. Then, the articular capsule was removed to expose the interior meniscus, and the meniscus was completely removed. Following total meniscectomy of the right knee joint, the rabbits were randomly divided into vehicle- or 2ccPA-treated group. Intra-articular treatment was initiated 7 days after surgery.
Vehicle (200 μL saline; Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan) or 50 μg/mL 2ccPA (200 μL) was injected into the joint cavity twice per week over five consecutive weeks (at 7, 11, 14, 18, 21, 25, 28, 32, 35, and 39 days after surgery). The animals were individually caged (48.5 × 30 × 35 cm3), received tap water ad libitum, and were fed a standard diet (CR-3, 150 g/day; CLEA, Japan Inc.) throughout the trials. During the experiments, the SPF room temperature was 18°C with unidirectional airflow systems; lighting was provided for 12 h daily (7:00 AM–19:00 PM). The experimental schedule is shown in Figure 2A.[](https://www.ncbi.nlm.nih.gov/mesh/C558317)
## Measurement of pain
*In the **Measurement of pain** section:*
Changes in hind paw weight distribution between the right (OA model) and left (contralateral control) limbs were measured as a pain index . Hind-paw weight distribution was measured once a week (at 7, 14, 21, 28, 32, and 35 days after surgery). The percentage of weight distribution of the right hind paw was calculated with the following equation:
## Measurement of swelling
*In the **Measurement of swelling** section:*
At 42 days after meniscectomy, articular swelling was measured with a digital vernier caliper. The maximum widths of the right and left hind paw were measured and recorded. The percentage of swelling was calculated as follows:
## Histopathological assessment of OA
*In the **Histopathological assessment of OA** section:*
At 42 days after meniscectomy, the rabbits were euthanized by exsanguination immediately after pentobarbital (32.4 mg/kg) administration. The femoral condyle and tibial plateau were resected and immediately fixed in 10% formalin buffer. After decalcification with ethylenediaminetetraacetic acid (EDTA), the samples were cut into 4-μm sections, then stained with HE for general morphology, or Saf-O for proteoglycan, and were observed using a 2.0 MP microscope (H-Micron, Hyogo, Japan, Figure 3A) and an optical microscope BX51TF (OLYMPUS, Tokyo, Japan, Figure 3B, 3C1 and 3C2). Images of 3–6 microscopic fields were incorporated into one image for Figure 3A.[](https://www.ncbi.nlm.nih.gov/mesh/D010424)
## Cell culture and measurement of hyaluronic acid, IL-6, and MMP-1, -3, and -13 produced by synoviocytes and chondrosarcoma SW1353 cells
*In the **Cell culture and measurement of hyaluronic acid, IL-6, and MMP-1, -3, and -13 produced by synoviocytes and chondrosarcoma SW1353 cells** section:*
All procedures were specifically approved by the ethics committee of Ochanomizu University (Ethics approval number: 24–12) and the National Institute of Biomedical Innovation, Japanese Collection of Research Bioresources Cell Bank (previously Health Science Research Resources Bank, Ethics approval number: 36); the patient gave full written informed consent for tissue donation. Synovial tissue was excised from the knee joint of a 60-year-old female patient with OA during replacement surgery. The patient-derived synoviocytes (Japanese Collection of Research Bioresources Cell Bank, HT91989516, Lot. 07042011) were plated at 1.5 × 104 cells/well with Dulbecco’s modified Eagle medium (DMEM) containing 10% fetal bovine serum (FBS, Life Technologies Corporation, CB) on 12-well plates and incubated overnight at 37°C in humidified 95% air and 5% CO2 atmosphere. Chondrosarcoma SW1353 cells were obtained from American Type Culture Collection (ATCC) (no. HTB-94) and were plated at 1.5 × 104 cells/well with DMEM containing 10% FBS on 12-well plates and incubated overnight as well as synoviocytes. The medium was replaced with serum-free DMEM, and the cells were serum-starved for 16 h.
To measure hyaluronic acid, 2ccPA was added after serum starvation at final concentrations of 1, 3, and 10 μM and incubated. The culture media were collected at 24, 48, and 72 h, then the concentration of hyaluronic acid was measured by ELISA (R&D Systems, Inc. MN).[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
To measure IL-6 and MMPs, the cells were treated with either 10 μM of ibuprofen (IBF), diclofenac sodium (DCF) (Wako Pure Chemical Industries, Ltd. Osaka, Japan), or various concentrations of 2ccPA for 24 h in the presence of 10 ng/mL of IL-1β (R&D Systems, Inc.). Culture media were collected at 24 h and the concentrations of IL-6 and MMPs were measured by ELISA kit according to the manufacturer’s instructions (RayBiotech, Inc., GA). For treatment with Ki16425 (Cayman Chemicals, MI), the selective antagonist for LPA1R and LPA3R, the cells were plated at 1.5 × 104 cells/well on 12-well plates and incubated for 30 min with 10 μM of Ki16425 before adding 2ccPA (1, 3, or 10 μM) in the presence of 10 ng/mL IL-1β. Culture media were collected at 24 h and the concentration of each MMP was measured by ELISA kit.[](https://www.ncbi.nlm.nih.gov/mesh/D007052)
## Quantitative real-time reverse transcription-polymerase chain reaction (RT-PCR)
*In the **Quantitative real-time reverse transcription-polymerase chain reaction (RT-PCR)** section:*
To quantitate the mRNA levels of inflammatory cytokines (IL-1β, IL-6, IL-8, TNF-α), MMPs (MMP-1, -3, -13), and LPA receptors, real-time RT-PCR was performed with SYBR Premix Ex Taq (Takara Bio, Inc., Shiga, Japan). Total RNA was extracted from cultured synoviocytes and SW1353 cells using ISOGEN reagent (Nippon Gene, Tokyo, Japan) according to the manufacturer’s instructions. cDNA was synthesized with the PrimeScript RT reagent kit (Takara Bio, Inc.). mRNA levels were quantified on a lightCycler 96 system (Roche) instrument. Gene-specific primer sets for IL-1β, IL-6, IL-8, MMP-3, and -13 were described in .
Gene-specific primer sets for TNF-α and MMP-1 were described in . Gene-specific primer sets for LPA1R, LPA4R, LPA5R, LPA6R, and p2y10 were described in . The following primer sets were used: GAPDH, 5′-GTGAAGGTCGGAGTCAACG-3′ (F) and 5′-TGAGGTCAATGAAGGGGTC-3′ (R); LPA2R, 5′-GAGGCCAACTCACTGGTCA-3′ (F) and 5′-GGCGCATCTCAGCATCTC-3′ (R); LPA3R, 5′-GAAGCTAATGAAGACGGTGATGA-3′ (F) and 5′-AGCAGGAACCACCTTTTCAC-3′ (R); and GPR87, 5′-AAATCCAGCAGGCAATTCAT-3′ (F) and 5′-CCCTGATGCTCTGGTTATGTT-3′ (R).
The data were calculated based on the Cq values, and the expression of each gene was normalized to GAPDH.
## Statistical analyses
*In the **Statistical analyses** section:*
All values are reported as means ± standard error. The data were analyzed using the Student’s t-test. A P value less than 0.05 was considered statistically significant.
## Abbreviations
*In the **Abbreviations** section:*
BSA: Bovine serum albumin; ctl: Control; cPA: Cyclic phosphatidic acid; DCF: Diclofenac sodium; DMEM: Dulbecco’s modified Eagle medium; DMOADs: Disease-modifying osteoarthritis drugs; ECM: Extracellular matrix; EDTA: Ethylenediaminetetraacetic acid; ELISA: Enzyme-linked immunosorbent assay; FBS: Fetal bovine serum; HE: Hematoxylin and eosin; IBF: Ibuprofen; IL: Interleukin; i.v.: Intravenous; LC-MS: Liquid chromatography-mass spectroscopy; LPA: Lysophosphatidic acid; LPAR:
Lysophosphatidic acid receptor; MMP: Metalloproteinase; OA: Osteoarthritis; PBS: Phosphate-buffered saline; Saf-O: Safranine-O; Vehi: Vehicle; 2ccPA: 2-carba-cyclic phosphatidic acid.[](https://www.ncbi.nlm.nih.gov/mesh/D010712)
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interests.
## Authors’ contributions
*In the **Authors’ contributions** section:*
MG, AN, RT participated in the experimental designing, collection and analyses of data, and drafted the manuscript. KO was in charge of histological analysis. TM and HM participated in the design of the study and analysis of the data. KMM elaborated a study plan, and was in charge of the overall adjustment of the experimental design and coordination of the whole study. All authors read and approved the final manuscript.
## Acknowledgments
*In the **Acknowledgments** section:*
We are very grateful to Dr. Hisako AKIYAMA (Brain Science Institute, RIKEN) for experimental support and to Mr. Noboru YAMAWAKI (Kyocera Medical Corporation) for kind and helpful advice.
This work was supported in part by the Princess Takamatsu Cancer Research Fund, the Rational Evolutionary Design of Advanced Biomolecules (REDS3) Project, the Central Saitama Area in the Program for Fostering Regional Innovation (City Area Type), and a Grant-in-Aid for Scientific Research (KAKENHI, No. 26860144) from the Ministry of Education, Culture, Sports, Science and Technology.
|
# Introduction
Mammalian Cytochrome P450-Dependent Metabolism of [Polychlorinated Dibenzo-p-dioxins](https://www.ncbi.nlm.nih.gov/mesh/D000072317) and Coplanar [Polychlorinated Biphenyls](https://www.ncbi.nlm.nih.gov/mesh/D011078)
# Abstract
*In the **Abstract** section:*
Polychlorinated dibenzo-p-dioxins (PCDDs) and coplanar polychlorinated biphenyls (PCBs) contribute to dioxin toxicity in hu[mans and wildlife after bioaccumu](https://www.ncbi.nlm.nih.gov/mesh/D000072317)la[tion ](https://www.ncbi.nlm.nih.gov/mesh/D000072317)through the foo[d chain from the environm](https://www.ncbi.nlm.nih.gov/mesh/D011078)en[t. T](https://www.ncbi.nlm.nih.gov/mesh/D011078)he authors exami[ned hu](https://www.ncbi.nlm.nih.gov/mesh/D004147)man and rat cytochrome P450 (CYP)-dependent metabolism of PCDDs and PCBs. A number of human CYP isoforms belonging to the CYP1 and CYP2 families showed remarkable activities toward [low-c](https://www.ncbi.nlm.nih.gov/mesh/D000072317)hlori[nate](https://www.ncbi.nlm.nih.gov/mesh/D011078)d PCDDs. In particular, human CYP1A1, CYP1A2, and CYP1B1 showed high activities toward monoCDDs, diCDDs, and triCDDs but no d[etect](https://www.ncbi.nlm.nih.gov/mesh/D000072317)able activity toward 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-tetraCDD). Lar[ge amino](https://www.ncbi.nlm.nih.gov/mesh/D004147) a[cids l](https://www.ncbi.nlm.nih.gov/mesh/D000072317)ocated[ at put](https://www.ncbi.nlm.nih.gov/mesh/D000072317)ative substrate-recognition sites a[nd the F-G loop in rat CYP1A1 contr](https://www.ncbi.nlm.nih.gov/mesh/D000072317)ib[uted to the succ](https://www.ncbi.nlm.nih.gov/mesh/D000072317)essful me[tabolism of](https://www.ncbi.nlm.nih.gov/mesh/D000596) 2,3,7,8-tetraCDD. Rat, but not human, CYP1A1 metabolized 3,3',4,4',5-pentachlorobiphenyl (CB126) to two hydroxylated metabo[lites. These met](https://www.ncbi.nlm.nih.gov/mesh/D000072317)abolites are probably less toxic than is [CB126, due to their higher solu](https://www.ncbi.nlm.nih.gov/mesh/C023035)bi[lity.](https://www.ncbi.nlm.nih.gov/mesh/C023035) Homology models of human and rat CYP1A1s and CB126 docking studies indicated that tw[o ami](https://www.ncbi.nlm.nih.gov/mesh/C023035)no acid differences in the CB126-binding cavity were important for CB126 metabo[lism.](https://www.ncbi.nlm.nih.gov/mesh/C023035) In this review, the importance of C[YPs in the](https://www.ncbi.nlm.nih.gov/mesh/D000596) metabolism of dioxi[ns an](https://www.ncbi.nlm.nih.gov/mesh/C023035)d PCBs in mammals and the species-b[ased ](https://www.ncbi.nlm.nih.gov/mesh/C023035)differences between humans and rats are described. In addition, the autho[rs reve](https://www.ncbi.nlm.nih.gov/mesh/D004147)al th[e mo](https://www.ncbi.nlm.nih.gov/mesh/D011078)lecular mechanism behind the binding modes of dioxins and PCBs in the heme pocket of CYPs.[](https://www.ncbi.nlm.nih.gov/mesh/D004147)
## 1. Introduction
*In the **1. Introduction** section:*
Dioxins containing polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) are generated naturally through processes such as forest fires and waste incineration; dioxins also are generated as byproducts of industrial processes (Figure 1). In contrast, coplanar polychlorinated biphenyls (PCBs) have been produced commercially (Figure 1). They are considered environmental contaminants because of the extreme toxicity of some family members, with 2,3,7,8-tetrachlorodibenzo-p-dioxin (tetraCDD) being considered the most toxic. The World Health Organization (WHO) recommends a maximum total daily intake of 1–4 pg I-TEQ/kg body weight [1]. Human exposure to dioxins appears to come predominantly from meat, fish, and dairy products [2].
Human breast milk can also be contaminated with dioxins [3].[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
Structures of polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and polychlorinated biphenyls (PCBs).[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
Toxicity evaluations of chemicals are carried out by using experimental animals, such as mice, rats, and dogs, because humans are not exposed to chemicals, the safety of which is unclear. Safety data obtained from experimental animals are extrapolated to humans using safety factors. Usually, a safety factor of 100 is applied to the No Observed Adverse Effect Level (NOAEL) dose derived from animal experiments. This figure is a product of factors for inter-species (10) and intra-species (10) differences. These factors are important to take into account when considering safety differences between experimental animals and humans; however, there are often large differences between mammals in terms of their sensitivity toward chemicals.
The metabolism of PCDDs has been studied in vivo by using experimental animals [4,5,6,7,8,9,10], and in vitro by using liver slices and liver microsomal fractions [9]. The major metabolites are hydroxylated products, glucuronide conjugates, and sulfate conjugates [5,7]. PCDD metabolism involves the initial insertion of a single oxygen atom into the PCDD molecule to form an epoxide, probably by cytochrome P450 (CYP)-dependent monooxygenases (Figure 2). Hu and Bunce
[9] studied the metabolism of PCDDs in microsomal fractions prepared from 3-methylcholanthrene (MC)-treated rats and suggested that CYP1A1 and CYP1A2 play an important role in the metabolism of PCDDs. Their findings strongly suggest that CYP-dependent monooxygenases are key enzymes for the metabolism of PCDDs and that phase II enzymes, including UDP-glucuronosyltransferase (UGT) and sulfotransferase (SULT), also play important roles in the metabolism of PCDDs.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
Cytochrome P450 (P450, CYP) monooxygenases. RH, substrate; and CPR, NADPH-P450 oxidoreductase.
Ten years ago, Ukrainian President Victor Yushchenko experienced severe 2,3,7,8-tetraCDD poisoning. The 2,3,7,8-tetraCDD level in his blood serum was 108,000 pg/g lipid weight, which was more than 50,000-fold higher than the levels in the general population. Two metabolites of 2,3,7,8-tetraCDD were detected in his feces [11]; however, there were no published reports identifying the human CYP isoforms responsible for this 2,3,7,8-tetraCDD metabolism, although CYP1A1, CYP1A2, and CYP1B1 are known to be induced by 2,3,7,8-tetraCDD in humans [12,13].[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
PCBs were once widely used as materials for industrial products such as transformers and condensers because of their characteristic non-flammability, low electrical conductivity, and chemical stability. The PCBs in materials are a mixture of chlorinated biphenyls containing one to 10 chlorine atoms on the biphenyl rings, and there are 209 isoforms, which are termed congeners. Among these congeners, coplanar PCBs, which are non-ortho and mono-ortho, exhibit a dioxin-like toxicity toward mammals [14]. Coplanar PCBs have a planar structure similar to that of dioxins and bind aryl hydrocarbon (AhR) receptors, resulting in the transcriptional activation of genes responsible for the expression of dioxin-like toxicity [15]. Although PCBs have been banned in most (if not all) industrialized countries, not just Japan, PCBs have continued to cause environmental pollution due to their leakage from products and their persistency in the environment [16].
Environmental contamination by PCBs has been detected even in the polar regions as well as in industrial areas due to the ability of PCBs to migrate long distances. The PCBs bio-accumulate in the adipose tissues of wildlife via the food chain, with high accumulation leading to dioxin-like toxicity [17].[](https://www.ncbi.nlm.nih.gov/mesh/D011078)
CYP-dependent monooxygenases are also involved in the metabolism of some PCBs. MC-treated rats that received 3,3',4,4',5-pentaCB (CB126), which is the most toxic of the PCB congeners, metabolized CB126 to 4-OH-3,3',4',5,5'-pentaCB, which was detected in their feces [18]. Haraguchi et al. detected five different metabolites, including 4- and 5-hydroxy metabolites, from rats treated with CB126 [19].
Rat, but not human, CYP1A1 metabolized CB126 to hydroxylated metabolites [20]. This in vitro experiment revealed that CYPs were responsible for the hydroxylation of CB126. In rats, hamsters, and guinea pigs treated with the CYP inducers phenobarbital (PB) and MC, 2,2',3,4',5,5',6-heptaCB (CB187) was metabolized to different hydroxylated metabolites depending on the animal species [21]. These results showed that there are species-specific differences with regard to the production of metabolites, indicating that the expression levels of CYP genes and the activities of the gene products differ among mammals.[](https://www.ncbi.nlm.nih.gov/mesh/D011078)
## 2. Yeast Expression System for Mammalian Cytochrome P450 (CYP) Isoforms
*In the **2. Yeast Expression System for Mammalian Cytochrome P450 (CYP) Isoforms** section:*
Three decades ago, heterologous expression of mammalian CYP genes using a Saccharomyces cerevisiae expression system succeeded [22]. The expressed rat CYP1A1 was localized on the yeast endoplasmic reticulum membrane and received electrons from yeast NADPH–P450 oxidoreductase (reductase) to demonstrate monooxygenase activity [23]. Yeast cells appear to have a machinery similar to that of mammalian cells with regard to the co-translational localization of mammalian microsomal CYPs. Although the S. cerevisiae genome contains three microsomal CYP genes [24], the gene products have no detectable activity toward xenobiotics, such as drugs, dioxins, and PCBs. In vitro studies revealed that the efficiency of electron transfer from yeast reductase to mammalian CYPs is nearly the same as that from mammalian reductase
[25]. Human CYP1A1, CYP1A2, CYP1B1, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C18, CYP2C19, CYP2D6, CYP2E1, and CYP3A4 genes have successfully been expressed in S. cerevisiae [26,27]. This system is thus useful for identifying CYPs involved in drug metabolism. One of the advantages of the yeast expression system is that whole cells of recombinant S. cereviasiae producing human CYPs can be used as biocatalysts for the biosynthesis of metabolites. This system was used to predict the metabolism of PCDDs and coplanar PCBs in humans.[](https://www.ncbi.nlm.nih.gov/mesh/D004147)
## 3. Metabolism of Polychlorinated Dibenzo-p-dioxins (PCDDs) by Human and Rat CYPs
*In the **3. Metabolism of Polychlorinated Dibenzo-p-dioxins (PCDDs) by Human and Rat CYPs** section:*
## 3.1. Metabolism of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (tetraCDD) in Humans
*In the **3.1. Metabolism of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (tetraCDD) in Humans** section:*
As mentioned in the Introduction, in 2004, Ukrainian President Victor Yushchenko experienced severe 2,3,7,8-tetraCDD poisoning. The 2,3,7,8-tetraCDD level in his blood serum was 108,000 pg/g lipid weight (50,000-fold above the normal level). Sorg et al. [11] monitored the levels of 2,3,7,8-tetraCDD and its metabolites in blood serum, subcutaneous fat, feces, sweat, and urine for 3 years; 8-OH-2,3,7-triCDD and 2-OH-1,3,7,8-tetraCDD were detected as metabolites in the feces, and the amount of the former was approximately two times higher than that of the latter. Trace amounts of these metabolites were detected in the serum and urine.
The half-life of 2,3,7,8-tetraCDD in Victor Yushchenko’s body was estimated to be 15.4 months. Of note, these two metabolites of 2,3,7,8-tetraCDD have also been detected in dogs [6]. The authors believe that CYPs were responsible for the hydroxylation of 2,3,7,8-tetraCDD for the following reasons: (1) The contributions of other enzymes (i.e., not CYPs) to the oxidation of PCDDs may be excluded [28]; (2) Spectral analysis strongly suggested that human CYP1A1 binds 2,3,7,8-tetraCDD in its substrate-binding pocket [29]; and (3) Human recombinant CYP1A2-dependent activity was inhibited by 2,3,7,8-tetraCDD [30].[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
Furthermore, the authors concluded that CYPs belonging to the CYP1 family, that is, CYP1A1, CYP1A2, and/or CYP1B1, could metabolize 2,3,7,8-tetraCDD, although their activity was too subtle to detect under in vitro procedures. The metabolites 8-OH-2,3,7-triCDD and 2-OH-1,3,7,8-tetraCDD that were detected in Victor Yushchenko were likely produced by CYPs belonging to the CYP1 family whose expression was induced by 2,3,7,8-tetraCDD (Figure 3A) [31].[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
Metabolic pathways for 2,3,7,8-tetraCDD (A) and 2-monoCDD (B) involving CYPs.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
## 3.2. CYP-Dependent Metabolism of PCDDs
*In the **3.2. CYP-Dependent Metabolism of PCDDs** section:*
Recombinant yeast microsomal fractions containing all the human CYPs were examined for their participation in the metabolism of PCDDs. Multiple CYP isoforms showed remarkable metabolic processes toward dibenzo-p-dioxin (DD), 1-monoCDD, 2-monoCDD, 2,3-diCDD, 2,7-diCDD, and 2,3,7-triCDD [29] (Figure 3B). These metabolic processes included multiple reactions such as hydroxylation at an unsubstituted position, hydroxylation with migration of a chlorine substituent, and hydroxylation with elimination of a chlorine substituent. Clear differences were observed among the CYP1, CYP2, and CYP3 families. The CYP1 family showed high activities toward DD and mono-, di-, and triCDDs and remarkable activities toward 2,7-diCDD and 2,3,7-triCDD.
In contrast, the CYP2 family showed activities toward DD, 1-monoCDD, 2-monoCDD, and 2,3-diCDD but no activities toward 2,7-diCDD and 2,3,7-triCDD. CYP3A4, which is the most important CYP in drug metabolism, showed no activities toward PCDDs. None of the CYPs showed detectable activity toward 2,3,7,8-tetraCDD.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
## 3.3. Binding of Tetra- and PentaCDDs to the Substrate-Binding Pocket of CYPs
*In the **3.3. Binding of Tetra- and PentaCDDs to the Substrate-Binding Pocket of CYPs** section:*
As described earlier, none of the CYPs showed activity toward 2,3,7,8-tetraCDD [29]. So, can CYPs bind 2,3,7,8-tetraCDD or not? To answer this question, tetraCDD-induced difference spectra were measured. Addition of 2,3,7-triCDD or 2,3,7,8-tetraCDD to microsomal fractions containing human CYP1A1 induced typical type I spectra, indicating a change in the heme iron of CYP1A1 from a low-spin state to a high-spin state upon binding of 2,3,7-triCDD or 2,3,7,8-tetraCDD. The value of the dissociation constant Kd for 2,3,7-triCDD was estimated to be 0.65 µM. Although the authors did not determine the Kd value for 2,3,7,8-tetraCDD, the affinity for 2,3,7,8-tetraCDD appeared to be not particularly different from that for 2,3,7-triCDD [30].
These results indicate that CYP1A1 can bind 2,3,7,8-tetraCDD in its substrate-binding pocket but shows no detectable activity toward it.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
Staskal et al. reported that human recombinant CYP1A2-dependent activity was inhibited by 2,3,7,8-tetraCDD, 1,2,3,7,8-pentaCDD, 2,3,7,8-tetraCDF, and 2,3,4,7,8-pentaCDF, with Ki values of less than 1 μM [30]. These results indicate that CYP1A2 can bind to these dioxins with high affinity. On the basis of these results, PCDDs and PCDFs with five chlorine substituents could dock into the substrate-binding pocket of CYPs belonging to the CYP1 family.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
## 3.4. Further Metabolism of PCDDs after CYP-Dependent Hydroxylation
*In the **3.4. Further Metabolism of PCDDs after CYP-Dependent Hydroxylation** section:*
In vivo studies using experimental animals have demonstrated that the major metabolites of PCDDs are hydroxylated products, glucuronide conjugates, and sulfate conjugates [5,7]. The formation of glucuronides or sulfates is an indispensable step in the detoxification of lipophilic compounds, which subsequently are transformed into more hydrophilic metabolites and excreted in bile or urine. Glucuronidation of steroids, bile acids, bilirubin, hormones, drugs, and environmental toxicants is catalyzed by UDP glucuronosyltransferases (UGTs), whose gene family in the human genome contains 19 species with different substrate specificities [32]. No human UGT species that catalyzes the glucuronidation of PCDDs has yet been identified, although some human UGT species appear to be involved in the metabolism of PCDDs.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
## 3.5. Metabolism of 8-OH-2,3,7-triCDD by Human Recombinant UDP Glucuronosyltransferases (UGTs)
*In the **3.5. Metabolism of 8-OH-2,3,7-triCDD by Human Recombinant UDP Glucuronosyltransferases (UGTs)** section:*
The additional metabolism of dioxins by UGTs that occurs after the CYP-dependent reaction was examined. 8-OH-2,3,7-triCDD, a major metabolite of 2,3,7-triCDD formed via the activity of CYP1A1, CYP1A2, and CYP1B1, served as a substrate for UGT [33]. Because 8-OH-2,3,7-triCDD is a major metabolite of 2,3,7,8-tetraCDD in mammals [6,34], it is important to understand the metabolism of 8-OH-2,3,7-triCDD to fully understand the metabolism of 2,3,7,8-tetraCDD in mammals. The metabolism of 8-OH-2,3,7-triCDD by 12 species of human UGTs was examined by using recombinant UGTs expressed in baculovirus-infected insect cells (Supersomes, BD Sciences, San Jose, CA, USA). Surprisingly, 2,3,7-triCDD glucuronide was detected in the recombinant system containing each of UGT1A1, UGT1A3, UGT1A7, UGT1A8, UGT1A9, UGT1A10, UGT2B7, UGT2B15, and UGT2B17, yet it was not detected in systems containing each of UGT1A4, UGT1A6, and UGT2B4.
Thus, 8-OH-2,3,7-triCDD is a good substrate for most human UGT species.[](https://www.ncbi.nlm.nih.gov/mesh/D004147)
## 3.6. Successive Metabolism of 2,3,7-triCDD by CYPs and UGTs in Human Liver Microsomes
*In the **3.6. Successive Metabolism of 2,3,7-triCDD by CYPs and UGTs in Human Liver Microsomes** section:*
As described in the previous sections, heterologous expression systems for CYPs and UGTs are very useful for predicting the metabolism of dioxins. In addition, commercially available human liver microsomes are also useful for understanding the metabolism of dioxins in the human body. CYP-dependent 2,3,7-triCDD hydroxylation activities in human liver microsomes from 10 individual human livers in the presence of NADPH were measured [33]. A significant difference in hydroxylation activity was observed among the different microsomal preparations from the 10 human livers, and the activity range was 15-fold. Although both human CYP1A1 and CYP1B1 have high 2,3,7-triCDD 8-hydroxylation activity, their expression levels in human liver are quite low, whereas CYP1A2 is a major CYP isoform in the liver [35].
Therefore, the authors predicted that CYP1A2 was the major CYP that catalyzed the hydroxylation of 2,3,7-triCDD in human liver. Because phenacetin O-deethylation is specifically metabolized by CYP1A2 [36], the authors examined the correlation between 2,3,7-triCDD 8-hydroxylation and phenacetin O-deethylation in human liver microsomes. As expected, a good correlation (r = 0.92) was observed between the two reactions. These results strongly suggest that CYP1A2 is responsible for 2,3,7-triCDD 8-hydroxylation in human liver, and the significant difference in the activity among the microsomal preparations from the 10 different human livers likely reflected differences in the CYP1A2 contents of the liver microsomes.
In contrast, glucuronidation activity toward 8-OH-2,3,7-triCDD was not significantly different among the different microsomal preparations from the 10 human livers; the activity range was only 1.9-fold. A time course of 2,3,7-triCDD metabolisms on human liver microsomes in the presence of NADPH and UDP-glucuronic acid was also examined [33]. Initially, 8-OH-2,3,7-triCDD levels increased linearly with time, and then reached a plateau. Glucuronide formation was detected, but after 8-OH-2,3,7-triCDD accumulation following a lag phase, glucuronide levels then increased linearly with time. The time courses of these metabolites thus represent the typical sequential conversion by a two-enzyme system.[](https://www.ncbi.nlm.nih.gov/mesh/D004147)
## 3.7. Species-Based Difference in CYP-Dependent Metabolism of PCDDs between Humans and Rats
*In the **3.7. Species-Based Difference in CYP-Dependent Metabolism of PCDDs between Humans and Rats** section:*
Human and rat CYP1A-dependent metabolism of PCDDs by CYP1A subfamily members by using recombinant yeast microsomes was compared. Considerable species differences between humans and rats were observed for both CYP1A1- and CYP1A2-dependent metabolism of dioxins. Among four CYPs, rat CYP1A1 showed the highest activity toward DD, mono-, di-, and triCDDs [37].[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
## 3.8. Generation of 2,3,7,8-tetraCDD-Metabolizing CYPs by Modifying Rat CYP1A1 through Site-Directed Mutagenesis
*In the **3.8. Generation of 2,3,7,8-tetraCDD-Metabolizing CYPs by Modifying Rat CYP1A1 through Site-Directed Mutagenesis** section:*
The extremely high toxicity of 2,3,7,8-tetraCDD derives from its high affinity for the Ah receptor and its nearly undetectable metabolism in the mammalian body [9]. Given that 2,3,7-triCDD is a good substrate for rat CYP1A1, the authors hypothesized that enlargement of the space for the putative substrate-binding pocket and of the substrate access channel of rat CYP1A1 might generate catalytic activity toward 2,3,7,8-tetraCDD. Large amino acid residues such as Phe, Tyr, Leu, and Ile that are involved in substrate binding and substrate entry were substituted for alanine by using site-directed mutagenesis. Among eight mutants examined, the mutant in the putative F–G loop, F240A, catalyzed the conversion of 2,3,7,8-tetraCDD to 8-OH-2,3,7-triCDD [38]. Because the affinity of 8-OH-2,3,7-triCDD for Ah receptor was less than 0.001% of that for 2,3,7,8-tetraCDD, this metabolic event resulted in remarkable detoxification of 2,3,7,8-tetraCDD.
A docking model of 2,3,7,8-tetraCDD with rat CYP1A1 on the basis of the crystal structure of human CYP1A2 was constructed [20,39,40] using homology modeling. Figure 4 shows the docking model of 2,3,7,8-tetraCDD in the active site of rat CYP1A1. 2,3,7,8-tetraCDD is accommodated in the planar active site. This model concurs well with the experimental result that rat CYP1A1 can bind 2,3,7,8-tetraCDD as described in Section 3.2. The figure shows the three-dimensional overall structure of rat CYP1A1 Phe240 located in the F–G loop, which is associated with the membrane and involved in substrate entry.
As mentioned in the Section 3.1, 8-OH-2,3,7-triCDD is a major metabolite of 2,3,7,8-tetraCDD in mammals [6,34]. These in vivo studies thus appear to be inconsistent with our results showing that native CYPs have no catalytic activity toward 2,3,7,8-tetraCDD. However, it is possible that native CYP1A may be able to convert 2,3,7,8-tetraCDD to 8-OH-2,3,7-triCDD.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
A docking model of 2,3,7,8-tetraCDD into rat CYP1A1. The cyan ribbon represents the F–G loop. The yellow shaded region is the substrate-binding cavity.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
## 4. Metabolism of PCBs by Human and Rat CYPs
*In the **4. Metabolism of PCBs by Human and Rat CYPs** section:*
## 4.1. In Vitro Metabolism of CB126 with Microsomal Fractions from Recombinant Yeast
*In the **4.1. In Vitro Metabolism of CB126 with Microsomal Fractions from Recombinant Yeast** section:*
Microsomal fractions containing human and rat CYP1A1s were prepared from recombinant S. cerevisiae expressing the corresponding genes [22]. With NADPH as an electron donor and the NADPH-recycling system using gluose-6-phosphate and glucose-6-phosphate dehydrogenase, CB126 was mixed with the microsomal fractions. After a 2-h incubation at 37 °C with shaking, 13C-hydroxylated PCBs, as internal standards, were added to the reaction mixture. CB126 and its metabolites were extracted with hexane from the reaction mixture and subjected to methylation [41]. The residues dissolved in hexane were subjected to high-resolution gas chromatography–high-resolution mass spectrometry.
Rat CYP1A1 showed two NADPH-dependent peaks, whereas no peaks were detected with human CYP1A1 [20]. The vector control also showed no detectable peaks.[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
## 4.2. Identification of CB126 Metabolites
*In the **4.2. Identification of CB126 Metabolites** section:*
Koga et al. reported that rats that had orally been administered CB126 excreted 4-OH-3,3',4',5,5'-pentaCB as a major metabolite in their feces [18]. However, the CYP isoforms involved in this reaction were not identified. The authors showed that rat CYP1A1 produced two hydroxylated metabolites, identified as 4-OH-3,3',4',5-tetraCB and 4-OH-3,3',4',5,5'-pentaCB on the basis of the retention times of the authentic standards, their isotope ratios, and their fragment ion patterns [20]. Much more of the newly identified metabolite 4-OH-3,3',4',5-tetraCB than of 4-OH-3,3',4',5,5'-pentaCB was produced in vitro. These results suggest that CB126 is detoxified by rat CYP1A1 because PCB congeners that are less chlorinated and hydroxylated generally have less hydrophobicity (i.e., increased body clearance) and decreased binding affinity for the Ah receptor that is responsible for dioxin toxicity
[18]. In contrast, human CYP1A1 did not produce detectable metabolites, and rat CYP1A1 in the reaction mixture without NADPH did not yield either peak for the hydroxylated metabolites. These results strongly suggest that CB126 is more toxic for humans than for rats due to its bioaccumulation in humans. Dioxin toxicity, including PCB toxicity, has been defined by estimations obtained largely from in vivo experiments using animals. The species-specific differences between humans and rats in terms of PCB metabolism could lead to misinterpretation of toxicity evaluations.[](https://www.ncbi.nlm.nih.gov/mesh/C023035)
## 4.3. Molecular Modeling of Human and Rat CYP1A1s
*In the **4.3. Molecular Modeling of Human and Rat CYP1A1s** section:*
Human CYP1A1 shares 79% amino acid sequence homology with rat CYP1A1. Four amino acid residues—Ser116, Ser122, Asn221, and Leu312 for humans and the corresponding amino acids Ala120, Thr126, Ser225, and Phe316 for rats—are not conserved among the amino acid residues that comprise the substrate-binding cavity (Table 1). These residues contribute to the differences in the metabolism of CB126. To clarify the influence of these amino acid differences on metabolic activities, 3D structures of human and rat CYP1A1s were constructed. The 3D structure of human CYP1A1 was based on human CYP1A2 [39], and that of rat CYP1A1 was based on the constructed human CYP1A1 [40].
In the rat CYP1A1 model, there was a steric crush error between Tyr263 on the G-Helix and Phe316 on the I-Helix, whereas no errors occurred in the human CYP1A1 model (Figure 5). Therefore, it was thought that the side chain of Phe316 in rat CYP1A1 flips into the substrate-binding cavity given that Ala120 is sufficiently small enough to allow the flip. This creates a smaller cavity volume (510 Å) compared with that of human CYP1A1 (600 Å). Leu312 in human CYP1A1 does not conflict with Tyr259, corresponding to Tyr263 in rat CYP1A1, because the side chain of Leu is smaller than that of Phe. Leu is mostly conserved in animals, including dog, guinea pig, monkey, mouse, and rabbit, although the rat has Phe (Table 1).
Furthermore, Ala120 in rat CYP1A1 is unique among these animals. The combination of Ala120 and Phe316 confers unique metabolic activities to rat CYP1A1 by decreasing the cavity volume.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Structures of the amino acid residues in the substrate-binding pocket in mammalian CYP1A1s.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Mammalian amino acids corresponding to Ala120, Tyr263, and Phe316 of rat CYP1A1 are represented.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Docking models of CB126 into rat and human CYP1A1s. Orientation numbers of CB126 in a substrate-binding cavity (A) and conformation numbers of CB126 within 5 Å of the heme (B). Left and right panels show rat and human CYP1A1s, respectively. Yellow and green shades indicate substrate-binding cavities of rat and human CYP1A1s, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/C023035)
## 4.4. Construction of Docking Models with CB126 and CYP1A1s
*In the **4.4. Construction of Docking Models with CB126 and CYP1A1s** section:*
Docking models revealed that CB126 was more stable in the cavity of rat CYP1A1 than in that of human CYP1A1 [20] (Figure 5A). This stability was due to less orientation of CB126 in the cavity of rat CYP1A1 relative to that in human CYP1A1. In contrast, CB126 was more accessible to the heme of rat CYP1A1 than to that of human CYP1A1 (Figure 5B); three conformations were predicted to be close to the heme (less than 5 Å) of rat CYP1A1, whereas no conformation was in human CYP1A1. Less than 5 Å between the four-position of the carbon in CB126 and the iron of the heme in CYPs is necessary for the reaction to occur.
These results reveal that the orientation of PCBs in the cavity of CYPs and the ability to be close to the iron of the heme of CYP1A1 stem from the differences in the CYP amino acids in different species and their subsequent effects on PCB metabolism.[](https://www.ncbi.nlm.nih.gov/mesh/C023035)
## 5. Conclusions
*In the **5. Conclusions** section:*
Here, the authors described the metabolism of PCDDs and PCBs via CYP-dependent hydroxylation. Molecular modeling of CYPs and the construction of docking models with these compounds explained the reaction mechanisms based on differences in amino acid sequence. Appropriate mutations at these amino acids made it possible to create novel CYPs with higher metabolic activity. This approach can be applied to the efficient remediation of PCDDs and PCBs. It is important to understand the metabolism of these compounds with respect to their toxicity for humans, but it is unacceptable to administer these compounds to humans.
Therefore, toxicity data from experimental animals are extrapolated to estimate toxicity toward humans. However, there are metabolic differences between experimental animals and humans, which complicate such toxicity determinations. Our approach to in vitro CYP-dependent metabolism of PCDDs and PCBs is a promising method to more accurately evaluate toxicity for humans.[](https://www.ncbi.nlm.nih.gov/mesh/D000072317)
# Author Contributions
*In the **Author Contributions** section:*
H.I. and T.S. prepared the manuscript; K.Y., T.I., and S.-I.I. made critical contributions to the research described and provided comments during manuscript preparation and revision.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
|
# Introduction
Characterization of [VPS34-IN1](https://www.ncbi.nlm.nih.gov/mesh/C000593572), a selective inhibitor of Vps34, reveals that the [phosphatidylinositol 3-phosphate](https://www.ncbi.nlm.nih.gov/mesh/C055525)-binding SGK3 protein kinase is a downstream target of class III phosphoinositide 3-kinase
# Abstract
*In the **Abstract** section:*
The Vps34 (vacuolar protein sorting 34) class III PI3K (phosphoinositide 3-kinase) phosphorylates PtdIns (phosphatidylinositol) at endosomal membranes to generate PtdIns(3)P that regulates membrane trafficking processes via its ability to recruit a subset of proteins possessing PtdIns(3)P-binding PX [(phox ](https://www.ncbi.nlm.nih.gov/mesh/D010716)ho[mology) and FYVE dom](https://www.ncbi.nlm.nih.gov/mesh/D010716)ains. In the present study, we descri[be a highl](https://www.ncbi.nlm.nih.gov/mesh/C055525)y selective and potent inhibitor of Vps34, termed VPS34-IN1, that inhibits Vps34 with 25 nM IC50 in vitro,[ but does ](https://www.ncbi.nlm.nih.gov/mesh/C055525)not significantly inhibit the activity of 340 protein kinases or 25 lipid kinases tested that include all isoforms of class I as well as [class II ](https://www.ncbi.nlm.nih.gov/mesh/C000593572)PI3Ks. Administration of VPS34-IN1 to cells induces a rapid dose-dependent dispersal of a specific PtdIns(3)P-binding probe from endosome membranes, within 1 min, without affecting the ability of class I PI3K to regulate Akt. Mo[reover, w](https://www.ncbi.nlm.nih.gov/mesh/C000593572)e explored whether SGK3 (serum- and glucocorticoid-regulated kina[se-3), the](https://www.ncbi.nlm.nih.gov/mesh/C055525) only protein kinase known to interact specifically with PtdIns(3)P via its N-terminal PX domain, might be controlled by Vps34. Mutations disrupting PtdIns(3)P binding ablated SGK3 kinase activity by suppressing phosphorylation of the T-loop [PDK1 (phosphoinos[itide-depe](https://www.ncbi.nlm.nih.gov/mesh/C055525)ndent kinase 1) site] and hydrophobic motif (mammalian target of rapamycin site) r[esidues. V](https://www.ncbi.nlm.nih.gov/mesh/C055525)PS34-IN1 induced a rapid ~50–60% loss of SGK3 phosphorylation within 1 min. VPS34-IN1 did not inhibit activity of the SGK2 isoform that does not possess a PtdIns(3)P-binding PX domain. Furthermore, clas[s I PI3K ](https://www.ncbi.nlm.nih.gov/mesh/C000593572)inhibitors (GDC-0941 and BKM120) that do not inhibit Vps34 suppresse[d SGK3 ac](https://www.ncbi.nlm.nih.gov/mesh/C000593572)tivity by ~40%. Combining VPS34-IN1 and GDC-0941 reduced SGK3 activity[ ~80–90%. ](https://www.ncbi.nlm.nih.gov/mesh/C055525)These data suggest SGK3 phosphorylation and hence activity[ is cont](https://www.ncbi.nlm.nih.gov/mesh/C532162)rolle[d by t](https://www.ncbi.nlm.nih.gov/mesh/C571178)wo pools of PtdIns(3)P. The first is produced through phosphorylation of[ PtdIns b](https://www.ncbi.nlm.nih.gov/mesh/C000593572)y Vps[34 at th](https://www.ncbi.nlm.nih.gov/mesh/C532162)e endosome. The second is due to the conversion of class I PI3K product, PtdIns(3,4,5)P3 into PtdIns(3)P, via the sequent[ial action](https://www.ncbi.nlm.nih.gov/mesh/C055525)s of the PtdIns 5-phosphatases [SHIP1/2 (Src homolo[gy 2-d](https://www.ncbi.nlm.nih.gov/mesh/D010716)omain-containing inositol phosphatase 1/2)] and PtdIns 4-phosphatase [INPP4B (inositol p[olyphosphate 4-](https://www.ncbi.nlm.nih.gov/mesh/C060974)phosph[atase type](https://www.ncbi.nlm.nih.gov/mesh/C055525) II)]. VPS34-IN1 will be a useful probe to delineate physiological roles of the Vps34. Monitoring SGK3 phosphorylation and activity could be employed as a biomarker of Vps34 activity, in an analogous manner by [which Akt](https://www.ncbi.nlm.nih.gov/mesh/C000593572) is used to probe cellular class I PI3K activity. Combining class I (GDC-0941) and class III (VPS34-IN1) PI3K inhibitors could be used as a strategy to better analyse the roles and regulation of the elusive class II PI3K.[](https://www.ncbi.nlm.nih.gov/mesh/C532162)
We characterize VPS34-IN, a potent and selective inhibitor of class III Vps34 PI3K. Using VPS34-IN1, we demonstrate that PtdIns(3)P, produced by Vps34 controls phosphorylation and activity of the SGK3 protein kinase.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## INTRODUCTION (cont.)
*In the **INTRODUCTION (cont.)** section:*
PI3Ks (phosphoinositide 3-kinases) phosphorylate the 3′-position hydroxy group of the D-myo-inositol head group of PtdIns (phosphatidylinositol) to generate 3-phosphoinositides that are critical in switching on downstream signalling pathways. These orchestrate a wide range of biological responses including controlling cell growth, proliferation and intracellular trafficking. Work in this area has taken on added urgency, as it is now clear that understanding disruptions in diverse PI3K signalling pathways lie at the centre of understanding major diseases such as cancer, inflammation heart failure and diabetes.[](https://www.ncbi.nlm.nih.gov/mesh/D007294)
There are three different classes of PI3Ks termed class I, class II and class III. These enzymes act on most cellular membranes. The most studied enzymes are the four members of class I PI3Ks (p110α, p110β p110γ and p110δ) that phosphorylate PtdIns(4,5)P2 at the plasma membrane to generate the signalling second messenger PtdIns(3,4,5)P3 in response to agonists that trigger activation of growth factors, Ras or G-protein-coupled receptors. PtdIns(3,4,5)P3, as well as its immediate breakdown product PtdIns(3,4)P2, trigger downstream signalling responses by interacting specifically with a subgroup of signalling proteins that possess a PH (pleckstrin homology) domain that bind with high affinity to these 3-phosphoinositides. One of the best characterized signalling pathways activated by class I PI3K proteins are the Akt protein kinases that control critical processes such as metabolism and growth by phosphorylating multiple targets including FOXO (forkhead box O) transcription factors, TSC2 (tuberous sclerosis complex 2) and GSK3 (glycogen synthase kinase 3).
Following activation of class I PI3Ks, Akt and one of its upstream protein kinases termed PDK1 (phosphoinositide-dependent kinase 1) are recruited to the plasma membrane via their PtdIns(3,4,5)P3/PtdIns(3,4)P2-binding PH domains. This results in PDK1 phosphorylating the T-loop Thr308 residue thereby partially activating Akt. Akt is also phosphorylated by mTORC2 (mammalian target of rapamycin complex-2) at Ser473 located within the hydrophobic motif at the C-terminal non-catalytic region. Phosphorylation of Ser473 also serves to promote the phosphorylation of Akt at Thr308 by PDK1. PtdIns(3,4,5)P3 can then be converted into PtdIns(3)P via the sequential actions of the PtdIns 5-phosphatases [SHIP1/2 (Src homology 2-domain-containing inositol phosphatase 1/2)] and PtdIns 4-phosphatase [INPP4B (inositol polyphosphate 4-phosphatase type II)].
PtdIns(3)P is rapidly dephosphorylated in cells to PtdIns by a family of myotubularin PtdIns 3-phosphatases.[](https://www.ncbi.nlm.nih.gov/mesh/D019269)
Class II and class III PI3Ks phosphorylate PtdIns to generate PtdIns(3)P. The class II PI3K subfamily has three members in vertebrates (PI3KC2α, PI3KC2β and PI3KC2γ), but the roles that these perform and how they are regulated are poorly understood. The single class III PI3K isoform termed Vps34 (vacuolar protein sorting 34) play an important role in controlling vesicular protein sorting, a phenomenon that was first discovered in yeast. In all eukaryotes, Vps34 forms a core complex with two other protein subunits termed Vps15 (a serine/threonine protein kinase) and beclin-1 (also known as Vps30 or ATG6). This core complex then interacts with a growing list of proteins to form distinct complexes controlling membrane endosomal trafficking processes and endosome–lysosome maturation as well as autophagy.[](https://www.ncbi.nlm.nih.gov/mesh/D010716)
The most thoroughly characterized PtdIns(3)P-binding domains are a subset of FYVE and PX (Phox homology) domains. There is a pool of PtdIns(3)P that is highly enriched on early endosomes and in the internal vesicles of multivesicular endosomes. PtdIns(3)P-binding PX and FYVE domains display relatively low PtdIns(3)P-binding affinity which, in combination with high myotubularin PtdIns 3-phosphatase activity, permits rapid and highly dynamic localization and responses. Consistent with a key role of Vps34 and PtdIns(3)P in regulating vesicular trafficking, several proteins containing PX and FYVE domains specific for PtdIns(3)P have been identified that play critical roles in protein sorting pathways.[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
In the present study, we characterize a Vps34 inhibitor termed VPS34-IN1. We show that this compound inhibits recombinant Vps34 with nanomolar potency, but does not significantly inhibit other protein kinases or lipid kinases tested, including class I or class II PI3Ks. We demonstrate that VPS34-IN1 rapidly reduces endosomal PtdIns(3)P levels within 1 min of drug treatment. Furthermore, we utilize VPS34-IN1 to demonstrate that Vps34 plays a role in regulating the phosphorylation and activity of the SGK3 (serum- and glucocorticoid-regulated protein kinase-3), which is the only protein kinase known to possess a selective PtdIns(3)P-binding PX domain. Our data suggest that SGK3 activity is partially controlled by a pool of PtdIns(3)P, produced via phosphorylation of PtdIns by Vps34 and another pool of PtdIns(3)P that could be derived from dephosphorylation of PtdIns(3,4,5)P3 to PtdIns(3)P through the sequential actions of SHIP1/2 and INPP4B PtdIns phosphatases.
VPS34-IN1 will be a useful probe for defining the roles that the class III PI3K plays. Moreover, analysing SGK3 phosphorylation and/or activity could be used as a downstream biomarker for Vps34 activity in the same way in which Akt phosphorylation is currently deployed to monitor class I PI3K pathway.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## MATERIALS AND METHODS
*In the **MATERIALS AND METHODS** section:*
## Materials
*In the **Materials** section:*
Protein G–Sepharose was from GE Healthcare. [γ-32P]ATP was from PerkinElmer. Agarose-conjugated anti-FLAG M2 antibody, Triton X-100, EDTA, EGTA, sodium orthovanadate, sodium glycerophosphate, sodium fluoride, sodium pyrophosphate, 2-mercaptoethanol, sucrose, benzamidine, Tween 20, Tris/HCl, sodium chloride, magnesium acetate and doxycyclin were from Sigma. PMSF was from Melford. Tissue culture reagents, Novex 4–12% Bis-Tris gels and NuPAGE LDS sample buffer was from Invitrogen.
Ampicillin was from Merck. P81 phosphocellulose paper was from Whatman. Methanol and chloroform were from VWR Chemicals. Inhibitors GDC-0941 (Axon Medchem), GSK2334470 (Tocris), AZD8055 (Selleck) and BKM120 (Chemie Tek) were purchased from the indicated suppliers. VPS34-IN1 (1-[{2-[(2-chloropyridin-4yl)amino]-4′-(cyclopropylmethyl)-[4,5′-bipyrimidin]-2′-yl}amino]-2-methyl-propan-2-ol) was synthesized as described in patent WO 2012085815 A1
[Cornella Taracido, I., Harrington, E.M., Honda, A. and Keaney, E. (2012) Preparation of bipyrimidinamine derivatives for use as Vps34 inhibitors; method for synthesis of this compound is described on page 73, Table 4, example 16a. VPS34-IN1 has a CAS registry number 1383716-33-3].[](https://www.ncbi.nlm.nih.gov/mesh/D012685)
## General methods
*In the **General methods** section:*
Recombinant DNA procedures were performed using standard protocols. Mutagenesis was performed using the QuikChange site-directed mutagenesis (Stratagene) with KOD polymerase (Novagen). DNA constructs were purified from Escherichia coli DH5α using a Maxi prep kit (Qiagen) according to the manufacturer's instructions. Verification of the constructs was performed by the Sequencing Service (MRC Protein Phosphorylation Unit, College of Life Sciences, University of Dundee, U.K.; http://www.dnaseq.co.uk). DNA for bacterial protein expression was transformed into E. coli BL21-CodonPlus (DE3)-RIL cells (Stratagene).
All recombinant proteins generated for the present study have an assigned [DU] number and are described on the reagents website (http:s://mrcppureagents.dundee.ac.uk/). Recombinant proteins used in the present study were as follows: Vps34/15 [DU8692], GST–TAPP1-(195–315) [DU17323], GST–Grp1-(241–399) [DU3464], GST–PLCδ-(1–178) [DU12981], 3×FLAG–SGK3-(1–162) wild-type [DU44877], 3×FLAG–SGK3-(1–162) R50A
[DU44923] and 3×FLAG–SGK3-(1–162) R90A [DU44883].
## Cell culture, transfection and cell lysis
*In the **Cell culture, transfection and cell lysis** section:*
U20S cell line was kindly provided by John Rouse (MRC Protein Phosphorylation and Ubiquitylation Unit, University of Dundee, Dundee, U.K.). Cells were cultured in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% (v/v) FBS, 2 mM L-glutamine, 100 U/ml penicillin and 0.1 mg/ml streptomycin. cDNA for mouse Hrs (residues 147–223 ×2 joined by a linker) were cloned into a pBABE.puro vector. The construct was co-transfected into HEK (human embryonic kidney)-293FT cells with GAG/POL and VSV-G (vesicular stomatitis virus glycoprotein) expression plasmids (Clontech) for retrovirus production using Lipofectamine 2000 (Life Technologies) in accordance with the manufacturers’ instructions. The virus was harvested 48 h after transfection and applied to U20S cells in the presence of 10 μg/ml polybrene.
Cells were selected with 10 μg/ml puromycin before single cell colony selection to identify low expression GFP–FYVEHrs clones. U2OS Flp/In cell line was kindly provided by John Rouse. Cells were cultured in McCoy medium supplemented with 10% (v/v) FBS, 2 mM L-glutamine, 100 U/ml penicillin and 0.1 mg/ml streptomycin. Stable cell lines expressing doxycycline-inducible proteins were generated using the Flp-In™ T-REx™ system (Invitrogen).
Stable transfected clones were selected using 0.01 mM hygromycin in cell medium. Protein expression was induced by adding 0.01 mg/ml doxycycline in the medium for 24 h prior to cell lysis. Inhibitor treatment was carried out as described in the Figure legends. The cells were lysed in buffer containing 50 mM Tris/HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM sodium orthovanadate, 10 mM sodium glycerophosphate, 50 mM sodium fluoride, 10 mM sodium pyrophosphate, 0.27 M sucrose, 0.1% 2-mercaptoethanol, 1 mM benzamidine and 0.1 mM PMSF. Lysates were clarified by centrifugation at 16000 g for 10 min at 4°C.
Protein concentration was calculated using Bradford assay (Thermo Scientific). Immunoblotting and immunoprecipitation were performed using standard procedures. The signal was developed using the ECL Western Blotting Detection kit (GE Healthcare) on Hyperfilm ECL film (GE Healthcare).[](https://www.ncbi.nlm.nih.gov/mesh/D005973)
## Antibodies
*In the **Antibodies** section:*
The following antibodies were raised in sheep, by the DSTT (Division of Signal Transduction Therapy) at the University of Dundee, and affinity-purified against the indicated antigens: anti-Akt1 (S695B, third bleed; raised against residues 466–480 of human Akt1: RPHFPQFSYSASGTA), anti-PRAS40 (proline-rich Akt substrate 40 kDa) (S115B, first bleed; raised against residues 238–256 of human PRAS40: DLPRPRLNTSDFQKLKRKY), anti-(phospho-PRAS40 Thr246) (S114B, second bleed; raised against residues 240–251 of human PRAS40: CRPRLNTpSDFQK), anti-NDRG1 (N-Myc downstream-regulated gene-1) (S276B third bleed; raised against full-length human NDRG1) and anti-SGK3 [S037D second bleed; raised against human SGK3 PX domain comprising residues 1–130 of SGK3]. Anti-phospho-Akt Ser473 (9271), anti-phospho-Akt Thr308 (4056), anti-phospho-NDRG1 Thr346 (5482), anti-GAPDH (2118), anti-phospho-4E-BP1 (eukaryotic initiation factor 4E-binding protein 1) Thr37/46 (9459), anti-phospho-4E-BP1 S65 (9451), anti-4E-BP1 (9452) and anti-phospho-SGK3 Thr320 (5642) antibodies were purchased from Cell Signaling Technology (note we found that this antibody did not recognize T-loop of SGK1 or SGK2). Anti-(phospho-SGK hydrophobic motif [Ser486 in SGK3]) antibody (sc16745) was from Santa Cruz Biotechnology and total SGK1/2 antibody was from Sigma (5188) as was the mouse anti-FLAG antibody.
Secondary antibodies coupled to HRP (horseradish peroxidase) were obtained from Thermo Scientific.[](https://www.ncbi.nlm.nih.gov/mesh/D011392)
## In vitro Vps34 PI3K assay
*In the **In vitro Vps34 PI3K assay** section:*
The PI3K activity of recombinant Vps34–Vps15 complex expressed in insect cells was assayed in vitro via a radioactive liposome kinase assay. Liposomes were formed by extrusion of crude liver PtdIns (Avanti Polar Lipids, 840042) through a 100 nm filter membrane and incubated with 50 ng of recombinant Vps34/15 for 30 min at 1000 rev./min and 4°C. Assay buffer was 20 mM Tris/HCl, pH 7.4, 67 mM NaCl, 10 mM MnCl2, 0.02% CHAPS and 1 mM DTT in the presence of 5 μM ATP and 3 μCi [γ-32P]ATP per reaction. Reactions were terminated by the addition of 500 μl of chloroform/methanol/HCl (100:200:3.5, by vol.) and subsequently 180 μl of chloroform and 300 μl of 0.1 M HCl to phase split reactions.
The lower lipid-containing phase was retained and dried before the addition of chloroform. Reactions were spotted on to a Silica 60 TLC plate [activated with 1% potassium oxalate, 5 mM EDTA and 50% (v/v) methanol] and separated by a solvent comprising methanol/chloroform/water/ammonium bicarbonate (47:60:11.2:2, by vol.). Radioactive incorporation was analysed by Fujifilm Image reader FLA-2000 and quantified by AIDA image analysis, IC50 values were determined by non-linear regression using Prism software.[](https://www.ncbi.nlm.nih.gov/mesh/D010716)
## AstraZeneca lipid kinase profiling
*In the **AstraZeneca lipid kinase profiling** section:*
All compounds or DMSO for the PI4Kα, PI4Kβ, PIP5Kγ and PI3Kα biochemical, and IP1 (inositol phosphate) cell-based, assays were dispensed from source plates containing compounds at 10 mM in 100% (v/v) DMSO or 100% DMSO, directly into assay plates using an ECHO 555 Acoustic dispenser (Labcyte™).[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
PI4Kα and PI4Kβ were assayed using the ADP-Glo™ Kinase Assay Kit (Promega), in Greiner 384-well white low-volume plates in a 5 μl reaction volume consisting of 20 mM Bis-Tris propane (pH 7.5), 10 mM MgCl2, 0.5 mM EGTA, 1 mM DTT, 0.075 mM Triton X-100, 1.5% DMSO with or without an inhibitor at varying concentrations, D-myo-PtdIns substrate (Echelon Biosciences), ATP and purified enzyme. The PI4Kα assay was performed with 1 nM PI4KCA (Millipore), 72 μM ATP (KMappATP) and 25 μM PtdIns (KMappPI). The PI4Kβ assay was performed with 2 nM PI4KCB (SignalChem), 220 μM ATP (KMappATP) and 50 μM PtdIns (<KMappPI). The assay was allowed to proceed for 45 min at ambient temperature before stopping the reaction by the addition of 5 μl of ADP-Glo reagent. Plates were then covered and incubated for 40 min at ambient temperature.
Kinase detection reagent (10 μl) was then added and the plates were incubated for 30 min before the luminescence signal was read with a PHERAstar plate reader (BMG Labtech).[](https://www.ncbi.nlm.nih.gov/mesh/C034249)
PIP5Kγ assay was performed with the ADP-Glo™ Kinase Assay Kit (Promega) in Greiner 384-well white low-volume plates in a 5 μl reaction volume consisting of 20 mM Bis-Tris propane (pH 7.0), 10 mM MgCl2, 0.5 mM EGTA, 1 mM DTT, 0.024 mM Triton X-100, 1.5% DMSO with or without an inhibitor at varying concentrations, 14 μM D-myo-PIP (PtdIns 4-phosphate) substrate [<KMappPI(4)P] (Echelon Biosciences), 20 μM ATP (KMappATP) and PIP5K1C (expressed and purified by CRT, used at a 1:750 dilution). Assay incubations, additions and reads were performed as described for the PI4Kα and PI4Kβ ADP-Glo assays.[](https://www.ncbi.nlm.nih.gov/mesh/C034249)
The PI3Kα assay was performed using the Kinase-Glo® Plus Luminescence Assay Kit (Promega) in Greiner 384-well white low-volume plates. PIK3CA (expressed and purified by AZ; 20 nM) in phosphorylation buffer consisting of 50 mM Tris/HCl (pH 7.4), 0.05% CHAPSO {3-[(3-cholamidopropyl)dimethylammonio]-2-hydroxy-1-propanesul-fonic acid}, 10 mM MgCl2 and 2.1 mM DTT was pre-incubated with or without an inhibitor at varying concentrations for 20 min, before adding 80 μM PtdIns(4,5)P2 substrate (Cayman Chemicals) and 8 μM (<KMappATP) ATP, to give a final assay volume of 6 μl [2% (v/v) DMSO assay final]. The assay was allowed to proceed for 80 min at ambient temperature, before the assay was stopped with the addition of 4 μl of kinase Glo® Plus reagent. Plates were then covered and incubated for 20 min before the luminescence signal was read with a PHERAstar plate reader (BMG Labtech).[](https://www.ncbi.nlm.nih.gov/mesh/D006851)
The IP1 cell-based assay was performed using the HTRF IP-One Tb kit (CisBio) in Greiner 384-well tissue culture-treated white low-volume plates. NIH 3T3 cells, stably transfected using SV40 with the PDGFRβ receptor, were cultured in DMEM, 10% FBS, 1% Glutamax™ and 2 ng/ml puromycin. Cells were starved 24 h before use in DMEM, 1% charcoal-stripped FBS and 1% Glutamax™. After harvesting, cells were resuspended at a final concentration of 1.25×106 cells/ml in IP1 cell stimulation buffer, and 8 μl was dispensed into each well of the assay plate pre-dosed with compound or DMSO. After incubation for 30 min at 37°C, cells were then stimulated with 150 nM PDGF (Sigma–Aldrich) dispensed using an ECHO 555.
A standard curve IP1 calibration plate was prepared according to the manufacturer's instructions. After incubation for 30 min at 37°C, cells were lysed with the addition of 3 μl of lysis buffer with IP-One d2 (1:20 dilution), and 3 μl of lysis buffer with anti-IP-One Tb cryptate (1:20 dilution). Plates were covered and incubated for 2 h at ambient temperature, before a HTRF read at 615 nm and 665 nm on an EnVision plate reader (PerkinElmer). Assay values were then normalized to the IP1 calibration curve.[](https://www.ncbi.nlm.nih.gov/mesh/D007295)
## Immunoprecipitation and assay of SGK3 and SGK2
*In the **Immunoprecipitation and assay of SGK3 and SGK2** section:*
In vitro kinase activity of SGK3 and SGK2 was assayed by measuring [γ-32P]ATP incorporation into Crosstide substrate peptide [GRPRTSSFAEGKK]. SGK3 with a C-terminal FLAG-tag was immunoprecipitated from doxycycline-induced U2OS Flp/In cell lines. Immunoprecipitates were washed in sequence with lysis buffer containing high salt concentration (500 mM NaCl), lysis buffer and buffer A (50 mM Tris/HCl, pH 7.5, and 0.1 mM EGTA). Reactions were carried in 40 μl of total volume containing 0.1 mM [γ-32P]ATP (400–1000 c.p.m./pmol), 10 mM magnesium acetate and 30 μM Crosstide peptide.
Reactions were terminated by adding 10 μl of 0.1 mM EDTA and spotting 40 μl of the reaction mixture on P81 paper which were immediately immersed into 50 mM orthophosphoric acid. Papers were washed several times in 50 mM orthophosphoric acid, rinsed in acetone and air dried. Radioactivity was quantified by Cerenkov counting. One unit of enzyme activity was defined as the amount of enzyme that catalyses incorporation of 1 nmol of [γ-32P]ATP into the substrate over 1 min.[](https://www.ncbi.nlm.nih.gov/mesh/C000615311)
## Immunofluorescence and time-lapse microscopy
*In the **Immunofluorescence and time-lapse microscopy** section:*
Cells were cultured on glass coverslips and processed for immunocytochemistry using standard protocols. For experiments investigating the subcellular distribution of SGK3 C-terminally tagged GFP wild-type and mutant proteins and GFP–PHAkt cells were fixed with 4% paraformaldehyde for 15 min at room temperature and processed for imaging without permeabilization. For experiments quantifying co-localization of SGK3–GFP with EEA1 (early endosome antigen 1) and localization GFP–FYVEHrs in response to inhibitor treatment, cells were permeabilized via one freeze–thaw cycle in liquid nitrogen and subsequently fixed. The following primary antibodies were used: rabbit anti-EEA1 (1:500 dilution; Life Technologies) and mouse anti-GFP (1:1000 dilution; Life Technologies) antibodies. Fluorophore-conjugated secondary antibodies (Alexa Fluor® 594 and Alexa Fluor® 488) were obtained from Life Technologies.
Imaging was conducted using a Nikon Eclipse Ti-S microscope (×40 objective) or a Zeiss LSM710 laser scanning confocal microscope [×63 Plan-Apochromat oil immersion objective; 1.40 NA (numerical aperture)] as described in the Figure legends. Time-lapse microscopy was carried out on the same microscope using a heated incubator, heated stage plate and CO2 controller (Pecon) to maintain the cells.[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
Images were acquired every 0.5 min up to 60 min. Inhibitors were added after acquiring two images, except for VPS34-IN1 that was added after the first image taken. Post-acquisition, image analysis was conducted with Volocity 6.3 (PerkinElmer).[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## Lipid overlay assay
*In the **Lipid overlay assay** section:*
The assay was performed as described previously. Briefly, 500 pmol of each lipid [Ptdins, PtdIns(3)P, PtdIns(4)P, PtdIns(5)P, PtdIns(3,4)P2, PtdIns(4,5)P2, Ptdins(3,5)P2 and PtdIns(3,4,5)P3] were spotted on to Hybond-C membrane (GE Healthcare) and let dry for 1 h. The membrane was blocked in blocking buffer (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 0.1% Tween-20 and 2 mg/ml fatty acid-free BSA) for 1 h. The membrane was incubated overnight at 4°C in a 10 nM solution of each recombinant protein. After washing in TBST buffer (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 0.1% Tween 20), the signal was detected by using HRP-conjugated anti-FLAG or HRP-conjugated anti-GST antibodies (Roche) at a 1:3000 dilution.
All lipids were purchased from Echelon and dissolved in 1:1 solution of methanol and chloroform.[](https://www.ncbi.nlm.nih.gov/mesh/D010716)
## ITC (isothermal titration calorimetry)
*In the **ITC (isothermal titration calorimetry)** section:*
ITC experiments were carried out on a Microcal VP-ITC. A series of injections of peptide were made into an isolated chamber containing the protein at a constant temperature of 30°C. Heat changes within the cell were monitored during each injection of peptide and recorded as the total heat change per second over time. A binding isotherm was then fitted to data and expressed as the heat change per mol of peptide against the peptide/protein ratio. Injectant was PtdIns and PtdIns(3)P (Echelon) dissolved in buffer (50 mM Hepes, pH 7.4, and 150 mM NaCl).
The chamber contained recombinant 3×FLAG–SGK3-(1–162), 3×FLAG–SGK3-(1–162) R50A and 3×FLAG–SGK3 R90A in a buffer (50 mM Hepes, pH 7.4, and 150 mM NaCl). During each experiment, 40×6 μl doses of peptide were injected into the chamber of protein, which was stirred constantly at 300 rev./min. Each injection was followed by a 3 min period to ensure equilibration of the solution. All experiments were repeated and where possible using different concentrations of separately prepared protein. Data were analysed using Origin 7.0 with the Microcal software patch installed.
Each experimental condition had a blank run with protein in the chamber replaced with buffer. These data were then subtracted from the run with protein present to take into account any energy of dilution or metal/buffer reaction. A binding isotherm was then fitted to the data using a least squares calculation to yield a χ2 value. The model that produced the lowest χ2 value was taken as the best fit.[](https://www.ncbi.nlm.nih.gov/mesh/D010716)
## Statistical analysis
*In the **Statistical analysis** section:*
All experiments in the present study were performed at least twice and similar results were obtained. Data were analysed using one-way ANOVA followed by Bonferroni's post-hoc test (P<0.05) using Prism software. The error bars indicate S.D.
## RESULTS
*In the **RESULTS** section:*
## VPS34-IN1 is a potent and selective inhibitor of Vps34 class III PI3K
*In the **VPS34-IN1 is a potent and selective inhibitor of Vps34 class III PI3K** section:*
The structure of the bipyrimidinamine VPS34-IN1 compound is shown in Figure 1(A). This compound was originally reported in a patent (WO 2012085815 A1). We found that VPS34-IN1 inhibited phosphorylation of PtdIns by recombinant insect cell expressed Vps34–Vps15 complex with an IC50 of ~25 nM employing both an in-house assay (Figure 1B) or using an external assay undertaken by Life Technologies (Supplementary Figure S1).[](https://www.ncbi.nlm.nih.gov/mesh/D011743)
VPS34-IN1, a selective inhibitor of Vps34 kinase[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
(A) Chemical structure of the VPS34-IN1. (B) Insect cell expressed recombinant human Vps34–Vps15 complex was assayed by measuring phosphorylation of PtdIns in a 32P-radioactive kinase assay in the absence or presence of the indicated concentrations of VPS34-IN1. Reactions were chromatographed on a Silica 60 TLC plate and 32P-radioactivity associated with the spot comprising PtdIns(3)P was visualized (top panel) and quantified (bottom panel) by phosphoimager analysis on a Fujifilm Image reader FLA-2000 employing the AIDA image analysis software. Data are shown as the mean kinase activity±S.D. for three independent experiments, relative to DMSO-treated sample. The IC50 histogram was generated using Prism Software with non-linear regression analysis. (C) Protein kinase profiling of the VPS34-IN1 at a single concentration of 1 μM carried out against the Dundee panel of 140 protein kinases at the International Centre for Protein Kinase Profiling. Results for each kinase are presented as the mean kinase activity±S.D. for an assay undertaken in triplicate relative to a control kinase assay in which the inhibitor was omitted. Abbreviations and assay conditions used for each kinase are defined at http://www.kinase-screen.mrc.ac.uk.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
To evaluate the selectivity of VPS34-IN1, we studied the effect that this compound had on the activity of two protein kinase panels namely the Dundee panel (140 kinases) (Figure 1C, and Supplementary Table S1) and the ProQinase panel (300 kinases) (Supplementary Figure S2). These data revealed that VPS34-IN1 was remarkably selective and at a concentration of 1 μM, did not significantly inhibit the activity of any of the protein kinases assessed.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
We next evaluated VPS34-IN1 against three different lipid kinase profiling panels namely the Dundee panel (Figure 2A, 19 lipid kinases, includes class I PI3Ks), the AstraZeneca panel (Figure 2B, eight lipid kinases, includes class I) and ProQuinase panel (Figure 2C and Supplementary Figure S3, 13 lipid kinases, includes class I and class II PI3Ks). This revealed that VPS34-IN1 did not significantly inhibit any of the lipid kinases tested including class I (p110α, p110β p110γ and p110δ) and all three members of the class II PI3Ks (PI3KC2α, PI3KC2β and PI3KC2γ) enzymes.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
VPS34-IN1 selectively inhibit class III PI3K[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
(A) Lipid kinase profiling of the VPS34-IN1 at 1 μM was carried out against a panel of 19 lipid kinases at the International Centre for Protein Kinase Profiling. (B) Summary of the measurements of VPS34-IN1 for the indicated kinases and IP1 levels in cells (see the Materials and methods section for assay conditions). (C) Lipid kinase profiling of VPS34-IN1 inhibitor was carried out against ProQinase panel of 13 lipid kinases. Summary of IC50 measurements for each kinase is presented in the Table. Abbreviations used for each kinase and assay conditions used are defined at http://www.proqinase.com. Dose–response curves for data marked with an asterisk are presented in Supplementary Figure S3.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## VPS34-IN1 inhibits PtdIns(3)P levels at endosomes in a dose-dependent manner
*In the **VPS34-IN1 inhibits PtdIns(3)P levels at endosomes in a dose-dependent manner** section:*
The major pool of cellular PtdIns(3)P generated in cells by Vps34 accumulates on the surface of endosome membranes, which can be visualized using a GFP-tagged probe fused to a tandem repeat of the FYVE domain of the endocytic pathway Hrs protein. Consistent with previous work, we observed that GFP–2×FYVEHrs when stably expressed in U2OS osteosarcoma cells localized to discrete punctate cytoplasmic structures characteristic of endosomes (Figure 3). Treatment with Vps34-IN1 for 1 h induced a marked dose-dependent suppression of endosomal localization of GFP–2×FYVEHrs with ~80% loss observed at 1.0 μM and ~50% loss at 0.1 μM (Figure 3). We did not observe significant reduction in binding when employing PI3K class I inhibitor GDC-0941.[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
VPS34-IN1 reduces GFP–2×FYVEHrs probe localization on endosomes in a dose-dependent manner[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
U2OS cell line stably expressing GFP–2×FYVEHrs was treated with indicated concentration of VPS34-IN1 inhibitor for 1 h. The cells were permeabilized by freeze–thaw in liquid nitrogen and fixed in 4% paraformaldehyde. The GFP signal was enhanced by using mouse anti-GFP primary and anti-mouse Alexa Fluor® 488 secondary antibody. The top panel shows representative cell images of GFP–2×FYVEHrs localization and the bottom panel shows corresponding DAPI staining for each condition. The histogram displays average cell fluorescence±S.D. compared with DMSO-treated control. Similar result was obtained in at least one other experiment. The average cell fluorescence was calculated by dividing total fluorescence of each field by the number of the cells in the field. For each condition, ten random fields were chosen containing 20–25 cells/field. Images were taken using Nikon Eclipse Ti-S microscope using ×40 objective. Analysis was performed using NIS-Elements BR3.1 program. Scale bar, 20 μm. *P≤0.05, ***P≤0.001. ns, not significant.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## VPS34-IN1 rapidly reduces PtdIns(3)P levels at endosomes
*In the **VPS34-IN1 rapidly reduces PtdIns(3)P levels at endosomes** section:*
We also undertook a live imaging analysis monitoring the effect of 1 μM VPS34-IN1 on localization GFP–2×FYVEHrs with time. The data demonstrated that the effect of VPS34-IN1 in dispersing punctate endosomal localization of GFP–2×FYVEHrs was rapid with significant effects observed within 1 min (the earliest time-point that can be reliably analysed) and near maximal effects within 2 min (Figure 4A). Low levels of endosomal localization of GFP–2×FYVEHrs were maintained for up to 60 min, the longest time point we analysed (Figure 4). As expected treatment of U2OS cells with 0.5 μM GDC-0941, a class I PI3K inhibitor that does not inhibit Vps34, had no effect on endosomal localization of GFP–2×FYVEHrs even after 60 min treatment (Figure 4A). Combination of VPS34-IN1 and GDC-0941 dispersed endosomal localization of GFP–2×FYVEHrs similarly to VPS34-IN1 alone (Figure 4A).
It should be noted that low levels of residual PtdIns(3)P were still observed in cells treated with both VPS34-IN1 and GDC-0941, which could represent PtdIns(3)P generated through class II PI3K which was not inhibited by either inhibitor. We also found that 1 μM VPS-34-IN1 had no impact on the ability of IGF1 (insulin-like growth factor 1) to induce plasma membrane recruitment of a GFP probe encompassing the isolated PtdIns(3,4,5)P3/PtdIns(3,4)P2 binding PH domain of Akt1 (GFP–PHAkt1) (Figure 4B). In parallel experiments the class I PI3K GDC-0941 inhibitor blocked IGF1-induced translocation of the GFP–PHAkt1 probe to the plasma membrane (Figure 4B).[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
VPS34-IN1 rapidly reduces GFP–2×FYVEHrs probe localization on endosomes[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
(A) The U2OS cell line stably expressing GFP–2×FYVEHrs was recorded for approximately 1 min, before adding either no inhibitor (top panel), 1 μM VPS34-IN1 (second panel), 0.5 μM GDC-0941 (third panel) or a combination of 1 μM VPS34-IN1 and 0.5 μM GDC-0941 (bottom panel). Images were taken starting at 1.5 min from the time that the inhibitor was added (the first time point we could reliably measure) and subsequently at 0.5 min intervals up to a period of 1 h. Time-lapse microscopy was performed on Zeiss 710 microscope using ×63 objective. Similar results were obtained in at least two separate experiments. Scale bar, 20 μm. (B) As in (A) except the U2OS cell line stably expressing GFP–PHAkt1 was starved of serum overnight and treated with either no inhibitor (left panel), 1 μM VPS34-IN1 (middle panel) or 0.5 μM GDC-0941 (right panel) for 1 h before stimulation with IGF (100 ng/ml for 15 min). The cells were fixed with 4% paraformaldehyde and images were taken using Zeiss 710 microscope at ×63 objective. Representative images are shown and similar results were obtained in two experiments. Scale bar, 20 μm.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## PtdIns(3)P binding to PX domain localizes SGK3 to endosomes
*In the **PtdIns(3)P binding to PX domain localizes SGK3 to endosomes** section:*
We speculated that one downstream target of Vps34 might comprise the serum- and glucocorticoid-induced protein kinase SGK3, which is unique among kinases in that it possesses an N-terminal PX domain that binds specifically to PtdIns(3)P. Using isothermal calorimetry, we confirmed previous work that the isolated PX domain of SGK3 (residues 1–162, PXSGK3) interacted with PtdIns(3)P (Kd1 of 1.8 μM and Kd2 of 5.2 μM), but not with unphosphorylated PtdIns (results not shown). Molecular modelling of the non-complexed crystal structure of PXSGK3 suggested that the conserved Arg50 and Arg90 residues located within the PX domain might form ionic interactions with the 3′-phosphate of PtdIns(3)P. We therefore mutated Arg50 or Arg90 to alanine and found that these mutations abolished binding of PXSGK3 to PtdIns(3)P in either ITC (Figure 5A) or protein lipid overlay binding assays (Figure 5B). We also corroborated previous work, that PXSGK3 binds specifically to PtdIns(3)P and not to a range of other phosphoinositides tested [PtdIns, PtdIns(4)P, PtdIns(5)P, PtdIns(4,5)P2, PtdIns(3,4)P2 or PtdIns(3,4,5)P3] under conditions which Grp1 PH domain interacted with PtdIns(3,4,5)P3, the TAPP1 PH domain interacted with PtdIns(3,4)P2 and the phospholipase Cδ PH domain bound PtdIns(4,5)P2 (Figure 5B).[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
SGK3 binds to PtdIns(3)P via its PX domain[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
(A) ITC was performed by gradually titrating 0.5 mM PtdIns(3)P into the reaction chamber containing wild-type 3×FLAG–SGK3-(1–162) PX domain (left panel) or mutant 3×FLAG–SGK3-(1–162) R50A PX domain (middle panel) and 3×FLAG–SGK3-(1–162) R90A PX domain (right panel). The top panel illustrates enthalpic heat released during titration at 30°C and the bottom panel presents integrated binding isotherms and the best fit curves. Kd are indicated on the graphs. The measurement was performed on a VP-ITC MicroCalorimeter MicroCal machine. Kd were calculated using Origin 7.0 with ITC data analysis disc program. (B) The ability of the indicated GST fusion proteins to bind various phosphoinositides was analysed. The indicated phosphoinositides (500 pmol) were spotted on to nitrocellulose membranes, which were then incubated with 10 nM of the wild-type and indicated mutants of 3×FLAG–SGK3-(1–162) PX domain or GST-fusion PH domains [PLCδ-(1–178), GRP1-(241–399), TAPP1-(195–315) proteins]. The membranes were washed and the PX domain protein bound to the membrane by virtue of its interaction with lipid was detected using by using HRP-conjugated anti-FLAG antibodies (SGK3 PX domain) or anti-GST (PH domains).[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
## PtdIns(3)P binding localizes SGK3 to early endosomes
*In the **PtdIns(3)P binding localizes SGK3 to early endosomes** section:*
We next analysed the cellular location in U2OS cells of full-length wild-type SGK3 possessing a C-terminal GFP tag. This revealed clear-cut punctate endosomal localization of wild-type SGK3 that overlapped with localization of the early endosomal EEA1 marker (Figure 6A). In contrast, the non-PtdIns(3)P-binding SGK3(R50A) or SGK3(R90A) mutants were diffusely dispersed throughout the cytosol (Figure 6A). Using time-lapse microscopy we also observed that treatment of cells with 1 μM VPS34-IN1 resulted in the rapid dispersal of SGK3 from the endosomes within 1 min. In contrast, treatment with the class I PI3K inhibitor GDC-0941 (0.5 μM) did not change significantly the SGK3 localization on endosomes (Figure 6B).
Dual treatment with VPS34-IN1 and GDC-0941 inhibitors revealed that the effect of inhibitors on PtdIns(3)P production was as rapid as with VPS34-IN1 alone resulting in reduced SGK3 punctate appearance.[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
PtdIns(3)P binding localizes SGK3 to endosomes[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
(A) U2OS stably expressing wild-type or indicated mutant of full-length SGK3 with a C-terminal GFP-tagged were fixed with 4% (v/v) paraformaldehyde and GFP distribution in cell was visualized. Images were taken using a Nicon Eclipse Ti-S microscope with ×40 objective (upper panel). SGK3 co-localization with early endosomal marker EEA1 marker was visualized using rabbit anti-EEA1 primary and anti-rabbit Alexa Fluor® 594 secondary antibody (lower panel). Pictures were taken with Zeiss 710 microscope using ×100 objective. Mander's correlation coefficient in 112 cells was calculated using the Volocity 6.3 program. Scale bar, 20 μm. (B) The U2OS cell line stably expressing full-length SGK3 with an C-terminal GFP-tagged were recorded for approximately 1 min, before adding either no inhibitor (top panel), 1 μM VPS34-IN1 (second panel), 0.5 μM GDC-0941 (third panel) or a combination of 1 μM VPS34-IN1 and 0.5 μM GDC-0941 (bottom panel). Images were taken starting at 1.5 min from the time that the inhibitor was added (the first time point we could reliably measure) and subsequently at 0.5 min intervals up to a period of 1 h. Time-lapse microscopy was performed on Zeiss 710 microscope using ×63 objective. Representative images are shown and similar results were obtained in two experiments. Scale bar, 20 μm.[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
## PtdIns(3)P binding controls SGK3 activity and hydrophobic motif phosphorylation
*In the **PtdIns(3)P binding controls SGK3 activity and hydrophobic motif phosphorylation** section:*
SGK3, similar to many other AGC kinases including Akt isoforms, is activated following phosphorylation of its catalytic domain T-loop motif (Thr320) by PDK1 and its non-catalytic C-terminal hydrophobic motif (Ser486) by mTOR. Previous work with SGK1 has shown that mTORC2-mediated phosphorylation of the hydrophobic motif markedly enhances phosphorylation of the T-loop residue by creating a binding motif for PDK1. Consistent with previous work, we observed that full-length wild-type SGK3 when stably expressed in U2OS cells displayed considerable kinase activity towards the Crosstide peptide (assessed after immunoprecipitate kinase assay) and was also significantly phosphorylated at its T-loop and hydrophobic motifs (assessed employing phosphospecific antibodies recognizing these sites) (Figure 7). Strikingly, the non-PtdIns(3)P-binding SGK3(R50A) or SGK3(R90A) mutants displayed only low kinase activity and T-loop and hydrophobic motif phosphorylation, suggesting PtdIns(3)P binding to the PX domain plays a critical role in enabling SGK3 to become phosphorylated at its T-loop and hydrophobic motif and therefore become activated (Figure 7). As expected, treatment of cells with a selective PDK1 inhibitor (GSK2334470) suppressed SGK3 activity, as well as T-loop and hydrophobic motif phosphorylation, to near basal levels, similar to those observed following mutation of the PX domain (Figure 7).[](https://www.ncbi.nlm.nih.gov/mesh/D013912)
SGK3 phosphorylation and kinase activity is controlled by its ability to bind PtdIns(3)P[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
(A) U2OS stably expressing wild-type or indicated mutant of full-length SGK3 with a C-terminal FLAG-tag were treated in the absence or presence of the indicated PDK1 inhibitor (1 μM GSK2334470) for 1 h. The cells were lysed and SGK3 immunoprecipitated with anti-FLAG antibody. The immunoprecipitates were subjected to immunoblot analysis with the indicated antibodies (lower panel) after being assayed for SGK3 kinase activity by measuring phosphorylation of the Crosstide substrate peptide in the presence of 0.1 mM [γ-32P]ATP in a 30 min reaction (upper panel). Kinase reactions are presented as means±S.D. for triplicate reaction. The similar result was obtained in at two separate experiments. ****P≤0.0001.[](https://www.ncbi.nlm.nih.gov/mesh/C555257)
## Class I and class III PI3K inhibitors rapidly suppress SGK3 activity and T-loop/hydrophobic motif phosphorylation
*In the **Class I and class III PI3K inhibitors rapidly suppress SGK3 activity and T-loop/hydrophobic motif phosphorylation** section:*
We next treated U2OS cells stably expressing SGK3 with increasing doses of VPS34-IN1 for 1 h and tested how this affected SGK3 activity and T-loop phosphorylation. This revealed that VPS34-IN1 induced a dose-dependent reduction in SGK3 activity that was maximally lowered by ~60% at 1 μM and ~40% at 0.1 μM (Figure 8A). The suppression of SGK3 activity induced by VPS34-IN1 treatment was accompanied by a commensurate decrease in T-loop and hydrophobic motif phosphorylation (Figure 8A).[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
Evidence that SGK3 phosphorylation and kinase activity is controlled by Vps34 and class I PI3K
(A) U2OS stably expressing wild-type or indicated mutant of full-length SGK3 with a C-terminal FLAG-tag were treated in the absence or presence of the indicated concentrations of VPS34-IN1 inhibitor or class I PI3K inhibitor (0.5 μM GDC-0941) PDK1 inhibitor (1 μM GSK2334470), mTOR inhibitor (0.1 μM AZD8055) for 1 h. The cells were lysed and SGK3 immunoprecipitated with anti-FLAG antibody. The immunoprecipitates were subjected to immunoblot analysis with the indicated antibodies (lower panel) after being assayed for SGK3 kinase activity by measuring phosphorylation of the Crosstide substrate peptide in the presence of 0.1 mM [γ-32P]ATP in a 30 min reaction (upper panel). Kinase reactions are presented as means±S.D. for triplicate reaction. (B and C) As in (A) except cells were treated with the indicated doses of the BKM120 class I PI3K inhibitor. Similar results were obtained in at least two separate experiments for all data shown in this Figure. ***P≤0.001, **P≤0.01 and *P≤0.05. ns, not significant.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
We also observed that addition of the selective class I PI3K inhibitor GDC-0941 induced a ~40% inhibition of SGK3 activity accompanied by a similar reduction in T-loop and hydrophobic motif phosphorylation (Figure 8A). Moreover, a structurally distinct selective class I PI3K inhibitor that is in clinical trials termed BKM120 induced a comparable ~40% inhibition of SGK3 (Figure 8B). Consistent with GDC-0941 (Figure 8A) and BKM120 (Figure 8B) inhibiting a class I PtdIns kinase, they suppressed the Akt T-loop (Thr308) and hydrophobic motif (Ser473) phosphorylation, as well as Akt substrate PRAS40 phosphorylation (Thr246), to near basal levels. At the doses used neither GDC-0941 nor BKM120 was judged to significantly suppress mTOR activity based on the lack of effect that these compounds had on the phosphorylation of the 4EBP1 at sites phosphorylated by mTORC1 (Thr37, Thr46 and Ser65) (Figure 8B).
Treatment of cells with the AZD8055 mTOR inhibitor in parallel experiments led to marked dephosphorylation of 4E-BP1 at residues phosphorylated by mTOR (Figure 8B).[](https://www.ncbi.nlm.nih.gov/mesh/C532162)
Combining 1.0 μM VPS34-IN1 and 0.5 μM GDC-0941 inhibitors reduced SGK3 activity and phosphorylation by over 80% to a similar extent as observed with treatment with AZD8055 mTOR inhibitor (Figure 8A).[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
A time course analysis revealed that the effect of 1 μM VPS34-IN1 on SGK3 activity and phosphorylation was extremely rapid and activity reduced by 50% within 15 s and maximally suppressed within 1 min (Figure 9A). In contrast, the effect of GDC-0941 on SGK3 activity and phosphorylation was slower with little effect observed at a 1 min time point, but activity declining significantly thereafter with near maximal inhibition observed after 2 min of GDC-0941 treatment (Figure 9B).[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
Class I and class III PI3K inhibitors rapidly inactivate SGK3 and evidence that SGK2 is not regulated by Vps34
(A and B) U2OS stably expressing wild-type full-length SGK3 with a C-terminal FLAG-tag were treated in the absence or presence of the 1 μM VPS34-IN1 (A) or 0.5 μM GDC-0941 (B) for the indicated times. The cells were lysed and SGK3 immunoprecipitated with anti-FLAG antibody. The immunoprecipitates were subjected to immunoblot analysis with the indicated antibodies (top panel) after being assayed for SGK3 kinase activity by measuring phosphorylation of the Crosstide substrate peptide in the presence of 0.1 mM [γ-32P]ATP in a 30 min reaction (bottom panel). Kinase reactions are presented as means±S.D. for triplicate reaction. (C) As above except U2OS stably expressing wild-type full-length SGK2 with a C-terminal FLAG-tag were treated in the absence or presence of the indicated inhibitors for 1 h. Similar results was obtained in at least two separate experiments for all data shown in this Figure. ***P≤0.001, **P≤0.01 and *P≤0.05. ns, not significant.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
Consistent with VPS34-IN1 not supressing class I PI3Ks, we observed that the inhibitor had no effect on Akt phosphorylation at Ser473 and Thr308 or phosphorylation of the Akt substrate PRAS40 (Figure 8A). We also monitored NDRG1 phosphorylation at Thr346 a site that can be phosphorylated by SGK or Akt isoforms. In U2OS cells, as observed in some other cell types, it seems that phosphorylation of NDRG1 is mediated mainly through Akt as its phosphorylation is significantly suppressed by GDC-0941, but only affected moderately by VPS34-IN1 (Figure 8). Moreover, in agreement with the notion that the ability of SGK3 to be inhibited by VPS34-IN1 is dependent on its PtdIns(3)P-binding PX domain, we observed that VPS34-IN1 did not inhibit the activity of SGK2 isoform that does not possess a PX domain (Figure 9C).[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## DISCUSSION
*In the **DISCUSSION** section:*
VPS34-IN1 is the first highly selective cell permeable Vps34 inhibitor to be reported. Previously utilized Vps34 inhibitors have included wortmannin, LY294002, 3-methyladenine, KU55933 and Gö6976 that are not selective inhibitors. Wortmannin and LY294002 when used at concentrations that inhibit Vps34 also suppress class I PI3Ks and several kinases that belong to the PI3K-related kinase (mTOR and DNA-dependent protein kinase) in addition to many other protein and lipid kinases. KU55933 potently inhibits the DNA damage repair ATM protein kinase and to our knowledge has not been thoroughly profiled against a panel of lipid kinases including class I and class II PI3Ks so the off-target effects of this compound are unknown. The protein kinase C inhibitor Gö6976 is poorly selective inhibiting many protein kinases (http://www.kinase-screen.mrc.ac.uk/screening-compounds/341062).
To our knowledge extensive protein and lipid kinase profiling has not been undertaken for 3-methyladenine, but in our judgment this simple adenine derivative is unlikely to comprise a selective Vps34 inhibitor. In contrast, the selectivity data for VPS34-IN1 indicates that it is a remarkably selective Vps34 inhibitor. We have profiled VPS34-IN1 against 340 unique protein kinases (Figure 1 and Supplementary Figure S2) and 25 lipid kinases that include all four members of the class I PI3K and three members of the class II PI3Ks (Figure 2, and Supplementary Figures S1 and S3). We have found that, other than Vps34, no kinase is markedly inhibited at a concentration of 1 μM, which is 40-fold higher than the IC50 value that VPS34-IN1 inhibits Vps34 and the highest concentration of compound we have used in cells. It is reassuring that VPS34-IN1 does not inhibit any of the three members of the class II PI3Ks, as these poorly understood enzymes, which are similar to Vps34, also generate PtdIns(3)P. This provides confidence that the impact that VPS34-IN1 has on endosomal PtdIns(3)P levels and SGK3 activity in vivo is indeed due to inhibition of PtdIns(3)P generated through the action of Vps34.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
The evidence that VPS34-IN1 suppresses Vps34 activity in vivo is supported by the observation that treatment of cells with VPS34-IN1 induces a strikingly rapid dose-dependent dispersal of the localization of the PtdIns(3)P-binding GFP–2×FYVEHrs from endosomal membranes (Figures 3, 4 and 6), where PtdIns(3)P is produced by Vps34 resides. Consistent with the in vitro IC50 of VPS34-IN1 for Vps34 being 25 nM (Figure 1B and Supplementary Figure S1), treatment of cells with 0.01 μM VPS34-IN1 induced a weak effect on GFP–2×FYVEHrs localization, whereas 0.1 μM and 1 μM VPS34-IN1 induced ~50% and ~80% respectively dispersal of GFP–2×FYVEHrs endosomal localization (Figure 7). The effect of VPS34-IN1 in dispersing the endosome-localized GFP–2×FYVEHrs (Figure 4) and GFP–SGK3 (Figure 6) is rapid and essentially complete within 60–120 s, the earliest time point we can reliably assess in live cell imaging experiments. In future work it would be interesting to explore whether there is an interplay between class II and class III PI3Ks and whether inhibition of Vps34 results in the activation of a class II PI3K-driven compensatory mechanism to produce PtdIns(3)P. Moreover, we observed a marked reduction in SGK3 activity and phosphorylation within 15 s treatment with VPS34-IN1 which was nearly maximal by 30 s. These observations are consistent with previous work showing that the localization of GFP–2×FYVEHrs at the endosome is highly dynamic and responsive to inhibition of Vps34.
It has been argued that the relatively low affinity of PX and FYVE domains for PtdIns(3)P coupled with high myotubularin PtdIns 3-phosphatase activity accounts for rapid and highly dynamic localization and response. Our results highlight that full-length SGK3 is similarly dynamic and responsive to changes PtdIns(3)P as the well-characterized GFP–2×FYVEHrs probe. The finding that VPS34-IN1 does not inhibit the localization of the PtdIns(3,4,5)P3/PtdIns(3,4)P2 binding GFP–PHAkt1 to the plasma membrane or inhibit Akt phosphorylation at Thr308 or Ser473 establishes that VPS34-IN1 is not suppressing the activity of class I PI3Ks in vivo, consistent with the in vitro lipid kinase profiling data. The finding that VPS34-IN1 does not suppress activity of SGK2 that does not possess a PtdIns(3)P-binding PX domain (Figure 9C) is also consistent with the notion that the ability of VPS34-IN1 to suppress SGK3 is dependent on its ability to bind PtdIns(3)P.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
SGK3 was first identified as a novel isoform of SGK1 that played a role in enabling the IL (interleukin)-3-dependent survival of haemopoietic cells. Subsequent studies revealed that SGK3 possesses a PtdIns(3)P-binding N-terminal PX domain. To our knowledge SGK3 is unique among kinases in possessing a PtdIns(3)P-binding domain. Consistent with previous work, we show that endosomal localization, as well as SGK3 activity and phosphorylation of regulatory T-loop and hydrophobic motif, is markedly suppressed following mutation of critical PtdIns(3)P-binding PX domain residues. Moreover, inhibiting SGK3 binding to the endosomes by treating cells with VPS34-IN1 induced a rapid 50–60% decrease in SGK3 activity accompanied by a commensurate decrease in the T-loop and hydrophobic motif phosphorylation.
In future work it would be important to explore the mechanism by which PtdIns(3)P binding and endosomal localization controls SGK3 activity by regulating phosphorylation of its T-loop by PDK1 and hydrophobic motif by mTOR. It would be interesting to explore whether PtdIns(3)P binding induces a conformational change in SGK3 promoting phosphorylation by PDK1 and/or mTOR. An analogous mechanism operates for Akt isoforms, in which interaction of the PH domains of these kinases with PtdIns(3,4,5)P3/PtdIns(3,4)P2 at the plasma membrane induces a conformational change that markedly enhances phosphorylation and activation by PDK1. It is also possible that PtdIns(3)P binding serves to anchor SGK3 to the endosome, as well as other membranes where a pool of PDK1 and mTOR reside, and this increases the efficiency by which these enzymes are brought together on a two-dimensional membrane surface.[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
Despite mutations inhibiting PtdIns(3)P binding suppressing SGK3 activity by over 90%, and VPS34-IN1 treatment dispersing the vast majority of GFP–2×FYVEHrs and GFP–SGK3 from the endosome, we find that VPS34-IN1 only partially reduced SGK3 activity by 50–60%. Consistent with this, a previous study has found that knockout of Vps34 in mouse embryonic fibroblasts resulted in ~60% loss of PtdIns(3)P levels, which correlates with the inhibition of SGK3 activity observed in U2OS cells treated with VPS34-IN1. The activity of SGK3 that remains following VPS34-IN1 treatment is likely to be controlled by additional pool(s) of PtdIns(3)P generated independently of Vps34 that are able to recruit SGK3 and hence promote activation via mTOR and PDK1. Previous work has revealed that a pool of PtdIns(3)P can be generated in vivo through the dephosphorylation of PtdIns(3,4,5)P3 via sequential actions of the PtdIns 5-phosphatases (SHIP1/2) and the PtdIns 4-phosphatase (INPP4B). PtdIns(3)P produced through this route is then rapidly dephosphorylated to PtdIns by a family of myotubularin PtdIns 3-phosphatases.
Our finding that treatment of cells with structurally diverse class I selective PI3K inhibitors (GDC-0941 and BKM120) that do not inhibit Vps34 induce a ~40% reduction in SGK3 activity at doses that suppress Akt activity significantly (Figure 8) suggests that PtdIns(3)P generated through dephosphorylation of PtdIns(3,4,5)P3 could be controlling SGK3. However, we cannot rule out that the effect of GDC-0941 and BKM120 on SGK3 activity in these studies is not at least partially mediated through the ability of these inhibitors to suppress the activity of mTORC2, which is potentially also regulated by class I PI3K (Figure 10).[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
Model of how SGK3 is regulated by two distinct pools of PtdIns(3)P[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
Our data suggest that SGK3 is controlled by two pools of PtdIns(3)P. One pool is produced via phosphorylation of PtdIns by class III PI3K termed Vps34 at the endosome and the other pool of PtdIns(3)P is generated as a result of dephosphorylation of PtdIns(3,4,5)P3 to PtdIns(3)P through the sequential actions of SHIP1/2 and INPP4B PtdIns phosphatases. Our findings suggest that binding of PtdIns(3)P to the PX domain triggers SGK3 activation by promoting phosphorylation of the T-loop by PDK1 and the hydrophobic motif by mTOR. Further work is required to understand how binding of PtdIns(3)P promotes phosphorylation of SGK3 by PDK1 and mTOR.[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
It is also interesting to note that the impact of GDC-0941 on SGK3 activity is slower than that observed for VPS34-IN1 (compare Figure 9A with 9B), which may be consistent with the several steps between activation of class I PI3K and PtdIns(3)P. The observation that combining VPS34-IN1 and GDC-0941 reduces SGK3 activity by 80–90% comparable with the effect of mTOR inhibitors is consistent with the notion that the class I and class III PI3K controlled pools of PtdIns(3)P are the key in controlling SGK3 (Figure 10).[](https://www.ncbi.nlm.nih.gov/mesh/C532162)
In future work it would also be interesting to explore whether residual SGK3 kinase activity (Figure 8A) and the presence of low levels of GFP–2×FYVEHrs probe and GFP–SGK3 still present on endosomes after dual treatment with class I and class III PI3K inhibitors are due to pools of PtdIns(3)P generated by the poorly characterized class II PI3Ks. Previous work has suggested that class II PI3Ks play a role in intracellular vesicular trafficking as the kinases were found to reside on endosomes, clathrin-coated vesicles and trans-Golgi network.[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
Further work is required to pinpoint the location of the class I PI3K-controlled PtdIns(3)P pool in cells. We do not see a marked localization of GFP–2×FYVEHrs or GFP–SGK3 at the plasma membrane where the class I PI3K controlled the pool of PtdIns(3)P might be expected to reside. However, PtdIns(3)P at the plasma membrane may be much more dispersed than at the endosome membrane, where PtdIns(3)P levels are very high, thereby making GFP–2×FYVEHrs or GFP–SGK3 localization at the plasma membrane harder to visualize.[](https://www.ncbi.nlm.nih.gov/mesh/C055525)
There is increasing interest in understanding the physiological roles that Vps34 as well as SGK3 play. Our data indicate that VPS34-IN1 will be a useful tool to probe the cellular functions of Vps34 and that monitoring SGK3 activity and/or phosphorylation could serve as a biomarker for Vps34 activity in vivo. There is also increasing evidence that a significant number of human tumours that are insensitive to class I PI3K inhibition display elevated levels of SGK3 and that this plays an important role in driving survival and proliferation. In future work it will be important to learn more about the roles that SGK3 plays at the endosome and potentially other membranes. It will be important to identify physiological substrates for SGK3 and explore whether inhibiting Vps34 or SGK3 could be deployed as a therapeutic strategy to treat tumour cells displaying elevated SGK3 activity.
Finally, our data indicate that combining GDC-0941 and VPS34-IN1 might be used as a strategy to inhibit pools of PtdIns(3)P generated by class I and III PI3Ks that would better allow the study of the roles and regulation of the elusive class II PI3Ks to be visualized.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## Online data
*In the **Online data** section:*
## AUTHOR CONTRIBUTION
*In the **AUTHOR CONTRIBUTION** section:*
Ruzica Bago carried out the experiments in Figures 3, 4, 5, 6, 7 and 8. Nazma Malik carried out the experiments in Figure 9. Michael Munson performed the experiment in Figure 1(B) and he also made the U20S cell line stably expressing GFP–FYVEHrs probe. Data in Figures 1(C) and 2(A) were obtained by staff in the Dundee International Kinase profiling centre, Figure 2(B) was provided by staff in AstraZeneca and Figures 2(C), and Supplementary Figures S2 and S3 were from staff in ProQuinase. Alan Prescott took and analysed images for the experiment in Figure 6(B) and helped with live-imaging experiments.
Paul Davies advised and helped analyse the ITC experiments. Natalia Shprio synthesized VPS34-IN1. Ian Ganley, Richard Ward and Darren Cross provided critical advice. Eeva Sommer undertook some of the initial studies in this project. Ruzica Bago and Dario Alessi planned the experiments, analysed the results and wrote the paper.[](https://www.ncbi.nlm.nih.gov/mesh/C000593572)
## FUNDING
*In the **FUNDING** section:*
This work was supported by the , the for microscopy [grant number ] and flow cytometry [grant number ] and the pharmaceutical companies supporting the (AstraZeneca, Boehringer-Ingelheim, GlaxoSmithKline, Merck, Janssen Pharmaceutica and Pfizer).
|
# Introduction
Induction, Purification and Characterization of a Novel Manganese Peroxidase from Irpex lacteus CD2 and Its Application in the Decolorization of Different Types of Dye
# Abstract
*In the **Abstract** section:*
Manganese peroxidase (MnP) is the one of the important ligninolytic enzymes produced by lignin-degrading fungi which has the great application value in the field of environmental biotechnology. Searching for new MnP with stronger tolerance to metal ions an[d orga](https://www.ncbi.nlm.nih.gov/mesh/D008031)nic solvents is important for the maximization of potential of MnP in the biodegradation of recalcitrant xenobiotics. In this study, it was found tha[t oxa](https://www.ncbi.nlm.nih.gov/mesh/D008670)lic acid, veratryl alcohol and 2,6-Dimehoxyphenol could stimulate the synthesis of MnP in the white-rot fungus Irpex lacteus CD2. A novel manganese peroxidase named [as CD2-MnP ](https://www.ncbi.nlm.nih.gov/mesh/D019815)wa[s purified and c](https://www.ncbi.nlm.nih.gov/mesh/C042197)harac[terized from this ](https://www.ncbi.nlm.nih.gov/mesh/C010120)fungus. CD2-MnP had a strong capability for tolerating different metal ions such as Ca2+, Cd2+, Co2+, Mg2+, Ni2+ and Zn2+ as well as organic solvents such as methanol, ethanol, DMSO, ethylene glycol, isopropyl alcohol, butanediol and[ glyc](https://www.ncbi.nlm.nih.gov/mesh/D008670)erin. The diff[eren](https://www.ncbi.nlm.nih.gov/mesh/D002118)t [type](https://www.ncbi.nlm.nih.gov/mesh/D002104)s [of d](https://www.ncbi.nlm.nih.gov/mesh/D003035)ye[s in](https://www.ncbi.nlm.nih.gov/mesh/D008274)cl[udin](https://www.ncbi.nlm.nih.gov/mesh/D009532)g the[ azo](https://www.ncbi.nlm.nih.gov/mesh/D015032) dye (Remazol Brilliant Violet 5R, Di[rect Red](https://www.ncbi.nlm.nih.gov/mesh/D000432) 5[B), ant](https://www.ncbi.nlm.nih.gov/mesh/D000431)hr[aqui](https://www.ncbi.nlm.nih.gov/mesh/D004121)no[ne dye (Remazol](https://www.ncbi.nlm.nih.gov/mesh/D019855) B[rilliant Blue R),](https://www.ncbi.nlm.nih.gov/mesh/D019840) i[ndigo dye ](https://www.ncbi.nlm.nih.gov/mesh/D002072)(Indi[go Carmi](https://www.ncbi.nlm.nih.gov/mesh/D005990)ne) and triphenylmethane dye (Methyl Green) [as well](https://www.ncbi.nlm.nih.gov/mesh/D001391) a[s simulated textile wastewa](https://www.ncbi.nlm.nih.gov/mesh/C111247)te[r could be ef](https://www.ncbi.nlm.nih.gov/mesh/C542892)fic[iently decolo](https://www.ncbi.nlm.nih.gov/mesh/D000880)rized [by CD2-MnP. CD2-MnP also](https://www.ncbi.nlm.nih.gov/mesh/C024758) ha[d a st](https://www.ncbi.nlm.nih.gov/mesh/D007203)rong a[bility of deco](https://www.ncbi.nlm.nih.gov/mesh/D007203)lorizi[ng different dye](https://www.ncbi.nlm.nih.gov/mesh/C046945)s with[ the coexist](https://www.ncbi.nlm.nih.gov/mesh/D008739)ence of metal ions and organic solvents. In summary, CD2-MnP from Irpex lacteus CD2 could effectively degrade a broad range of synthetic dyes and exhibit a great potential for [envir](https://www.ncbi.nlm.nih.gov/mesh/D008670)onmental biotechnology.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Manganese peroxidase (MnP, EC 1.11.1.13) is the heme-containing glycoprotein which is mainly produced by white-rot fungi such as Phanerochaete chrysoporium, Ceriporiopsis subvervispora, Dichomitus squalens, Pleurotus ostreatus, Pleurotus pulmonarius, Pleurotus eryngii. The MnP, which is the important component of extracellular ligninolytic enzymes of lignin-degrading fungi, can catalyze the H2O2-dependent oxidation of Mn2+ into Mn3+, and then chelates of Mn3+ with fungal organic acid cause one-electron oxidation of various compounds (A schematic representation of the enzyme reaction was shown in Fig.S1). MnP has the strong ability of oxidizing and depolymerizing natural and synthetic lignins –. Besides the use in the conversion of lignin and lignocelluloses , MnP has great application potential in the field of environmental biotechnology and degradation of some recalcitrant organopollutants that are very harmful to human health, such as polycyclic aromatic hydrocarbons , , chlorophenols , industrial dyes – and nitroaromatic compounds . The great value of MnP in the application in bioremediation results in more and more attention to this enzyme.[](https://www.ncbi.nlm.nih.gov/mesh/D008031)
The unique degradative ability of MnP makes this enzyme valuable for various biotechnological applications. Thus, in recent years, some MnPs have been purified and characterized from different fungal strains such as Agrocybe praecox , Dichomitus squalens , Irpex lacteus , , Stereum ostrea , Phanerochaete chrysosporium , Lentinula edodes , Schizophyllum . The enzymatic properties of these purified MnPs from different sources have been studied. Previous research has demonstrated that some azo and anthraquinone dyes, polycyclic aromatic hydrocarbons (phenanthrene, anthracene, fluoranthene, and pyrene), 2,4,6-trinitrotoluene can be efficiently degraded by the purified MnPs from Dichomitus squalens , Stereum ostrea , Irpex lacteus and Phlebia radiate . The ability of nanoclay-immobilized MnP from Anthracophyllum discolor to degrade polycyclic aromatic hydrocarbons and the capability of sol–gel matrix immobilized MnP from Ganoderma lucidum for decolorization of different dye effluents have also been evaluated .[](https://www.ncbi.nlm.nih.gov/mesh/D000880)
Although there have been some reports about the properties of purified MnPs and their application in the enzymatic degradation of environmental pollutants as described above, some other factors have to be considered for the more efficient application of MnP in the area of biodegradation. For example, the dye effluents discharged by textile industry usually contain high level of different metal ions and organic solvents. Thus, the ability of MnP to tolerate different metal ions or organic solvents is very important for the efficient application of this enzyme in the treatment of wastewater. However, to our knowledge, few studies have been performed to evaluate the capability of purified MnP for tolerating different metal ions and organic solvents. Most previous research mainly focused on the enzymatic and kinetic properties of MnP purified from different fungi , , , .
Therefore, searching for new MnP with stronger tolerance to metal ions and organic solvents is important for the maximization of potential of MnP in the biodegradation of recalcitrant xenobiotics.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
The white-rot fungi Irpex lacteus has been shown to demonstrate a significant potential for the various biotechnological applications such as bioremediation of organopollutants in water and soil environments, degradation of different lignocellulosic substrates yielding higher sugar recoveries compared to other fungal treatments. The great application values of Irpex lacteus are attributed to the extracellular peroxidase including manganese peroxidase, versatile peroxidase and dye-decolorizing peroxidase , . In the previous research of our laboratory, a new white-rot fungi strain Irpex lacteus CD2 has been isolated and characterized from the Shennongjia Nature Reserve of Hubei Province in China –. The effect and mechanism of biopretreatment of cornstalks by Irpex lacteus CD2 have been intensively studied in our laboratory –. For the purpose of better use of this fungus and its manganese peroxidase in the field of environmental biotechnology, in this work, the properties of the purified manganese peroxidase (named as CD2-MnP) from Irpex lacteus CD2 and its ability to decolorize different types of dyes and simulated textile wastewater were investigated.
We also focused on the evaluation of the capability of this MnP for tolerating different metal ions and organic solvents. In addition, the capability of CD2-MnP to decolorize different dyes with the coexistence of metal ions and organic solvents was further assessed.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Materials and Methods
*In the **Materials and Methods** section:*
## Dyes and Chemicals
*In the **Dyes and Chemicals** section:*
The different types of dyes used in this study were purchased from Aldrich-Sigma (USA). All of other chemicals were of analytical grade and obtained from Sinopharm Chemical Reagent Company (China).
## Strains and culture conditions
*In the **Strains and culture conditions** section:*
The white rot fungus Irpex lacteus CD2 was characterized in the previous work of our laboratory –. It was maintained at 4°C on potato dextrose agar (PDA) slant. The inoculum was grown in potato dextrose broth (PDB) medium for 7 days at 28°C, then cultures were transferred into the basal liquid medium as a 10% (v/v) inoculum and incubated at 28°C in a shaking incubator (150 rpm). The basal liquid medium contained (g/L): Glucose 20 g, Yeast extract 2.5 g, KH2PO4 1 g, Na2HPO4 0.05 g, MgSO4·7H2O 0.5 g, CaCl2 0.01 g, FeSO4·7H2O 0.01 g, MnSO4·4H2O 0.001 g, ZnSO4·7H2O 0.001 g, CuSO4·5H2O 0.002 g .[](https://www.ncbi.nlm.nih.gov/mesh/D000362)
## Measurement of MnP activity and protein contents
*In the **Measurement of MnP activity and protein contents** section:*
Manganese peroxidase activity was measured by monitoring the formation of Mn3+-malonate complexes at 270 nm as described previously . The assay mixture contained 1 ml of 4 mM MnSO4, 1 mL of 20 mM malonate buffer (pH 5.0), 0.5 mL of 0.4 mM H2O2 and 0.1 mL of enzyme solution. One unit of enzyme activity was defined as the amount of enzyme that oxidized 1 µmol of Mn2+ per min at 30°C. Protein contents were determined by the method of Bradford using BSA as the standard.[](https://www.ncbi.nlm.nih.gov/mesh/D008314)
## Induction of manganese peroxidase produced by Irpex lacteus CD2
*In the **Induction of manganese peroxidase produced by Irpex lacteus CD2** section:*
The fungus was grown at 28°C with shaking at 150 rpm for 5 days. Then the following inducers including oxalic acid, veratryl alcohol, 2,6-dimethoxyphenol were respectively added into the actively growing 5-day-old cultures of Irpex lacteus CD2 at the final concentration of 100 mg/L. After adding the inducers, the fungal cultures were then grown at 28°C with shaking at 150 rpm continuously. Samples were withdrawn every day, centrifuged, and the clear supernatant was used for measuring the extracellular MnP activity.[](https://www.ncbi.nlm.nih.gov/mesh/D019815)
## Purification of manganese peroxidase named as CD2-MnP from Irpex lacteus CD2
*In the **Purification of manganese peroxidase named as CD2-MnP from Irpex lacteus CD2** section:*
The liquid cultures of Irpex lacteus CD2 at the peak of MnP activity were collected and centrifuged at 5000 g for 20 min. Then the culture supernatant was concentrated by 80% ammonium sulfate at 4°C. The sodium acetate buffer (20 mM, pH 4.8) was used to dissolve the pellets. The enzymatic crude extract was dialyzed to remove ammonium sulfate and then applied to a DEAE Sepharose Fast Flow anion exchange column (GE) equilibrated with sodium acetate buffer (20 mM, pH 4.8). The MnP was eluted with a linear gradient of 0–1 M NaCl in the same buffer at a flow rate of 1 ml/min.
The proteins in the eluted fractions was detected by recording the absorbance at 280 nm continuously. Active fractions containing MnP activity were pooled, desalted, filter-sterilized, and stored at 4°C. The purified MnP was verified by SDS-PAGE using 10% polyacrylamide gel. The molecular mass of the purified MnP was estimated by protein ladder molecular weight markers.[](https://www.ncbi.nlm.nih.gov/mesh/D000645)
## Characterization of purified CD2-MnP
*In the **Characterization of purified CD2-MnP** section:*
Kinetic studies were performed in 20 mM malonate buffer (pH 4.5) at 30°C using 5–150 µM Mn2+ (in the presence of 0.08 mM H2O2), 4–80 µM hydrogen peroxide (in the presence of 1.6 mM Mn2+) as substrates. The Lineweaver–Burk plot method was used to determine Km and Vmax of the purified CD2-MnP.[](https://www.ncbi.nlm.nih.gov/mesh/D008314)
The UV-visible spectrum of purified CD2-MnP, in 20 mM malonate buffer (pH 5.0), was measured in the range from 300 nm to 800 nm (UV-1600PC Spectrophtometer, Apada).[](https://www.ncbi.nlm.nih.gov/mesh/D008314)
The effect of temperature on MnP activity was measured in 20 mM malonate buffer (pH 4.5) at 20–80°C. The effect of pH on MnP activity was determined in 20 mM malonate buffer within a pH range of 3.0–7.0 at 30°C. The maximum activity of MnP was set as 100%.[](https://www.ncbi.nlm.nih.gov/mesh/D008314)
To evaluate the thermal stability, the purified MnP was incubated at 40–70°C for 5 h. To evaluate the pH stability, the purified MnP was incubated in different pH (3–6) for 6 h and 24 h. Then the residual MnP activity was calculated based on the original activity before incubation. The initial activity of MnP was set as 100%.
Al3+, Ca2+, Cd2+, Co2+, Mg2+, Ni2+ and Zn2+ (at concentration of 0.4 mM, 2 mM, 4 mM and 40 mM) were used to study the effect of metal ions on the activity of purified MnP. The residual activity was calculated based on the control without adding any metal compound (set as 100%).[](https://www.ncbi.nlm.nih.gov/mesh/D000535)
Methanol, ethanol, DMSO, ethylene glycol, isopropyl alcohol, butanediol, glycerin and acetonitrile (at concentration of 10%, 20% and 30%) were used to study the effect of organic solvents on the activity of purified MnP. The residual activity was calculated based on the control without adding any organic solvent (set as 100%).[](https://www.ncbi.nlm.nih.gov/mesh/D000432)
To evaluate the effect of different metal ions on the stability of purified MnP, the purified MnP was incubated with different concentrations of metal ions (0.4 mM and 4 mM) at 25°C for 12 h and 24 h respectively. Then the MnP activity was measured. The residual activity was calculated based on the control without adding any metal compound (set as 100%).[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
To evaluate the effect of different organic solvents on the stability of purified MnP, the purified MnP was incubated with different concentrations of organic solvents (10% and 30%) at 25°C for 12 h and 24 h respectively. Then the MnP activity was measured. The residual activity was calculated based on the control without adding any organic solvent (set as 100%).
All of above experiments were performed in triplicate.
## Decolorization of different types of dyes by purified CD2-MnP
*In the **Decolorization of different types of dyes by purified CD2-MnP** section:*
To evaluate the dye decolorization capability of the purified MnP, the purified CD2-MnP was used to decolorize four types of synthetic dyes including azo dye Remazol Brilliant Violet 5R and Direct Red 5B, triphenylmethane dye Methyl Green, anthraquinone dye Remazal Brilliant Blue R and indigo dye Indigo Carmine. The reaction mixture in a total volume 1 ml contained (final concentration): malonate buffer (20 mM, pH 4.5), Mn2+ (1.6 mM), H2O2 (0.08 mM), purified CD2-MnP (0.25 U/ml) and dye (50 mg/L). Decolorization was monitored by measuring the absorbance of the reaction mixture at 556 nm for Remazol Brilliant Violet 5R, 510 nm for Direct Red 5B, 640 nm for Methyl Green, 600 nm for Remazal Brilliant Blue R, 610 nm for Indigo Carmine. The decolorization of dye, expressed as dye decolorization (%), was calculated as the following formula: decolorization (%) = [(Ai-At)/Ai]*100, where Ai is the initial absorbance of the dye and At is the absorbance of the dye at time t .[](https://www.ncbi.nlm.nih.gov/mesh/D001391)
To evaluate the effect of different metal ions on the decolorization of dyes by purified CD2-MnP, the reaction mixture in a total volume of 1 ml contained (final concentration): malonate buffer (20 mM, pH 4.5), Mn2+ (1.6 mM), H2O2 (0.08 mM), purified CD2-MnP (0.25 U/ml), dye (50 mg/L), Ca2+, Co2+, Mg2+, Zn2+ (4 mM). Then decolorization was monitored and calculated by the method described above.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
To evaluate the effect of different organic solvents on the decolorization of dyes by purified CD2-MnP, the reaction mixture in a total volume of 1 ml contained (final concentration): malonate buffer (20 mM, pH 4.5), Mn2+ (1.6 mM), H2O2 (0.08 mM), purified CD2-MnP (0.25 U/ml), dye (50 mg/L), methanol, DMSO, ethylene glycol, glycerin (20%). Then decolorization was monitored and calculated by the method described above. All of the decolorization experiments were performed in triplicate.[](https://www.ncbi.nlm.nih.gov/mesh/D008314)
## Decolorization of simulated textile wastewater by purified CD2-MnP
*In the **Decolorization of simulated textile wastewater by purified CD2-MnP** section:*
Simulated textile wastewater containing Remazol Brilliant Violet 5R, Direct Red 5B, Remazal Brilliant Blue R and Indigo Carmine was prepared as described in reference . The stimulated textile wastewater containing different dyes was prepared as follows: 0.5 g L−1 dye, 30 g L−1 NaCl, 5 g L−1 Na2CO3 and 1.5 mL L−1 35% w/v NaOH, the pH was adjusted to 4.5. The reaction mixture in a total volume of 1 ml contained (final concentration): malonate buffer (20 mM, pH 4.5), Mn2+ (1.6 mM), H2O2 (0.08 mM), purified CD2-MnP (0.5 U/ml), simulated textile wastewater (10%, 30%, 50%). Then decolorization was monitored and calculated by the method described above. The decolorization of simulated textile waste was measured by monitoring the decrease in maximum absorbance of each wastewater in a UV-vis spectrophotometer.[](https://www.ncbi.nlm.nih.gov/mesh/C111247)
## Statistical analysis
*In the **Statistical analysis** section:*
To evaluate the effects of metal ions and organic solvents on MnP activity and decolorization of dyes, the ANOVA, the analysis of variance, was performed using the software SPSS (*significant difference, p-value<0.05; **highly significant difference, p-value<0.01).[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
## Results
*In the **Results** section:*
## Induction and purification of manganese peroxidase from Irpex lacteus CD2
*In the **Induction and purification of manganese peroxidase from Irpex lacteus CD2** section:*
It is known that the aromatic compounds play an important role in the regulation of the ligninolytic enzymes synthesis . Besides, organic acids are oxidised by MnP to produce extracellular hydrogen peroxide, which can stimulate the manganese peroxidase gene transcription . To enhance the production of extracellular MnP by Irpex lacteus CD2, the effect of different lignin monomer analogs and organic acids on the activity of extracellular MnP was studied. Time course of MnP activity after addition of various inducers was shown in the Fig. S2.
As shown in Fig. S2, the maximum activity of MnP occurred at the 7th day after addition of inducers. Oxalic acid and veratryl alcohol could significantly enhance the synthesis of extracellular MnP produced by Irpex lacteus CD2. The highest MnP activity was observed in the cultures supplemented with oxalic acid (640.7 U/L), veratryl alcohol (549.0 U/L) and 2,6-dimethoxyphenol (273.5 U/L) (Fig.S2).[](https://www.ncbi.nlm.nih.gov/mesh/D010755)
The MnP secreted by Irpex lacteus CD2 was then purified as described in Table 1. This MnP named as CD2-MnP was purified over 29.3-fold with a terminal specific activity of 24.9 U/mg protein. The result of SDS-PAGE suggested that this enzyme was purified to homogeneity. The molecular mass of CD2-MnP was about 42 kDa as determined by SDS-PAGE (Fig. 1).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
SDS-PAGE analysis of the purified MnP from Irpex lacteus CD2.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
lane M: molecular mass marker; lane 1 and lane 2: purified MnP.
Purification of manganese peroxidase from Irpex lacteus CD2.
## Kinetic studies on the purified CD2-MnP
*In the **Kinetic studies on the purified CD2-MnP** section:*
The kinetic parameters of CD2-MnP with respect to hydrogen peroxide and Mn2+ were determined. The Km values of CD2-MnP were 20.72 µM for H2O2 and 49.41 µM for Mn2+.[](https://www.ncbi.nlm.nih.gov/mesh/D006861)
## UV-visible spectrum of the purified CD2-MnP
*In the **UV-visible spectrum of the purified CD2-MnP** section:*
Like heme peroxidase including horseradish peroxidase and lignin peroxidase, the catalytic cycle of MnP included the native ferric enzyme and the reactive intermediate compound I, II . The identification of oxidized states of MnP compounds I, II was reported by different absorption maxima . As shown in Fig.2, the absorption spectrum of purified CD2-MnP from Irpex lacteus CD2 showed maxima at 419 nm, 529 nm and 556 nm, which suggested that CD2-MnP was a heme protein with iron protoporphyrin IX as compound II.[](https://www.ncbi.nlm.nih.gov/mesh/C448299)
UV-visible spectrum of the purified CD2-MnP.
## Effect of pH on the MnP activity and stability of CD2-MnP
*In the **Effect of pH on the MnP activity and stability of CD2-MnP** section:*
The optimal pH for CD2-MnP was 4.5. CD2-MnP was completely inactive when the pH was above 6.0 (Fig.3A). As shown in Fig.3B, CD2-MnP exhibited high stability in pH ranging from 3.5 to 6.0. The residual MnP activity of CD2-MnP after 24 h incubation at pH 3.5, 4, 4.5, 5, 5.5, 6 was 62.4%, 88.7%, 99.1%, 98.7%, 99.2%, 94.2% of the original activity before incubation, respectively.
Effect of pH and temperature on the activity and stability of purified CD2-MnP from Irpex lacteus CD2.
A: Effect of pH on MnP activity. The activity of 100% was that which was measured at the optimal pH. B: Effect of pH on the stability of CD2-MnP. The initial MnP activity before incubation was set as 100%. C: Effect of temperature on MnP activity. The activity of 100% was that which was measured at the optimal temperature. D: Effect of temperature on the stability of CD2-MnP. The initial MnP activity before incubation was set as 100%.
## Effect of temperature on the MnP activity and stability of CD2-MnP
*In the **Effect of temperature on the MnP activity and stability of CD2-MnP** section:*
The optimal temperature of CD2-MnP was determined to be 70°C (Fig.3C). As shown in Fig.3D, CD2-MnP could respectively retain 72.0%, 68.4% and 53.1% of MnP activity after 5 h incubation at 40°C, 50°C and 60°C. When the temperature increased to above 70°C, the thermostability of CD2-MnP significantly decreased.
## Effect of different metal ions on the MnP activity and stability of CD2-MnP
*In the **Effect of different metal ions on the MnP activity and stability of CD2-MnP** section:*
As shown in Fig.4A, low concentrations of metal ions such as Ca2+, Cd2+, Co2+, Mg2+, Ni2+ and Zn2+ had little effect on the MnP activity of CD2-MnP. When the concentration of Ca2+, Cd2+, Co2+, Mg2+, Ni2+ and Zn2+ was 4 mM, the MnP activity of CD2-MnP was 106.8%, 70.0%, 119.1%, 114.1%, 110.0%, 104.5% of the control without adding any metal compound (Fig.4A). But higher concentrations (40 mM) for all metal ions (other than Ca2+ and Mg2+) resulted in the reduced MnP activity (Fig.4A). It suggested that higher concentrations of metal ions such as Cd2+, Co2+, Ni2+ and Zn2+ had an inhibitory effect on the MnP activity of CD2-MnP.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
Effect of metal ions and organic solvents on the activity of purified CD2-MnP.
A: The effect of different metal ions on MnP activity. The MnP activity of the control without adding any metal compound was set as 100%. B: The effect of different organic solvents on MnP activity. The MnP activity of the control without adding any organic solvent was set as 100%.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
The relative MnP activities of CD2-MnP at different metal ions (final concentration: 40 mM) were compared. As shown in Fig. S3, compared with the relative activity of CD2-MnP at 40 mM Ca2+ and Mg2+ (111.5% and 107.1%), the relative activity of CD2-MnP at 40 mM Al3+, Cd2+, Co2+, Ni2+ was much lower (8.6%, 39.5%, 49.5%, 37.4%). The MnP activities of CD2-MnP at 40 mM Al3+, Cd2+, Co2+, Ni2+ were significantly lower than that of CD2-MnP at 40 mM Ca2+ and Mg2+ (p-value<0.01) (Fig.S3-A,B). Thus the data obtained by the statistical analyses suggested that CD2-MnP showed stronger tolerance to Ca2+ and Mg2+ compared to other metal ions.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
As shown in Fig. S3, the relative MnP activity of control (without adding any metal compound) was also significantly lower than that of CD2-MnP at 40 mM Ca2+ and Mg2+ (p-value<0.01) (Fig.S3-A,B). It suggested that higher concentration of Ca2+ and Mg2+ (40 mM) had no inhibitory effect on the MnP activity of CD2-MnP. In contrast, higher concentration of Ca2+ and Mg2+ could enhance the MnP activity of CD2-MnP.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
As shown in Table 2, CD2-MnP exhibited good stability in different metal ions.
The MnP activity of control without adding any metal compound was set as 100%.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
Effect of metal ions on the stability of purified CD2-MnP from Irpex lacteus CD2.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
CD2-MnP was stable in all of the metal ions tested here when the concentration was 0.4 and 4 mM. It remained about 95% or even higher residual activity after incubation with different metal ions for 24 h (Table 2).[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
## Effect of different organic solvents on the MnP activity and stability of CD2-MnP
*In the **Effect of different organic solvents on the MnP activity and stability of CD2-MnP** section:*
As shown in Fig.4B, when the concentration of organic solvents was 10% and 20%, different organic solvents such as methanol, ethanol, DMSO, ethylene glycol, isopropyl alcohol, butanediol, glycerin had little effect on the MnP activity of CD2-MnP. When the concentration of methanol, ethanol, DMSO, ethylene glycol, isopropyl alcohol, butanediol, glycerin was 20%, CD2-MnP could retain 86.5%, 91.4%, 92.5%, 94.1%, 100.6%, 90.0%, 102.2% residual activity relative to control, respectively (Fig.4B). Acetonitrile had an slight inhibitory effect on the MnP activity of CD2-MnP. CD2-MnP especially exhibited strong tolerance to glycerin, DMSO, ethylene glycol and isopropyl alcohol. When the concentration was increased to 30%, the activity of CD2-MnP could still retain 103.7% (glycerin), 91.0% (DMSO), 93.6% (ethylene glycol) and 86.0% (isopropyl alcohol) relative to the control without adding any organic solvent (Fig.4B).[](https://www.ncbi.nlm.nih.gov/mesh/D000432)
The stability of CD2-MnP in organic solvents was also studied and showed in Table 3. CD2-MnP remained stable in all of the organic solvents tested here at the concentration of 10%. After 24 h incubation with methanol, ethanol, DMSO, ethylene glycol, isopropyl alcohol, butanediol, glycerin and acetonitrile (final concentration: 10%), the residual activity of CD2-MnP retained 82.6%, 91.7%, 90.5%, 72.5%, 77.2%, 119.3%, 85.4% and 80.0%, respectively. When the concentration of organic solvents was increased to 30%, the stability of CD2-MnP decreased. But CD2-MnP still remained relative stable in ethanol, DMSO, butanediol, and glycerin.
The residual activity of CD2-MnP still retained over 80% after 24 h incubation with 30% of ethanol, DMSO, butanediol, and glycerin (Table 3).[](https://www.ncbi.nlm.nih.gov/mesh/D000432)
The MnP activity of control without adding any organic solvent was set as 100%.
Effect of organic solvents on the stability of purified CD2-MnP from Irpex lacteus CD2.
In summary, above results suggested that CD2-MnP had a strong ability to tolerate many organic solvents and metal ions. From the viewpoint of practical applications, the strong resistance to different metal ions and organic solvents was a very valuable advantage of CD2-MnP.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
## Decolorization of different dyes by the purified CD2-MnP
*In the **Decolorization of different dyes by the purified CD2-MnP** section:*
The different types of dyes including the azo dye (Remazol Brilliant Violet 5R, Direct Red 5B), triphenylmethane dye (Methyl Green), anthraquinone dye (Remazol Brilliant Blue R) and indigo dye (Indigo Carmine) were used to evaluate the dye decolorization capability of CD2-MnP. As shown in Fig.5, CD2-MnP showed a strong decolorization capability for a broad range of dyes. Remazol Brilliant Violet 5R, Remazol Brilliant Blue R and Indigo Carmine (50 mg/l) could be respectively decolorized up to 92.8%, 87.1% and 91.5% by the purified CD2-MnP within 5 h (Fig.5A,C,D). Direct Red 5B and Methyl Green (50 mg/l) could be respectively decolorized up to 82.4% and 32.0% by the purified CD2-MnP within 36 h (Fig.5B,E).[](https://www.ncbi.nlm.nih.gov/mesh/D001391)
Decolorization of different types of dyes by the purified CD2-MnP with the coexistence of metal ions.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
The reaction mixture in a total volume 1 ml contained (final concentration): malonate buffer (20 mM, pH 4.5), Mn2+ (1.6 mM), H2O2 (0.08 mM), purified CD2-MnP (0.25 U/ml), dye (50 mg/L) and Ca2+, Co2+, Mg2+, Zn2+ (4 mM). CK (MnP+H2O2) was the control without addition of any metal compound except Mn2+. H2O2 (no MnP) was the negative control without addition of purified CD2-MnP. (A): Decolorization of RBV5R; (B): Decolorization of DR5B; (C): Decolorization of RBBR; (D): Decolorization of IC; (E): Decolorization of MG. RBV5R: Remazol Brilliant Violet 5R, DR5B: Direct Red 5B, RBBR: Remazol Brilliant Blue R, IC: Indigo Carmine, MG: Methyl Green. The negative control (H2O2 was added into the decolorization mixture in the absence of purified CD2-MnP) showed no significant decolorization of different dyes.[](https://www.ncbi.nlm.nih.gov/mesh/D008314)
CD2-MnP especially exhibited stronger ability to decolorize Indigo Carmine, Remazol Brilliant Violet 5R and Remazol Brilliant Blue R. As shown in Fig.5A,C,D, Indigo Carmine, Remazol Brilliant Violet 5R and Remazol Brilliant Blue R could be decolorized up to 90.5%, 75.3% and 72.1% by CD2-MnP within only 1 h. Compared with the monoazo dye (Remazol Brilliant Violet 5R), anthraquinone dye (Remazol Brilliant Blue R) and indigo dye (Indigo Carmine), it was found that the disazo dye (Direct Red 5B) and triphenylmethane dye (Methyl Green) was harder to be decolorized by CD2-MnP. There was limited decolorization in the negative control (H2O2 was added into the decolorization mixture in the absence of purified CD2-MnP). For example, as shown in Fig.5A,C,D, Remazol Brilliant Violet 5R, Remazol Brilliant Blue R and Indigo Carmine were respectively decolorized up to 12.0%, 10.3% and 6.1% in the negative control within 5 h.[](https://www.ncbi.nlm.nih.gov/mesh/D007203)
## Decolorization of different dyes by the purified CD2-MnP with the coexistence of metal ions
*In the **Decolorization of different dyes by the purified CD2-MnP with the coexistence of metal ions** section:*
In order to investigate the capability of CD2-MnP for decolorizing different dyes at the conditions of high concentrations of metal ions and organic solvents, CD2-MnP was further used to decolorize different dyes with the coexistence of metal ions or organic solvents.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
The decolorization of different dyes in the presence of metal ions were tested. As shown in Fig.5A and Fig.S4, the maximum decolorization of RBV5R with the coexistence of Ca2+ (within 5 h) was 97.6%, which was higher than that of control without adding any metal compounds (96.6%) (significant difference, p-value<0.05). The maximum decolorization of RBV5R with the coexistence of Zn2+ (within 5 h) was 98.0%, which was higher than that of control (96.6%) (highly significant difference, p-value<0.01). Above results suggested that Ca2+ and Zn2+ had a promotion effect on the ability of CD2-MnP to decolorize RBV5R. As shown in Fig.5A and Fig. S4, the maximum decolorization of RBV5R when Mg2+ was present or absent were very similar (p-value>0.05).
It suggested that Mg2+ had no inhibitory effect on the ability of CD2-MnP to decolorize RBV5R. As shown in Fig.5A and Fig.S4, the maximum decolorization of RBV5R with the coexistence of Co2+ (within 5 h) was 67.7%, which was lower than that of control (96.6%) (highly significant difference, p-value<0.01). It suggested that Co2+ had an inhibitory effect on the ability of CD2-MnP to decolorize RBV5R.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
As shown in Fig.5B and Fig.S4, the maximum decolorization of DR5B with the coexistence of Co2+ (within 36 h) was 91.4%, which was higher than that of control without adding any metal compounds (82.6%) (highly significant difference, p-value<0.01). The maximum decolorization of DR5B with the coexistence of Zn2+ (within 36 h) was 90.0%, which was higher than that of control (82.6%) (significant difference, p-value<0.05). Above results suggested that Co2+ and Zn2+ had a promotion effect on the ability of CD2-MnP to decolorize DR5B. As shown in Fig.5B and Fig.S4, DR5B with the coexistence of Ca2+ and Mg2+ could be decolorized up to 85.5% and 85.8% within 36 h, respectively. Compared with the control (82.6%), the decolorization percentages were not significantly different (p-value>0.05). Thus Ca2+ and Mg2+ had no inhibitory effect on the ability of CD2-MnP to decolorize DR5B.[](https://www.ncbi.nlm.nih.gov/mesh/C542892)
As shown in Fig.5C,D,E and Fig.S4, compared with the control, decolorization of Methyl Green, Remazol Brilliant Blue R and Indigo Carmine with the coexistence of different metal ions were not significantly different (p-value>0.05). Thus the data obtained by the statistical analyses demonstrated that different metal ions such as Ca2+, Co2+, Mg2+ and Zn2+ had no inhibitory effect on the capacity of CD2-MnP for decolorizing Methyl Green, Remazol Brilliant Blue R and Indigo Carmine.[](https://www.ncbi.nlm.nih.gov/mesh/D008739)
## Decolorization of different dyes by the purified CD2-MnP with the coexistence of organic solvents
*In the **Decolorization of different dyes by the purified CD2-MnP with the coexistence of organic solvents** section:*
The decolorization of different dyes in the presence of organic solvents were also performed. As shown in Fig.6A and Fig. S5, the maximum decolorization of RBV5R with the coexistence of methanol, DMSO, ethylene glycol and glycerin (within 5 h) was 96.0%, 94.0%, 96.4% and 96.0%, which were not significantly different from the control in the absence of any organic solvent (96.6%) (p-value>0.05). It suggested that these organic solvents had no inhibitory effect on the ability of CD2-MnP to decolorize RBV5R.[](https://www.ncbi.nlm.nih.gov/mesh/C111247)
Decolorization of different types of dyes by the purified CD2-MnP with the coexistence of organic solvents.
The reaction mixture in a total volume 1 ml contained (final concentration): malonate buffer (20 mM, pH 4.5), Mn2+ (1.6 mM), H2O2 (0.08 mM), purified CD2-MnP (0.25 U/ml), dye (50 mg/L) and methanol, DMSO, ethylene glycol, glycerin (20%). CK (MnP+H2O2) was the control without addition of any organic solvent. (A): Decolorization of RBV5R; (B): Decolorization of DR5B; (C): Decolorization of RBBR; (D): Decolorization of IC. (E): Decolorization of MG. RBV5R: Remazol Brilliant Violet 5R, DR5B: Direct Red 5B, RBBR: Remazol Brilliant Blue R, IC: Indigo Carmine, MG: Methyl Green.[](https://www.ncbi.nlm.nih.gov/mesh/D008314)
As shown in Fig.6B and Fig. S5, the maximum decolorization of Direct Red 5B with the coexistence of methanol, DMSO, ethylene glycol and glycerin (within 36 h) were 71.6%, 67.1%, 70.0%, 63.5%, which were much lower than that of control (82.6%) (highly significant difference, p-value<0.01). It suggested that the tested organic solvents had an inhibitory effect on the ability of CD2-MnP to decolorize Direct Red 5B.[](https://www.ncbi.nlm.nih.gov/mesh/C542892)
As shown in Fig.6C and Fig. S5, when methanol, ethylene glycol and glycerin were present, the maximum decolorization of RBBR within 5 h were 96.0%, 96.7% and 97.0% respectively, which were not significantly different from the control (98.0%) (p-value>0.05). It suggested that methanol, ethylene glycol and glycerin had no inhibitory effect on the ability of CD2-MnP to decolorize RBBR. But the maximum decolorization of RBBR with the coexistence of DMSO (within 5 h) was only 73.2%, which was much lower than that of control (98.0%) (highly significant difference, p-value<0.01). It suggested that the decolorization of RBBR by CD2-MnP was significantly inhibited by DMSO.[](https://www.ncbi.nlm.nih.gov/mesh/D000432)
As shown in Fig.6D (Indigo Carmine), Fig.6E (Methyl Green) and Fig. S5, the decolorization percentages were not significantly different (between when the organic solvents were present and when they were absent) (p-value>0.05). It suggested that the organic solvents such as methanol, ethylene glycol and glycerin had no inhibitory effect on the ability of CD2-MnP to decolorize Indigo Carmine and Methyl Green.[](https://www.ncbi.nlm.nih.gov/mesh/D007203)
Based on above results, CD2-MnP had a strong capability for decolorizing some dyes such as RBV5R, RBBR, Indigo Carmine and Methyl Green with the coexistence of organic solvents.[](https://www.ncbi.nlm.nih.gov/mesh/C111247)
## Decolorization of simulated textile wastewater by the purified CD2-MnP
*In the **Decolorization of simulated textile wastewater by the purified CD2-MnP** section:*
Considering the high concentration of salts and high ionic strength in textile effluents, the purified CD2-MnP was further evaluated for the decolorization of simulated textile wastewater containing different dye (details are described in Materials and methods). As shown in Table 4, purified CD2-MnP could effectively decolorize different simulated textile wastewater. The simulated textile wastewater containing Remazol Brilliant Violet 5R (10%, 30%), simulated textile wastewater containing Direct Red 5B (10%, 30%), simulated textile wastewater containing Remazol Brilliant Blue R (10%, 30%) and simulated textile wastewater containing Indigo Carmine (10%, 30%) could be decolorized up to 90.1%, 94.9%, 91.8%, 77.0%, 70.0%, 40.1%, 69.0%, 80.6% within 72 h by CD2-MnP, respectively (Table 4). The maximum decolorization of various simulated textile wastewater decreased with the increase of the initial concentration of simulated textile wastewater.[](https://www.ncbi.nlm.nih.gov/mesh/D012492)
RBV5R: Remazol Brilliant Violet 5R; DR5B: Direct Red 5B; RBBR: Remazol Brilliant Blue R; IC: Indigo Carmine.[](https://www.ncbi.nlm.nih.gov/mesh/C111247)
Decolorization of simulated textile wastewater (10%, 30%, 50%) by purified CD2-MnP for 72 h.
## Discussion
*In the **Discussion** section:*
There existed some differences between the properties of CD2-MnP from Irpex lacteus CD2 and that of MnP from other organisms. For example, the optimal temperature of CD2-MnP was determined to be 70°C (Fig.2C), which was higher than MnP from Phanerochaete chrysosporium BKMF-1767 (30°C) , MnP from Lentinula edodes (40°C) and MnP from Schizophyllum sp.F17 (35°C) . Especially, UV-visible absorbance spectra of CD2-MnP suggested that this MnP was different from MnP of other organism. The absorption spectrum of CD2-MnP from Irpex lacteus CD2 showed maxima at 419 nm, 529 nm and 556 nm, which suggested that CD2-MnP was a heme protein with iron protoporphyrin IX as compound II. However, Shin et al. reported that the absorption spectrum of another MnP from Irpex lacteus strain KR 35W showed maxima at 407, 500, and 640 nm .
It indicated that MnP from Irpex lacteus strain KR 35 W was compound I by spectroscopical characterization , .[](https://www.ncbi.nlm.nih.gov/mesh/C448299)
As shown in Fig.2D, there was evidence that the enzyme activity reduced slowly with time even at lower temperatures (such as 40°C). One reason for this phenomenon was that there still existed protein denaturation even at lower temperature , . Previous research has reported that MnP from Phanerochaete chrysosporium was inactivated rapidly at temperature above 40°C . Previous research also indicated that MnP was more susceptible to denaturation by temperature than LiP . Another possible reason was the biphasic first-order model proposed by Liing and Lund based on the presence of two groups with distinct thermal stabilities-a heat labile fraction that inactivates rapidly and a heat resistant fraction which cannot be inactivated completely , .
Thus we assumed that the heat-labile fraction of CD2-MnP may not tolerate 40°C. The enzyme activity reduced slowly with time even at lower temperatures (40°C).
In this study, it was found that the MnP activity of CD2-MnP was significantly inhibited by high concentration of Cd2+ (40 mM). Cd2+ in general was the inhibitor of enzymes. The Mn binding site of MnP was more flexible and allowed a broad range of metal ions to bind to its active site , . Previous research about the kinetic analysis of the effect of cadmium on the activity of manganese peroxidase suggested that Cd2+ could bind to the Mn2+-binding sites, which prevented the oxidation of Mn2+ . Therefore, Cd2+ was considered as a strong inhibitor of MnP.
The possible reason for better tolerance to Mg and Ca than other ions was described as follows. Calcium was a component of binding sites of manganese peroxidase, and it has also been reported that calcium could maintain the structural stability of peroxidases . Mg2+ was a cofactor of the enzyme peroxidase. Previous research also suggested that the binding of Mg2+ may stabilize and activate the manganese peroxidase . In this study, it was found that the MnP activity of of CD2-MnP decreased when the concentration of organic solvents was increased to 30% (Fig.4B).
It has been reported that the organic solvent had an inhibitory effect on the enzyme stability, because the organic solvent could affect the hydration shell of the enzyme molecule which was necessary for maintaining the native conformation . Therefore, the inhibition of the activity of CD2-MnP by high concentration of organic solvents may be caused by the deformation of enzyme structure due to the hydrophobic effects.[](https://www.ncbi.nlm.nih.gov/mesh/D002104)
Our results suggested that Ca2+, Mg2+ and Zn2+ could stimulate the MnP activity of CD2-MnP, which agreed with the previous research about the effect of metal ions on the activity of MnP purified from Stereum ostrea , Rhizoctonia and Schizophyllum . Although there have been some reports about the effect of different metal ions on the activity of MnP, few study has been performed to evaluate the influence of metal ions on the stability of MnP. Our research suggested that CD2-MnP from Irpex lacteus CD2 showed high stability in different metal ions (Table 2). To our knowledge, this is the first report about the effect of different metal ions on the stability of MnP. This character may be very valuable for the application of CD2-MnP in the treatment of wastewaters containing different metal ions.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Most previous research focused on studying the enzymatic and kinetic properties of MnP from different sources. However, the effect of different organic solvents on the MnP activity was rarely studied. Boer et al. have reported that the MnP purified from Lentinula edodes showed a high percentage of activity in reaction mixtures containing 10% (v/v) of different organic solvents such as acetone, isopropanol and ethanol. But the effect of higher concentration of organic solvents on the MnP activity was not investigated. The stability of MnP in different organic solvents was also not reported .
Our results showed that higher concentration of organic solvents such as glycerin, DMSO, ethylene glycol, isopropyl alcohol and butanediol (30%, v/v) had no inhibitory effect on the activity of CD2-MnP (Fig.3B). More importantly, our research indicated that CD2-MnP showed strong tolerance to many organic solvents especially ethanol, DMSO, butanediol and glycerin (Table 3). Thus, compared with the MnP from Lentinula edodes , CD2-MnP purified from Irpex lacteus CD2 in this study appeared to be more resistant to different organic solvents.[](https://www.ncbi.nlm.nih.gov/mesh/D000096)
Previous research suggested that some azo dyes and anthraquinone dyes could be decolorized by the purified MnP from different fungal strains such as Dichomitus squalens , Schizophyllum , Stereum ostrea and Bjerkandera adusta . But as far as we know, no study has been performed to evaluate the decolorization capability of MnP in the presence of different metal ions or organic solvents. In this study, we found that the purified CD2-MnP from Irpex lacteus CD2 had a strong ability to decolorize different dyes with the coexistence of different metal ions or organic solvents (Fig.4 and Fig.6). This important property may contribute to the efficient use of MnP in the treatment of dye effluents.[](https://www.ncbi.nlm.nih.gov/mesh/D001391)
## Conclusions
*In the **Conclusions** section:*
In this study, we found that the synthesis of manganese peroxidase in the white-rot fungus Irpex lacteus CD2 could be significantly enhanced by oxalic acid, veratryl alcohol and 2,6-Dimehoxyphenol. A novel manganese peroxidase named as CD2-MnP was purified and characterized from this fungus. CD2-MnP exhibited strong tolerance to different metal ions and organic solvents. The different types of dyes including the azo dye (Remazol Brilliant Violet 5R, Direct Red 5B), anthraquinone dye (Remazol Brilliant Blue R), indigo dye (Indigo Carmine) and triphenylmethane dye (Methyl Green) as well as simulated textile wastewater could be efficiently decolorized by CD2-MnP. CD2-MnP also had a strong capability for decolorizing different dyes with the coexistence of metal ions and organic solvents. In summary, the manganese peroxidase CD2-MnP from Irpex lacteus CD2 showed a great potential for the enzymatic degradation of different industrial dyes and textile dye effluents.[](https://www.ncbi.nlm.nih.gov/mesh/D019815)
## Supporting Information
*In the **Supporting Information** section:*
# References
*In the **References** section:*
|
# Introduction
A study of the ultrasound-targeted microbubble destruction based triplex-forming oligodexinucleotide delivery system to inhibit tissue factor expression
# Abstract
*In the **Abstract** section:*
The efficiency of cellular uptake of triplex-forming oligodexinucleotides (TFO), and the inhibition of tissue factor (TF) is low. The aim of the present study was to improve the absorption o[f TFO, and increase the inhibition o](https://www.ncbi.nlm.nih.gov/mesh/D009841)f [TF ](https://www.ncbi.nlm.nih.gov/mesh/D009841)induced by shear stress both in vitro and in vivo, by using an ultrasound-targeted microbubble destruction (UTMD)-[bas](https://www.ncbi.nlm.nih.gov/mesh/D009841)ed delivery system. TFO-conjugated lipid ultrasonic microbubbles (TFO-M) were first constructed and characterised. The absorption of TFO was observed by a fluorescence-based[ me](https://www.ncbi.nlm.nih.gov/mesh/D009841)thod, and th[e inh](https://www.ncbi.nlm.nih.gov/mesh/D008055)ibition of TF by immunoflu[ore](https://www.ncbi.nlm.nih.gov/mesh/D009841)scence and quantitative polymerase chain reaction. ECV304 human [umb](https://www.ncbi.nlm.nih.gov/mesh/D009841)ilical vein endothelial cells were subjected to fluid shear stress for 6 h after treatment with TFO conjugated lipid ultrasonic microbubbles without sonication (TFO-M group); TFO alone; TFO conjugated lipid ultrasonic microbubbles, plus immediate so[nic](https://www.ncbi.nlm.nih.gov/mesh/D009841)ation (TFO+U[ grou](https://www.ncbi.nlm.nih.gov/mesh/D008055)p and TFO-M+U group); or mock treated with 0.[9% ](https://www.ncbi.nlm.nih.gov/mesh/D009841)NaCl only ([SSR](https://www.ncbi.nlm.nih.gov/mesh/D009841)E group)[. T](https://www.ncbi.nlm.nih.gov/mesh/D009841)he in vivo e[xperi](https://www.ncbi.nlm.nih.gov/mesh/D008055)ments were established in a similar manner to the in [vit](https://www.ncbi.nlm.nih.gov/mesh/D009841)ro experiment[s, ](https://www.ncbi.nlm.nih.gov/mesh/D009841)except that TFO or TFO-M was injec[ted into ](https://www.ncbi.nlm.nih.gov/mesh/D000077330)rats through the tail vein. Six hours after the preparation of a carotid stenosis model, the rats were humanely sacrificed. The t[ran](https://www.ncbi.nlm.nih.gov/mesh/D009841)sfection efficiency of TFO in the TFO-M+U group was higher as compared with the TFO-M and TFO+U group (P<0.01). The protein and mRNA expression of TF in the TFO-M+U group was s[ign](https://www.ncbi.nlm.nih.gov/mesh/D009841)ificantl[y d](https://www.ncbi.nlm.nih.gov/mesh/D009841)ecreased both in vitro and in vivo (P<0.01)[, a](https://www.ncbi.nlm.nih.gov/mesh/D009841)s compa[red](https://www.ncbi.nlm.nih.gov/mesh/D009841) with the TFO-M, TFO+U and SSRE groups. The UTMD-based TFO deliv[ery](https://www.ncbi.nlm.nih.gov/mesh/D009841) system promoted the absorption of TFO and the inhibition of TF, and was therefore considered to[ be](https://www.ncbi.nlm.nih.gov/mesh/D009841) fav[ora](https://www.ncbi.nlm.nih.gov/mesh/D009841)ble for preventing thrombosis induc[ed ](https://www.ncbi.nlm.nih.gov/mesh/D009841)by shear stress.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Triplex-forming oligonucleotides (TFOs) are a useful tool in anti-gene therapy due to their sequence-specific DNA binding capacity (1–4). The formation of triplexes with a targeted promoter inhibits the transcription of the target gene(5–7), resulting in control of gene expression. Although this strategy holds great potential, the rate of transcriptional inhibition remains a challenge.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
Ultrasound-targeted microbubble destruction (UTMD) is a promising approach for effective thrombosis therapy. Previous studies (8–10) have confirmed that UTMD can enhance gene transfection efficiency. As a gene delivery system, UTMD can penetrate the endothelial barriers of the capillary walls, avoid initiating an immune response, and penetrate the nuclear membrane (11). Microbubble sonoporation has improved intracellular gene delivery (12–14) through the creation of transient pores in vascular endothelial cells, disruption of vascular endothelial integrity, and stimulation of endocytic cellular uptake (15). Co-administration of microbubbles and ultrasound, in combination with pharmaceutical thrombolysis ex vivo, can further enhance thrombolytic activity (16).
Furthermore, the combination of microbubbles and ultrasound, without the use of fibrinolytics, increases the effect of ultrasound on thrombolysis in vivo. Administration of microbubbles has been shown to accelerate clot lysis during continuous 2-MHz ultrasound monitoring in stroke patients treated with intravenous tissue plasminogen activator (17). Previous studies (18–20) have further demonstrated that UTMD holds significant potential for thrombosis gene therapy.
Tissue factor (TF) (21–23) is a membrane-bound glycoprotein that is expressed or exposed at sites of vascular injury, and is essential to hemostasis. Binding of circulating factor VII/VIIa to TF initiates the clotting cascade, which promotes the formation of fibrin and platelet plugs. Activation of the TF gene endothelial domain (24–31) is usually induced in the narrow, curved, and divergent areas of brain blood vessels, and atherosclerotic plaques (32–34), suggesting that hemodynamic factors, including shear stress (SS), have an important role in cerebral atherosclerotic thrombosis and distribution (35–40).
In our previous studies, TFO (41–43) blocked the activation of the shear stress responsive element (SSRE) (44–48) in the TF gene promoter (44, 49–51) and resulted in the failure of TF gene transcription; however, the inhibition level of TF transcription by TFO still needs to be improved. The rate of TFO uptake in the ECV304 endothelial cell line, and the inhibition of TF in endothelial cells of the rat common carotid artery, 6.5 h after the injection of TFO, was 11.65 and 23%, respectively (41–43). The present study aimed to overcome the difficulty of delivering TFO into the cell nucleus by using a UTMD-based delivery system. TFO-conjugated lipid ultrasonic microbubbles were delivered for the first time using this method, both in vitro and in vivo.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
## Materials and methods
*In the **Materials and methods** section:*
## TFO design
*In the **TFO design** section:*
The TFO sequence targeted to the SSRE in the promoter of the human TF gene in vitro was, 3′-GGGTGGTGTGGTGGGGGTGGG-5′. The TFO sequence targeted to the SSRE in the promoter of the rat TF gene in vivo was, 3′-GGGGGGTGGGGTGTGTGTGT-5′. The sequences were designed, synthesized and modified by a phosphorothioate method, and then labeled with fluorescein isothiocyanate (FITC). The synthesis, purification, modification, and fluorescence labeling was completed by Shanghai Shenggong Biological Engineering Technology & Services Co., Ltd. (Shanghai, China).[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
## Preparation of TFO conjugate lipid ultrasonic microbubble complexes
*In the **Preparation of TFO conjugate lipid ultrasonic microbubble complexes** section:*
A suspension of lipid ultrasonic microbubbles, containing 7×109 microbubbles/ml, was obtained from Xinqiao Hospital, The Third Military Medical University (Chongqing, China). TFO-FITC (10 μl, 100 μmol/l) and lipid ultrasonic microbubbles (100 μl, 7×109 μg/ml) were gently agitated in phosphate-buffered saline (PBS), and the resulting transfection complexes were transferred to a polystyrene tube and incubated at 4°C overnight. The shape of the complexes was observed using a light microscope, and the fluorescence labeling was detected using a fluorescence microscope. The particle size, diameter, and surface potential was detected using a Coulter events-per-unit-time meter and a Malvern laser particle size analyzer (Zetasizer 3000; Malvern, Westborough, MA, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
## Antibodies and cell culture
*In the **Antibodies and cell culture** section:*
A polyclonal rabbit anti-rat TF primary antibody and rhodamine labeled anti-rabbit secondary antibody were obtained from Boster Biological Technology Co., Ltd. (Wuhan, China.)[](https://www.ncbi.nlm.nih.gov/mesh/D012235)
ECV304 human umbilical vein endothelial cells were obtained from the China Center for Type Culture Collection (Wuhan, China) and incubated in M199 medium with 10% fetal bovine serum at 37°C in a humidified environment of 5% CO2 and 95% air. The initial cell viability was determined for further experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
## In vitro experimental protocol
*In the **In vitro experimental protocol** section:*
The TFO-conjugated lipid ultrasonic microbubbles were centrifuged at low speed prior to the experiment. The suspension was diluted with 0.9% NaCl to a final concentration of 0.2 μmol/l and a final volume of 60 μl. The ECV304 cells were randomly divided into four groups; a blank control group (SSRE group), in which the ECV304 cells were mock treated with 0.9% NaCl without TFO, microbubble or ultrasound; a TFO and ultrasound group (TFO+U), in which the ECV304 cells were added to the TFO mixture and immediately sonicated using a therapeutic ultrasound transducer (Xinqiao Hospital, The Third Military Medical University) with parameters set at 1 MHz, 1 W/cm2, 30 S, and a duty cycle of 0.5%; a TFO conjugated-lipid ultrasonic microbubble group (TFO-M), in which the ECV304 cells were added to the TFO conjugated lipid ultrasonic microbubbles; and a TFO-conjugated lipid ultrasonic microbubble injection plus ultrasound (U) group (TFO-M+U), in which the ECV304 cells were added to the TFO conjugated lipid ultrasonic microbubbles and exposed to ultrasound with the same irradiation parameters as the TFO+U group. The ECV304 cells were then subjected to fluid shear stress of 12 dyn/cm for 6 h.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
The efficacy of the gene transfection was measured as the number of fluorescent cells per region, and then normalized to the total number of cells per unit area. All of the fluorescent cells within the region of insonation were counted, as well as the total number of cells present by both phase contrast and fluorescence microscopy. Six OptiCells® (China Center for Type Culture Collection, Wuhan, China) were used per treatment group and the experiment was repeated at least twice on separate days.
The inhibition of TF gene expression was measured 6 h after application of fluid shear stress of 12 dyn/cm. The expression of TF protein was detected using an immunofluorescence method. The samples were washed with PBS and fixed with cold 4% paraformaldehyde. The samples were then incubated with polyclonal rabbit anti-rat TF primary antibodies (1:400) at 4°C overnight, followed by incubation with rhodamine-labeled anti-rabbit secondary antibodies (1:200) at 37°C for 1 h; all antibodies were obtained from Boster Biological Technology Co. Ltd (Wuhan, China). A laser scanning confocal microscope was then used to examine the expression and distribution of fluorescence in the ECV304 cells.
Image Pro Plus (MediaCybernetics, Inc., Rockville, MD, USA) analysis system was used to determine the average gray scale of positive expression. The expression of TF mRNA was analyzed by quantitative polymerase chain reaction (qPCR). Total RNA was isolated from the cultured ECV304 cells using TRIzol (Invitrogen Life Technologies, Carlsbad, CA, USA) reagent according to the manufacturer’s instructions. Primer sequences for TF were; forward, 5′-GAACCCAAACCCGTCAAT-3′ and reverse, 5′-GAAGACCCGTGCCAAGTA-3′. The reverse transcription was performed at 42°C for 40 min and the cDNA (2 μl) was amplified under standard PCR reaction conditions.
The qPCR reaction was performed as follows: 5 min at 94°C (one cycle), 30 sec at 94°C, 30 sec at 55°C, 30 sec at 72°C, plate reading (38 cycles), and then 10 min at 72°C. The PCR amplification was performed on a thermal cycler over 27 cycles. The average gray value was analyzed using Gel Pro Analyzer (MediaCybernetics) software and a raw data value for TF expression in each group was normalized to GAPDH.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## Establishment of a Sprague Dawley (SD) rat model of carotid artery stenosis
*In the **Establishment of a Sprague Dawley (SD) rat model of carotid artery stenosis** section:*
SD rats were housed in a constant room temperature of 24°C under a 12 h light-dark cycle, and fed ad libitum. All experiments were performed with the approval of the Third Military Medical University Animal Ethics Committee. Efforts were made to minimize animal suffering and to keep the number of animals used to a minimum. Twenty-four male SD rats aged 6 months old and weighing 300±8.5 g were randomly placed into one of the four groups (n=6). The rats were anaesthetized by intraperitoneal injection of 3% pentobarbital sodium at a dose of 45 mg/kg until the eyelash reflex disappeared.
The rats were then fixed in a dorsal position. Using aseptic techniques, a 2cm incision was made in the median neck of each rat. Following layer separation, ~1 cm of the left common carotid artery was separated from the paratracheal carotid sheath, set into a longitudinally-split silica gel tube with an inner diameter of 0.5 mm and a length of 3 mm, tightly ligated twice using no. 4 silk thread, and sewn onto the skin following repositioning.[](https://www.ncbi.nlm.nih.gov/mesh/D010424)
## In vivo experimental protocol
*In the **In vivo experimental protocol** section:*
The TFO-conjugated lipid ultrasonic microbubble was centrifuged at low speed prior to the experiment and the suspension was diluted with 0.9% NaCl to a concentration of 1.0 mg ml−1. The rats were anesthetized using 3% sodium pentobarbital and were fixed on the experimental table. All the TFOs or mock complexes (0.5 mg kg−1) were administered through the tail vein. The carotid stenosis animal model was generated 0.5 h after treatment.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
The SD rats were intravenously injected with or without ultrasound (U) treatment to derive the following four groups: 0.9% NaCl only without TFO, microbubble, or ultrasound (blank control SSRE group); half mg kg−1 of TFO plus immediate sonication with a therapeutic ultrasound transducer set at 347 KHz and 2.4 MPa for 2 min (TFO+U group); TFO conjugated lipid ultrasonic microbubbles (TFO-M group); and TFO-conjugated lipid ultrasonic microbubble plus ultrasound with the same sonication parameters as above (TFO-M+U group). Half an hour after the treatment, the carotid stenosis model was generated and six hours after model preparation, the rats were humanely sacrificed. Serial sections of the left common carotid artery were perfused with 0.9% NaCl solution at a velocity of 3 ml/min under low pressure until the outflow of liquid was transparent. The liquid was then changed to 100 ml paraformaldehyde (4%), diethylpyrocarbonate (0.1%), and PBS (0.1 M) under low pressure to perform in situ perfusion and fixation. Subsequently, the stenosis segment was dissected from the left common carotid artery and embedded in embedding medium.
Frozen 5-μm sections were then subjected to immunofluorescence. The samples were incubated with polyclonal rabbit anti-rat TF primary antibody and rhodamine labeled anti-rabbit secondary antibody. A laser scanning confocal microscope was used to examine the expression and distribution of fluorescence in the frozen sections and the Image Pro Plus (MediaCybernetics) was used to determine the average gray scale of expression.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## Statistical analysis
*In the **Statistical analysis** section:*
Statistical analyses were performed using SPSS version 13.0 (SPSS, Inc., Chicago, IL, USA). All values are expressed as means ± standard deviation. Analysis of variance was used to determine significant differences through multiple comparisons. A P<0.05 was considered to indicate a statistically significant difference.
## Results
*In the **Results** section:*
## Preparation of TFO conjugated lipid ultrasonic microbubble complexes
*In the **Preparation of TFO conjugated lipid ultrasonic microbubble complexes** section:*
The lipid ultrasonic microbubble with FITC-labeled TFO appeared as a pale green suspension, and had a smooth round surface, even size and light density as observed under a light microscope (Fig. 1A). The microbubble concentration was ~7×109/ml. The microbubble surfaces appeared green under the fluorescence microscope (Fig. 1B), while the lipid microbubbles without FITC-TFO were not visible, indicating that FITC-TFO was packaged on the microbubble lipid membrane (Fig. 2). The analysis of the particle size and diameter indicated that the mean intensity, volume and mean diameter were 2092.8, 2114.2, and 2166.9 nm, respectively. The surface potential analysis was −46.0±1.6 mm.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
## UTMD-based TFO delivery system in vitro
*In the **UTMD-based TFO delivery system in vitro** section:*
The absorption rate of TFO into ECV304 cells was measured by fluorescence microscopy. The green fluorescence of FITC-labeled TFO was detected in the ECV304 cells (Fig. 3A), and was visible in the three experimental groups. The positive cells were most abundant in the TFO-M+U group as compared with the TFO-M and TFO+U group. The green fluorescence signal in the TFO-M and TFO+U groups was weak and mainly distributed in the cytoplasm, whereas the TFO-M+U group exhibited brighter green fluorescence.
The transfection efficiency of TFO in the TFO-M+U group (38.83±6.52%) was significantly higher as compared with the TFO-M (9.50±2.88%) and the TFO+U group (12.66±3.01%, P<0.01) (Fig. 3B). There was no significant difference between the TFO-M and TFO+U groups (P>0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
The expression of TF protein was detected by immunofluorescence as fine red particles (Fig. 4). The TF protein was observed mainly in the cytoplasm and membrane of the ECV304 cells in the SSRE group, with a small amount in the nucleus. The intensity of the red fluorescence was greater in the SSRE group as compared with the TFO+U, TFO-M and TFO-M+U groups; the intensity in the TFO-M+U group was significantly lower as compared with the TFO+U and TFO-M groups. The TF protein content in the TFO+U (36.83±8.34), the TFO-M (40.77±9.40) and the TFO-M+U groups (13.98±6.39) was significantly lower as compared with the SSRE group (74.00±16.67) (P<0.01). The gray value in the TFO-M+U group was significantly lower as compared with both the TFO+U or TFO-M group (P<0.01); and there was no significant difference between the TFO+U and TFO-M groups (P>0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
TF mRNA expression was determined by qPCR (Fig. 5). There was a marked amplification of TF in the SSRE group. Based on image analysis, the TF mRNA was significantly lower (P<0.01) in the TFO+U (0.36±0.07), the TFO-M (0.38±0.07) and the TFO-M+U groups (0.11±0.02), as compared with the SSRE group (0.71±0.08). The TF mRNA expression in the TFO-M+U group was significantly lower as compared with the TFO+U and TFO-M group (P<0.01), however there was no significant difference between the TFO+U and TFO-M group.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
## UTMD-based TFO delivery system in vivo
*In the **UTMD-based TFO delivery system in vivo** section:*
A rat model of carotid stenosis was successfully generated. The expression of TF protein in endothelial cells of carotid arteries in the four different groups was detected by immunofluorescence (Fig. 6A). The number of positive cells and the degree of staining was significant in the SSRE group, as compared with the other groups. The amount of red fluorescence in the TFO+U, TFO-M and the TFO-M+U group was lower; with the amount of fluorescence in the TFO-M+U group being significantly lower as compared with the TFO+U and TFO-M groups. Image analysis identified that the TF protein content in the TFO+U (51.22±5.69), TFO-M (55.22±6.47) and the TFO-M+U groups (20.59±4.38) was significantly lower (P<0.01) as compared with the SSRE group (71.78±7.10) (Fig. 6B).
The fluorescence in the TFO-M+U group was significantly lower as compared with the TFO+U and TFO-M groups (P<0.01) and there was no significant difference between the TFO+U and TFO-M groups.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
## Discussion
*In the **Discussion** section:*
To improve the efficiency of TFO delivery, a UTMD-based delivery system was used to deliver TFO both in vitro and in vivo. It was first observed that FITC-labeled TFO had been successfully packaged onto the lipid microbubble membrane, and surface potential analysis showed that the FITC-labeled TFO microbubble measured −46.0±1.6 mm. The average size of the microbubble contrast agent has a key role in function. The microbubbles must be small enough to pass through the capillary wall endothelial barriers and be less than the diameter of a human red blood cell (7.8 μm) (52). The typical diameter range of the microbubbles was 0.5–10 μm, with some microbubbles reaching nanoscale.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
The absorption of TFO and the inhibition of TF by the UTMD-based delivery system was observed in vitro. These results indicated that UTMD efficiently delivered TFO to the cells, resulting in a down regulation of TF protein and mRNA expression. The decrease in TF expression was associated with increased TFO in ECV304 cells in vitro. The increased TFO transfection efficiency was associated with a decrease in TF expression, suggesting that TFO was effective in inhibiting TF expression induced by shear stress in ECV304 cells. The UTMD-based TFO gene delivery system could significantly increase the absorption rate of TFO into cells and subsequently strengthen the inhibition of TF expression in vitro.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
The expression of TF induced by shear stress in endothelial cells of rat carotid arteries, was observed following UTMD-based TFO delivery in vivo. The results from the present study were in accordance with previous studies: TFO formed triplexes with the TF endothelial cell gene promoter region in vitro (53) and in vivo (26). TFO uptake and TF inhibition in the ECV304 endothelial cell line and of the rat common carotid artery, 6.5 h after the injection of TFO, were ~11.65% and 23% TF transcription, respectively (41–43). TFO absorption in the TFO-M+U group was increased to 38.83% in vitro and the inhibition level of TF in the TFO-M+U group was decreased to 71.31%. These findings suggest that TF expression, induced by fluid shear stress of the cells of the carotid artery, could be inhibited by TFO.
The UTMD-based TFO gene delivery system could promote inhibitive effects, and may be favorable for preventing shear stress-induced thrombosis in vivo.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
Taken together, the UTMD-based TFO delivery system increased TFO delivery and decreased TF expression in vitro and in vivo. However, the amount of TFO loaded on the lipid ultrasonic microbubbles is limited, and the optimization of ultrasound parameters and strategies, in order to increase TFO absorption, may be time-consuming. The lipid ultrasonic microbubble may be used as a carrier for TFO, and the UTMD-based TFO delivery system provides a promising strategy for cerebral thrombosis gene therapy.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
# References
*In the **References** section:*
Visualisation of FITC-labeled TFO lipid microbubbles. (A) Lipid microbubbles labeled with FITC-TFO were observed under a light microscope (magnification, ×100). The arrow indicates one of the lipid ultrasonic microbubbles labeled with FITC-TFO. (B) Lipid microbubbles labeled with FITC-TFO under a fluorescence microscope (magnification, ×100). The red arrow indicates one of the lipid ultrasonic microbubbles labeled with FITC-TFO. FITC, fluorescein isothiocyanate; TFO, triplex-forming oligonucleotides.[](https://www.ncbi.nlm.nih.gov/mesh/D016650)
Model of TFO conjugated lipid ultrasonic microbubbles. TFO, triplex-forming oligonucleotides.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
Cell absorption rate of TFO. (A) The rate of TFO absorption by ECV304 cells was detected by fluorescence microscopy (magnification, ×400). The red arrows indicate positive green fluorescence of FITC-labeled TFO in ECV304 cells. (B) The rate of TFO absorption by ECV304 cells in the TFO-M, TFO+U and TFO-M+U groups. The values represent the means ± standard deviation, n=6 per group. *P<0.01 as compared with the TFO-M+U group. TFO, triplex-forming oligonucleotides; M, microbubble; U, ultrasound.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
Immunofluorescent analysis of tissue factor protein in vitro (magnification, ×400). The arrows indicate positive immunofluorescent detection of tissue factor protein. TFO, triplex-forming oligonucleotides; M, microbubble; U, ultrasound; SSRE, shear stress response element.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
Analysis of tissue factor mRNA expression by quantitative polymerase chain reaction. Lane 1, TFO-M-U group; Lane 2, TFO-M group; Lane 3, TFO-U group; Lane 4, SSRE group. TFO, triplex-forming oligonucleotides; M, microbubble, U, ultrasound; SSRE, shear stress response element; TF, tissue factor; bp, base pairs.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
Immunofluorescence assay of tissue factor protein in vivo. (A) In vivo expression of tissue factor protein was detected by immunofluorescence microscopy (magnification, ×400). The arrows indicate positive immunofluorescence of tissue factor protein in endothelial cells of the carotid artery. (B) Average gray value in the SSRE, TFO-M, TFO+U and TFO-M+U groups. The values represent the means ± standard deviation, n=6 per group. #P<0.01 as compared with the SSRE group; *P<0.01 as compared with the TFO-M+U group. TFO, triplex-forming oligonucleotides; M, microbubble; U, ultrasound; SSRE, shear stress response element.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
|
# Introduction
Synthesis and characterization of a new photoinduced switchable [β-cyclodextrin](https://www.ncbi.nlm.nih.gov/mesh/C031215) dimer
# Abstract
*In the **Abstract** section:*
Summary
This paper reports an efficient preparation of bridged bis-β-CD AZO-CDim 1 bearing azobenzene as a linker and exhibiting high solubility in water. T[he photo](https://www.ncbi.nlm.nih.gov/mesh/C031215)i[somerizati](https://www.ncbi.nlm.nih.gov/mesh/C009850)on proper[ties were ](https://www.ncbi.nlm.nih.gov/mesh/C009850)studied by UV–vis and HPLC and supported by ab [initi](https://www.ncbi.nlm.nih.gov/mesh/D014867)o calculations. The cis/trans ratio of AZO-CDim 1 is 7:93 without irradiation and 37:63 after 120 min of irradiation at 365 nm; the [reaction i](https://www.ncbi.nlm.nih.gov/mesh/C009850)s reversible after irradiation at 254 nm. The photoinduced, switchable binding behavior of AZO-CDim 1 was evaluated by ITC, NMR and molecular modeling in the presence of a ditopic adam[antyl gues](https://www.ncbi.nlm.nih.gov/mesh/C009850)t. The results indicate that AZO-CDim 1 can form two different inclusion comple[xes with ](https://www.ncbi.nlm.nih.gov/mesh/D000218)an adamantyl dimer depending on it[s photoind](https://www.ncbi.nlm.nih.gov/mesh/C009850)uced isomers. Both cavities of cis-AZO-CDim 1 are co[mplexed s](https://www.ncbi.nlm.nih.gov/mesh/D000218)imultaneously by two adamantyl units of the guest forming a 1:1[ complex while](https://www.ncbi.nlm.nih.gov/mesh/C009850) trans-AZO-CDim 1 seems to lead to th[e formati](https://www.ncbi.nlm.nih.gov/mesh/D000218)on of supramolecular polymers with an n:n stoich[iometry.](https://www.ncbi.nlm.nih.gov/mesh/C009850)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
α-, β- or γ-Cyclodextrins (CDs) are cyclic oligosaccharides composed of 6, 7 or 8 α-D-1,4 glucopyranose moieties, respectively. They are natural compounds produced from starch by the reaction of 4-α-glucanotransferases [1]. Their toroidal shape, with C6-primary hydroxy groups on the narrow rim and C2 and C3 secondary groups on the wider rim, enables encapsulation of hydrophobic molecules inside their cavity. Since the 1950s, it has been demonstrated that CDs can form non-covalent-force complexes in water due to their unique spatial arrangement. In particular, β-cyclodextrin (β-CD) is known to form supramolecular inclusion complexes with molecules, and such inclusion usually enhances the solubility of water-insoluble substances [1–4].
Pharmaceutical companies already use these cyclodextrins or their derivatives in their formulations [5–6]. In fact, they have a well-defined structure, low toxicological or pharmacological activity, and good solubility in water. For example, the inclusion of active substances in CDs can reduce their undesirable[ storage or metabolism de](https://www.ncbi.nlm.nih.gov/mesh/D047391)gr[ada](https://www.ncbi.nlm.nih.gov/mesh/D003505)tion, which h[as led research ](https://www.ncbi.nlm.nih.gov/mesh/D009844)on CDs to foc[us on controlled drug delivery ](https://www.ncbi.nlm.nih.gov/mesh/D005936)[4]. In the food industry, CDs enable the fixation or retention of[ volat](https://www.ncbi.nlm.nih.gov/mesh/D013213)ile flavors, as well as the removal of undesirable flavors from food [7–8].[](https://www.ncbi.nlm.nih.gov/mesh/C031356)
In comparison with CD monomers, bridged bis-cyclodextrins can improve the original binding ability of native CDs through the cooperative binding of both cavities located close to the guest molecules [9–10]. These cyclodextrins linked by ester [11], thioether [12–16], urea [17–19], or triazole [20] moieties have been previously described. In addition, aromatic azobenzenes are excellent candidates as molecular switch linkers as they have two forms, namely cis (Z) and trans (E) isomers, which can be interconverted by both photochemical and thermal means [21]. This transformation by external stimuli induces a molecular movement and a significant geometric change
[22–23]. CDs and azobenzene derivatives can form inclusion complexes controlled by photoisomerization of the guests and this property has been widely applied to molecular shuttles, motors, information storage [24–25] and catalysis [26].[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
Some examples of azobenzene-linked CD dimers can be found in the literature but they generally suffer from arduous purification steps and very low yields [27–28]. As an exception, Vargas et al. [29] described the synthesis of 1,2,3-triazole-linked azobenzene-cyclodextrin derivatives producing rather good yields but the photoisomerization and inclusion complex properties were not investigated. Here, we report an efficient preparation of a new bis-β-CD with azobenzene dicarboxylate and the influence of photoisomerization of the linker on the conformation and binding behavior of the CD dimer.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
## Results and Discussion
*In the **Results and Discussion** section:*
The AZO-CDim 1 synthesis was performed as follows: 4,4’-azobenzenedicarboxylic acid was first obtained by reductive coupling of 4-nitrobenzoic acid with a yield of 49% (Scheme 1) [30]. Then, the carboxylic groups were activated by N-hydroxysuccinimide (NHS) and condensed with mono-6-amino-6-deoxy-β-cyclodextrin (β-CD-NH2) [31] in anhydrous DMF at room temperature. Flash chromatography (C18 column, H2O/MeOH 90:10 to 10:90 v/v in 20 min) afforded pure AZO-CDim 1 with 62% yield.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Synthesis pathway of the dimer AZO-CDim 1.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
It should be noted that AZO-CDim 1 exhibits surprisingly high solubility in water, reaching 220 mM at 293 K, even though the linker is hydrophobic and the solubility of β-CD is only 16 mM under the same conditions. Concentration-variable NMR analysis revealed a strong upfield shift and a broadening of the proton signals. No significant chemical shift variations were observed at concentrations below 1 mM, which is consistent with a critical aggregation concentration of around 1·10−3 M.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
The UV–vis absorption spectrum of azobenzene presents three characteristic absorption bands (250, 320 and 450 nm) corresponding to π–π* and n–π* electronic transitions, respectively. For the trans isomer, the absorption band π–π* at 320 nm is very intense while the other two bands (π–π*) at 250 nm and (n–π*) at 420 nm are much weaker. For the cis isomer, the absorption band π–π* is shifted slightly to a shorter wavelength and is significantly less intense at 320 nm. Because the n–π* transition is possible in the cis isomer, this band increases in intensity [22–23]. A molecular switch is based on the light-induced, reversible transformation of chemical species between two molecular states with different absorption spectra.
Thus, the trans/cis isomerization can be reversibly controlled through UV light irradiation as depicted in Fig. 1. As shown in Fig. 1, when a sample containing AZO-CDim 1 in pure water at room temperature was UV irradiated at 365 nm, it switched from its trans to its cis form resulting in a marked change in the UV–vis spectra. As the irradiation continued, the absorption band at around 320 nm gradually decreased while the bands at 420 nm and 250 nm slightly increased.
This change is clearly due to the simple, but partial, isomerization of the azo groups from the trans photoisomer to the cis photoisomer [21]. The maximum isomerization yield was obtained after 120 min of irradiation at 365 nm.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Overlaid UV spectra of the irradiation of AZO-CDim 1 (a) from 0 to 120 min at 365 nm and then (b) from 0 to 90 min at 254 nm; (c = 10−4 M, water, 6 W lamp).[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
The reaction is reversible and when irradiated at 254 nm (Fig. 1), the cis isomer of AZO-CDim 1 gradually returned to its trans form, and the maximum isomerization yield was obtained after 90 min of irradiation. Both isomers could be separated by HPLC (Dionex, H2O/MeCN 90:10) and the cis/trans ratio of AZO-CDim 1 before irradiation (7:93) and after 2 h of irradiation (37:63) at 365 nm was determined (Fig. 2). Although each isomer could not be obtained in pure form, as is often the case for many azoderivatives [32], the isomerization efficiency is better than the cis/trans ratio of 14:86 after irradiation described by Liu [27].[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
HPLC quantification of the cis/trans ratio of AZO-CDim 1 before irradiation (left) and after irradiation at 365 nm (2 h, 6 W lamp) (right).[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Both isomers of AZO-CDim 1 have an appreciable resistance to fatigue thus the irradiation cycle could be carried out several times without causing side effects, as shown in Fig. 3.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Percentage of cis isomer of AZO-CDim 1 produced during photoisomerization cycles (c = 10−4 M, water). A cycle consists of irradiation at 365 nm for 2 h followed by irradiation at 254 nm for 1 h.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Ab initio calculations were also performed but the cis/trans transition was not observed since molecular modeling methods are unable to break bonds. In order to collect data on this phenomenon, the two configurations of the system had to be taken into account separately. Thus, ab initio calculations were performed on the two configurations of the azobenzene linker, the so-called 4,4’-bis(N-methylcarboxamide)azobenzene linker, to determine the structures of minimal energy (Fig. 4 and Fig. 4). Once optimized, the measured C4–C4’ distances were 9.1 and 6.6 Å for the trans and cis configurations, respectively, and the C–N=N–C dihedral angles were 180° and −10°, respectively. These results are comparable to those obtained by Koshima et al.
[33–34] on crystal structures where intermolecular packing effects might be important. From these calculations, the geometrical force field parameters needed for molecular dynamics simulations were derived.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Representation of the most stable structures obtained for the azobenzene linker (a) for the trans configuration and (b) for the cis one and for the AZO-CDim 1 systems (c) for the trans configuration and (d) for the cis one.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Molecular dynamics simulations performed on the two configurations of AZO-CDim 1 highlighted the rigidity of the linker, which governs the relative position of the two CD cavities. The trajectories, corresponding to 50,000 snapshots, were clustered into thirty representative conformations. The most stable structure of these thirty representative conformations for each configuration is shown in Fig. 4 and Fig. 4. It should be kept in mind that although the linker is quite rigid, the two CD cavities can rotate and move around the azobenzene axis.
Throughout the simulations, the C4–C4’ distances and the C–N=N–C dihedral angles did not fluctuate much. The C–C average value was 8.9 ± 0.1 and 5.8 ± 0.3 Å and the C–N=N–C dihedral angle was 175.1 ± 4.8 and −6.3 ± 5.4° for the trans and cis configurations, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
In these conditions, using azobenzene as a linker between two β-CD can lead to a modulation of the inclusion properties, such as a cooperative effect. The cavity of each CD is available to form an inclusion complex with a hydrophobic guest molecule. Among these, adamantane is known to be an excellent guest for the β-CD cavity, with an association constant K a ranging from 2·104 M−1 to 4·104 M−1 [35]. This is because the adamantyl residue fits perfectly inside the β-CD cavity.
In this present work, we investigated how the affinity between the dimer of adamantane and switchable AZO-CDim 1 may be influenced by the cis/trans ratio of the host molecule. For this purpose, we synthesized the adamantyl dimer EDTA bis-1-aminoadamantyldiamide disodium salt, ADAdim 4, as described by Vasquez Tato et al. [36] (Fig. 5). These authors showed that the single interaction between one binding site of the ditopic guest ADAdim 4 and one binding site of a particular β-CD dimer, bearing a terephthalic acid linker, was independent of the number of binding sites, that is, no cooperative effect was observed and a supramolecular polymer was formed.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Structure of the ditopic guest ADAdim 4.[](https://www.ncbi.nlm.nih.gov/mesh/D003516)
ITC is one of the most interesting methods to characterize the interaction of CDs with guests in solution [37–38]. It enables the enthalpy, entropy and equilibrium constants involved in complexation processes to be determined in a single experiment. Moreover, the guest:host molar ratio (i.e., the stoichiometry of the complex) can be measured. First, enthalpies of dilution of the monotopic hosts β-CD and β-CD-NH2, the ditopic host AZO-CDim 1 and the guest ADAdim 4 were measured in separate experiments to determine the maximum concentration to use for the ITC experiments. Enthalpies of dilution of β-CD and β-CD-NH2 were negligible over a broad concentration range, whereas enthalpies of dilution of AZO-CDim 1 were negligible only for concentrations lower than 1 mM, which is in agreement with NMR data.
As observed by Vásquez Tato and coworkers [39], ADAdim 4 can be considered as a surfactant. However, by using a maximum concentration of 4 mM in the ITC experiment, the effect of any heat resulting from a deaggregation process can be avoided. First, the interactions between β-CD or β-CD-NH2 and ADAdim 4 (added to the CD solution) were studied by ITC. After integrating the heat signal as a function of the molar ratio between the guest and the host, the isotherm was fitted to the one-site binding model as shown in Fig. 6.
The average values for the thermodynamic parameters are given in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
Titration of (a) β-CD (c = 0.8 mM) and (b) β-CD-NH2 (c = 0.8 mM) by ADAdim 4 (c = 4 mM). (c) Dilution of ADAdim 4 (c = 0.6 mM) in water at 298 K. Titration of AZO-CDim 1 (c = 0.1 mM) by ADAdim 4 (c = 0.6 mM) (d) before irradiation at 298 K and (e) after irradiation in water at 365 nm for 120 min at 278 K.[](https://www.ncbi.nlm.nih.gov/mesh/C031215)
Thermodynamic parameters deduced from ITC experiments for the different host–guest systems studied (temperature 298 K in water).[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
n a: guest:host molar ratio in the complex. The values in parentheses are theoretical values for the formation of complexes with all the binding sites of both guest and host occupied. bValues corresponding to the model of one set of sites. cValues corresponding to the model of two sets of sites. dValues corresponding to the two sets of sites model after irradiation at 365 nm, for 120 min at 278 K. The units of the experimental enthalpy value are kJ·mol−1 of the titrating species (i.e. the guest species).
For β-CD and β-CD-NH2, the experimental n values (0.50 and 0.43, respectively) for the complexes correspond to one ADAdim 4 for two cyclodextrins (stoichiometry 1:2). Considering the slightly different experimental conditions, all the data obtained are in very good agreement with the literature [36,39]. The same study was performed with AZO-CDim 1 exhibiting a cis/trans ratio of 7:93 and the experimental values are presented in Table 1. As depicted in Fig. 6, two jumps can be observed in the calorimetric titration curves revealing two independent interactions.
The experimental curve was well-fit to the two-sites binding model and the following results were achieved: the first jump corresponds to a very strong interaction between a small fraction of AZO-CDim 1 and ADAdim 4 with n = 0.13 and K a = 8·108 M−1. The second jump corresponds to a weaker interaction between a major fraction of AZO-CDim 1 and ADAdim 4 with n = 0.77 and K a = 8·105 M−1. Assuming a 1:1 stoichiometry in both cases, the first jump involves 14% of the mixture and the second 86% in relatively good agreement with the cis/trans ratio measured by HPLC (7:93). It is therefore tempting to assign the first jump to the complexation between ADAdim 4 and the cis isomer of AZO-CDim 1 and the second jump to the complexation between ADAdim 4 and the trans isomer of AZO-CDim 1.[](https://www.ncbi.nlm.nih.gov/mesh/C031215)
To confirm this hypothesis, the same ITC study was performed after UV irradiation at 365 nm for 120 min (Fig. 6). To improve the stability of cis-AZO-CDim 1, the titration was carried out at 278 K under the same experimental conditions as previously performed. The average values for the thermodynamic parameters are summarized in Table 1. Again, two jumps can be observed in the enthalpogram (Fig. 6), the stoichiometry of which has been shifted due to the UV irradiation. Neither the 20 °C difference nor the changing stoichiometry is thought to greatly affect the formation constants of the complexes.
We therefore attempted a fit of both experimental curves with the same values for association constants and enthalpies. The following parameter values afforded a reasonable fit: K 1 = 8·108 L·mol−1, ΔH 1 = −95 kJ·mol−1, K 2 = 8·105 L·mol−1, ΔH 2 = −67 kJ·mol−1, with stoichiometries n 1 = 0.26 and n 2 = 0.42 (vs experimental stoichiometries n 1 = 0.27 and n 2 = 0.44) after irradiation. Assuming a 1:1 stoichiometry in both cases, the first jump involves 38% of the mixture and the second 62%, again matching the cis/trans ratio measured by HPLC (37:63). This unambiguously proves that the two jumps detected by ITC correspond to the complexation of both isomers of AZO-CDim 1.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Interestingly, the association constants measured for the ditopic host (about 109 M−1 for the cis isomer and 106 M−1 for the trans isomer) are orders of magnitude larger than the association constant for the monotopic β-CD (K a = 5·104 M−1). This means that the complexation is highly cooperative, particularly in the case of the cis isomer. Although it was not demonstrated, such an additional interaction could explain why the association constant between the trans isomer of the ditopic host AZO-CDim 1 and the ditopic guest ADAdim 4 is significantly larger than that between β-CD and ADAdim 4. This, in turn, hints at particularly well matched conformations, as shown by the molecular simulation (see below).[](https://www.ncbi.nlm.nih.gov/mesh/C031215)
The 1H NMR spectra of AZO-CDim 1 with a cis/trans ratio of 7:93 were obtained in the absence or presence of an equimolar concentration of ADAdim 4 (Fig. 7). As previously stated, each cis and trans isomer could not be isolated in pure form, complicating the NMR study. In the presence of the ditopic guest ADAdim 4, a strong broadening of all signals was observed, indicating the presence of large objects in solution. A ROESY experiment was also carried out and although the presence of cross-correlation peaks between protons of AZO-CDim 1 and ADAdim 4 supports an inclusion complex, the strong signal broadening impeded any assignment (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
(a) 1H NMR spectra of AZO-CDim 1 (500 MHz, D2O, 2.5 mM) in the absence (bottom) and presence of ADAdim 4 (2.5 mM, top); (b) DOSY spectra of AZO-CDim 1 (red), ADAdim 4 (blue) and an equimolecular mixture of both (purple), (1 mM, D2O).[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Diffusion-ordered spectroscopy (DOSY) is a solution-based NMR method used to discriminate signals arising from different species by their diffusion rates. This method is very helpful and convenient for characterizing molecular aggregates or inclusion complexes [40]. The diffusion coefficient (D) is directly related to the molecular mass of the observed species in solution. In other words, when the molecular mass increases, the diffusion rate decreases. The DOSY spectra for ADAdim 4, AZO-CDim 1, and an equimolecular mixture of both were recorded in D2O (Fig. 7) and the measured D values were 3.540·10−10 m2·s−1, 2.293·10−10 m2·s−1 and 1.33·10−10 m2·s−1, respectively.
The D value of ADAdim 4 is smaller than that of AZO-CDim 1, in accordance with their molecular masses. It is usually assumed that the D value of an inclusion complex is the same as that of the host molecule alone [40], however, this was not observed in our case, which is in agreement with the formation of larger objects in solution.[](https://www.ncbi.nlm.nih.gov/mesh/D003516)
The experimental data strongly suggests that AZO-CDim 1 is a switchable host which forms two different inclusion complexes with this ditopic guest. The structural analyses of the molecular dynamics trajectories of the two configurations of the AZO-CDim 1 systems enable us to draw some conclusions as to how the adamantyl units could be contained in one or both cavities. To further support our assumptions, the corresponding systems were built and minimized. The following main conclusions can be drawn. For the first, in the cis configuration, both cavities of the ditopic host AZO-CDim 1 are available for complexation and their orientation favors the simultaneous inclusion of both adamantyl units of ADAdim 4, forming a 1:1 chelate-type complex depicted in Fig.
8. The chelate effect has been extensively studied by Breslow et al. [41–42] among others and a higher stability constant is expected due to the strong cooperative effect. Regarding the second main conclusion, the size and rigidity of the linker in AZO-CDim 1 prevent the trans configuration from forming ditopic 1:1 complexes upon complexation with only one molecule of ADAdim 4. Nevertheless, the two cavities remain available for complex formation through their wider rim with two adamantyl units belonging to two different ADAdim 4 molecules, leading to the formation of supramolecular polymers with an n:n stoichiometry.
This situation has already been encountered in the complex of ADAdim 4 and a β-CD dimer bearing a terephthalic moiety as the linker [36]. At this stage, based on the molecular dynamics study, at least two supramolecular polymers can be considered: the first is linear, as often described in the literature [43–44] (Fig. 8), and the second is cyclic (Fig. 8). Furthermore, it is possible that such linear or cyclic polymers are aggregated into larger objects stabilized by hydrogen bonds between cyclodextrin moieties and by π-stacking interactions between azobenzene linkers.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Proposed structures of inclusion complexes with the ditopic host AZO-CDim 1 and the ditopic guest ADAdim 4 after minimization by molecular modeling methods: (a) 1:1 chelate-type complex with AZO-CDim 1 in its cis configuration; linear (b) and cyclic (c) supramolecular polymers with AZO-CDim 1 in its trans configuration.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
Finally, according to the computational data, the complexes of AZO-CDim 1 with ADAdim 4 are further stabilized by intramolecular interactions between CD subunits that are more favorable in the cis than in the trans configurational arrangement, which is in agreement with the ITC data.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
## Experimental
*In the **Experimental** section:*
## Materials and methods
*In the **Materials and methods** section:*
All solvents were used as purchased, unless otherwise noted. All starting materials were used without purification. β-CD was purchased from Roquette Frères (Lestrem, France) and β-CD-NH2 was synthesized as previously described [31] or purchased from Biocydex (Poitiers, France). Analytical TLC was performed using Silica Gel 60 F254 plates (Merck, Germany). Eluents were mixtures of dichloromethane/methanol or cyclohexane/ethyl acetate.
Ratios are specified in each case in the experimental section. Products were illuminated under UV light (λ = 254 nm) followed by charring with vanillin/H2SO4.[](https://www.ncbi.nlm.nih.gov/mesh/C031215)
UV analyses were performed on a UV–vis Cary VARIAN spectrophotometer coupled with an optic fiber. A 6 Watt mercury lamp was used (λ = 365 nm) for the irradiation of aqueous AZO-CDim 1 solutions (c = 10−4 M).[](https://www.ncbi.nlm.nih.gov/mesh/D008628)
Stepwise control of the reactions was readily achieved by ESIMS in the positive ion mode using a ZQ 4000 quadrupole mass spectrometer (Waters-Micromass, Manchester, UK). High resolution electrospray mass spectra (HRMS–ESI) operated in the positive ion mode were obtained on a Q-TOF Ultima Global instrument (Waters-Micromass, Manchester, UK). Data acquisition and processing were performed with MassLynx 4.0 software. High resolution mass spectra were recorded in the positive mode on a ZabSpec TOF (Micromass, UK) tandem hybrid mass spectrometer with EBETOF geometry. The compounds were individually dissolved in a 1:1 water/CH3CN mixture at a concentration of 10 μg·cm−3 and then infused into the electrospray ion source at a flow rate of 10 mm3·min−1 at 333 K.
The mass spectrometer was operated at 4 kV while scanning the magnet over a typical range of 4000–100 Da. The mass spectra were collected as a continuum profile data. Accurate mass measurement was achieved using polyethylene glycol as the internal reference mass with a resolving power set to a minimum of 10,000 (10% valley).[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
NMR experiments were performed at 300.13 and 600.13 MHz using Bruker AVANCE DPX300 and AVANCE 600 spectrometers equipped with a Z-gradient unit for pulsed-field gradient spectroscopy. Me4Si was used as an external standard and calibration was performed using the signal of the residual protons or of the carbon of the solvents as a secondary reference. Measurements were performed at 300 K with careful temperature regulation. The length of the 90° pulse was approximately 7 μs. 1D NMR data spectra were collected using 16K data points. 2D experiments were run using 1K data points and 512 time increments.
The phase-sensitive (TTPI) sequence was used and processing resulted in a 1K·1K (real-real) matrix. The DOSY experiments were performed using the ledbpgp2s sequence from the Bruker library, with stimulated echo, longitudinal eddy current compensation, bipolar gradient pulses and two spoil gradients using 16 different gradient values varying from 2 to 95% of the maximum gradient strength. A 100 ms diffusion time and a 2.2 ms gradient length were used.[](https://www.ncbi.nlm.nih.gov/mesh/C073196)
Isothermal Titration Calorimetry (ITC) was performed using a VP-ITC microcalorimeter at 298 K or 278 K in pure water. Briefly, titration was carried out with 60 injections of 5 µL every 6 min. Control experiments were performed by the dilution of the guest solution in water and showed small heats of dilution. Thus, these results were subtracted from each titration to remove guest heats of dilution. The experimental data were fitted to a theoretical titration curve using Origin 7.0, Microcal software with the one set or two sets of sites models. During this fitting, enthalpy (ΔH), stoichiometry (n) and association constants (K a) were adjustable parameters.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Molecular modeling
*In the **Molecular modeling** section:*
Initial geometries of the β-cyclodextrin dimers studied in this work were built using the LEaP program from the AmberTools 1.4 distribution, following the strategy and methodology previously established [45]. Except for the linker, the CD fragments were taken from the R.E.DD.B. database [46] under project F-85 (http://q4md-forcefieldtools.org/REDDB/). Both the linker fragment and the cation were defined and parameterized according to the strategy previously developed using the RED program [47] along with the RED server [48].[](https://www.ncbi.nlm.nih.gov/mesh/C031215)
Molecular dynamics (MD) simulations were performed using the SANDER module of the AMBER10 program suite to perform MD simulations on the aforementioned complexes [49]. The systems were solvated in a truncated octahedral box with a buffer distance of 10.0 Å. The q4md-CD force field parameters were used to model the β-CD systems [45]. The parameters used for water were taken from the TIP3P model [50]. Classic MD simulations of 50 ns were then performed using the NPT ensemble at a pressure of 1 atm and a temperature of 300 K. In order to obtain representative ensembles of conformations for the two bis-CD systems, molecular configurations from MD trajectories were clustered.[](https://www.ncbi.nlm.nih.gov/mesh/C031215)
Ab initio calculations were performed with the Gaussian 09 program [50] to perform quantum chemical calculations. The structures corresponding to the two configurations of the linker were optimized at the B3LYP level of theory using the 6-31+G* basis set.
## Synthesis
*In the **Synthesis** section:*
4,4’-Dicarboxyazobenzene (2): p-nitrobenzoic acid (7.72 g, 46.2 mmol, 1 equiv) and sodium hydroxide (14.24 g, 356 mmol, 7.8 equiv) in 100 mL of water were heated at 338 K for 1 h. Then, 120 mL of 60 wt % glucose solution in water was added dropwise in two portions separated by 1 h of stirring. The mixture was heated at 353 K for 3 h and dropped into a large crystallizer where a precipitate appeared on the surface after several hours. The solid was filtered, dissolved in hot water, acidified with 100 mL of acetic acid, the precipitate filtered again and dried under reduced pressure. 4,4’-Dicarboxyazobenzene (2) was obtained as a pink solid (m = 6.11 g) with a yield of 49%.
The analyses are in full agreement with the literature [30]. Mp > 523 K (dec); 1H NMR (DMSO-d 6, 300.13 MHz) δ 8.16 (d, J = 8.4 Hz, 4H), 8.01 (d, J = 8.4 Hz, 4H) ppm; 13C NMR (DMSO-d 6, 75.77 MHz) δ 166.9, 154.3, 133.7, 130.9, 123.0 ppm; ESIMS (m/z): [M − H]− calcd for C14H9N2O4, 269.1; found, 268.9.[](https://www.ncbi.nlm.nih.gov/mesh/D001391)
4,4’-Dicarboxyazobenzene bis(N-hydroxysuccinimide ester) (3): Compound 2 (1.0 g, 3.70 mmol, 1 equiv), N-hydroxysuccinimide (1.87 g, 16.28 mmol, 4.4 equiv) and DMAP (90 mg, 0.74 mmol, 0.2 equiv) were dissolved in 10 mL of DMF at room temperature. After 10 min of stirring, EDCI (2.13 g, 11.1 mmol, 3 equiv) was added, then the solution was stirred at room temperature for 16 h under an inert atmosphere. The mixture was extracted by 100 mL of DCM and 100 mL of HCl (0.1 M). The aqueous phase was extracted twice with 50 mL of DCM.
Organic phases were combined, dried over Na2SO4 and purified over a plug-in of silica with DCM/MeOH (99:1 v/v) as eluent to obtain 3 as a red solid (m = 1.41 g) with a yield of 83%. Mp > 523 K (dec); 1H NMR (CDCl3, 300.13 MHz) δ 8.33 (d, J = 8.4 Hz, 4H), 8.07 (d, J = 8.4 Hz, 4H), 2.95 (s, 8H) ppm; 13C NMR (CDCl3, 75.77 MHz) δ 169.3, 161.4, 155.9, 132.0, 127.7, 123.6, 25.9 ppm; HRMS–ESI (m/z): [M + Na]+ calcd for C22H16N4O8Na, 487.0866; found, 487.0876.[](https://www.ncbi.nlm.nih.gov/mesh/D001391)
N,N’-Bis[6I-deoxy-β-cyclodextrin-6I-yl]carboxamide-4,4’-azobenzene, AZO-CDim (1): β-CD-NH2 (2.02 g, 1.78 mmol, 2 equiv) and 3 (404 mg, 0.87 mmol, 1 equiv) were dissolved in 5 mL of dried distilled DMF. After 16 h of stirring at room temperature, the mixture was concentrated and the product precipitated by addition of acetone, then dried under reduced pressure. The product 1 was obtained as an orange powder (m = 4.44 g) with a quantitative yield and an HPLC purity over 98%. Then, the product was purified by flash chromatography (20 min, H2O/MeOH from 90:10 to 10:90 (v/v), 40 mL·min−1) to afford compound 1 (2.75 g, 62%). Mp 423 K (dec); 1H NMR (D2O, 600.13 MHz)
δ 7.94 (d, 3 J H10–H11 = 8.1 Hz, H10, 4H), 7.90 (d, 3 J H11–H10 = 8.1 Hz, H11, 4H), 4.97–5.17 (m, H1 I–VII, 14H), 3.14–4.25 (m, H2 I–VII–H3 I–VII–H4 I–VII–H5 I–VII-H6 I–VII–H6 I–VII, 84H) ppm; 13C NMR (D2O, 150.76 MHz) δ 168.19 (C=O), 153.56 (C12), 135.35 (C9), 128.21 (C10), 122.94 (C11), 101.28–101.93 (C1 I–VII), 83.69 (C4 I), 80.26–81.05 (C4 II–VII), 71.63–73.54 (C2 I–VII–C3 I–VII–C5 II–VII), 70.44 (C5 I), 59.16–60.42 (C6 II–VII), 41.15 (C6 I); HRMS–ESI (m/z): [M + Na]+ calcd for C98H148N4O70Na, 2523.8042; found, 2523.8125.[](https://www.ncbi.nlm.nih.gov/mesh/C009850)
EDTA bis-1-adamantanylamine disodium salt, ADAdim (4): Adamantine (1.01 g, 6.68 mmol, 2.1 equiv) was dissolved in 30 mL of dried DMF and 10 mL of Et3N. The mixture was cooled to 273 K under an inert atmosphere and EDTA anhydride (0.82 g, 3.20 mmol, 1 equiv) was added portionwise. After 16 h of stirring under inert atmosphere, the solvent was removed under vacuum, and 10 mL of water was added and the solution was neutralized by HCl. The precipitate was washed with water, dried under reduced pressure then recrystallized in MeOH to obtain the diacidic compound. The diacid (814 mg, 1.46 mmol, 1 equiv) was suspended in water (10 mL) and NaOH (116 mg, 2.90 mmol, 2 equiv) was added.
The mixture was sonicated for 10 min and the product precipitated by addition of 100 mL of acetone. The solid was filtered, washed with acetone and dried under reduced pressure to obtain ADAdim 4 as a white powder (m = 420 mg), with a yield of 52% over the two steps. The analyses are in full agreement with the literature [36]. Mp 505–506 K; 1H NMR (DMSO-d 6, 300.13 MHz)
δ 7.47 (NH, 2H), 3.37 (Hf, 4H), 3.15 (Hd, 4H), 2.72 (He, 4H), 1.99 (Hb, 6H), 1.91 (Ha, 12H), 1.60 (Hc, 12H); 13C NMR (DMSO-d 6, 75.77 MHz) δ 172.5, 169.3, 58.8, 55.9, 52.4, 50.8, 41.1, 36.1, 29.0; HRMS–ESI (m/z): [M + Na]+ calcd for C30H46N4O6Na, 581.3315; found, 581.3299.[](https://www.ncbi.nlm.nih.gov/mesh/D003516)
|
# Introduction
Inclusion of [trans-resveratrol](https://www.ncbi.nlm.nih.gov/mesh/C059514) in methylated [cyclodextrins](https://www.ncbi.nlm.nih.gov/mesh/D003505): synthesis and solid-state structures
# Abstract
*In the **Abstract** section:*
Summary
The phytoalexin trans-resveratrol, 5-[(1E)-2-(4-hydroxyphenyl)ethenyl]-1,3-benzenediol, is a well-known, pote[nt antioxid](https://www.ncbi.nlm.nih.gov/mesh/C011991)a[nt having a varie](https://www.ncbi.nlm.nih.gov/mesh/C059514)ty[ of possible biomedical applications. However, its ](https://www.ncbi.nlm.nih.gov/mesh/C059514)adverse physicochemical properties (low stability, poor aqueous solubility) limit such applications and its inclusion in cyclodextrins (CDs) has potential for addressing these shortcomings. Here, various methods of the attempt[ed synthesis ](https://www.ncbi.nlm.nih.gov/mesh/D003505)of[ in](https://www.ncbi.nlm.nih.gov/mesh/D003505)clusion complexes between trans-resveratrol and three methylated cyclodextrins (permethylated α-CD, permethylated β-CD and 2,6-dime[thylated β-CD) ar](https://www.ncbi.nlm.nih.gov/mesh/C059514)e described. Isolation[ of the corre](https://www.ncbi.nlm.nih.gov/mesh/D003505)sp[onding crystalline](https://www.ncbi.nlm.nih.gov/mesh/D047391) 1[:1 inclusion compo](https://www.ncbi.nlm.nih.gov/mesh/D047392)unds [enabled their full st](https://www.ncbi.nlm.nih.gov/mesh/D047392)ructure determination by X-ray analysis for the first time, revealing a variety of guest inclusion modes and unique supramolecular crystal packing motifs. The three crystalline inclusion complexes were also fully characterized by thermal analysis (hot stage microscopy, thermogravimetric analysis and differential scanning calorimetry). To complement the solid-state data, phase-solubility studies were conducted using a series of CDs (native and variously derivatised) to establish their effect on the aqueous solubility of trans-resve[rat](https://www.ncbi.nlm.nih.gov/mesh/D003505)rol and to estimate association constants for complex formation.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
The naturally occurring phytoalexin trans-resveratrol (5-[(1E)-2-(4-hydroxyphenyl)ethenyl]-1,3-benzenediol; trans-3,5,4′-trihydroxystilbene, RSV) (1, Fig. 1), is a triphenolic species which is known to have potent antioxidant activity and consequently a wide range of pharmacological activities [1–2]. In recent years the list of potential medicinal benefits exhibited by RSV (including, e.g., anti-inflammatory effects, cardiovascular protection, and anticancer activity [3]) has increased considerably. Its low aqueous solubility, however, is one of the factors that limits its utility [4] and various methods have been employed to address this shortcoming [5], among them inclusion complexation with cyclodextrins (CDs), which are well-known solubilisers of lipophilic molecules
[6].[](https://www.ncbi.nlm.nih.gov/mesh/C011991)
Chemical structure of trans-resveratrol (1).[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
In addition to enhancing the solubility of guest molecules, CDs can confer chemical stability on bioactive molecules through inclusion of sensitive guest moieties within their hydrophobic cavities [6].[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
There have been numerous reports of the enhancement in the aqueous solubility of RSV as a result of its inclusion in CDs, most of them based on phase-solubility studies, e.g., [7–8]. In addition, NMR spectroscopic studies have yielded information on solution-state complexation (stoichiometries, association constants) [9]. The latter study was complemented by attempts to characterize putative solid inclusion complexes between CDs and RSV using thermoanalytical, Fourier-transform Infrared (FTIR) spectroscopic and powder X-ray diffraction (PXRD) techniques [9]. In the case of the interaction between α-CD and RSV, for example, the disappearance of the melting endotherm for RSV in the differential scanning calorimetric (DSC) trace of the product was cited as evidence for the formation of a α-CD·RSV complex [9].
The putative inclusion complex between β-CD and RSV, prepared by either the suspension method or using microwave irradiation, yielded highly amorphous products, evident from their PXRD traces [9]; in this case, the absence of characteristic peaks for RSV and the reduction in the degree of crystallinity of the product were considered as indirect proof of complexation. It should be noted that, in general, solid-state characterization using the latter techniques is limited, the evidence for genuine inclusion complex formation not always being definitive because the preparative method may result in one or both components becoming amorphous, or an unexpected solid phase (e.g., a hydrate of the guest compound) might be generated during attempted complexation. Loss of crystallinity of CD inclusion compounds also results when they dehydrate, rendering their PXRD traces less informative.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
There is, hitherto, a distinct lack of information on the structural nature of solid-state inclusion complexes between RSV and CDs, despite the fact that such complexes have strong potential for incorporation into tablets or capsules when formulated for medicinal use. A search of the Cambridge Crystallographic Database [10] revealed that no CD·RSV crystal structures have been reported to date.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
In this study, various preparative methods were explored in an attempt to generate CD·RSV inclusion complexes with a series of methylated CDs (permethylated α-CD, permethylated β-CD and 2,6-dimethylated β-CD). Here, CD–RSV interaction products were prepared by physical mixing, kneading or co-crystallization from different solutions, by co-evaporation using a rotavapor, or by exposure to microwave radiation. Characterization of the products was achieved using DSC and simultaneous thermogravimetric analysis (TGA/DSC), with support from FTIR spectroscopy and PXRD, where necessary.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
An important goal was the isolation of RSV·CD inclusion complexes in crystalline form so that definitive details of the mode of inclusion of the RSV molecule and the packing of complex units could be established by single crystal X-ray analysis. In view of its extended shape and apparent rigidity, the RSV molecule was expected to be partially inserted in the CD host cavities and hence to produce somewhat different supramolecular arrangements in the crystals from those observed with guest molecules having greater conformational freedom. As reported in detail below, the successful isolation of the target inclusion compounds as single crystals enabled their complete structural elucidation, revealing several novel supramolecular features which are relevant for future studies of the antioxidant RSV. Availability of well-defined, crystalline inclusion complexes also ensured that their characterization using thermoanalytical methods could be interpreted on a sound basis. Since a primary application of CDs is enhancement of the solubility of poorly soluble guest molecules [6], phase solubility studies [11] were conducted, and the results, including estimates of complex formation constants, are also reported here.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
## Results and Discussion
*In the **Results and Discussion** section:*
## Screening for CD–RSV interactions and product characterization
*In the **Screening for CD–RSV interactions and product characterization** section:*
Screening for new solid forms of RSV via its interaction with CDs using different preparative methods required preliminary characterization of RSV itself and assessment of the effects on pure RSV of procedures used to prepare the binary systems. Differential scanning calorimetry (DSC) indicated that the commercial product melted at T peak,m = 266.3(4) °C (T onset = 265.1(3) °C; ΔH m = 279(2) J g−1) (Fig. 2, curve (a)). Thermogravimetric analysis (TGA) revealed mass loss only at 275 °C attributable to sample decomposition (curve not shown). It was ascertained that the kneading treatment (KN) and exposure to irradiation with microwave radiation (MW) had no significant effect on RSV.
(Further experimental data on this aspect and all other methods employed in this study are provided in Supporting Information File 1).[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
DSC traces of RSV (a), TMA (b), TMA–RSV physical mixture (PM) (c), TMA–RSV preparation by kneading (KN) (d).[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
On DSC analysis, permethylated α-CD [(hexakis(2,3,6-tri-O-methyl)-α-CD; TRIMEA; TMA] yielded an endotherm of fusion only (T peak,m = 217.6(1) ° C, ΔH m = 40(3) J g−1) (Fig. 2, curve (b)). The physical mixture (PM) of TMA and RSV instead showed a new endothermic peak at ca.
175 °C, due to the melting of a new crystalline phase (curve (c)). The same endothermic peak was present in the KN (curve (d) (and MW, curve not shown) products, preceded by a small exothermic effect at 120 °C, confirming the TMA–RSV interaction and formation of a new thermally-induced solid phase. Comparison of FTIR spectra of the starting components with those of the binary systems showed that several bands shifted to significantly higher frequencies in the treated products, supporting the interpretation based on the thermal data.[](https://www.ncbi.nlm.nih.gov/mesh/D047391)
With the host TMB (heptakis(2,3,6-tri-O-methyl)-β-CD (TMB)), which displayed in DSC a sharp melting endotherm at T peak,m = 158.2(7) °C with ΔH m = 38(2) J g−1, the TMB–RSV combinations PM and MW yielded products with virtually featureless DSC traces, from which it was deduced that they were amorphous. Given that both the TMB and RSV samples employed were crystalline, with well-defined melting behaviour, it was interesting to note that even physical mixing appeared to yield a significantly amorphous product. (Powder X-ray diffraction of the PM sample confirmed its essentially amorphous nature, though a few prominent peaks due to RSV, of low intensity, were still evident above the general ‘halo’). It was therefore inferred that solid-state interaction had occurred to some extent on physical mixing. For the MW product (amorphous from the PXRD trace), no RSV was evident in the PXRD pattern and solid-state interaction between TMB and RSV was further confirmed from the FTIR spectrum, which showed several peaks displaced to slightly higher wavenumbers.[](https://www.ncbi.nlm.nih.gov/mesh/C000609658)
For the DMB (heptakis(2,6-di-O-methyl)-β-CD)–RSV binary combinations, an endotherm at T peak = 207.4(5) °C for the preparation PM reflected definite solid-state interaction, but the KN and MW products were effectively amorphous, based on the lack of distinct thermal events. An attempt to recrystallize the PM from MeOH/H2O (1:1 v/v) yielded a sample which displayed a distinct endo–exothermic effect in the DSC trace, attributed to inclusion complex formation. The FTIR spectrum of the KN product lacked two characteristic peaks of RSV, suggesting its inclusion in the cavity of DMB. It is noted that the DSC trace from ground single crystals of the phase later identified as the inclusion complex DMB·RSV·4H2O instead showed different features from those reported above for DMB–RSV combinations, the most prominent endotherm appearing at ca. 233 °C.
We infer that the nature of the inclusion complex formed depends on the preparative method employed.[](https://www.ncbi.nlm.nih.gov/mesh/C038119)
## Thermal characterization of crystalline CD·RSV inclusion complexes obtained by co-precipitation
*In the **Thermal characterization of crystalline CD·RSV inclusion complexes obtained by co-precipitation** section:*
The co-precipitation method using small amounts of ethanol to aid dissolution of the RSV produced high-quality single crystals of each of the three inclusion complexes. The host–guest stoichiometries of the inclusion complexes between RSV and the three methylated CDs were all found to be 1:1 from 1H NMR spectra of solutions of single crystals of the respective complexes (Supporting Information File 1).[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
TGA and DSC techniques were used primarily to estimate the water content and/or possible guest loss upon heating and to identify complex melting and other phase changes respectively, with hot stage microscopic (HSM) observations facilitating the interpretation of thermal events. Representative data are shown for the TMA·RSV complex (Fig. 3), where a TG mass loss of 7.5 ± 1.3% (n = 3) over the temperature range 30–100 °C yielded an estimated 6.6 ± 1.2 water molecules per 1:1 complex unit.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
TG (red) and DSC (blue) traces for the hydrated TMA·RSV complex (top), and hot stage micrographs showing the crystals at various temperatures (bottom).[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Water loss is evident in the HSM micrograph recorded at 112 °C with the crystal immersed in silicone oil and the DSC trace shows a corresponding broad endotherm accompanying the dehydration. However, a sharp endotherm subsequently developed, peaking at ca. 110 °C, interpreted as commencement of complex fusion which overlaps the dehydration process. This coincides with the melting observed in HSM at 120 °C. A phase change of the anhydrous complex is evident in the HSM at 136 °C, where microcrystallites appear within the melt, the small endotherm at ca.
145 °C being attributed to subsequent melting of the new phase. In HSM, the sample is completely molten at 177 °C. Finally, the TG trace indicates complex decomposition onset at ca. 280 °C.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
A summary of the results for the TMB·RSV and DMB·RSV complexes follows (for their TG, DSC and HSM figures, see Supporting Information File 1). The TG trace of the hydrated complex TMB·RSV yielded an initial mass loss of 5.3 ± 0.1% (n = 2), equivalent to 5.2 water molecules per 1:1 complex unit. The endotherm observed over the range of 30–120 °C appears sharper than expected for solvent loss alone, suggesting simultaneous melting of the complex. The HSM photographs confirm that dehydration is accompanied by complex fusion, the latter spanning a wide temperature range, with the sample fully molten at 120 °C. Complex decomposition commences at ca.
280 °C. In contrast to TMB·RSV, the thermal behaviour of DMB·RSV is distinctly more complicated (Supporting Information, File 1). The TG trace shows an initial mass loss of 4.4 ± 0.2 % (n = 3) over the range 30–110 °C, yielding 4.0 ± 0.2 water molecules per 1:1 complex unit. This loss is reflected in a broad endotherm recorded in the DSC over the same temperature range and is evident in the HSM images from fracturing of the crystal at 130 °C. Between 150 and 200 °C there is negligible mass loss and the anhydrous complex appears to undergo more than one phase transition.
A second mass loss appears in the TG trace corresponding to partial guest loss and the DSC shows a sharp but small melting endotherm at ca. 233 °C, the remaining sample decomposing soon after, at ca. 320 °C.[](https://www.ncbi.nlm.nih.gov/mesh/C000609658)
## X-ray analysis
*In the **X-ray analysis** section:*
Table 1 lists the crystal data, as well as data-collection and refinement parameters for the new hydrated inclusion complexes TMA·RSV, TMB·RSV and DMB·RSV. The remarkably low R 1 -factors (range 0.04–0.07) and the relatively small residual electron densities are exceptional for CD structures of this complexity, given also the presence of guest disorder in two cases. An account of the key features of the inclusion of the RSV molecule within the respective host cavities as well as descriptions of the crystal packing arrangements follows.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Crystal data, data collection parameters and refinement details.
The asymmetric unit of the complex TMA·RSV·6.25H2O, namely two TMA molecules, two RSV molecules and 12.5 water molecules, is shown in Fig. 4. In both 1:1 host–guest complex units the guest phenyl ring bearing one phenolic group (the 4-hydroxyphenyl residue) is fully immersed in the host cavity, being located at the primary side, while the ring bearing two phenolic groups (the 1,3-benzenediol residue) protrudes significantly from the host secondary side, where its phenol groups engage in hydrogen bonding with water molecules. Crystallographic atomic nomenlature for the host is shown in Fig. 4.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
The two symmetry-independent complex units of TMA·RSV·6.25H2O (A and B), with only the major component of disorder shown for RSV in host B (a), and the non-H atom and methylglucose ring nomenclature illustrated for host A as representative (b). For clarity, host H atoms have been omitted.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
The full description of the guest molecules is provided in Fig. 5, where the ordered structure of guest molecule A is contrasted with the twofold-disordered model (components B, C) for the second guest molecule. Several of the host B atoms were disordered over two positions and were modelled accordingly. These included, on the primary side, two C6–O6–C9 chains, a methoxy group and an O5 atom, and on the secondary side, three methoxy groups. Full geometrical analyses that included nine metrical parameters describing the host molecule conformations was performed (Supporting Information, File 1).
This revealed that host molecules A and B adopt the expected elliptical shape [12], the longer axis of each macrocycle being approximately parallel to the planes of the respective included 4-hydroxyphenyl rings.
Representative atomic labelling for the ordered RSV molecule A (blue) present in host A and the two disorder components B (orange, s.o.f. = 0.56) and C (green, s.o.f. = 0.44) of the RSV molecule included in host molecule B.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
In addition, the crystallographically independent TMA host molecules adopt somewhat different conformations given the fact that their contents differ, owing to the disorder described in Fig. 5. In particular, the average extent of ‘tilt’ of each glucose ring relative to the mean O4-plane is small for host molecule A [range 3.24(3)–6.44(5)°], indicating a relatively open primary side, whereas for host B, the average tilt angle is significantly larger [range 5.72(4)–10.31(9)°], reflecting a more ‘closed’ primary side.[](https://www.ncbi.nlm.nih.gov/mesh/C121468)
Regarding the mode of guest inclusion, the angle between the mean plane of the RSV molecule and the mean O4-plane of the host molecule A is ca. 85.6°, with that between the RSV major disorder component B and the mean O4-plane of host molecule B being virtually the same (ca. 86.8°). While the RSV molecule in its own crystal structure ([10], refcode DALGON) is planar, it is notable that the RSV molecules in the TMA complex deviate significantly from planarity and to different extents; in the case of the ordered RSV guest molecule A, the interplanar angle between the two phenyl residues is 51.6(3)°, and for the major disorder component of RSV which is included in host molecule B, the corresponding angle is 23.1(4)°. Thus, the significant host conformational differences coupled with the significant guest conformational differences reflected in the parameters reported above clearly indicate a mutual induced fit when TMA forms an inclusion complex with RSV.
This phenomenon of mutual induced fit has recently been cited as a frequent occurrence in biological systems, but a rare one for synthetic host–guest systems [13]. However, its occurrence in CD inclusion complexes is known and was in recent years prominently manifested in CD complexes of rocuronium salts [14].[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Fig. 6 illustrates the three-component supramolecular systems A and B occurring in the crystal. Each consists of a TMA molecule, a RSV molecule and a decorative motif (here referred to as a ‘crown’) of three hydrogen bonded water molecules (H atoms not shown), the terminal water molecules forming hydrogen bonds with the phenolic groups. For the ordered RSV guest in complex A, for example, the four O···O distances are in the range of 2.700(6)–2.863(6) Å. It is noteworthy that the ‘crown’ feature is a robust motif, occurring in all three inclusion complexes described here. Furthermore, this motif is unique to the trans-resveratrol inclusion complexes described here: no analogous motifs were found on searching the Cambridge Structural Database
[10]. It is also important to note that for the TMA·RSV complex, the major stabilising host–guest interaction is that between the phenolic group of the 4-hydroxyphenyl ring and the primary rim of the host TMA molecule, which is mediated by a bridging water molecule. In ordered complex unit A, for example, the linkage is RSV(4-OH)···O(water)···O6(primary methoxy), with respective O···O distances of 2.731(6) and 2.829(7) Å.[](https://www.ncbi.nlm.nih.gov/mesh/C121468)
Space-filling representations of the two independent complex units A (a) and B (b) of the complex TMA·RSV·6.25H2O with a cutaway view of the host to illustrate the details of guest inclusion. For the RSV molecules, the atoms are colour coded blue (C), green (O) and yellow (H). For clarity, only the major RSV disorder component is shown in (b).[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
A complex network of hydrogen bonds stabilises the crystal structure; these include host–guest O–H···O and C–H···O hydrogen bonds, host–host C–H···O hydrogen bonds, guest–water and water–water O–H···O hydrogen bonds.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Crystal packing is shown in Fig. 7. The complex units pack in a head-to-tail manner in columns parallel to the crystal b-axis. Columns of complex units A propagate as rows parallel to the a-axis, alternating with analogous columns of B complex units.
Crystal packing for the complex TMA·RSV·6.25H2O projected down [010].[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Structural analysis of the inclusion complex between permethylated β-CD (TMB) and RSV, with formula TMB·RSV·5.6H2O, revealed twofold disorder of the RSV molecule. The symmetry of the disorder model (Fig. 8) is, however, clearly different from that in the TMA complex (Fig. 5) but the close proximity of the chemically equivalent phenolic groups of the A and B components in principle enables them to engage in similar hydrogen bonding schemes.[](https://www.ncbi.nlm.nih.gov/mesh/D047392)
The components of the disorder model for RSV in its inclusion complex with TMB (s.o.f. = 0.73 for the major component A (blue) and 0.27 for the minor component B (green)).[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
The conformational flexibility of the RSV molecule is again evident in this complex, the interplanar angles between the phenyl rings being 17.7(1)° for the major component and 23.9(3)° for the minor component, thus extending the range of guest conformational flexibility encountered in the TMA complex.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
The crystal asymmetric unit contains the equivalent of one RSV molecule, one TMB molecule and 5.6 water molecules (Fig. 9). The molecule of RSV is included within the TMB cavity with the 4-hydroxyphenyl group located at the host primary side, being anchored directly via a hydrogen bond [RSV(4-OH)···O611] to a partial oxygen atom (s.o.f. = 0.65) of a primary methoxy group. This differs from the situation in the TMA complex, where the host–guest link is mediated by a bridging water molecule.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
The asymmetric unit in the crystal of TMB·RSV·5.6H2O (a), and the non-H atom and methylglucose ring nomenclature illustrated for the host TMB (b). Only the major RSV disorder component is shown in (a) for clarity.[](https://www.ncbi.nlm.nih.gov/mesh/C000609658)
The major disorder component of the guest engages in a geometrically more favourable hydrogen bonding interaction, such that the O···O distance in O1A–H1A···O611 is 2.73(1) Å, whereas for the minor guest component, the corresponding O···O distance in O1B–H1B···O611 is 2.95(1) Å. The situation is slightly more complicated since each of the phenolic groups (–O1A–H1A and –O1B–H1B) engages in bifurcated H-bonding, the second acceptor being a disordered water oxygen atom O7W, located at distances 2.60(1) Å and 2.80(1) Å from O1A and O1B respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Another important feature of the inclusion geometry relates to the guest inclination in the host cavity: here the mean plane of the RSV molecule is inclined at ca. 45° to the mean O4-plane of the TMB molecule (Fig. 10), effectively resting on the surface of one side of the host molecule, in strong contrast to the situation in the TMA complex where the equivalent angle is ~86° (Fig. 4). As is usually observed, the primary methoxy groups of the host TMB are generally directed towards the centre of the macrocycle, and effectively close the primary side, presenting a bowl-shaped surface to the RSV molecule. Instead, the secondary side of the host molecule is open and a portion of the 1,3-benzenediol residue protrudes from that side, where the two phenolic groups are again linked by a ‘crown’ of three hydrogen bonded water molecules, analogous to that observed in the TMA complex.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Space-filling model of the inclusion complex TMB·RSV·5.6H2O showing the inclusion of the RSV molecule in the host TMB (left) and a cutaway view (right) emphasising the shallow inclination of the guest molecule in the cavity. Only the major guest disorder component is illustrated for clarity. Water hydrogen atoms are also omitted.[](https://www.ncbi.nlm.nih.gov/mesh/C000609658)
The higher quality of the diffraction data for the TMB·RSV complex enabled location of the hydrogen atoms of the water molecules in difference Fourier syntheses. Both disorder components of the RSV molecule engage in equivalent hydrogen bonds with the host molecule. Stabilisation of the crystal structure of TMB·RSV·5.6H2O is effected by a complex network of attractive interactions, including host–guest hydrogen bonds (both O–H···O and C–H···O), several host–host C–H···O interactions and numerous O–H···O hydrogen bonds that involve ordered and disordered water molecules (the 5.6 H2O molecules in the asymmetric unit being disordered over nine sites). The complex units stack in columns parallel to the a-axis in a head-to-tail fashion (Fig. 11), adjacent columns being related by the two-fold screw axis along 1/2, y, 1/2. The view down the columns (Fig. 11) reveals a channel-like arrangement of the host molecules in this direction.
Among the various isostructural classes of CD inclusion complexes [15], the one to which this complex belongs has relatively few members.[](https://www.ncbi.nlm.nih.gov/mesh/C000609658)
Packing arrangement in the crystal of TMB·RSV·5.6H2O viewed down [010] (a) and [100] (b). Hydrogen atoms have been omitted for clarity; water oxygen atoms in red.[](https://www.ncbi.nlm.nih.gov/mesh/C000609658)
The third complex whose X-ray structure is described here has the formula DMB·RSV·4.0H2O, the host molecule DMB being 2,6-dimethylated β-CD and consequently having properties that are intermediate between those of the native β-CD and fully methylated β-CD [16]. The formula unit corresponds to the crystal asymmetric unit, shown in Fig. 12. Despite the inclusion of the guest molecule, the DMB molecule retains its ‘round shape’ owing to the formation of the well-known ‘belt’ of intramolecular O2(n)···O3(n−1) hydrogen bonds that link contiguous glucose residues [17–18]. In this complex, the average O···O distance in the belt is 2.83 Å and the O–H···O angles span the range 165–173°.[](https://www.ncbi.nlm.nih.gov/mesh/C038119)
Structure of the host–guest complex DMB·RSV·4.0H2O (a), ring and atomic nomenclature for the host molecule DMB (b), and structure and atomic numbering of the included RSV molecule (c).[](https://www.ncbi.nlm.nih.gov/mesh/C038119)
As in the previous two complexes, the RSV molecule is again included with the 4-hydroxyphenyl ring located deep within the cavity with the phenolic group at the primary side, while the 1,3-benzenediol residue protrudes from the secondary rim of the DMB molecule and the two phenolic groups are again decorated by a ‘crown’ of three hydrogen bonded water molecules. In this complex, the included RSV molecule shows the highest degree of planarity, the phenyl ring planes intersecting at only 13.6(2)°. The topology of guest inclusion is shown in Fig. 13. The angle between the mean O4-plane of the DMB molecule and the mean plane through the RSV molecule is ca.
73°, intermediate between the corresponding values in the TMA and TMB complexes.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Space-filling model of the inclusion complex DMB·RSV·4.0H2O showing the encapsulation of part of the RSV molecule by the host DMB (left) and a cutaway view highlighting the location and orientation of the guest molecule in the host cavity (right).[](https://www.ncbi.nlm.nih.gov/mesh/C038119)
Closer examination of the binding of the 4-hydroxyphenyl ring to the host molecule reveals that its hydroxy group is linked to a methoxy oxygen atom (O6G7) on the primary rim of the host molecule via a bridging water molecule, the relevant hydrogen bond sequence being RSV(O1–H1)···O2W–H2WA···O6G7, with respective O···O distances 2.718(4) Å and 2.778(4) Å. The second hydrogen atom on the water molecule (H2WB) is in turn a donor to the atom O3G3i of a translated (i = −1 + x, y, z) DMB molecule, this hydrogen bond having a O2W···O3G3i distance of 2.857(3) Å and being responsible for cohesion between successive complex units along the crystal x-direction. Fig.
14 illustrates the principal hydrogen bonds associated with the two complex units referred to above.[](https://www.ncbi.nlm.nih.gov/mesh/D010100)
Stereoview of two DMB·RSV·4.0H2O complex units related by a unit translation along the crystal a-axis, illustrating the intramolecular hydrogen bonds which stabilise the host conformation as well as the hydrogen bonding role of the bridging water molecule that links complex units along the crystal x-direction.[](https://www.ncbi.nlm.nih.gov/mesh/C038119)
It is noteworthy that in the above motif, the two host molecules are fairly steeply inclined to the a-axis (which is approximately vertical) with the result that two primary methoxy groups of the uppermost molecule are partially included within the cavity of the translated molecule. In addition to the hydrogen bonds discussed above, the crystal structure of the DMB complex is stabilised by a series of host–host C–H··· O hydrogen bonds as well as numerous water–water O–H···O hydrogen bonds.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
The crystal packing is shown in Fig. 15. Complex units stack in columns parallel to the a-axis in a head-to-tail fashion with (as noted above) a small extent of host self-inclusion (Fig. 15). Fig. 15 illustrates the modified herringbone packing arrangement as viewed down the b-axis.
Projections of the crystal structure of the complex DMB·RSV·4H2O along [100] (a) and [010] (b). Hydrogen atoms are omitted for clarity.[](https://www.ncbi.nlm.nih.gov/mesh/C038119)
Regarding the phase purity of the three new inclusion complexes described above, we confirmed that their simulated powder X-ray diffraction patterns are in good agreement with those calculated from the single crystal X-ray data. This is an important verification that the single crystals selected are truly representative of the respective bulk materials (Supporting Information File 1).
## Phase-solubility analysis
*In the **Phase-solubility analysis** section:*
According to Higuchi and Connors [11], phase-solubility diagrams can be classified as being of types A and B. A-type behaviour corresponds to an increase in the solubility of the drug as the concentration of the CD is increased, as a result of soluble complex formation. A-type curves can further be distinguished depending on whether the solubility increases linearly (AL) as the CD concentration increases, or with a positive (AP-type) or negative (AN-type) deviation due to a change in the physical properties of the solution. B-type curves indicate the formation of an insoluble complex, where BS suggests the formation of a complex with limited solubility, while BI denotes the formation of an insoluble complex.[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
Fig. 16 shows the phase-solubility results for RSV with the native CDs β- and γ-CD. The phase-solubility profile resulting from the use of β-CD is of type AL and this host produces a guest solubility enhancement of 26-fold over the concentration range indicated. The results for the experiments with γ-CD were limited to a maximum CD concentration of 6 mM by inefficient filtration through the filter membrane that was employed. The precipitation of complex or aggregated CD particulates was physically observed during sample preparation.
Over this range the solubility plot appears to increase to a plateau, indicating an AN solubility profile, the negative deviation possibly being due to changes in the solubility of the complex and/or aggregation of the CD molecules. The solubility enhancement for RSV with γ-CD was only 3.4-fold.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Solubility of RSV as a function of [β-CD] (blue) and [γ-CD] (red) at 25 °C.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
The solubility enhancements for RSV in the presence of the derivatised CDs are significant (Fig. 17). With TMB, AL-type behaviour was observed with a solubility enhancement of 36 times that of the intrinsic solubility of the guest. Each of the remaining derivatised CDs shows two different solubility profiles over the common concentration range. Hydroxypropyl-β-CD (HP-β-CD) and randomly methylated β-CD (RMB) show relatively small initial solubility enhancements of RSV solubility (up to ca. 4 mM CD concentrations), with significant solubility increases thereafter (AL-type).
The changes in slope may indicate an increase in the complex order with respect to RSV. The solubility enhancement for RSV at the highest CD concentration employed is 44-fold in the presence of HP-β-CD and 63-fold in the presence of RMB.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Solubility of RSV as a function of the concentrations of TMB (light blue), DMB (red), HP-β-CD (green) and RMB (dark blue) at 25 °C.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
The results with DMB follow the opposite trend, with the solubility of the guest increasing linearly over the CD concentration range 0–8 mM, while above that concentration, the apparent solubility of RSV decreases. This is attributed to the formation of an insoluble complex, which removes RSV from the solution. The maximum solubility enhancement, occurring at a CD concentration of 8 mM is 45 times that of the intrinsic solubility of the guest.[](https://www.ncbi.nlm.nih.gov/mesh/C038119)
Values of the association constants for complex formation (K C) were estimated using the relationship (1) and the slopes of the recorded phase-solubility diagrams, assuming 1:1 host–guest complex formation [11].
Table 2 shows the approximate stability constants for complexation between each of the CDs investigated and RSV. Only the initial slopes were used to calculate K C (up to 6 mM for γ-CD, 8 mM for DMB and 4 mM for HP-β-CD and RMB).[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
The apparent stability constants (K C) for complexation between various CDs and RSV.[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
The values obtained indicate relatively weak interactions between RSV and γ-CD, initially weak interaction with HP-β-CD and RMB, fairly strong binding with β-CD and TMB, and the formation of a very stable complex between RSV and DMB up to the CD concentration of 8 mM.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Lu et al. [8] found a linear relationship between the concentration of RSV and the concentrations of both β-CD and HP-β-CD, reporting the derived K C values as 1815 M−1 for β-CD·RSV and 6778 M−1 for HP-β-CD·RSV. The conditions under which these experiments were carried out were slightly different from ours, however: an excess of RSV was added to 5 mL of CD solution; the solutions were shaken for 24 h and the suspensions were filtered with cellulose acetate. The results we obtained for the phase-solubility behaviour using β-CD are comparable to those obtained by Lu et al. [8], but the results obtained with HP-β-CD are quite different, probably due to the different preparation methods used.
However, a comparison of the data for the derivatised CDs TMB and DMB show a similarly strong interaction to that obtained by Lu et al. with HP-β-CD. The general trend, indicating that the derivatised CDs interact more strongly with RSV, is confirmed in the present study as well.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Regarding the reliability of the data based on this methodology when dealing with RSV solutions, we noted that the choice of filtration membrane in these studies can greatly affect the outcome of the experiment. We found that nylon filters remove RSV from the aqueous solution completely, allowing only some of the molecules which are protected by the CD to pass through the filter membrane. Of all the membranes tested the PTFE filters were found to give the most consistent results, although cellulose acetate was not tested in the present study.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
## Conclusion
*In the **Conclusion** section:*
A variety of methods (physical mixing, kneading, microwave irradiation) of effecting interaction between RSV and three CD hosts (TMA, DMB and TMB) was tested and a combination of thermal analysis and FTIR spectroscopy subsequently yielded evidence for the formation of interaction products, many of them amorphous in nature. For more definitive characterization of interaction products, crystalline inclusion complexes with the formulae TMA·RSV·6.25H2O, TMB·RSV·5.6H2O and DMB·RSV·4.0H2O were subsequently isolated using the co-precipitation method and fully characterized by thermal and single crystal X-ray diffraction methods. For the complexes containing the fully methylated hosts TMA and TMB, thermal analysis revealed dehydration overlapping with fusion of the anhydrous complex, followed by final decomposition, whereas the DMB complex displayed more intricate thermal events, namely dehydration followed by phase transitions and partial guest loss that preceded final decomposition.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
The X-ray studies reported here reveal, for the first time, the unique features of the mode of inclusion of the RSV molecule within CDs. The TMA complex contains two symmetry independent TMA·RSV complex units: in one of these, the guest is ordered while in the second the guest is disordered over two positions. This disorder was successfully modelled, as was the slightly modified twofold disorder of the RSV molecule in the TMB·RSV complex, and in all cases, the disorder never results in spatial interchange of the 4-hydroxyphenyl and 1,3-benzenediol units; instead, it brings the 4-hydroxyphenyl residues of the two disordered components into close proximity, and likewise the respective 1,3-benzenediol residues. DMB·RSV is the only complex in which there is no guest disorder.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
For all three hosts, complex formation involves insertion of the less sterically bulky 4-hydroxyphenyl ring of RSV deep within the CD cavity where it is located at the host primary side. In the TMA and DMB complexes, the phenolic group is linked to the host by hydrogen bonding of the type RSV(4-OH)···O(water)···O6(primary methoxy), whereas in the TMB complex, there is a direct host–guest linkage via a RSV(4-OH)···O6(primary methoxy) hydrogen bond. A common feature, however, is the significant extent of protrusion of the 1,3-benzenediol moiety from the secondary side of each of the three hosts, with the two phenolic groups being linked by a series of four hydrogen bonds RSV(1-OH)···O(water)···O(water)···O(water)···[HO(-3)-RSV]. This persistent supramolecular motif has not been observed previously in the solid state and since it may also exist in aqueous solution, it is worth consideration in molecular modelling studies that address CD-RSV interaction in that medium. Equally significant in the context of molecular modelling is our observation in the solid-state of the potentially more probable bridging role of water in mediating host–guest binding, this feature occurring in two of the three complex crystal structures investigated.
Finally, as far as further new insights from the X-ray studies are concerned, we conclude that CD-RSV inclusion in the more flexible hosts TMA and TMB involves a mutual induced fit. The evidence for this is the flexibility displayed by the RSV molecule, reflected in the wide range observed for the interplanar angle between the phenyl rings [17.7(1)–51.6(3)°] in the respective complex crystal structures, coupled with significant host distortions to accommodate the RSV molecule. In contrast, with the host DMB, whose round structure is maintained by intramolecular hydrogen bonds, the resulting unrestricted cavity volume enables the RSV molecule to be accommodated with very little adaptation, the interplanar angle between the phenyl rings being only 13.6(2)°.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Phase-solubility studies have been useful in confirming the generally higher solubility enhancements that derivatised CDs confer on RSV. Derived values of the association constants for 1:1 CD–RSV complexation in aqueous solution spanned the range of 410 M−1 for inclusion in γ-CD to 11 600 M−1 for inclusion in DMB.[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
## Experimental
*In the **Experimental** section:*
## Materials
*In the **Materials** section:*
The trans-resveratrol sample used for the preparation of binary mixtures was a generous gift from Denk Feinchemie GmbH (München, Germany). For co-precipitation experiments, the RSV used was supplied by Sigma-Aldrich (South Africa). Cyclodextrins were purchased from Wacker Chemie Italia Srl (Milan, Italy) and Cyclolab (Budapest, Hungary). All other materials and solvents used were of analytical reagent grade.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
## Preparation of the binary systems
*In the **Preparation of the binary systems** section:*
Each physical mixture (PM) (1:1 mol/mol) was prepared by gentle co-grinding of the powder components in a mortar with a pestle and passing the resultant material through a 250 μm sieve.
Kneaded products (KN) were prepared by wetting each PM in a mortar with ethanol/water 4:1 (v/v) and grinding thoroughly with a pestle, after which the product was dried to constant weight at 70 °C in an oven. The entire procedure was repeated in triplicate. The samples were then sieved through a 250 μm sieve.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
Co-evaporated products (CP) were prepared by dissolving each PM in the minimum amount of ethanol/water 4:1 (v/v) to obtain a clear solution. The solvent was removed using a rotavapor under reduced pressure at 80 °C, and the residue was gently ground in a mortar with a pestle, and passed through a 250 μm sieve.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
Microwave irradiation products (MP) were prepared by dissolving each PM in the minimum amount of ethanol/water 4:1 (v/v) to obtain a clear solution in a glass container, followed by microwave irradiation at 425 W (Pabish CM-Aquatronic) for a time sufficient to remove the solvent. The dried residue was gently ground in a mortar with a pestle, and passed through a 250 μm sieve.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## Differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA)
*In the **Differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA)** section:*
For the binary systems investigated, temperature and enthalpy values were measured with a Mettler STARe system (Mettler Toledo, Novate Milanese, MI, Italy) equipped with a DSC821e Module and an Intracooler device for sub-ambient temperature analysis (Julabo FT 900) on 2–4 mg (Mettler M3 Microbalance) samples in sealed aluminium pans with pierced lid [heating rate β = 10 K min−1, nitrogen atmosphere (flux 50 mL min−1), 30–350 °C temperature range)]. The instrument was previously calibrated with indium as standard reference. Measurements were carried out at least in triplicate. For co-precipitated, crystalline products, traces were recorded on a DSC-Q200 differential scanning calorimeter with samples in closed aluminium pans heated at 10 K min−1 and dry nitrogen purge gas flowing at 50 mL min−1. TG traces for these products were recorded on samples in alumina crucibles using a TA-Q500 instrument under similar conditions as for the DSC measurements.[](https://www.ncbi.nlm.nih.gov/mesh/D000535)
## Simultaneous thermogravimetric analysis (TGA/DSC)
*In the **Simultaneous thermogravimetric analysis (TGA/DSC)** section:*
Mass losses were recorded with a Mettler STARe system (Mettler Toledo, Novate Milanese, MI, Italy) TGA with simultaneous DSC (TGA/DSC1) on 4–6 mg samples in alumina crucibles with lid [β = 10 K min−1, nitrogen air atmosphere (flux 50 mL min−1), 30–350 °C temperature range]. The instrument was previously calibrated with indium as standard reference and measurements were carried out at least in triplicate.[](https://www.ncbi.nlm.nih.gov/mesh/D000537)
## Fourier transform infrared (FTIR) spectroscopy
*In the **Fourier transform infrared (FTIR) spectroscopy** section:*
Mid-IR (650–4000 cm−1) spectra were recorded on powder samples using a Spectrum One Perkin-Elmer FTIR spectrophotometer (resolution 4 cm−1) (Perkin Elmer, Wellesley, MA, USA) equipped with a MIRacleTM ATR device (Pike Technologies, Madison, WI, USA).
## Crystal preparation
*In the **Crystal preparation** section:*
trans-Resveratrol (20 mg) was dissolved in 0.5 mL of ethanol and was added to an equimolar amount of CD dissolved in water, according to Table 3 below. Turbid solutions were clarified by adding ethanol dropwise. Each solution was then filtered into a new vial, closed with a punctured polytop lid and was allowed to evaporate slowly on the benchtop or in an oven. The vial was sealed after crystals appeared.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
Masses of CDs employed, volumes of water added and temperatures at which complex crystals formed.[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
## X-ray diffraction analysis
*In the **X-ray diffraction analysis** section:*
All intensity data were collected on a Bruker KAPPA APEX II DUO diffractometer. In each case a single crystal was surface-dried, coated in paratone N oil (Exxon Chemical Co., TX, USA) and mounted on a cryoloop in a constant stream of nitrogen vapour (Oxford Cryostream, UK). Crystal systems and space groups for the CD complexes were deduced from the Laue symmetries and systematic absences, respectively. The structures were solved by direct methods (program SHELXD [19]) and refined by full-matrix least-squares (program SHELXH-97 [19]). In general, location of the host molecules from the E-map was followed by their refinement using isotropic thermal displacement parameters.
This was followed by location of the guest molecules from the resulting difference Fourier synthesis. Successive difference maps revealed the water molecules, which were modelled with appropriate site-occupancy factors (s.o.f.s) to reconcile the model with the thermogravimetric analytical data as far as possible. Disorder of the host and guest residues, where they occurred, were similarly treated using appropriate s.o.f.s. In the final cycles of refinement, anisotropic thermal displacement parameters were introduced for most or all of the non-H atoms. A large proportion of the H atoms were located in difference Fourier syntheses and were generally included in idealised positions in a riding model with U iso values in the range 1.2–1.5 times those of their parent atoms.
Further details of the refinements for the individual complexes appear in the Supporting Information File 1. CCDC 1020492–1020494 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge at http://www.ccdc.cam.ac.uk/products/csd/request/ [or from the Cambridge Crystallographic Data Centre (CCDC), 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 (0)1223-336033; email: deposit@ccdc.cam.ac.uk].[](https://www.ncbi.nlm.nih.gov/mesh/C008967)
## Phase-solubility analysis (cont.)
*In the **Phase-solubility analysis (cont.)** section:*
Phase-solubility studies were peformed according to the method described by Higuchi and Connors [11]. Six CDs [β-CD, γ-CD, TMB, DMB, HP-β-CD and RMB] were dissolved in water to yield solutions whose concentrations spanned the range 2.0–12.0 × 10−3 M. An excess of RSV (1.5–2.5 mg) was added to 2 mL of each CD solution and the solutions were allowed to stir at 25 ± 0.5 °C for 48 h. The solubility of RSV in the absence of CD (So) was determined by preparing solutions containing an excess of RSV in water, and stirring at 25 ± 0.5 °C for 48 h. Samples were subsequently filtered through 0.45 μm PTFE syringe filters and diluted appropriately. The concentration of RSV was determined using UV–vis spectrophotometry at a wavelength of 316 nm.
The UV spectra were recorded on a GCB Cintra 20 UV–vis spectrometer over a wavelength range of 200–500 nm at a scanning rate of 200 nm min−1. The extinction coefficient was determined for this wavelength by preparing a calibration curve of RSV in water. All measurements were recorded in triplicate. Each phase-solubility curve was prepared by plotting the concentration of RSV against the concentrations of the CD employed in the experiment.[](https://www.ncbi.nlm.nih.gov/mesh/D003505)
## Supporting Information
*In the **Supporting Information** section:*
Additional experimental data include thermal (HSM, DSC, TGA) and FTIR data for CD–RSV combinations, thermal data (HSM, DSC, TGA) for the crystalline complexes, 1H NMR peak integrations for complex stoichiometry determinations, details of X-ray structural refinements, geometrical data for CD host conformations, and comparative experimental and calculated PXRD patterns for the complexes.[](https://www.ncbi.nlm.nih.gov/mesh/C059514)
|
# Introduction
Endogenous [Prostaglandins](https://www.ncbi.nlm.nih.gov/mesh/D011453) and Afferent Sensory Nerves in Gastroprotective Effect of [Hydrogen Sulfide](https://www.ncbi.nlm.nih.gov/mesh/D006862) against Stress-Induced Gastric Lesions
# Abstract
*In the **Abstract** section:*
Hydrogen sulfide (H2S) plays an important role in human physiology, exerting vasodilatory, neuromodulatory and anti-inflammatory effects. H2[S has been impli](https://www.ncbi.nlm.nih.gov/mesh/D006862)ca[ted](https://www.ncbi.nlm.nih.gov/mesh/D006862) in the mechanism of gastrointestinal integrity but whether this gaseous mediator can affect hemorrhagic lesions indu[ced](https://www.ncbi.nlm.nih.gov/mesh/D006862) by stress has been little elucidated. We studied the effect of the H2S precursor L-cysteine, H2S-donor NaHS, the H2S synthesizing enzyme (CSE) activity inhibitor- D,L-propargylglycine (PAG) and the gastric H[2S ](https://www.ncbi.nlm.nih.gov/mesh/D006862)production [by CSE/CBS](https://www.ncbi.nlm.nih.gov/mesh/D003545)/3[-MS](https://www.ncbi.nlm.nih.gov/mesh/D006862)T activ[ity ](https://www.ncbi.nlm.nih.gov/mesh/C025451)in water immersion and restraint stress (WRS) ulcerogene[sis and the accompan](https://www.ncbi.nlm.nih.gov/mesh/C009055)yi[ng ](https://www.ncbi.nlm.nih.gov/mesh/C009055)changes in gastric[ bl](https://www.ncbi.nlm.nih.gov/mesh/D006862)ood flow (GBF). The role of endogenous pr[ostag](https://www.ncbi.nlm.nih.gov/mesh/D014867)landins (PGs) and sensory afferent nerves releasing calcitonin gene-related peptide (CGRP) in the mechanism of gastroprotection induc[ed by H2S was ](https://www.ncbi.nlm.nih.gov/mesh/D011453)ex[ami](https://www.ncbi.nlm.nih.gov/mesh/D011453)ned in capsaicin-denervated rats and those pretreated with capsazepine to inhibit activity of vanilloid receptors (VR-1). Rats [wer](https://www.ncbi.nlm.nih.gov/mesh/D006862)e pretreated with[ vehicle,](https://www.ncbi.nlm.nih.gov/mesh/D002211) NaHS, the donor of H2S and or L-cysteine, [the H2S pre](https://www.ncbi.nlm.nih.gov/mesh/C071423)cursor, with or without the concurrent treatment with 1) nonselective (indomethacin) an[d se](https://www.ncbi.nlm.nih.gov/mesh/C025451)lective cycloox[yge](https://www.ncbi.nlm.nih.gov/mesh/D006862)nase (CO[X)-1 (SC-5](https://www.ncbi.nlm.nih.gov/mesh/D003545)60) or[ CO](https://www.ncbi.nlm.nih.gov/mesh/D006862)X-2 (rofecoxib) inhibitors. The expression of mRNA and protein for COX-1 an[d COX-2 were](https://www.ncbi.nlm.nih.gov/mesh/D007213) analyzed in gastric mucosa pretreated w[ith Na](https://www.ncbi.nlm.nih.gov/mesh/C115461)HS with or w[ithout PA](https://www.ncbi.nlm.nih.gov/mesh/C116926)G. Both NaHS and L-cysteine dose-dependently attenuated severity of WRS-induced gastric lesions and significantly inc[reas](https://www.ncbi.nlm.nih.gov/mesh/C025451)ed GBF. These eff[ect](https://www.ncbi.nlm.nih.gov/mesh/C009055)s were [sign](https://www.ncbi.nlm.nih.gov/mesh/C025451)ifica[ntly reduc](https://www.ncbi.nlm.nih.gov/mesh/D003545)ed by pretreatment with PAG and capsaicin denervation. NaHS increased gastric H2S production via CSE/CBS but not 3-MST activity. Inhibition of COX-1 and COX-2 activ[ity](https://www.ncbi.nlm.nih.gov/mesh/C009055) sign[ificantly](https://www.ncbi.nlm.nih.gov/mesh/D002211) diminished Na[HS- ](https://www.ncbi.nlm.nih.gov/mesh/C025451)and L-cysteine-induced protection and hyperemia. NaHS increased expression of COX-1, COX-2 mRNAs and proteins and raised CGRP mRNA exp[ress](https://www.ncbi.nlm.nih.gov/mesh/C025451)ion. T[hese effec](https://www.ncbi.nlm.nih.gov/mesh/D003545)ts of NaHS on COX-1 and COX-2 prote[in c](https://www.ncbi.nlm.nih.gov/mesh/C025451)ontents were reversed by PAG and capsaicin denervation. We conclude that H2S exerts gastroprotection agains[t WR](https://www.ncbi.nlm.nih.gov/mesh/C025451)S-induced gastric lesions by the mechanism involving e[nha](https://www.ncbi.nlm.nih.gov/mesh/C009055)nceme[nt in gas](https://www.ncbi.nlm.nih.gov/mesh/D002211)tric microcirculation mediated [by ](https://www.ncbi.nlm.nih.gov/mesh/D006862)endogenous PGs, sensory afferent nerves releasing CGRP and the activation of VR-1 receptors.[](https://www.ncbi.nlm.nih.gov/mesh/D011453)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Hydrogen sulfide (H2S) is a gaseous mediator, which plays an important role in human physiology. Like other endogenous gasotransmitters, nitric oxide (NO) and carbon monoxide (CO), H2S can modulate vascular tone. H2S is mostly generated via L-cysteine metabolism and the activity of two pyridoxal-5`-phosphate (P5P, vitamin B6) dependent enzymes: cystathionine β-synthase (CBS) and cystathionine γ-lyase (CSE). However, this molecule may be synthesized by another pathway, mainly within mitochondria, that involves the activity of 3-mercaptopyruvate sulfotransferase (3-MST) and cysteine aminotransferase.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
H2S can evoke anti-inflammatory and pro-inflammatory effects depending on lower and higher concentration, respectively. The vasodilatory effects of the H2S donor, sodium hydrosulfide (NaHS) in blood vessels was confirmed by Kubo et al. and Zhao et al. who demonstrated that this donor results in a direct vasorelaxation of vascular smooth muscle acting as KATP channel opener.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Morsy et al. demonstrated that treatment with H2S donors reduced acetaminophen-induced hepatotoxicity in mice, which was correlated with a decreased expression of tumor necrosis factor-α (TNF-α) and enhanced expression of cyclooxygenase (COX)-2. H2S may play an important role within the gastrointestinal (GI) tract because Wallace et al. observed that treatment with L-cysteine and H2S donors accelerated the healing of experimental chronic gastric ulcers. Additionally, endogenous synthesis of H2S is significantly increased in the colon of animals with experimental colitis, compared to healthy colonic tissue.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
However, relatively little is known concerning the contribution of H2S to the mechanism of gastroprotection against acute gastric lesions, in particular, those induced by stress. Therefore, we attempted to determine the effect of pretreatment with NaHS and L-cysteine against gastric lesions induced by water immersion and restraint stress (WRS) and accompanying changes in the gastric blood flow (GBF) and the gastric mucosal production of H2S assessed by CSE/CBS/3-MTS activity. We examined the mechanism of the potential protective action of H2S released from NaHS or that generated from its precursor L-cysteine against stress ulcerogenesis with special reference to endogenous prostaglandins (PGs) and sensory afferent nerves releasing calcitonin gene-related peptide (CGRP) by using animals with non-selectively and selectively inhibited COX-1 and COX-2 activity and those with functionally ablated sensory neurons by capsaicin, respectively. Finally, the expression of proinflammatory cytokine TNF-α and its plasma levels as well as the mucosal expression of mRNAs for CGRP, COX-1 and COX-2 and the COX-1 and COX-2 protein concentration were analyzed in rats treated with NaHS with or without capsaicin denervation in order to determine the relationship between PG and sensory nerves releasing vasodilatatory mediators.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
## Materials and Methods
*In the **Materials and Methods** section:*
## Animals & Ethics Statement
*In the **Animals & Ethics Statement** section:*
Male Wistar rats with an average weight of approximately 250 g were fasted for 24 h, with free access to water before being exposed to 3.5 h of WRS. The Ethics Committee for Animal Research of Jagiellonian University Medical College approved all procedures, and experiments were run according to the principles of Helsinki Declaration.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Chemicals and drugs application, determination of the number of lesions and gastric blood flow
*In the **Chemicals and drugs application, determination of the number of lesions and gastric blood flow** section:*
Animals were selected into the groups pretreated with: A) vehicle (saline; 1 ml/rat), B) L-cysteine (2–80 mg/kg i.g.) and C) NaHS (0.1–5 mg/kg i.g.) with or without the combination with D, L-propargylglycine (PAG 30 mg/kg i.g.), an inhibitor of CSE activity. All chemicals were of the highest purity grade and were purchased from Sigma-Aldrich, Schnelldorf, Germany. In a separate groups of rats, NaHS and L-cysteine were administered with or without the co-treatment with 16,16 dmPGE2 or with the non-selective COX inhibitor—indomethacin (5 mg/kg i.p., Sigma-Aldrich, Schnelldorf, Germany) or selective to COX-1 inhibitor—SC-560 (5 mg/kg i.p., Cayman Chemical, Ann Arbor, USA) and COX-2 inhibitor—rofecoxib (30 mg/kg i.g., Pfizer, Illertissen, Germany). In rats of series D, capsazepine (10 mg/kg i.g., Sigma-Aldrich, Schnelldorf, Germany) was administered in order to inhibit vanilloid receptor (VR-1) according to the evidence published previously.[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
In rats of series E, capsaicin (Sigma-Aldrich, Schnelldorf, Germany) was administered in a large dose of 125 mg/kg s.c. to induce the functional ablation of sensory nerves as reported in our previous studies. Briefly, 2 weeks before the experiment capsaicin was administered subcutaneously in three doses 25, 50 and 50 mg/kg (total dose: 125 mg/kg). To check the effectiveness of the capsaicin denervation, a drop of 0.1 mg/ml solution of capsaicin was instilled into the eye of each rat and the protective wiping movement was counted.[](https://www.ncbi.nlm.nih.gov/mesh/D002211)
All COX-1 and COX-2 inhibitors, PGE2 and capsazepine were administered 30 min prior to the subsequent application of NaHS or L-cysteine, followed 30 min later by 3.5 h of WRS. For this purpose rats were immobilized in individual Bolman’s cages and immersed in the water (22°C) to the level of the xyphoid cartilage as described in our previous studies. At the termination of WRS animals were anesthetized with ketamine (10 mg/kg i.p.) and their abdomen was opened and the stomach was exposed for the GBF measurement by H2-gas clearance technique as described in detail previously. The GBF was measured in fundic part of the gastric mucosa not involving mucosal lesions.
Average values of three measurements were determined and expressed as a percentage change of the value determined in intact gastric mucosa. Gastric lesions number was determined with computerized planimetry (Morphomat, Carl Zeiss, Berlin, Germany) as described before. The blood samples from vena cava and gastric tissue were collected and stored in −80°C for further biochemical and molecular analysis.[](https://www.ncbi.nlm.nih.gov/mesh/D015232)
## H2S production in gastric mucosa determined by CSE/CBS/3-MST activity
*In the **H2S production in gastric mucosa determined by CSE/CBS/3-MST activity** section:*
The ability of gastric mucosa to produce H2S via CSE/CBS or 3-MST pathway was measured in homogenized tissue in the presence of exogenous substrates using a previously described zinc (Zn)-trapping assay. Briefly, gastric mucosa was quickly isolated, snap-frozen, and stored at −80°C. The gastric tissue was homogenized in ice-cold 50 mM potassium phosphate buffer, pH 8.0 (12% w/v). The homogenate (0.5 ml) and buffer (433 μl) were then cooled on ice for 10 min before L-cysteine (10 mM) and P5P (2 mM) or α-ketoglutarate (α-KG; 100 μM) were added (up to 1 ml of total volume). A smaller 1,5-ml tube containing a piece of filter paper (0.5×1.5 cm) soaked with zinc acetate (1%; 0.3 ml) was put inside the larger vial.
The vials were then flushed with nitrogen gas for 20 s and capped with an airtight serum cap. The vials were then incubated in a shaking water bath at 37°C for 90 min. Trichloroacetic acid (TCA; 50%; 0.5 ml) was then injected into the reaction mixture through the serum cap. The mixture was left to stand for another 60 min in 50°C to allow for the trapping of evolved H2S by the Zn acetate. The serum cap was then removed and N, N-dimethyl-p-phenylenediamine sulfate (20 mM; 50 μl) in 7.2 M HCl and FeCl3 (30 mM; 50 μl) in 1.2 M HCl were added to the inner tube. After 20 min, absorbance at 670 nm was measured with a microplate reader (Biotek Instruments, ELx808, VT, USA).
The calibration curve of absorbance vs. H2S concentration was obtained by using NaHS solution of varying concentrations. Addition of the substrate, L-cysteine (10 mM), was necessary for detection of H2S synthesis. H2S biosynthesis via CSE and CBS required the presence of P5P (2 mM unless otherwise stated), while that via 3-MST required α-KG.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
## Expression of mRNA for COX-1, COX-2, TNF-α and CGRP in the rat gastric mucosa determined by reverse transcriptase-polymerase chain reaction (RT-PCR)
*In the **Expression of mRNA for COX-1, COX-2, TNF-α and CGRP in the rat gastric mucosa determined by reverse transcriptase-polymerase chain reaction (RT-PCR)** section:*
Biopsy samples of gastric mucosa weighing about 200 mg were scraped off from oxyntic mucosa using a slide glass and immediately snap frozen in liquid nitrogen, and stored at −80°C until analysis. The total RNA was extracted from the mucosal samples by a guanidium isothiocyanate/phenol chloroform method using a kit from Stratagene (Heidelberg, Germany) according to methods described by Chomczynski and Sacchi. The concentration of RNA in RNase-free Tris EDTA buffer was measured at absorption of 260 nm wavelengths by spectrophotometry.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
Five μg of total cellular RNA single-stranded cDNA was generated using StrataScript reverse transcriptase and oligo(dT) primers (Stratagene). The polymerase chain reaction mixture was amplified in a DNA thermal cycler (Perkin-Elmer-Cetus, Norwalk, CT). The nucleotide sequences of the primers used in PCR are presented in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
Footnotes: CGRP—calcitonin gene-related peptide; COX—cyclooxygenase, TNF—tumor necrosis factor
Sense and antisense primers used in the assessment of mRNA expression for β-actin, CGRP, COX-1, COX-2 and TNF-α by reverse transcriptase polymerase chain reaction (RT-PCR).
PCR products were separated by electrophoresis in 2% agarose gel containing 0.5 μg/mL ethidium bromide and then visualized under UV light as described previously.[](https://www.ncbi.nlm.nih.gov/mesh/D012685)
Gastric mucosa of intact rats and those treated with vehicle (saline) and NaHS applied i.g. in different doses with or without PAG was evaluated for the expression of COX-1, COX-2, TNF-α and CGRP mRNAs and the signal intensity was analyzed by densitometry (Gel-Pro Analyzer, Fotodyne Incorporated, Hartland, WI, USA) as described before.[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## Assessment of COX-1, COX-2, PGE2 and TNF-α protein concentration
*In the **Assessment of COX-1, COX-2, PGE2 and TNF-α protein concentration** section:*
The concentration of COX-1, COX-2 and PGE2 in the gastric mucosa and TNF-α in plasma were determined using direct rat ELISA kits (MyBioSource, San Diego, USA) according to the procedures recommended by a manufacturer. Briefly, tissue homogenates of gastric mucosa were rinsed in ice-cold PBS and homogenized by two freeze-thaw cycles. Homogenates were centrifuged for 15 minutes at 1500×g (or 5000 rpm) and supernatant was stored at −20°C for further analysis. To assess COX-1, COX-2 and PGE2 protein concentrations in gastric mucosa homogenates or TNF-α in plasma, 50 μl of each sample and 100 μl of HRP-conjugate reagent was added to the each testing sample well. The plate was incubated for 1 h at 37°C and washed 4 times with wash buffer.
Subsequently, Chromogen Solution A (50 μl) and Chromogen Solution B (50 μl) were added to each well and then the microplate was incubated for 15 minutes at 37°C. Then the reaction was stopped with stopping solution and the optical density was read at 450 nm using microplate reader (Biotek Instruments, ELx808, VT, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D015232)
## Statistical analysis
*In the **Statistical analysis** section:*
Results are expressed as mean ± SEM. Statistical comparisons of two groups were performed with Student’s T-test, where appropriate. Comparison involving more than two groups was performed by ANOVA with Tukey post-hoc test. A difference with p <0.05 was considered statistically significant. Results are mean ± S.E.M of 6–8 rats per each experimental group.
## Results
*In the **Results** section:*
## Effect of graded doses of NaHS or L-cysteine on WRS-induced gastric lesions and changes in the GBF
*In the **Effect of graded doses of NaHS or L-cysteine on WRS-induced gastric lesions and changes in the GBF** section:*
The representative photomicrograph showing mucosal hemorrhagic lesions induced by WRS is presented in Fig. 1A. The exposure of rat to 3.5 h of WRS caused multiple hemorrhagic erosions mainly in the fundic part of the oxyntic mucosa. Pretreatment with NaHS (5 mg/kg i.g.) and L-cysteine (10 mg/kg i.g.) reduced the macroscopic lesions in the gastric mucosa induced by WRS and only a few gastric lesions were observed (Fig. 1B, 1C).[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
(A, B, C): Macroscopic appearance of rat gastric mucosa of animals exposed to 3.5 h of water immersion and restraint stress (WRS).[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
Rats were pretreated with vehicle (A), NaHS (5 mg/kg i.g.) (B) or L-cysteine (10 mg/kg i.g.) (C). Note, numerous gastric hemorrhagic erosions in gastric mucosa pretreated with vehicle (saline) (Panel A) and significant reduction in the number of gastric lesions in gastric mucosa pretreated with NaHS (Panel B) or L-cysteine (Panel C).[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
As shown in Fig. 2, pretreatment with NaHS dose-dependently attenuated WRS- induced gastric lesions, while producing a significant increase in the GBF with a dose of NaHS which inhibited WRS lesions by 50% being approximately 5 mg/kg. PAG (30 mg/kg i.g.) failed itself to affect the WRS-induced gastric lesions (Fig. 2). A NaHS-induced decrease in the lesion number and an accompanying rise in the GBF were completely reversed by PAG combined with NaHS. Likewise, L-cysteine administered in graded doses ranging from 0.1 mg/kg to 10 mg/kg (Fig. 3), dose-dependently reduced WRS-induced lesions and raised the GBF; the dose which inhibited number of these lesions by 50%, was approximately 10 mg/kg. PAG alone administered in a dose of 30 mg/kg i.p. had no significant influence on WRS lesions, but reversed the decrease in lesion number and an increase in the GBF induced by L-cysteine (10 mg/kg i.g.).[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Mean lesion number and the GBF in gastric mucosa of rats pretreated with NaHS.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Mean lesion number and the GBF in gastric mucosa of rats exposed to 3.5 h of WRS and pretreated with various doses of NaHS with or without combination with D,L- propargylglycine (PAG, 30 mg/kg i.g.). Control group (Vehicle) was pretreated with saline. Intragastric (i.g.) administration of sodium hydrosulfide (NaHS) in three different doses was ascribed to appropriate groups. One group received NaHS (5 mg/kg i.g.) with and without combination with PAG. Results are mean ±S.E.M of seven rats per each group. Significant change (p<0.05) as compared with the respective values in control group was indicated by asterisk. Cross was used to indicate significant change (p<0.05) comparing to the values obtained in PAG (30 mg/kg i.g.) treated group.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Mean lesion number and the GBF in rats pretreated with L-cysteine.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Rats received pretreatment with L-cysteine with or without the combination with PAG, before exposure to 3.5 h of WRS. Control group (Vehicle) was pretreated with placebo (saline). Intragastric (i.g.) administration of L-cysteine in three different doses was ascribed to appropriate groups. One group received D,L- propargylglycine (PAG, 30 mg/kg i.g.) with or without the combination with L-cysteine (10 mg/kg i.g.). Results are mean ±S.E.M of eight rats per each group. Significant change (p<0.05) as compared with the respective values in control group was indicated by asterisk. Cross indicates a significant change (p<0.05) comparing to the values obtained with PAG alone.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Assessment of H2S production determined by CSE/CBS and 3-MST activity in gastric mucosa of rats pretreated with vehicle or NaHS alone with or without the combination with PAG
*In the **Assessment of H2S production determined by CSE/CBS and 3-MST activity in gastric mucosa of rats pretreated with vehicle or NaHS alone with or without the combination with PAG** section:*
Fig. 4 (upper and lower panels) shows the effect of vehicle or NaHS (5 mg/kg i.g.) with or without the combination with PAG (30 mg/kg i.g.) on the concentration of H2S as determined by activity of CSE/CBS or 3-MST in gastric mucosa of rats exposed to WRS. The concentration of H2S produced in gastric mucosa determined by the enzymatic activity of CSE/CBS was significantly increased (p<0.05) in vehicle-control rats exposed to WRS as compared with the concentration of H2S in intact gastric mucosa. This increase in the H2S concentration in rats exposed to WRS was further significantly elevated (p<0.05) in rats pretreated with NaHS comparing to vehicle-pretreated rats (Fig. 4, upper panel). This increase in the H2S concentration was significantly inhibited (p<0.02) when rats received the combination of PAG and NaHS (Fig.
4, upper panel). As shown in the Fig. 4 (lower panel), the concentration of H2S determined by 3-MST activity was not significantly altered in rats treated with either vehicle or NaHS with or without the combination with PAG as compared with the concentration of H2S measured in intact gastric mucosa.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
H2S production in gastric mucosa determined by CSE/CBS/3-MST activity.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
The effect of vehicle or NaHS (5 mg/kg i.g.) with or without the combination with PAG (30 mg/kg i.g.) on the concentration of H2S determined as the activity of CSE/CBS (upper panel) or the activity of 3-MST (lower panel) in gastric mucosa of rats exposed to WRS. Results are mean ± S.E.M of six rats per each group. Significant change (p<0.05) as compared with the respective values in intact rats was indicated by asterisk. Cross indicates significant change (p<0.05) comparing to the values obtained in vehicle (control) group. Cross and asterisk indicate a significant change (p<0.05) comparing to the values obtained in group treated with NaHS (5 mg/kg i.g.) only.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
## The effect of COX-1 and COX-2 inhibitors on the lesion number and changes in the GBF in rats treated with NaHS and L-cysteine
*In the **The effect of COX-1 and COX-2 inhibitors on the lesion number and changes in the GBF in rats treated with NaHS and L-cysteine** section:*
As shown in Fig. 5, pretreatment with NaHS (5 mg/kg i.g.) and L-cysteine (10 mg/kg i.g.) resulted in a similar reduction in the lesion number and an increase in the GBF as presented in Figs. 2 and 3. Indomethacin, SC-560 or rofecoxib when administered prior to the onset of WRS, significantly increased the mean lesion number and also significantly decreased the GBF (p<0.05) compared with the pretreated vehicle-control (Fig. 5). The reduction of WRS lesion number and accompanying increase in the GBF induced by NaHS (5 mg/kg i.g.) and L-cysteine (10 mg/kg i.g.) were completely reversed by concurrent treatment with indomethacin, SC-560 and rofecoxib (p<0.05) (Fig. 5).[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Mean lesion number and GBF in gastric mucosa of rats pretreated with NaHS or L-cysteine combined with COX-1 and COX-2 inhibitors.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Before exposure to 3.5 h of WRS, NaHS and L-cysteine (L-cyst) were administered i.g. in doses of 5 mg/kg and 10 mg/kg, respectively. Indomethacin (INDO), SC-560 and rofecoxib (ROF) were administered to appropriate groups with and without combination with NaHS and L-cysteine. Results are mean ± S.E.M of seven rats per each group. Significant change (p<0.05) as compared with the respective values in control group was indicated by asterisk. Asterisk and cross indicate a significant (p<0.05) increase in mean lesion number and decrease in GBF as compared to respective values obtained in vehicle-control animals. Cross indicates significant change (p<0.05) comparing to the values obtained in group treated with NaHS (5 mg/kg i.g.) and L-cysteine (10 mg/kg i.g.) without any combination with COX-1 and COX-2 inhibitors.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
## The effect of capsaicin-induced deactivation of sensory nerves and blockade of VR-1 receptors by capsazepine on NaHS- and L-cysteine-induced protection
*In the **The effect of capsaicin-induced deactivation of sensory nerves and blockade of VR-1 receptors by capsazepine on NaHS- and L-cysteine-induced protection** section:*
As shown in Fig. 6 pretreatment with NaHS or L-cysteine applied i.g. in a dose of 5 mg/kg or 10 mg/kg, respectively, significantly reduced the mean lesion number of WRS-induced gastric lesions and caused the similar significant increase in the GBF as presented in Fig. 5. This attenuation in WRS lesion number and an accompanying increase in the GBF evoked by NaHS and L-cysteine were significantly reversed in capsaicin-denervated animals (p<0.05) (Fig. 6). Fig.
7 shows the effect of pretreatment with capsazepine (5 mg/kg i.g.) alone or that combined with L-cysteine or NaHS on WRS-induced gastric damage and the alterations in the GBF. NaHS and L-cysteine administered alone resulted in similar reduction of WRS-induced gastric lesions and an increase in the GBF as shown in Figs. 5 and 6 and these protective effects of NaHS or L-cysteine were completely lost in the presence of capsazepine (Fig. 7).[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Effect of capsaicin denervation on mean lesion number and GBF in rats exposed to WRS.[](https://www.ncbi.nlm.nih.gov/mesh/D002211)
Rats with intact and capsaicin-denervated sensory neurons were pretreated with vehicle (saline), NaHS (5 mg/kg i.g.) or L-cysteine (10 mg/kg i.g.) and exposed 30 min later to 3.5 h of WRS. Results are mean ±S.E.M of six rats per each group. Significant change (p<0.05) as compared with the respective values in vehicle-control group was indicated by asterisk. Cross indicates significant change (p<0.05) comparing to the values obtained in group treated with saline, NaHS and L-cysteine without denervation.[](https://www.ncbi.nlm.nih.gov/mesh/D002211)
Mean lesion number and the GBF in rats pretreated with capsazepine.[](https://www.ncbi.nlm.nih.gov/mesh/C071423)
WRS-exposed rats were initially administered with vehicle (saline), NaHS (5 mg/kg i.g.) or L-cysteine (10 mg/kg i.g.) with or without combination with capsazepine (5 mg/kg i.g.). Results are mean ±S.E.M of seven rats per each group. Significant change (p<0.05) as compared with the respective values in control group was indicated by asterisk. Cross indicates significant change (p<0.05) comparing to the values obtained in group treated with NaHS and L-cysteine without pretreatment with capsazepine.[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## The effect of NaHS administered alone or combined with PGE2 synthetic analog on WRS ulcerogenesis and plasma TNF-α levels in rats with intact sensory nerves and those with functional ablation of sensory nerves by capsaicin
*In the **The effect of NaHS administered alone or combined with PGE2 synthetic analog on WRS ulcerogenesis and plasma TNF-α levels in rats with intact sensory nerves and those with functional ablation of sensory nerves by capsaicin** section:*
As shown in Fig. 8, a similar degree of protection as reflected by the decrease in the mean lesion number was observed in NaHS-pretreated rats exposed to WRS as presented in Figs. 6 and 7 NaHS given in a dose of 5 mg/kg i.g. significantly reduced the plasma concentration of TNF-α (p<0.05) as compared with the value of TNF-α detected in vehicle-pretreated WRS-exposed animals. The combined administration of synthetic analog of PGE2 and NaHS further significantly diminished the mean lesion number and plasma TNF-α levels (both at p<0.05) as compared with the respective values of lesion number and plasma concentration of TNF-α in WRS rats pretreated with NaHS alone (Fig. 8). Both, the mean lesion number and plasma TNF-α levels were significantly increased (p<0.05) in rats with capsaicin denervation over the values in vehicle-pretreated rats with intact sensory nerves (Fig.
8). The NaHS-induced reduction in the mean lesion number and the accompanying significant fall in the plasma concentration of TNF-α were significantly decreased (p<0.05) in rats with capsaicin denervation as compared with the respective groups of animals with intact sensory nerves (Fig. 8). When 16,16 dmPGE2 was combined with NaHS in rats with sensory denervation, the significant reduction of mean lesion number and plasma TNF-α levels (p<0.05) were recorded, though these reductions failed to reach the similar level as observed in rats with intact sensory nerves treated concomitantly with NaHS and synthetic PGE2 analog (Fig. 8).[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Effect of pretreatment with NaHS alone and in combination with PGE2 on mean lesion number and TNF-α plasma levels in WRS rats with or without capsaicin denervation.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Rats with intact and capsaicin-denervated sensory neurons were pretreated with vehicle (saline), NaHS (5 mg/kg i.g.) or NaHS combined with 16,16 dmPGE2 (5 μg/kg i.g.) and exposed 30 min later to 3.5 h of WRS. Results are mean ± S.E.M of six rats per each group. Significant change (p<0.05) as compared with the respective values in vehicle-control group was denoted with asterisk. Double asterisks indicate a significant change (p<0.05) as compared with NaHS (5 mg/kg i.g.)-pretreated group without capsaicin-denervation. Cross indicates a significant change (p<0.05) comparing to the values obtained in respective control groups treated with saline and NaHS without capsaicin-denervation. Cross and asterisk represent significant change (p<0.05) as compared with vehicle- or NaHS-pretreated group without 16,16 dmPGE2 in rats with capsaicin denervation.[](https://www.ncbi.nlm.nih.gov/mesh/D002211)
## The effect of vehicle and NaHS with and without combination with PAG on the mRNA expression of COX-1, COX-2, TNF-α and CGRP
*In the **The effect of vehicle and NaHS with and without combination with PAG on the mRNA expression of COX-1, COX-2, TNF-α and CGRP** section:*
Fig. 9 A-E shows the effect of pretreatment with vehicle, NaHS applied alone in graded doses ranging from 1 mg/kg up to 10 mg/kg and NaHS (5 mg/kg i.g.) combined with PAG on the expression of mRNA for β-actin, COX-1, COX-2, TNF-α and CGRP in gastric mucosa of intact rats and those exposed to WRS. The strong signal for expression of COX-1 mRNA was recorded in intact mucosa and in those pretreated with vehicle. In rats pretreated with NaHS applied in the doses of 5 mg/kg and 10 mg/kg, the signal intensity was significantly enhanced as compared to the intact gastric mucosa or those pretreated with vehicle (Fig. 9, left panel).
This increase in signal intensity achieved in rats given NaHS in a dose of 5 mg/kg was significantly inhibited in group of rats treated with NaHS in the presence of PAG. Ratio of COX-1 mRNA over β-actin mRNA confirmed that the COX-1 mRNA was upregulated by NaHS and that this effect of NaHS (5 mg/kg i.g.) was significantly inhibited by PAG (p<0.05) (Fig. 9, right panel). The expression of COX-2 mRNA was detected as a strong signal in vehicle-pretreated gastric mucosa of WRS animals and this effect was further enhanced by NaHS applied in graded concentrations ranging from 1 mg/kg up to 10 mg/kg. The signal for increased expression of COX-2 by NaHS administered at a dose of 5 mg/kg was significantly inhibited by PAG. The ratio of COX-2 mRNA over β-actin mRNA confirmed that the expression of COX-2 mRNA was significantly increased in the vehicle-control gastric mucosa and this effect was further significantly elevated in rats pretreated with NaHS administered in higher doses of 5 mg/kg and 10 mg/kg (p<0.05) (Fig.
9, right panel). The ratio of COX-2 mRNA over β-actin mRNA confirmed that the expression of COX-2 mRNA was significantly inhibited (p<0.05) in rats treated with combination of NaHS and PAG (Fig. 9, right panel). As shown in Fig. 9 (left panel), the signal of expression of TNF-α mRNA was negligible in intact gastric mucosa but it appeared as detectable signal in vehicle-pretreated gastric mucosa and this effect persisted with almost the same signal intensity in gastric mucosa of rat administered with the dose of 1 mg/kg of NaHS. When higher doses of NaHS, 5 mg/kg and 10 mg/kg were applied, the signal intensity of TNF-α mRNA was gradually inhibited. This inhibition of TNF-α expression by NaHS applied at the dose of 5 mg/kg was reversed in part by PAG combined with NaHS.
The semiquantitive ratio of TNF-α over β-actin mRNA confirmed that NaHS significantly inhibited the expression of TNF-α (p<0.05), and that concurrent treatment with PAG abolished in part, the inhibitory effect of NaHS on TNF-α mRNA expression (Fig. 9, right panel). The CGRP mRNA was expressed in intact gastric mucosa as a strong signal. In contrast, the vehicle-pretreated gastric mucosa showed weaker signal intensity for CGRP mRNA. In rats pretreated with NaHS applied in graded concentrations ranging from 1 mg/kg up to 10 mg/kg, strong signals for CGRP mRNA were observed (Fig. 9E, left panel).
The increase in the signal intensity of CGRP mRNA evoked by NaHS was dramatically reduced by PAG administered in combination with NaHS. The ratio of CGRP mRNA over β-actin mRNA confirmed that NaHS dose-dependently increased the expression of mRNA for CGRP (p<0.05) and this effect of NaHS applied at a dose of 5 mg/kg was significantly diminished by concurrent treatment with PAG (p<0.05) (Fig. 9, right panel).[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
A, B, C, D, E: Expression of mRNA for COX-1, COX-2, TNF-α and CGRP in gastric mucosa.
The expression of mRNA for COX-1, COX-2, TNF-α and CGRP was determined in intact rats and those pretreated with vehicle (saline) or H2S-donor sodium hydrosulfide (NaHS) with and without combination with D, L- propargylglycine (PAG) and exposed 30 min later to 3.5 h of WRS (panel A). Semi-quantitative densitometry analysis of mRNA expression of COX-1, COX-2, TNF-α, CGRP transcripts normalized to β-actin mRNA expression (panel B). Results are mean ± S.E.M of four determinations. Significant change (p<0.05) as compared with the respective values in vehicle-control gastric mucosa was indicated by asterisk. Cross indicates a significant change (p<0.05) as compared to the value obtained in intact gastric mucosa. Asterisk and cross indicate a significant change (p<0.05) as compared to the values obtained in NaHS (5 mg/kg i.g.) alone pretreated group.[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## The effect of pretreatment with vehicle and NaHS with and without PAG administration or capsaicin-induced sensory nerves inhibition on gastric mucosal protein levels of COX-1, COX-2 and PGE2 biosynthesis
*In the **The effect of pretreatment with vehicle and NaHS with and without PAG administration or capsaicin-induced sensory nerves inhibition on gastric mucosal protein levels of COX-1, COX-2 and PGE2 biosynthesis** section:*
As shown in Fig. 10A, the gastric mucosal COX-1 protein concentration tended to decrease in the gastric mucosa of vehicle-pretreated animals exposed to WRS comparing to intact gastric mucosa but this effect failed to reach statistical significance. In contrast, the COX-1 protein concentration significantly increased (p<0.02) in rats pretreated with NaHS and this increase was significantly attenuated by concurrent administration of PAG (p<0.05) (Fig. 10 A). The COX-1 protein content was significantly decreased in capsaicin denervated animals pretreated with vehicle and exposed to WRS as compared to those with intact sensory nerves pretreated with vehicle and further exposed to WRS (Fig. 10 A). The increase in COX-1 protein concentration achieved in rats with intact sensory nerves pretreated with NaHS was significantly decreased (p<0.05) in those pretreated with NaHS with capsaicin denervation (Fig. 10 A).
Fig. 10 B shows the effect of pretreatment with vehicle or NaHS with or without combination with PAG (30 mg/kg i.p.) or the inhibition of sensory nerves on COX-2 gastric mucosal protein content in rats with intact sensory nerves or in those with capsaicin denervation. The significant increase in COX-2 protein content (p<0.05) was observed in vehicle-pretreated rats exposed to WRS as compared with the COX-2 concentration in the intact gastric mucosa (Fig. 10 B). In rats pretreated with NaHS before onset of WRS, the significant rise in the COX-2 protein content (p<0.02) was observed over that recorded in vehicle-treated gastric mucosa exposed to WRS and this increase in COX-2 protein content was significantly decreased (p<0.05) in rats treated with combination of PAG and NaHS (Fig. 10 B).
The protein COX-2 content was significantly decreased (p<0.05) in capsaicin denervated rats pretreated with vehicle or NaHS as compared with the respective values of COX-2 protein in rats with intact sensory nerves pretreated with vehicle or NaHS (Fig. 10 B).[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
A, B: The effect of pretreatment with vehicle and NaHS with and without PAG administration or capsaicin-induced sensory nerves inhibition on gastric mucosal protein levels of COX-1 and COX-2.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Rats with and without capsaicin-denervated sensory neurons were pretreated with vehicle (saline), NaHS (5 mg/kg i.g.) combined or not with PAG (30 mg/kg i.g.) and exposed 30 min later to 3.5 h of WRS. Intact group of rats did not undergo any procedures. Results are mean ±S.E.M of six rats per each group. Asterisk indicates significant change (p<0.05) as compared with the respective values in vehicle-control group (panel A) or with Intact rats (panel B). Cross indicates significant change (p<0.05) as compared with NaHS (5 mg/kg i.g.) pretreated group without capsaicin-denervation (panel A) or in vehicle-control group without capsaicin-denervation (panel B). Cross and asterisk represent significant change (p<0.05) as compared with NaHS pretreated group without capsaicin denervation.[](https://www.ncbi.nlm.nih.gov/mesh/D002211)
Table 2 presents the data with the gastric mucosal PGE2 content in intact rats and in those pretreated with vehicle (saline) or NaHS (5 mg/kg i.g.) with or without combination of PAG (30 mg/kg i.p.) or capsaicin sensory denervation. The PGE2 generation was insignificantly altered in vehicle-pretreated rats exposed to WRS as compared with intact gastric mucosa. In contrast, the gastric mucosal PGE2 generation was significantly increased (p<0.05) in animals pretreated with NaHS and this effect was significantly decreased in rats administered with the combination of PAG and NaHS (Table 2). The generation of PGE2 was significantly inhibited in gastric mucosa of capsaicin denervated rats (Table 2). The increase in the PGE2 content evoked by NaHS was significantly (p<0,05) diminished in rats with functional ablation of sensory nerves induced by capsaicin.[](https://www.ncbi.nlm.nih.gov/mesh/D015232)
Results are mean ± S.E.M of eight rats per each group.
* Asterisk denotes a significant change (p<0.05) as compared with the respective values in vehicle (control) group.
+ Cross indicates a significant change (p<0.05) as compared to the values obtained in NaHS alone pretreated group.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
Effect of pretreatment with vehicle (saline) or NaHS (5 mg/kg i.g.) applied alone or combined with PAG (30 mg/kg i.p.) or the inhibition of sensory nerves by capsaicin on the PGE2 concentration in the gastric mucosa of intact rats or those exposed to 3.5 hrs of WRS.[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## Discussion
*In the **Discussion** section:*
Current evidence indicates that H2S is a gaseous transmitter involved in the control of vascular tone and GI motility, however, its mechanism of gastroprotection has not been thoroughly studied. In particular, little is known whether donors of H2S could exert the protective action against the formation of acute gastric lesions such as those induced by stress and whether the activity of the afferent sensory nerves and endogenous PGs could be involved in this protection.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
We selected a stress animal model of WRS to determine the mechanism of action of H2S on the gastric mucosa, because of clinical relevance of stress as a potent risk factor of micro bleeding erosions and even peptic ulcers in humans. We reported that the fall in gastric microcirculation and the formation of bleeding gastric lesions are observed in rats under experimental conditions of stress. This animal model of WRS mimics hemorrhagic erosions that can appear as a consequence of stress bleedings under life threatening conditions in humans.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Herein, we demonstrated that pretreatment with L-cysteine, a H2S precursor, or NaHS, a H2S donor, protects the gastric mucosa against acute mucosal lesions induced by WRS, because the mean lesion number of WRS damage was decreased in rats treated with either H2S precursor or H2S donor and these protective effects of L-cysteine or NaHS were accompanied by an increase of the GBF. We demonstrated that exposure to WRS increased activity of key enzymatic CSE/CBS pathway resulting in an elevated H2S production in vitro in gastric mucosa suggesting that this gastric mucosa can compensate for the damaging effect of stress by an increase of H2S production to enhance the self-defense mucosal protective mechanism against the damage under stress conditions. This observation is in keeping with previous findings that the induction of an ulcer in the upper or lower gut results in a marked increase in H2S synthesis and production, possibly due to the compensatory protective mechanism in the gastrointestinal mucosa predisposed to the injury evoked by various ulcerogenes. Moreover, we found for the first time that NaHS further increased H2S production in the gastric mucosa, suggesting the “reciprocal” interaction of H2S donor with its endogenous product H2S endorsing its protective and hyperemic effects against stress ulcerogenesis. Our study supports the notion that CSE/CBS but not 3-MST pathway is responsible for the H2S production in gastric mucosa exposed to WRS because at our experimental conditions, NaHS administered with or without the combination with PAG failed to alter the 3-MST pathway.
The role of H2S in mucosal protection is supported by our present finding that L-cysteine, the H2S precursor, also exerted the protective and hyperemic activity against WRS ulcerogenesis. These observations clearly suggest that H2S-induced protection could be due to an increase in the GBF mediated by endogenous H2S and vasorelaxatory action of H2S donor and its precursor. These results are consistent with the findings of Fiorucci et al., who also reported that H2S donors reduced NSAID-induced leukocyte adherence in the gastric microcirculation. Neutrophil adherence induced by NSAIDs has been suggested to act as a primary mechanism for the gastric-damaging effect of these drugs.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Interestingly, NaHS can exert either vasodilatation or vasoconstriction depending on the dose employed. In the dose range used in our present study NaHS exhibited vasodilatory properties as documented by dose-dependent increase in GBF that accompanied the protective activity of this compound suggesting that H2S is responsible for an increase in the GBF observed in our study. Our data are corroborative with recent report that NaHS administered intraperitoneally prevented cold restraint stress-induced oxidative gastric damage by the mechanism involving the inhibition of gastric acid and attenuation of reactive oxygen metabolites. Lou et al. demonstrated that beneficial effect of H2S may depend upon hypothermia caused by H2S that reduced the concentration of the lipid peroxidation products in the gastric mucosa exposed to stress, thus attenuating stress ulcerations. Moreover, it was shown that NaHS protected gastric mucosal epithelial cells against H2O2-induced cell death.
H2S can inhibit gastric lesions caused by topical damaging agents including non-steroid anti-inflammatory drugs (NSAID), such as diclofenac and naproxen. In relevance to these findings, the new and safer NSAID releasing H2S such as H2S-releasing diclofenac (ATB-337) and H2S-releasing naproxen (ATB-346) were shown to reduce gastric lesions induced by conventional NSAID.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
We found that CSE inhibitor-PAG, which in the presence of NaHS markedly decreased the H2S production as analyzed in this study by CSE/CBS pathway and reversed the NaHS-induced protection against stress-induced gastric lesions and accompanying increase in the GBF. However, in another study, the administration of PAG resulted in protection against ethanol injury and this was explained by evident reduction in H2S concentration elevated in response to intragastric application of ethanol. Moreover, indomethacin inhibited the protective effects of PAG, thus suggesting the involvement of endogenous PGs in gastroprotection exerted by this CSE inhibitor. This apparent discrepancy in results with PAG between their results and our present study could be attributed to differences in mechanism of damage and protection against chemical and non-topical damage induced by absolute ethanol used in their report and by non-topical ulcerogenesis such as stress in our present study. Moreover, using similar experimental model of stress-induced gastric lesions, Aboubakr et al. demonstrated that the inhibition of key enzyme of H2S synthesis, CSE with beta-cyano-L-alanine (BCA) exacerbated stress-induced gastric damage.[](https://www.ncbi.nlm.nih.gov/mesh/C009055)
We clearly demonstrated that endogenous PGs and PG/COX system are involved in the mechanism of NaHS-induced gastroprotection because in rats pretreated with NaHS the upregulation of mRNA expression of COX-1 and COX-2 and the rise in the COX-1 and COX-2 protein concentrations were observed. Furthermore, we provided evidence that the pretreatment with PAG in combination with NaHS not only reversed protective and hyperemic effect of NaHS but also downregulated the expression of mRNA for COX-1 and COX-2 and NaHS induced increase in COX-1 and COX-2 protein contents.[](https://www.ncbi.nlm.nih.gov/mesh/D011453)
Previous studies in rats and humans demonstrated that PGE2 is present in the largest amounts in the gastric mucosa and exerts gastroprotective action within upper GI tract. That is why, we determined whether the co-administration of selective (SC-560, rofecoxib) and non-selective (indomethacin) inhibitors of COX-1 and COX-2, known to suppress the mucosal generation of endogenous PGE2, could affect the NaHS and L-cysteine-induced protection and changes in the GBF. We found that the concurrent treatment of COX-1 and COX-2 inhibitors with NaHS and L-cysteine almost completely reversed both protective and hyperemic activities exhibited by this H2S donor and the H2S precursor. Therefore, we conclude that H2S-induced gastroprotection and increase in gastric microcirculation may depend upon the activity of PG/COX system. We observed that this NaHS-induced protection against WRS-induced gastric lesions is accompanied by the rise in gastric mucosal PGE2 concentration suggesting that the protective and hyperemic actions of H2S can involve the activation of COX/PG system and its endogenous products, PG.[](https://www.ncbi.nlm.nih.gov/mesh/D015232)
The H2S-induced rise in COX-1 and COX-2 protein contents and mucosal PGE2 concentration were significantly attenuated in rats with capsaicin denervation suggesting that capsaicin—sensitive afferent neurons and sensory vasoactive neuropeptides such as CGRP may contribute to the beneficial protective effect of NaHS in rat stomach. Indeed, the afferent sensory fibers were originally proposed to play an important role in the mechanism of gastric mucosal integrity and gastroprotection. This is why we studied the effects of NaHS and L-cysteine in animals with capsaicin denervation and in those with blockade of VR-1 receptors with capsazepine. We confirmed that capsaicin denervation itself decreased the GBF and augmented WRS-induced ulcerogenesis. The protective and hyperemic activity of NaHS and L-cysteine were greatly attenuated in rats with capsaicin denervation.
Taken together, we propose that the mechanism of H2S-induced protection could be due to an increased activity of the afferent sensory nerves and a release of CGRP, which activates VR-1 receptors, thus causing vasodilatation. This is supported by the observation that NaHS and L-cysteine protection was accompanied by an increase of mRNA CGRP expression and this effect was attenuated when PAG combined with NaHS or L-cysteine. In order to look for the relation between PGs, altered activity of sensory nerves and proinflammatory cytokine TNF-α in gastroprotection by H2S, we tested the effect of NaHS in the presence or absence of synthetic analog of PGE2 against WRS ulcerogenesis in rats with or without capsaicin denervation. We found that NaHS-induced protection was potentiated by co-treatment with PGE2 analog, and that this effect of combined administration of NaHS and PGE2 persisted, though being significantly reduced, in rats with capsaicin denervation. Moreover, the NaHS protection was accompanied by the fall in the plasma TNF-α level, which was potentiated when exogenous PGE2 which was co-administered together with H2S donor.
In contrast, this decrease in the plasma TNF-α level observed in rats concomitantly treated with the combination of NaHS and PGE2 was diminished in those with capsaicin denervation suggesting the link between prostaglandins, neuropeptides released from sensory nerves and the release of proinflammatory cytokines such as TNF-α in gastroprotection elicited by H2S. We confirmed our previous observation that afferent sensory nerves ablation induced by application of capsaicin increased WRS-induced gastric lesions and decreased GBF as compared with rats with intact sensory nerves exposed to WRS. Moreover, SC-560 and COX-2 inhibitor, rofecoxib were shown to exert the same aggravatory effects as capsaicin denervation. Interestingly, SC-560 or rofecoxib administered to rats with capsaicin-induced sensory nerves ablation in our previous study even enhanced noxious effect of capsaicin denervation suggesting that the impairment of these two pathways can exert deleterious influence on gastric mucosa exposed to WRS. Indeed, as demonstrated in this study, capsaicin denervation and the selective and non-selective inhibition of COX-1/COX-2 pathways reversed the beneficial effects of H2S against WRS-induced gastric damage. Additionally, the afferent sensory nerves ablation by capsaicin decreased COX-1 and COX-2 protein contents and the rise in the mucosal generation of PGE2 observed in rats pretreated with NaHS.
H2S-induced gastroprotection may depend upon activity of two synergistic pathways, namely afferent sensory nerves activity releasing CGRP and endogenous PGs derived from the enhanced expression and activity of COX-1 and COX-2.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Interestingly, the systemic (i.v.) administration of L-cysteine and NaHS attenuated gastric injury induced by ischemia/reperfusion. The mechanism of this systemic action of NaHS and L-cysteine has been attributed to a decrease in plasma levels and downregulation of mRNA expression of proinflammatory cytokines IL-1β or TNF-α. These effects of H2S donor and H2S precursor were diminished by pretreatment with PAG. Herein, we present an evidence that NaHS administration decreased TNF-α mRNA expression and this effect was reversed when NaHS was combined with PAG. Thus, it is likely that gastroprotection induced by NaHS releasing H2S exerts the anti-inflammatory activity that also could be an important component of gastroprotection afforded by this gaseous molecule.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
In summary, H2S released from its donor, NaHS or that synthesized from L-cysteine, plays an important physiological role in gastric mucosal protection against stress-induced lesions. This protection induced by NaHS or L-cysteine is accompanied by an enhancement in in the gastric microcirculation possibly mediated by a significant local increase in the gastric mucosal production of H2S. The mechanism of H2S-induced gastroprotection involves activation of endogenous PGs/COX system, the rise in biosynthesis of PGE2, and afferent sensory fibers releasing CGRP acting via VR-1 receptors and by anti-inflammatory effect resulting in the inhibition of pro-inflammatory cytokines such as TNF-α.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
# References
*In the **References** section:*
|
# Introduction
Positive effects of the progestin [desogestrel](https://www.ncbi.nlm.nih.gov/mesh/D017135) 75 μg on migraine frequency and use of acute medication are sustained over a treatment period of 180 days
# Abstract
*In the **Abstract** section:*
Background
Premenopausal migraines frequently are associated with fluctuations of estrogen levels. Both, migraine and combined hormonal contraceptives (CHC) increase the risk of vascular events. Therefore progestagen-only contraceptives (POC) ar[e a safe](https://www.ncbi.nlm.nih.gov/mesh/D004967)r alternative. A previous short-term study demonstrated a positive impact of the oral POC desogestrel on migraine frequency. To study the effect of the POC desogestrel 75 μg on migraine frequency, intensity, use of acute medication and quality of life i[n a clinica](https://www.ncbi.nlm.nih.gov/mesh/D017135)l setting over the period of 180 days.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
Methods
Patients’ charts were screened for women with migraine, who had decided to use desogestrel for contraception. Charts were included, if routinely conducted headache diaries were complete for 90 days before treatment (baseline) and over a treatment period of 180 days. We also report about starters who stopped treatment early, because of adverse events. Baseline data (day 1–90 before treatment) were compared with first and second treatment period (treatment days 1–90 and days 91–180). Quality of life was evaluated using MIDAS questionnaires.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
Results
Days with migraine (5.8 vs 3.6), with any kind of headache (9.4 vs 6.6), headache intensity (15.7 vs 10.7), days with severe headache (5.4 vs 2.4) and use of triptans (12.3 vs7.8) were significantly reduced after 180 days. MIDAS score and grade improved significantly.[](https://www.ncbi.nlm.nih.gov/mesh/D014363)
Conclusion
Contraception with desogestrel 75 μg resulted in a significantly improved quality of life and a reduction of migraine days over the observation period of 180 days. A clinically meaningful 30% reduction in pain was observed in 25/42 (60%) participants. For counselling reasons it is of importance, that the major reduction in migraine frequency occured during the initial 90 days, however further improvement occurs with longer duration of use. Prospective studies are needed to confirm these results.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
## Background
*In the **Background** section:*
Epidemiological data suggest that combined hormonal contraceptives (CHC) initiate or worsen migraine and headache in predisposed women [1-5]. The incidence of migraine is highest during the reproductive years and more than 50% of women report an association between migraine attacks and their menstrual cycle [6,7]. The reproductive phase is also the life span in which most women need efficient contraception. Migraine with aura (MA) and to a lesser extent migraine without aura (MO) increase the risk for cardiovascular events, especially for stroke [8-11]. There is a substantial elevation of these risks in migraineurs using CHC [11-14].
The cardiovascular risk associated with CHC, has been mainly attributed to the estrogen component which exerts a strong effect on the coagulation system. Finding a well-tolerated estrogen-free form of contraception for headache patients therefore is an important issue.
Progestagen-only pills (POP) have so far not been found to be associated with an increased risk for thromboembolic or ischemic events [15]. Most guidelines recommend progestagen-only contraception as a safer option [16]. The POP desogestrel 75 μg (Cerazette®; MSD Merck Sharp & Dohme AG, Luzern, Switzerland) is used continuously and combines efficient inhibition of ovulation with maintenance of low estrogen levels [17,18].
Avoidance of estrogen peaks and withdrawal could contribute to good tolerability of this contraceptive in migraineurs. Recently we reported a benefit of desogestrel 75 μg on migraine and quality of life over a 3 month period of use [19,20]. The effect on frequency and quality of life was comparable to improvements observed with prophylactic agents. However, the observation interval was short. In the present study, we report effects of 6 cycles desogestrel contraception on headache frequency, intensity and use of pain medication.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
## Methods
*In the **Methods** section:*
This study was performed at the divison for family planning, unit of the Department of Reproductive Endocrinology, University Hospital Zürich, Switzerland where one of the authors (GM) runs an outpatient clinic for migraine patients with need for hormonal therapy. Migraine is diagnosed according to the IHS (International Headache Society) criteria by the referring neurologists from headache centres in Zürich, Bad Zurzach or by the author [21]. Reasons for referral were need for contraception in women with migraine, menstrual migraine or any form of hormonal therapy of headaches. To allow an exact diagnosis of the headache type and frequency according to the IHS our patients are principally instructed to conduct headache diaries for 3 cycles before their first visit and to continue after any intervention. MIDAS questionnaires are used before interventions and in intervals of 90 days thereafter.
The majority of our premenopausal patients have a need for efficient contraception. In the context of the discussions around the elevated risks for cardiovascular disease and stroke we advise against combined hormonal contraceptives as a first choice contraception in migraineurs and in women aged 35 years or more. Before starting a hormonal treatment women are informed about risks and potential side effects which include information about irregular bleeding and acne with the use of desogestrel.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
For the present study patients’ charts were screened for women with migraine, who had decided to use the POP desogestrel 75 μg and had conducted headache diaries 90 days before initiation and over 180 days of use of this medication. We included patients suffering from all types of migraine. The observation period was defined from July 2009 to December 2013. In a previous study we already reported 90 day treatment data of 16 included patients. Women had to be premenopausal and had to need effective contraception.
We report about all adverse events causing discontinuation earlier than 180 days. Exclusion criteria were: incomplete diaries, less than 10 headache episodes during the pretreatment period, initiation or change of prophylactic medications during the observation and postmenopause. This resulted in a drop-out rate of 26 out of 68 charts.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
The diaries include information on the number of migraine and headache days, the severity of headache, the use of triptans and other pain medication, the use of hormones and days with vaginal bleeding. Days of bleeding were assessed to allow an exact diagnosis of the migraine type according to the IHS criteria. Headache severity was rated in the diaries according to a 4-point scale (0 = no pain, 3 = severe pain). This score is easy to understand and has been proven to be useful in daily work with migraineurs. For ethical reasons all diaries were anonymised before data evaluation.
The evaluation of anonymised data in our setting was accepted by the ethical committee of the Kanton Zürich.[](https://www.ncbi.nlm.nih.gov/mesh/D014363)
Primary efficacy variables were the differences in number of migraine and headache days, the difference in pain score as well as MIDAS score and grade. Secondary outcomes included differences in the number of all pain medications and triptans used as well as differences in days with pain score three. In population-based studies of migraine- and headache sufferers in the US and UK the MIDAS questionnaire and the MIDAS summary scores proved to be a highly reliable means of assessment of the impact of the ailment on daily life [22,23]. The total MIDAS score strongly correlates with both the clinical evaluation of the severity of a patient’s headache problem, and the frequency of the episodes, determined from daily-based headache diaries [23].[](https://www.ncbi.nlm.nih.gov/mesh/D014363)
## Statistical analyses
*In the **Statistical analyses** section:*
Data were compared between baseline (BL) (day 1–90 before treatment) and treatment periods (TP): TP1 (day 1–90) TP2 (day 91–180). In addition all variables were compared between TP1 and TP2. Statistical analyses were done using IBM SPSS Statistics, version 22 (Armonk, New York, IBM Corp). Data are presented as mean (SD). Pain intensity score was calculated as the sum of headache intensities for baseline and each treatment period according to the above mentioned 4-point scale.
For each period this sum was divided by three to obtain a mean monthly pain score. To calculate monthly frequencies, the numbers for each observational episode was divided by three. Numbers of monthly migraine days, headache days, headache intensity, days with use of pain medication and questions of the MIDAS questionnaire were compared with Friedman’s test. Post-hoc comparisons between single time points were performed using Wilcoxon’s signed rank test with Bonferroni correction.
## Results
*In the **Results** section:*
Flow diagram of the study population.
Demographic and baseline characteristics of included charts (n = 42) and excluded charts (n = 26)
Changes in migraine, frequency, intensity and use of pain medication during use of the contraceptive pill desogestrel 75 μg over 180 days of use[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
A total of 68 women with migraine initiated contraception with desogestrel 75 μg. Headache diaries of 42 subjects were complete and eligible for analysis. Six patients had stopped desogestrel because of side effects within 42 or less days and were excluded (prolonged bleeding n = 3, increase of headache n = 2, acne = 1) (Figure 1). Demographics and characteristics of eligible women and drop-outs did not differ significantly (Table 1). Hormonal contraception was used by 50% (n = 21) of the included patients and 61% (n = 16) in the drop-out (p > 0.05).
One included woman had used a copper-device (drop-outs: n = 0). Chronic headaches (>15 /month) were found in 6 included patients and more than 8 triptans were used monthly by 9 included patients. Mean age of migraine onset was 22.4 years (SD 5.2). Two women suffered from endometriosis. Frequency of migraine, headache intensity, days with use of pain medication and triptans were significantly reduced during TP2 in comparison with BL (Table 2).
Days with severe pain declined from 5.4 (SD 4.2) to 2.4 (SD 3.5) (p < 0.001) (Table 2). The improvements were in large parts visible during TP1 and persisted during further follow-up. A according to the IHS clinically meaningful 30% reduction in pain was observed in 25/42 (60%) participants, whereas another 28% (12/42) experienced even a 50% reduction [24,25]. We found a 255 reduction in the sum of headache and migraine days in 55% (23/42) of the included migraineurs. Seven of 42 patients (16%) experienced 1–5 more headache/migraine days during TP2 in comparison to BL.
Interestingly, however, quality of life improved in five of these seven women. Further analyses to explain this seemingly contradictory result revealed a decrease in days with pain score 3 and a decrease in overall pain intensity in all these five patients. Two women with more migraine attacks and without improvement in the MIDAS score, decided to change to a non-hormonal contraception after 180 days.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
Changes in quality of life measured with the MIDAS after 90 days and 180 days contraception with desogestrel 75 μg[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
Changes in migraine and headache frequency during use of the progestin-only pill desogestrel 75 μg, comparison between MO (n=32) and MA (10) patients[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
Table 3 demonstrates the changes in quality of life. All MIDAS items improved significantly during 180 days of desogestrel use (TP2). Again significant improvement was already observed after TP1. Separate analyses for MO and MA women revealed no differences with regard to demographic parameters between the groups. In MO patients significant improvements of all features (except headache days) days were observed (Table 4).
The very small group of subjects with MA experienced significant reductions in the number of pain medications and triptans, MIDAS score and MIDAS grade.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
## Discussion
*In the **Discussion** section:*
In the present study we report the effects of 180 days of contraception with the progestin-only pill desogestrel 75 μg on headache and migraine. We observed a significant reduction in migraine frequency, migraine intensity, use of triptans and pain score. Quality of life measured by the MIDAS score improved by more than 50%. Mean MIDAS grades were diminished by point (Table 3). The majority of positive effects were apparent after 90 days and small further improvements were noted up to 180 days of use (Table 2).
To our knowledge, we report for the first time that hormonal treatment can reduce the use of triptans significantly. This might be of relevance for women at the boarder of medication overuse headaches. As different pathophysiologies underlie MA and MO, we performed subanalyses for both types of headaches. In women with MO, significant improvements for all variables except headache days were observed. In the group of patients suffering from MA (n = 10) migraine days decreased by two/month, what possibly as a result of the small group size was not significant.
Significant bettermends were observed with regard to use of pain medications, use of triptans and MIDAS score and grade.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
Migraine is a typical disorder with a high response rate to placebo in controlled trials. For ethical reasons placebo-controlled studies in the area of contraception are not acceptable. An important strength of the present study is the long run-in period and the evaluation not only of migraine frequency but also of additional parameters, like pain intensity, use of pain medications and quality of life. The combination of these data is a better reflection of the overall well-being as demonstrated in the detailed data analysis of the patients developing more migraine in our study. The run in period of 90 days allowed a balanced overview with regard to migraine frequencies which can vary markedly from month to month.
However, even if the headache diaries had been conducted prospectively our analyses could have generated selection or information bias. In particular, we assume that a prospective design might have resulted in a higher continuation rate and exclusion of less charts with incomplete diaries. A control group of women using other hormonal contraceptive methods would have been of advantage.
Our findings for MO patients are in accordance with a very recent retrospective diary-based study, demonstrating a significant reduction in migraine frequency, pain intensity and use of pain medication with 6 months use of desogestrel [26]. Triptan use did not decline, which contrasts with our result and might be related to the lower number of included patients. The comparison with a control group of users of a combined pill (COC) in a long-cycle in this study is of great interest, because both forms of contraception prevent hormone withdrawal [26]. While migraine attacks and pain intensity decreased significantly with the POP, headache frequency declined with the COC regimen only.
Desogestrel use failed to exert a significant effect on non-migrainous-headache in our sample and the comparative trial. This can be explained, by our earlier reported findings showing in an individual follow-up of both headache and migraine, that a temporary transformation of migraines to headaches occurs in some women [19]. Our present study with a longer observation period however indicates that, on the long-term, these headaches might decline as well (p = 0.05). Although our study includes only few women with MA, the findings are backed by Nappi et al. who reported a significant reduction in migraine frequency in MA patients, but did not investigate pain intensity and quality of life [27].[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
Even if there is still a lack of prospective controlled trials several diary-based studies indicate a positive impact of desogestrel on migraine without aura [19,26,27]. Continuous use of COCs exerts a positive impact on headaches and hormone-withdrawal migraines, however POP are much safer with regard to the cardiovascular and thromboembolic risks [26,28-32]. The benefit of desogestrel on migraine with aura, which is not typically associated with estrogen withdrawal, has to be confirmed in future studies. Many migraineurs are reluctant to use hormones as a consequence of previous bad experience. During counselling it is helpful to know that major improvements can be expected during 90 days of desogestrel use.
Furthermore, two trials indicate that migraines and pain tend to improve further beyond 3 months [19,26,27]. On the other hand, patients have to be informed that migraine rarely worsens. The clinically meaningful 30% reduction in pain (considered by the IHS as clinically meaningful) in 60% of our patients is another argument to prefer this contraceptive method in women with migraine [24,25]. Reduction in use of triptans and other pain medications might contribute of the prevention of medication-overuse headaches.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
In daily life the degree to which a reduction in headache frequency translates to decreased disability and improved quality of life is highly relevant. The MIDAS demonstrated highly reduced disability and significantly improved quality of life in our patients.
Use of POP can cause a variety of bleeding patterns including amenorrhoea, infrequent bleeding, frequent bleeding and prolonged bleeding episodes. Unfavorable bleeding patterns such as frequent bleeding and prolonged bleeding occur as result of the continuous progestin effect on the endometrium and can be a reason for withdrawal from this form of contraception [33]. Prolonged and frequent bleedings usually stop with longer duration of use and can be treated if not.
## Unanswered questions
*In the **Unanswered questions** section:*
New insights in the hormonal effects on the brain allow speculations about mechanisms underlying our observations. Avoidance of hormone withdrawal can only explain the decline of cycle-related headaches. In contrast to estrogens, progesterone seems to attenuate trigeminovascular nociception and reduces dural plasma protein extravasation following stimulation of the trigeminal ganglion [34-36]. Thus direct or receptor-mediated effects of the desogestrel on the trigeminovascular system can be postulated.
The variety of responses on desogestrel treatment could be a result of the genetic variability of estrogen receptors in women, with some polymorphisms being a significant risk factor for migraine [37]. The neurological basis of migraine auras has not yet been established, increasing evidence indicates that they are a clinical manifestation of a cortical spreading depression (CSD).[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
In mice the thresholds for cortical spreading depression (CSD) is lower in cycling females than in males. This would allow to hypothesise that maintenance of low estrogen levels induced by desogestrel might upregulate the threshold for CSD thus reduce MA attacks. A further mechanism could be that desogestrel or its metabolite etonogestrel, like progesterone and allopregnanolone decrease cortical excitability via the GABA -receptor [38-40].[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
At the moment we have no means to predict how an individual migraineur will react on desogestrel. Outside the study we achieved positive effects with higher dosages. However, this is off-label use and cannot be generally recommended before prospective trials have been conducted. Several trials highlight a positive effect of desogestrel on migraine. Among neurologists it is well known that headaches may be cycle-related, but they rarely consider to search advice for a hormonal treatment.
Vice versa gynaecologists are not always aware of the fact that hormonal treatment affects headache frequency in predisposed women. A closer collaboration between gynaecologic endocrinologists and headache specialists might provide better care and safety for young women, suffering from migraine during use of any hormones or in association with their natural cycle.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
## Conclusion
*In the **Conclusion** section:*
In conclusion our data indicate a positive impact of desogestrel 75 μg on migraine frequency, intensity, use of pain medication and quality of life. The major improvement was observed during the initial 90 days of use, which might be important for patients’ counselling. Randomised controlled trials are needed to substantiate our results.[](https://www.ncbi.nlm.nih.gov/mesh/D017135)
Competing interests
Gabriele S Merki-Feld and Bruno Imthurn had financial relationship (lecturer, member of advisory boards and// or consultant) with Bayer-Schering Pharma and MSD AG.
RL and AG declared no conflicts of interest.
Authors’ contributions
GM: participated in the design of the study, the acquisition and analysis of data and the drafting of the manuscript. BI: has been involved in drafting the manuscript and revising it critically. RL participated in the acquisition and interpretation of data and revision of the manuscript. BS has been involved in the analysis and interpretation of data and revision of the manuscript. AG participated in the design of the study, acquisition of data and drafting the manuscript. All authors read and approved the final version of the manuscript.
# References
*In the **References** section:*
|
# Introduction
[FL118](https://www.ncbi.nlm.nih.gov/mesh/C578515), a novel [camptothecin](https://www.ncbi.nlm.nih.gov/mesh/D002166) derivative, is insensitive to ABCG2 expression and shows improved efficacy in comparison with [irinotecan](https://www.ncbi.nlm.nih.gov/mesh/D000077146) in colon and lung cancer models with ABCG2-induced resistance
# Abstract
*In the **Abstract** section:*
Background
Irinotecan is a camptothecin analogue currently used in clinical practice to treat advanced colorectal cancer. However, acquired resistance mediated by the drug efflux pump ABCG2 is a recognized problem. We[ reported ](https://www.ncbi.nlm.nih.gov/mesh/D000077146)on a n[ovel camptot](https://www.ncbi.nlm.nih.gov/mesh/D002166)hecin analogue, FL118, which shows anticancer activity superior to irinotecan. In this study, we sought to investigate the potency of FL118 versus irinotecan or its active metabolite, SN-38, in both [in vitro and](https://www.ncbi.nlm.nih.gov/mesh/D002166) in vivo mo[dels ](https://www.ncbi.nlm.nih.gov/mesh/C578515)of human cancer with high ABCG2 activity. We a[lso sought](https://www.ncbi.nlm.nih.gov/mesh/D000077146) to assess the potency and ABCG2 affinity of several FL11[8 ana](https://www.ncbi.nlm.nih.gov/mesh/C578515)logues w[ith B-ring](https://www.ncbi.nlm.nih.gov/mesh/D000077146) substitutions.[](https://www.ncbi.nlm.nih.gov/mesh/D000077146)
Methods
Colon and lung cancer cells with and without ABCG2 overexpression were treated with FL118 in the presence and absence of Ko143, an ABCG2-selective inhibitor, or alternatively by genetically modulating ABCG2 expression. Using two distinct in vivo human tumor animal models, we further assessed whether FL118 could extend time to progression in comparison with irinotecan. Lastly, we investigated a series of FL118 analogues with B-ring substitutions for ABCG2 sensitivity.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
Results
Both pharmacological inhibition and genetic modulation of ABCG2 demonstrated that, in contrast to SN-38, FL118 was able to bypass ABCG2-mediated drug resistance. FL118 also extended time to progression in both in vivo models by more than 50% compared with irinotecan. Lastly, we observed that FL118 analogues with polar substitutions had higher affinity for ABCG2, suggesting that the nonpolar nature of FL118 plays a role in bypassing ABCG2-mediated resistance.[](https://www.ncbi.nlm.nih.gov/mesh/D000077146)
Conclusions
Our results suggest that in contrast to SN-38 and topotecan, FL118 is a poor substrate for ABCG2 and can effectively overcome ABCG2-mediated drug resistance. Our findings expand the uniqueness of FL118 and support continued development of FL118 as an attractive therapeutic option for patients with drug-refractory cancers resulting from high expression of ABCG2.[](https://www.ncbi.nlm.nih.gov/mesh/D000077146)
Electronic supplementary material
The online version of this article (doi:10.1186/s12943-015-0362-9) contains supplementary material, which is available to authorized users.
## Background
*In the **Background** section:*
Camptothecin analogues have been used clinically to treat cancer for almost 20 years. Irinotecan (also known as CPT-11) is used in combination with other antitumor agents as a first-line therapy for metastatic colorectal cancer and has a history of use as a second-line therapy in advanced gastric and non-small cell lung cancers (NSCLC). The second clinically used camptothecin analogue, topotecan, is approved for treatment of ovarian, cervical, and small cell lung cancers. It has been established in the literature that camptothecin analogues function through inhibition of the topoisomerase I (Top1) enzyme. Camptothecin-class compounds target the DNA-Top1 covalent complex, forming a ternary complex that prevents the dissociation of Top1.
This ternary complex inhibits replication and transcription and leads to the formation of double-strand DNA breaks.[](https://www.ncbi.nlm.nih.gov/mesh/D002166)
Unfortunately, resistance to irinotecan and topotecan is observed in the clinic. Failure of irinotecan- and topotecan-based regimens has been hypothesized to occur through a number of different mechanisms, though only a few are supported with clinical data. In vitro evidence and limited clinical observations suggest mutations in the Top1 gene decrease the affinity of the Top1 protein with clinically used camptothecin analogues. However, based on the literature, likely a more common cause of resistance to irinotecan and topotecan is the increased expression of ATP-binding cassette (ABC), subfamily G, isoform 2 protein (ABCG2, also known as breast cancer resistance protein, BCRP), a drug efflux pump and a member of the ABC transporter superfamily. A number of clinical studies revealed that failure of irinotecan and topotecan often correlates with increased ABCG2 expression.
Multiple in vitro studies have demonstrated that irinotecan, SN-38 (active metabolite of irinotecan), and topotecan are all substrates for ABCG2, and high expression of ABCG2 is associated with decreased intracellular accumulation of these compounds and consequentially a decrease in drug potency. Additionally, many other anticancer agents are known ABCG2 substrates, including methotrexate, many anthracyclines, and a variety of tyrosine kinase inhibitors.[](https://www.ncbi.nlm.nih.gov/mesh/D000077146)
Our lab recently reported on a novel camptothecin derivative, designated FL118. The chemical name of FL118 is 10,11-methylenedioxy-20(S)-camptothecin, also known as 10,11-MD-CPT, MDCPT, and 10,11-mCPT (Additional file 1: Figure S1). FL118 shows strong anticancer activity in several different cancer types in vitro and in vivo. We have demonstrated that although FL118 is not a better Top1 inhibitor than clinically used camptothecin analogues, FL118 is able to selectively inhibit the expression of several members of the Inhibitor of Apoptosis family (survivin, XIAP, and cIAP2) and the Bcl-2 family (Mcl-1), which was demonstrated to contribute to FL118 function and anti-cancer activity. More recent studies have further characterized the novel properties of FL118.
Induction of cancer cell senescence and cell death by FL118 employs both p53-dependent and p53-independent signaling pathways, and rapid induction of wild type p53 accumulation by FL118 is largely independent of the ATM-dependent DNA damage signaling pathway but dependent on E3-competent Mdm2. Our previous studies also revealed that, while mice showed continuing body weight loss after treatment with irinotecan, body weight rapidly recovers after the completion of FL118 treatment, suggesting that FL118 possesses a more favorable toxicity profile in comparison with irinotecan.[](https://www.ncbi.nlm.nih.gov/mesh/D002166)
In the present study we found that, although SN-38 and topotecan are ABCG2 substrates and fail to overcome ABCG2-mediated drug resistance, FL118 is insensitive to ABCG2 expression and effectively bypasses ABCG2 resistance. FL118 also demonstrates better antitumor efficacy than irinotecan in human xenografts with high ABCG2 expression. Additionally, we found that the relatively nonpolar nature of FL118 plays a role in bypassing ABCG2-induced resistance.[](https://www.ncbi.nlm.nih.gov/mesh/D000077146)
## Results
*In the **Results** section:*
## FL118 is a more potent anticancer agent than SN-38 in NSCLC and colon cancer cell lines
*In the **FL118 is a more potent anticancer agent than SN-38 in NSCLC and colon cancer cell lines** section:*
EC50 of FL118 and SN-38 in NSCLC and colorectal cancer cell lines, including Top1 inhibitor-resistant HCT116 sub-lines[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
Pharmacological inhibition of ABCG2 modulates the potency of SN-38, but not FL118. A and B, Western blot analysis of ABCG2 protein expression in HCT116 colon cancer cells, drug-resistant HCT116 sub-lines (A), and H460 and EKVX NSCLC cells (B). C and E, dose-response curves in the presence and absence of 1 μM Ko143, an ABCG2 inhibitor, after 72 hour treatments in HCT116 sub-lines (C) and NSCLC cell lines (E). D and F, dose-response curves in the presence and absence of 1 μM Ko143 after 72-hour treatments in HCT116 sub-lines (D) and NSCLC cell lines (F). Viability for each dose was determined using a ViCELL XR cell viability analyzer and normalized to that of DMSO control. Error bars = SEM, n = 3 independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D000077146)
The potency of FL118 versus SN-38 was compared in a panel of NSCLC and colon cancer cell lines. In each of the parental cell lines tested, FL118 was 5- to 10-fold more potent than SN-38, with EC50 values consistently below 1 nM (Table 1, Additional file 1: Figures S2, S3). In the four HCT116-derived camptothecin-resistant colon cancer sublines, each with mutations in Top1 was demonstrated to decrease potency of camptothecin analogues, FL118 showed greater potency than SN-38 overall. Intriguingly, FL118 showed much more potency than SN-38 in sublines SN50 and A2 in comparison with sublines SN6 and G7 (Table 1). SN50 and A2 sublines highly express ABCG2, while the SN6 and G7 sublines show undetectable ABCG2 expression (Figure 1A).
We assessed whether there was a difference between relative resistance (RR) for FL118 and SN-38 in HCT116 sublines. We back-transformed RR into LogRR for the purpose of statistical analysis, and found a statistically significant difference in three sublines tested (SN6, SN50, A2). Importantly, the difference between the LogRR of FL118 and SN-38 was statistically significant in both sublines with high ABCG2 expression, SN50 (p = 0.023) and A2 (p = 0.027) (Additional file 1: Table S1). Together, these data suggest that ABCG2 expression is an SN-38 resistance factor but not an FL118 resistance factor.
Therefore, we hypothesized that FL118 is a comparatively poor substrate of ABCG2, and thus the potency of FL118 is not affected by ABCG2 expression.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
## Pharmacological inhibition of ABCG2 does not modulate FL118 potency
*In the **Pharmacological inhibition of ABCG2 does not modulate FL118 potency** section:*
Many anticancer drugs are substrates of the ABC transporter ABCG2, which often contributes to drug treatment failure in the clinic. To determine whether FL118 potency is affected by ABCG2 activity, we utilized cell lines with varying ABCG2 protein expression to test drug sensitivity. In addition to the camptothecin-resistant HCT116 sublines discussed above, we also employed two NSCLC cell lines: H460, which express high levels of ABCG2 protein, and EKVX cells, which have undetectable levels of ABCG2 protein expression when assessed by Western blot (Figure 1B). Then, we employed a highly selective inhibitor of ABCG2, Ko143, to determine whether inhibition of ABCG2 would affect FL118 potency in these cells. As expected, inhibition of ABCG2 activity by Ko143 in HCT116-A2, HCT116-SN50, and H460 cell lines, which have high ABCG2 expression, resulted in a significant increase in potency for SN-38 (Figure 1C, E) and topotecan (Additional file 1: Figure S4), indicating that they are substrates of ABCG2.
In contrast, the potency of FL118 remained unchanged (Figure 1D, F) in these cell lines, suggesting that FL118 potency is unaffected by ABCG2 overexpression. Using HCT116-SN50 as a representative example, the EC50 of SN-38 alone was 135.1 nM, compared to 12.3 nM for SN-38 in the presence of Ko143 (p < 0.0001). In the same cell line, the EC50 of FL118 alone was 4.0 nM, compared to 3.0 nM for FL118 in combination with Ko143 (p = 0.22). In contrast, Ko143 did not alter the potency of either SN-38 or FL118 in HCT116, HCT116-G7, or EKVX cells that lack detectable ABCG2 (Figure 1).[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
## Genetic silencing or overexpression of ABCG2 does not affect the potency of FL118
*In the **Genetic silencing or overexpression of ABCG2 does not affect the potency of FL118** section:*
Genetic silencing or overexpression of ABCG2 demonstrates that the potency of FL118 is not affected by ABCG2. A and D, Western blot analysis of ABCG2 protein expression in HCT116-A2 cells that were stably transduced with a non-silencing shRNA control (ns) or anti-ABCG2 shRNA (sh1 = V3LHS_380805, sh2 = V3LHS_380806) (A), and HEK293 cells that were stably transfected with either an ABCG2 expression vector (HEK293/ABCG2) or a corresponding empty vector (HEK293/pcDNA3) (D). B and C, dose-response curves of SN-38 (B) and FL118 (C) after 72-hour treatments in HCT116-A2 cell lines. E and F, dose-response curves of SN-38 (E) and FL118 (F) after 72-hour treatments in HEK293 cell lines that were stably transfected with ABCG2 expression vector or empty vector. Viability was determined as in Figure 1. Error bars = SEM, n = 3 independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
We also employed a genetic approach through direct silencing of ABCG2 to test the hypothesis that FL118 potency is not affected by ABCG2 expression. Two ABCG2-specific shRNAs were validated in our studies that effectively knock down ABCG2 protein expression (Figure 2A). When the expression of ABCG2 was stably knocked down in HCT116-A2 cells, the EC50 of SN-38 was reduced to 19.6 nM (p < 0.0001) and 22.4 nM (p < 0.0001), compared to 97.0 nM in cells transduced with a non-silencing control shRNA (Figure 2B, Additional file 1: Table S2). In contrast, consistent with pharmacological inhibition, there was no change in potency for FL118 with or without ABCG2 silencing (Figure 2C, Additional file 1: Table S2). Next, we alternatively determined whether exogenous overexpression of ABCG2 would affect FL118 potency using a human embryonic kidney cell line, HEK293 that was stably transfected with either an ABCG2 expression vector (HEK293/ABCG2) or an empty vector (HEK293/pcDNA3) (Figure 2D).
As expected, overexpression of ABCG2 decreased the potency of SN-38 from 0.39 nM in HEK293/pcDNA3 cells to 62.95 nM in HEK293/ABCG2 cells (p = 0.002) (Figure 2E, Additional file 1: Table S2). In contrast, there was no significant difference in FL118 potency in HEK293/pcDNA3 cells compared to HEK293/ABCG2 (p = 0.09) (Figure 2F, Additional file 1: Table S2). These data confirm that ABCG2 expression does not mediate resistance to FL118, suggesting that FL118 could bypass ABCG2-mediated treatment resistance.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
## FL118 exhibits better antitumor activity than irinotecan and significantly extends time to progression in human xenograft models
*In the **FL118 exhibits better antitumor activity than irinotecan and significantly extends time to progression in human xenograft models** section:*
FL118 shows improved efficacy in two in vivo models of irinotecan-resistant cancer in comparison with irinotecan. SCID mice bearing HCT116-SN50 or H460 subcutaneous xenografts with an average volume of ~100 mm3 were treated via IP once per week with 100 mg/kg irinotecan or 1.5 mg/kg FL118, on a repeating treatment for 4 weeks followed by 1 week of rest. Individual animals were considered to have progressed and were removed from treatment as tumor volume reached 1500 mm3 or for a moribund condition. A and B, tumor growth curves for HCT116-SN50 (A) and H460 (B) xenografts during treatment with either FL118 or irinotecan. As individual mice progressed and were removed from treatment, their final tumor volume was included in the graphed average on subsequent dates. Each treatment is indicated with a caret. C and D, Kaplan-Meier chart of progression events for animals with HCT116-SN50 (C) and H460 (D) xenografts. Statistical difference in time to progression between the two treatments was determined using the log-rank test. E and F, average body weight over time during treatments. Error bars = SEM.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
Next, we determined antitumor activity of FL118 versus irinotecan in HC116-SN50 and H460 xenograft models of ABCG2-mediated drug-resistant cancer. Once tumors were established and reached an average volume of ~100 mm3, a repeating treatment was applied intraperitoneally (IP) weekly for 4 weeks, followed by 1 week of rest. The time to progression (TTP) was used as the primary endpoint for assessment of efficacy of FL118 versus irinotecan. In both models, FL118 controlled tumor growth better than irinotecan (Figure 3A, B) and led to significantly improved TTP (Figure 3C, D). Specifically, in the HCT116-SN50 model, median TTP was 58 days for animals treated with FL118, compared to 38.5 days for animals treated with irinotecan (p = 0.002), a 50.6% increase (Figure 3C).
In the H460 model, median TTP was extended from 21 days with irinotecan to 35 days with FL118 (p = 0.009), a 66.7% increase (Figure 3D). Consistent with our earlier reports, FL118 showed a tolerability profile similar to irinotecan (Figure 3E, F).[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
## The unique nonpolar structure of FL118 appears to play a role in FL118 bypassing ABCG2-mediated drug resistance.
*In the **The unique nonpolar structure of FL118 appears to play a role in FL118 bypassing ABCG2-mediated drug resistance.** section:*
Electronegative potentials influence the affinity of the Position 7-substituted FL118 analogues for ABCG2 binding. A, structures of FL118 analogues with B ring substitutions. B, correlation between electronegativity of chemical groups (χ, Pauling units, calculated using the method of Huheey) and the ratio of EC50 of drug alone / EC50 drug + Ko143.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
Electronegativities (χ, Pauling units) for added 7-position functional groups, and EC50 +/- K0143 in HCT116-SN50 cells[](https://www.ncbi.nlm.nih.gov/mesh/C541506)
We next investigated the potency of several analogues of FL118 with functional groups attached to the 7-position of the B ring (Figure 4A), a position where substitutions have been shown to favorably alter drug properties, in the irinotecan-resistant sub-line HCT116-SN50. We observed little change in potency between FL118 and the FL118-derived analogues with relatively nonpolar alkyl substitutions (methyl, ethyl, and allyl) (Table 2). In contrast, addition of more polar groups (bromomethyl, chloromethyl, and hydroxymethyl) resulted in a marked decrease in potency (Table 2). Thus, adding groups to the B ring that increase the polarity of the molecule would result in decreased potency. Previous reports indicated that the addition of polar residues to the A ring of camptothecin analogues increased affinity to ABCG2, but the effect of B ring substitutions with these types of residues remains unknown.
To assess whether the polarity-related decrease in potency was at least partially dependent on ABCG2, growth inhibition was subsequently assessed in the presence and absence of ABCG2 inhibitor Ko143. Although we observed no significant change in EC50 for 7-methyl-FL118, 7-ethyl-FL118, or 7-allyl-FL118 with the addition of Ko143, we were able to observe a significant decrease in EC50 for 7-bromomethyl-FL118 (87.0 nM to 17.0 nM, p < 0.0001), 7-chloromethyl-FL118, (26.8 nM to 12.9 nM, p = 0.016), and 7-hydroxymethyl-FL118 (24.2 nM to 4.8 nM, p < 0.0001) (Table 2). Thus, the potency of the latter three FL118 analogues is affected by ABCG2 activity. We next compared the electronegativity (χ, in Pauling units) of each added functional group to the ratio of EC50 / EC50 + Ko143 (Figure 4B).
Consistent with the data shown in Table 2, it was revealed that chemical groups with stronger electronegativity (bromomethyl, chloromethyl, and hydroxymethyl) show higher affinity to ABCG2 than the chemical groups with weaker or no electronegativity (methyl, ethyl, and allyl) (Figure 4B). Together, our work revealed that lack of polar functional groups on the B ring of FL118 plays a role in FL118 bypassing ABCG2-mediated drug resistance.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
## Discussion
*In the **Discussion** section:*
The present study expands the uniqueness of FL118 in mechanism of action to overcome drug resistance, and demonstrated that, in a panel of colon cancer and NSCLC cell lines, FL118 is more potent than SN-38. We found that, in contrast to the two clinically used camptothecin analogues (irinotecan, topotecan), which are ABCG2 substrates and unable to overcome ABCG2 resistance, FL118 potency is not affected by ABCG2 expression and can bypass ABCG2-mediated treatment resistance. This phenomenon was investigated and confirmed by multiple independent approaches to either inhibit (i.e., pharmacological and genetic inhibition) or enhance (i.e., overexpression) ABCG2 activity. These approaches demonstrated that high ABCG2 activity results in resistance to SN-38 and topotecan, but not FL118.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
We also assessed the efficacy of FL118 in two distinct in vivo xenograft models of ABCG2-mediated drug-resistant cancer. In both models, FL118 exhibited a better ability to decrease tumor growth in comparison with irinotecan, while maintaining a tolerable toxicity profile similar to that of irinotecan. Most importantly, mice treated with FL118 showed a significant increase in time to progression (TTP) compared to mice treated with irinotecan.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
Our group recently reported on the in vivo efficacy of FL118 in other models of human cancer, using an intravenous (IV)-compatible, Tween/polysorbate 80-free formulation. In that study, we compared three schedules (every day for five injections, every other day for five injections and once weekly for 4 injections) of FL118 administration via IV administration. We found that the optimum administration of FL118 appears to be every other day. However, for this study, we opted for a weekly schedule that mimics irinotecan administration in the clinic. Therefore, while our results in this study revealed that the weekly administration of FL118 was able to significantly extend TTP in comparison with irinotecan, we predict that treating with FL118 every other day for five treatments will show even better efficacy.
Additionally, in the current studies, we used IP routes instead of intravenous administration of FL118 for technical convenience. However, the Tween/polysorbate 80-free formulation of FL118 can be administrated via IV routes with increased maximum tolerated dose (MTD). This may also result in better in vivo tumor inhibition outcomes for FL118. We intend to perform these experiments with optimal routes and schedules in our follow-up studies as part of a broader goal of optimizing FL118 administration to prepare for clinical applications.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
It was reported that the affinity of camptothecin analogues to ABCG2 is influenced by polar additions to the A ring. We hypothesize that the polarity of B ring-substituted functional groups of FL118 may also affect ABCG2 binding. To test this hypothesis, we modeled the distinct electrostatic potential of six FL118 analogues with different chemical groups on the 7-position of the B ring versus FL118 itself, and assessed their potency using cell proliferation assays. As before, we used potency in the presence and absence of ABCG2 inhibition as a surrogate for assessing the affinity of individual drugs for ABCG2. We saw a significant change in EC50 in the presence of Ko143 for FL118 analogues with stronger electronegative groups, indicating that these FL118 analogues are substrates of ABCG2.
We also noted that some structures with less polar functional groups (e.g., 7-ethyl-FL118) still have regions of localized electronegativity in those groups (indicated by the red gradient) (Additional file 1: Figure S5), while some structures with more polar groups lack such regions of moderate localized electronegativity (e.g., 7-chloro-FL118, Additional file 1: Figure S5). Comparing these electrostatic potential maps to our results in Table 2, we propose that these localized electronegative regions are less predictive of a compound’s affinity for ABCG2 than the Pauling electronegativity of the molecule as a whole.[](https://www.ncbi.nlm.nih.gov/mesh/D002166)
We recently reported a novel IV-compatible Tween/polysorbate 80-free formulation for FL118, which could increase FL118 MTD 3-7 fold in comparison with FL118 formulated using the previous Tween/polysorbate 80-containing recipe. We found that in this new formulation, while FL118 showed solubility similar to 7-bromomethyl-FL118, 7-chloromethyl-FL118 and 7-hydroxymethyl-FL118, the solubility of 7-methyl-FL118, 7-ethyl-FL118 and 7-allyl-FL118 was much poorer than FL118. Similarly, FL118 analogues with other 7-substituted nonpolar groups (e.g., cycloalkyl, aryl) also showed poor solubility in our Tween/polysorbate 80-free formulation recipes with or without 5% DMSO. Therefore, future studies related to the structural modification of the FL118 scaffold should be based on the findings revealed from these compounds to further optimize potency and drug-like properties, for example by adding hydrophilic group on positions of 5, 9 or 12. Alternatively, we may generate the pro-drug for potent compounds, such as dipeptide derivatives to increase water solubility.
Working in these directions may allow us to generate compounds with even better therapeutic index (TI, i.e. ratio of antitumor activity versus toxicity) than the favorable TI of FL118.[](https://www.ncbi.nlm.nih.gov/mesh/D011136)
Here, it should be pointed out that the relationship between the affinity of camptothecin analogues to ABCG2 and the strength of electronegative charges of distinct chemical groups on the B ring has not been investigated previously, although the B ring of camptothecin is often modified in order to improve the anticancer potency and pharmacological properties. While there was a positive correlation between χ and ABCG2-induced resistance, the large loss of potency for 7-bromomethyl-FL118 compared to 7-chloromethyl-FL118 suggests that functional group electronegativity may be only one characteristic that influences ABCG2 affinity. Since ABCG2-mediated resistance to anticancer drugs, including irinotecan and topotecan, is a recognized problem in the clinic, further understanding of the structure-activity relationship of FL118 analogues for ABCG2 binding may lead to rational design of better anticancer agents for clinical application.[](https://www.ncbi.nlm.nih.gov/mesh/D002166)
Interestingly, cabazitaxel, a taxane derivative with poor affinity for another drug-efflux pump protein, P-glycoprotein 1 (P-gp, also known as multidrug resistance protein 1 and ABCB1), was recently approved for use in patients with castration-resistant prostate cancer who had previously failed docetaxel-based regimens. It is thought that cabazitaxel’s lack of affinity for P-gp plays an important role in its effectiveness in docetaxel-refractory cancer. In keeping with this rationale, the work presented here suggests that this strategy may be useful in other cancer types and with other classes of cytotoxic agents, including camptothecins. Due to FL118’s superior anticancer activity, favorable tolerability, and insensitivity to ABCG2, we posit that FL118 may become a better option for targeted cancer therapeutics to address the increasingly complex issue of drug failure by circumventing multiple mechanisms of drug resistance, including efflux pump-mediated resistance.[](https://www.ncbi.nlm.nih.gov/mesh/C552428)
In conclusion, the present study demonstrated that FL118 has additional mechanistic features in terms of its superior anticancer efficacy, which further distinguish it from irinotecan, SN-38 and topotecan. These findings suggest that FL118 is a poor substrate for the drug efflux pump ABCG2, and thus FL118 is able to overcome ABCG2-mediated resistance to SN-38, irinotecan and topotecan in vitro and in vivo. Additionally, this study also indicated that polar chemical groups on the B ring of FL118 analogues can contribute to ABCG2-mediated resistance, which provides one principle for new FL118 analogue design. Together, the new features of FL118 revealed in this study plus the other FL118 unique features reported in our previous studies warrants FL118 further development toward clinical application.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
## Materials and methods
*In the **Materials and methods** section:*
## Drug resource and preparation
*In the **Drug resource and preparation** section:*
Topotecan (Selleckchem Chemicals, Houston, TX), FL118 (in house), and FL118 analogues (RTI International) were prepared as stocks at 1 mM in DMSO (Merck KGaA, Darmstadt, Germany). The synthesis of FL118 and FL118 analogues (7-methyl-FL118, 7-ethyl-FL118, 7-allyl-FL118, 7-bromomethyl-FL118, 7-chloromethyl-FL118 and 7-hydroxymethyl-FL118) were reported previously. Stock SN-38 (Sigma-Aldrich Corporation, St. Louis, MO) was prepared at 2.5 mM in DMSO. Ko143 (Tocris Bioscience, Bristol, United Kingdom) was prepared as stock solutions at 10 mM in DMSO.[](https://www.ncbi.nlm.nih.gov/mesh/D019772)
## Cell culture
*In the **Cell culture** section:*
Human colorectal cancer cell lines HCT8, HCT116, and SW620 and NSCLC cell lines A549 and NCI-H460 (“H460”) were purchased from American Type Culture Collection (ATCC, Manassas, VA). The human NSCLC cell line EKVX (donated by Dr. Daniel Chan) was originally from the National Cancer Institute. Camptothecin resistant sub-lines of HCT116 with mutated Top1 (HCT116-SN6, HCT116-G7, HCT116-A2, and HCT116-SN50) were established and described previously by Drs. Gongora and Del Rio. Drug resistant cell lines were passaged in 10 nM SN-38, except for five days prior to all experiments, to maintain resistant phenotypes.
Human embryonic kidney HEK293 cells that were stably transfected with either an ABCG2 expression vector (HEK293/ABCG2) or an empty vector (HEK293/pcDNA3) were provided by Dr. Wendy Huss, which were originally a gift from Dr. Susan Bates (National Cancer Institute, Rockville, MD). All cell lines were maintained in RPMI-1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 100 U/mL penicillin, and 0.1 μg/mL streptomycin (“complete media”). Cells were cultured in 5 % CO2 at 37°C and passaged every 2-4 days.[](https://www.ncbi.nlm.nih.gov/mesh/D002166)
## Immunoblot analysis
*In the **Immunoblot analysis** section:*
Immunoblot analysis was performed as described previously, with minor modifications. Briefly, cells were lysed in radioimmunoprecipitation analysis buffer (50 mM Tris, 150 mM NaCl, 0.1% SDS, 0.5% sodium deoxycholate, 1% Nonidet P-40, 10 μg/mL PMSF, 20 μM leupeptin) and sonicated for 15 s using a sonic dismembrator 100 (Thermo Fisher Scientific, Waltham, MA) to homogenize lysate. Next, lysate was denatured with 5X Laemmli Sample Buffer (5X: 300 mM Tris-HCl pH 6.8, 10% SDS, 50% glycerol, 20% β-mercaptoethanol, 0.05% bromophenol blue) and equal amounts of protein were electrophoretically separated on 10-15% SDS-PAGE gels and electrotransferred onto 0.2 μm nitrocellulose membranes (Bio-Rad Laboratories, Inc., Hercules, CA). Membranes were blocked for 1 h at room temperature in 5% skim milk, then incubated with primary antibody (1:1,000 for ABCG2 and 1:5,000 for actin) in 5% bovine serum albumin in TBS-T overnight at 4°C. Membranes were washed with TBS-T, then incubated with species-specific anti-IgG antibodies conjugated to horseradish peroxidase (1:5,000, second antibody) at room temperature for 1 h in 5% milk.
Membranes were again washed with TBS-T. Chemiluminescence with ECL plus (PerkinElmer, Inc., Waltham, MA) was used to detect protein using X-ray film (Midsci, St. Louis, MO).[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
## Cell proliferation and viability assay
*In the **Cell proliferation and viability assay** section:*
Cells were plated in 6-well plates at densities ranging from 2 x 105 to 4 x 105 cells per well, depending on doubling time. On day 2, cells were treated with varying concentrations of indicated compounds in complete media. Final concentration of the vehicle, DMSO, was 0.1% in all treatments with or without drugs. On day 5, attached cells were harvested with 0.5 mL of 0.25% Trypsin-EDTA. Trypsin was deactivated by addition of 0.5 mL of complete media and cells were analyzed on a Vi-CELL XR Cell Viability Analyzer (Beckman Coulter Inc., Brea, CA).
EC50 values and coefficient of determination (R2) were calculated from seven doses of each analyzed compound, in addition to the vehicle control, using GraphPad Prism 6 (GraphPad Software, Inc., La Jolla, CA). Each experiment was performed at least 3 times.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
## Transduction of lentiviral particles containing ABCG2-specific or control shRNA
*In the **Transduction of lentiviral particles containing ABCG2-specific or control shRNA** section:*
Lentiviral particles containing ABCG2-specific shRNA, with sequences of AACTCTTGAATGACCCTGT (V3LHS_380805) or ATAATACTTGGTAACATCC (V3LHS_380806), and one control non-silencing shRNA (Dharmacon, Lafayette, CO) were prepared in the Rowell Park Cancer Institute shRNA Core Facility as previously described. HCT116-SN50 cells were plated in 6-well plates at a density of 7 x 105 cells/well. The next day, media was aspirated and 0.5 mL of fresh complete media and 0.5 mL of viral supernatant were added to each well. Cells were incubated with viral supernatant in 5% CO2 at 37°C for 16 h. Stably transduced cells were selected and maintained in 1 μg/mL puromycin (Invivogen, San Diego, CA).[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
## Electrostatic potential maps
*In the **Electrostatic potential maps** section:*
Electrostatic potential maps were generated using the PBEQ solver module in CHARMM-GUI in default conditions, using protein data bank (.pdb) files and Tripos Mol2 (.mol2) files created with MarvinSketch 14.7.7.0 (ChemAxon, Ltd., Budapest, Hungary). Maps were visualized with PyMOL 1.3r1 edu (Schrödinger, LLC, New York, NY).
## Pauling electronegativity
*In the **Pauling electronegativity** section:*
Electronegativities were taken from published studies. When electronegativities for certain functional groups were not reported in those studies, they were calculated using the methods developed by the authors in the cited works.
## Establishment of human xenograft models
*In the **Establishment of human xenograft models** section:*
Xenografts were established in female 12-week old severe combined immunodeficiency (SCID) mice. Cells (4 x 106 per injection) were suspended in 200 μL of a 1:1 solution of ice-cold serum-free RPMI 1640 media and matrigel (Corning Incorporated, Corning, NY) and injected subcutaneously into the left flank. When tumors reached an average volume of 100 mm3, animals were randomly assigned to one of two treatment groups. One group received 100 mg/kg irinotecan (Camptosar) (Pfizer, New York, NY), the MTD, IP once per week. The other group received the MTD of FL118, 1.5 mg/kg, IP once per week.
The treatment schedule, one treatment per week for 4 weeks, followed by 1 week of rest, repeated until progression, was selected to approximate the administration of camptothecin-class drugs in the clinic. Tumor volume and body weight were measured three times per week. Progression was defined as a tumor volume ≥ 1500 mm3 or a moribund condition. Tumor volume was calculated as V = 0.5*(length x width2), and was measured using digital calipers.[](https://www.ncbi.nlm.nih.gov/mesh/D007053)
## Statistical analysis
*In the **Statistical analysis** section:*
An extra sum-of-squares F test was used to compare dose response curves. Comparison of survival curves for xenograft models was done using the log-rank test. To assess the difference between calculated RR values in camptothecin-resistant HCT116 sublines, which had logarithmic error, the RR value was back-transformed to LogRR so that the error would be symmetric. The LogRR values for FL118 and SN-38 were then compared using Student’s t-test. A p-value of ≤ 0.05 was considered significant for all analyses.
Power analysis to determine appropriate group sizes for in vivo work was done with the following parameters: α = 0.05, power = 0.8.[](https://www.ncbi.nlm.nih.gov/mesh/D002166)
## Study approval
*In the **Study approval** section:*
All studies using animals were approved by the Institutional Animal Care and Use Committee (IACUC) at Roswell Park Cancer Institute.
## Additional file
*In the **Additional file** section:*
# Abbreviations
*In the **Abbreviations** section:*
NSCLC
non-small cell lung cancer
Top1
Topoisomerase I
ABC
ATP-binding cassette protein
IP
Intraperitoneal
IV
Intravenous
RR
Relative resistance
TI
Therapeutic index
Competing interests
FL118 will be further developed in Canget BioTekpharma LLC, a Roswell Park Cancer Institute spinoff company. XL, MW and FL are initial investors in Canget BioTekpharma.[](https://www.ncbi.nlm.nih.gov/mesh/C578515)
Authors’ contributions
DW participated in the conception and design of this study, performed and analyzed cell viability assays, lentiviral shRNA knockdown assays, and in vivo experiments, produced electrostatic potential maps, and participated in the preparation of this manuscript; XL conducted initial experiments which served as the basis for this study and provided training for in vivo experiments, and performed and analyzed cell viability assays; HL performed and analyzed cell viability and immunoblot assays; JW performed and analyzed cell viability and immunoblot assays and assisted in production of electrostatic potential maps; CJ provided experimental compounds, technical expertise, and critically evaluated early drafts of the manuscript; CG and MDR provided proprietary cell lines and offered technical advice on their use in this work; MW was involved in initially synthesizing most of the compounds used in this study, and provided overall technical expertise related to medicinal chemistry; FL participated in the conception and design of the study and preparation of the manuscript. All authors read and approved the final manuscript.
Authors’ information
Hong Lam and Jacob Welch were interns in the Roswell Park Cancer Institute Summer Research Program.
Dr. Wani is an emeritus member of RTI International.
# References
*In the **References** section:*
|
# Introduction
GABAergic inhibition is weakened or converted into excitation in the oxytocin and vasopressin neurons of the lactating rat
# Abstract
*In the **Abstract** section:*
Background
Increased secretion of oxytocin and arginine vasopressin (AVP) from hypothalamic magnocellular neurosecretory cells (MNCs) is a key physiological response to lactation. In the current study, we sought to test the hypothesis that the GABAA receptor-mediated inhibition of MNCs is altered in lactating rats.
Results
Gramicidin-perforated recordings in the rat supraoptic nucleus (SON) slices revealed that the reversal potential of GABAA receptor-mediated response (EGABA) of MNCs was significantly depolarized in the lactating rats as compared to virgin animals. The depolarizing EGABA shift was much larger in rats in third, than first, lactation such that GABA exerted an excitatory, instead of inhibitory, effect in most of the MNCs of these multiparous rats. [Immunohist](https://www.ncbi.nlm.nih.gov/mesh/D006096)ochemical analyses confirmed that GABAergic excitation was found in both AVP and oxytocin neurons within the MNC population. Pharmacological experiments indicated that the up-regulation of the Cl− importer Na+-K+-2Cl− cotransporter isotype 1 and the down-regulation of the Cl− extruder K+-Cl− cotransporter isotype 2 were responsible[ for](https://www.ncbi.nlm.nih.gov/mesh/D005680) the depolarizing shift of EGABA and the resultant emergence of GABAergic excitation in the MNCs of the multiparous rats.
Conclusion
We conclude that, in primiparous rats, the GABAergic inhibition of MNCs is weakened during the period of lactation while, in multiparous females, GABA becomes excitatory in a majority of the cells. This reproductive experience-dependent alteration of GABAergic transmission may help to increase the secretion of oxytocin and AVP during the period of lactation.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
GABA, a major neurotransmitter in the mammalian central nervous system (CNS), acts through ionotropic GABAA/GABAC or metabotropic GABAB receptors [1]. When GABA opens GABAA channel in a CNS neuron, chloride ion (Cl−) flows into the cell through this channel down its electrochemical gradient, inhibiting the neuron’s electrical activity [2]. However, there are conditions in which the electrochemical gradient for Cl− is set toward the extracellular side due to high intracellular Cl− concentration ([Cl−]i). In these conditions, Cl− efflux rather than influx occurs when GABAA channel is activated which can depolarize the membrane potential and even generate action potentials. For example, in pathological conditions such as hyperalgesia and hypertension, GABA exerts an excitatory effect in some neurons as a part of the etiology
[3, 4]. Moreover, in immature cortical neurons whose Cl−-extruding capability is low [5], GABA can depolarize the membrane potential to elicit action potentials [6]. This GABA-mediated excitation is thought to be important for various events occurring in CNS development including neural circuit formation. Lastly, in certain normal mature CNS neurons, GABA can elicit depolarizing/excitatory responses, but the physiological significance of this unusual GABAergic effect is unclear [7–11].[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
During the last several years, we have been interested in the question of whether GABAA receptor-mediated transmission in mature CNS neurons can switch between inhibition and excitation in response to physiological need. We have found that in the magnocellular neurosecretory cells (MNCs) of the paraventricular (PVN) and supraoptic nuclei (SON) of the rat, GABAergic inhibition is converted into excitation in a reversible fashion in response to chronic hyperosmotic stress given by 2 % NaCl as drinking solution. This switch enhances the secretion of the antidiuretic hormone arginine-vasopressin (AVP) and the natriuretic hormone oxytocin from these neurons [12]. In the current study, we investigated whether the GABAergic inhibition of MNCs is weakened or converted into excitation during the period of lactation. We envisioned that, if indeed such a change occurs, the secretion of oxytocin, a neurohormone with milk-ejecting function in lactating mammals [13], and AVP, another neurohormone which helps to prevent the mother from being dehydrated during lactation and thus maintain milk yield [14–16], would significantly increase, considering that the output of MNCs is regulated by the dense GABAergic innervation [17].
In testing this hypothesis, we utilized female rats in four conditions: 1) virgin, 2) lactating after giving the first successful birth (Lac1), 3) lactating after giving birth 3 times (Lac3) and 4) in dry period after giving birth 3 times (Dry3). We reasoned that the change in the strength or polarity of GABAergic transmission, if occurs, would be more marked in the Lac3, than Lac1, rats since milk yield and maternal behavior, both of which rely on oxytocin and AVP ([18–23], but also see [24, 25]), are known to be enhanced in multiparous mammals compared to primiparous ones [26–29].[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## Results
*In the **Results** section:*
## Milk yield was greater in multiparous rats
*In the **Milk yield was greater in multiparous rats** section:*
Difference between Lac1 and Lac3 rats in milk yield and pup growth. (a) Bar charts showing difference between Lac1 and Lac3 rats in milk yield at two different time points when the average body weights of their pups were 13 g and 29 g. The milk yield is defined in the Results section. Data from 128 pups nursed by 8 Lac1 and 8 Lac3 dams. Each dam nursed a litter of 8 pups for 1 h. (b) Line graphs showing the difference between the pups of Lac1 and Lac3 rats in their weight gain over the period of PND 0–15 (*: p < 0.05; **: p < 0.001; Student’s t-test). n: the number of pups. In this and the rest of the figures, values are shown as means (± SEM)
To determine if milk production is larger in Lac3 than Lac1 rats, we compared the milk yields of these groups of rats at two different time points, i.e., when the average body weights of their pups were 13 g and 29 g. To measure the milk yield, we let the dam nurse a litter of 8 pups for 1 h. The milk yield was defined as the difference in body weight of the pup measured 4 h before and immediately after the nursing. Milk yield was significantly greater in Lac3, than Lac1, rats at both time points (Fig. 1a). As an indirect index of milk production, the body weight of pup was also monitored over postnatal day (PND) 0–15. The body weight of Lac3 pups was significantly greater than that of Lac1 pups at PND 3–15 (Fig. 1b).
Collectively, these results indicated that milk yield was greater in multiparous rats.
## GABA was excitatory in most of the MNCs of rats in third lactation
*In the **GABA was excitatory in most of the MNCs of rats in third lactation** section:*
Conversion of GABAA receptor-mediated inhibition to excitation in the MNCs of lactating rats. (a) Spontaneously occurring fast GABAergic excitatory postsynaptic potentials (EPSPs, ●) and inhibitory postsynaptic potentials (IPSPs, ○) recorded in the presence of AP-5 (100 μM) and DNQX (20 μM) from the MNCs of Lac3 (upper trace) and virgin rats (lower trace). Note the action potentials that arise from the GABAergic EPSPs (upper trace). (b) Proportions of the MNCs having GABAergic EPSPs and IPSPs in different groups of rats. (**: p < 0.001, Chi-square test). n: the number of cells examined. These groups of rats were not significantly different from each other in the input resistance of MNC (Virgin: 308 ± 12 MΩ, n = 38; Lac1: 320 ± 15 MΩ, n = 30; Lac3: 303 ± 12 MΩ, n = 56; Dry3: 313 ± 12 MΩ, n = 22). (c) Depolarization and associated action potentials elicited by focal application of the GABAA agonist muscimol (10 μM; 10 ms) in an MNC of Lac3 rat[](https://www.ncbi.nlm.nih.gov/mesh/D015763)
Next, we examined fast GABAergic postsynaptic potentials (PSPs) occurring spontaneously in the MNCs sampled in the SON slices of the Virgin, Lac1, Lac3 and Dry3 groups of rats. These PSPs were recorded in the presence of DL-2-amino-5-phosphonopentanoic acid (AP-5, 100 μM) and 6,7-dinitroquinoxaline-2,3-dione (DNQX, 20 μM), with the use of gramicidin-perforated recording technique which preserves the [Cl−]i of the recorded cell [30]. The GABAergic PSPs were inhibitory in all MNCs studied in 13 virgin (n = 38 cells; Fig. 2a, lower panel) and 8 Dry3 rats (n = 22 cells), and in 28 of 30 cells examined in 12 Lac1 rats (Fig. 2b). On the other hand, they were excitatory in 37 of 56 MNCs recorded in 16 Lac3 rats (Fig. 2a, upper panel; b), and in 2 of 30 cells from Lac1 rats. The excitatory PSPs were blocked by the GABAA receptor antagonist bicuculline (30 μM; n = 5; example trace not shown) and mimicked by the GABAA receptor agonist muscimol (10 μM, 10 ms; n = 12; Fig. 2c).
Since GABAergic excitation occurs virtually in all of the MNCs of rats subjected to chronic hyperosmotic stress [12], we checked the plasma osmolality of Lac3 rats and found it to be in the normal range (310 ± 3.0 mOsm/kg H2O, n = 5). Collectively, these data indicate that GABAA receptor-mediated inhibitory synaptic transmission is converted into excitatory one in most MNCs of the multiparous lactating rats, but a similar change in GABAergic transmission does not occur in primiparous animals. In addition, they suggest that the GABAergic excitation in Lac3 rats does not result from osmotic stress, and that it is reversed to inhibition after the end of lactation.[](https://www.ncbi.nlm.nih.gov/mesh/D015763)
## A large depolarizing EGABA shift exceeding the action potential threshold was responsible for the emergence of GABAergic excitation in the MNCs of the rats in third lactation
*In the **A large depolarizing EGABA shift exceeding the action potential threshold was responsible for the emergence of GABAergic excitation in the MNCs of the rats in third lactation** section:*
Depolarizing shift of EGABA in the MNCs of multiparous lactating rats. (a) Estimation of EGABA with the use of the currents elicited by focally applied muscimol (10 μM, 10 ms) at various holding potentials (VH) in the MNCs of virgin (inset; upper left) and Lac3 rats (inset; lower right). These current traces were obtained after the blockade of fast sodium current and glutamatergic transmission with the cocktail of tetrodotoxin (0.5 μM), AP-5 (100 μM) and DNQX (20 μM). Peak amplitudes of the muscimol-elicited currents are plotted against VH. Linear regression was used to fit the data points. The intersections (,) of the regression lines with the abscissa were taken as the reversal potentials of the muscimol-elicited responses (i.e., EGABA’s). (b) Dot plots and bar graphs showing the ranges and means (±SEM) of the EGABA’s of the MNCs of Virgin, Lac1, Lac3 and Dry3 rat groups. Holm-Sidak all pairwise comparison tests performed after one-way ANOVA (p < 0.001) indicated that these rat groups were different from one another in the mean EGABA. This was denoted with different symbols associated with the bar graphs. n: the number of cells examined[](https://www.ncbi.nlm.nih.gov/mesh/D009118)
In order to identify the neurophysiological basis for the emergence of GABAergic excitation in the MNCs of Lac3 rats, we next estimated the EGABA in these neurons, with the use of the currents elicited at various holding potentials by focal application of muscimol (10 μM, 10 ms) in the presence of tetrodotoxin (0.5 μM) (Fig. 3a). As expected, the EGABA was found to be positive to the action potential threshold (which was about −45 mV) in most of the MNCs of Lac3 rats (37 of 56 cells from 16 rats); the EGABA of these cells was −45.1 ± 1.4 mV (n = 56), which was significantly less negative than the EGABA’s of the MNCs of virgin (−68.4 ± 1.2 mV; n = 38 neurons from 13 rats), Lac1 (−61.7 ± 2.0 mV; n = 30 neurons from 12 rats) and Dry3 rats (−55.4 ± 1.4 mV; n = 22 neurons from 8 rats) (Fig. 3b). Interestingly, EGABA was significantly less negative in the MNCs of Lac1 and Dry3, than virgin, groups of rats, and in the cells of Dry3, than Lac1, group (Fig. 3b). Taken together, these results indicate that a large depolarizing shift of EGABA surpassing the action potential threshold is responsible for the emergence of GABAergic excitation in the MNCs of multiparous lactating rats. In addition, they suggest that GABAergic inhibition is weakened in the MNCs of Lac1 rats and, to a greater extent, in the cells of Dry3 animals.[](https://www.ncbi.nlm.nih.gov/mesh/D009118)
## GABAergic excitation occurred in both the AVP and oxytocin neurons of multiparous lactating rats
*In the **GABAergic excitation occurred in both the AVP and oxytocin neurons of multiparous lactating rats** section:*
IHC identification of MNCs recorded in SON slices. (a, b) Double IHC staining for AVP-neurophysin and oxytocin-neurophysin in the MNCs of Lac3 rats injected with biocytin at the end of the recording. The biocytin-labeled cell expresses AVP-neurophysin in (a), whereas OXY-neurophysin in (b). Asterisks indicate recorded cells
To see whether the MNCs showing GABAergic excitation in Lac3 rats were AVP or oxytocin neurons, we performed double immunohistochemistry (IHC) for AVP and oxytocin neurophysin in 16 MNCs recorded in SON slices from 8 Lac3 rats. The recorded cells were marked with the use of biocytin in the internal solution. Our IHC analyses showed that the biocytin-labeled cells were positive for AVP (n = 11 cells) or oxytocin neurophysin (n = 5 cells) (Fig. 4), indicating that GABAergic excitation occurs in both AVP and oxytocin neurons.
## Depolarizing shift of EGABA and the resultant emergence of GABAergic excitation were due to the up-regulation of Na+-K+-2Cl− cotransporter isotype 1 (NKCC1) and down-regulation of K+-Cl− cotransporter isotype 2 (KCC2)
*In the **Depolarizing shift of EGABA and the resultant emergence of GABAergic excitation were due to the up-regulation of Na+-K+-2Cl− cotransporter isotype 1 (NKCC1) and down-regulation of K+-Cl− cotransporter isotype 2 (KCC2)** section:*
The up-regulation of NKCC1 and the down-regulation of KCC2 are responsible for the depolarizing shift of EGABA. (a-d) Graphs showing the impact of NKCC inhibitor bumetanide (BM; 10 μM) and the KCC2 blocker VU0463271 (VU; 5 μM) on the EGABA of the MNCs of the Virgin, Lac1, Lac3 and Dry3 groups of rats. The symbols connected by lines denote data from the same cells. Holm-Sidak all pairwise comparison tests were performed after one-way repeated measures ANOVA (p = 0.001-0.04). Asterisks indicate significant difference from the data obtained before the drug application, while double daggers denote significant difference from the data obtained during the drug application. (e) Bar charts summarizing the hyperpolarizing effects of bumetanide and the depolarizing effects of VU0463271 on the EGABA of the MNCs of different groups of rats. Holm-Sidak all pairwise comparison tests were performed after one-way ANOVA (p < 0.001). Asterisk indicates significant difference of the value of Lac3 group from those of other groups[](https://www.ncbi.nlm.nih.gov/mesh/D002034)
It is generally agreed that the [Cl−]i is a major determinant for EGABA and the polarity and strength of GABAA receptor-mediated PSP, and that NKCC1 and KCC2 play a pivotal role for Cl− homeostasis in CNS neurons [6]. In the next set of experiments, we examined whether and how these co-transporters contributed to the depolarizing shift of EGABA in the MNCs of Lac3 rats, by comparing virgin, Lac1, Lac3 and Dry3 rats in terms of the effects of the NKCC inhibitor bumetanide and the selective KCC2 inhibitor VU0463271 on the EGABA. Bumetanide (10 μM) and VU0463271 (5 μM) reversibly hyperpolarized and depolarized the EGABA, respectively, in all groups of rats (Figs. 5a-d).
However, the effects of these drugs were not uniform in magnitude across different rat groups; while bumetanide hyperpolarized the EGABA most in the Lac3 group (Fig. 5e, upper panel), VU0463271 depolarized the EGABA to a greater extent in the Virgin, Lac1 and Dry3, than Lac3, groups (Fig. 5e, lower panel). The Virgin, Lac1 and Dry3 groups were not significantly different from one another with regard to the magnitudes of the bumetanide and VU0463271 effects (Fig. 5e). Thus, these results collectively suggested that the depolarizing shift of EGABA in the MNCs of Lac3 rats arises from the combination of NKCC1 up-regulation and KCC2 down-regulation.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
## Discussion
*In the **Discussion** section:*
In the current study we examined GABAA receptor-mediated transmission in rat MNCs. We found that, during the period of lactation, the EGABA’s of MNCs depolarize significantly such that the GABAergic inhibition of these cells was weakened in primiparous animals while mostly converted into excitation in multiparous females. Furthermore, we obtained evidence that the inhibitory-to-excitatory switch in GABAergic transmission is reversed after the cessation of lactation. Thus, the results of the current study indicate that the GABAergic responses of rat MNCs are reversibly modulated during lactation by the depolarizing shift of EGABA and that this plastic change is more marked in animals with more reproductive experiences.
The switch from GABA inhibition to excitation in the MNCs would be expected to increase secretion of oxytocin and AVP in lactating females. Brussaard and his co-workers [31, 32] have found that, around parturition and during the period of lactation in the rat, the inhibitory effect of GABA on oxytocin neurons is attenuated by the postsynaptic alteration of GABAA receptor subunit composition and the consequent removal of the potentiating effects of neurosteroids on the function of this receptor. Thus, they postulated that the reduction of GABAergic restraint on oxytocin neurons underlies the increased oxytocin secretion at particular times of the reproductive cycle in the rat. On the other hand, Moos [33] has reported that, although GABA suppresses the baseline electrical activities of oxytocin neurons, it paradoxically facilitates the suckling-induced bursting activities of these cells to enhance their outputs. These interpretations will need to be re-examined in light of our observations of GABA-mediated excitation in the experienced lactating females.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
The present study demonstrates that the EGABA is significantly less negative in the cells of Dry3, than virgin and Lac1 rats. These findings indicate that the depolarizing shift of EGABA in multiparous lactating rats is not fully reversed after the end of lactation although the EGABA gets repolarized enough to prevent the occurrence of GABAergic excitation. We do not know why, after the end of lactation, the EGABA does not return to baseline levels. It is possible that the GABA excitation would fully reverse in time. We speculate that less negative EGABA in multiparous dry rats may help to speed up the change in GABAergic transmission that has to occur in the next lactation and that this plasticity may be a salient feature of the brain of the rat with multiple reproductive experiences
[34, 35].[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
The transmembrane Cl− gradient is a critical factor determining the strength and the polarity of GABAA receptor-mediated synaptic response. In CNS neurons, [Cl−]i is regulated by the Cl− importer NKCC1 and the Cl− extruder KCC2 [36]. In this study, we provided neurophysiological evidence that the up-regulation of NKCC1 and the down-regulation of KCC2 are responsible for the depolarizing shift of EGABA and the resultant emergence of GABAergic excitation in the MNCs of multiparous lactating rats. However, we did not identify the mechanisms linking lactation to the changes in NKCC1 and KCC2.
It is well established that AVP and oxytocin are released in the SON and PVN and this local release is enhanced during the period of lactation [37] when AVP can act as a paracrine signal to induce inward current in oxytocin neurons [38]. In addition, we have previously shown that the intracerebroventricular administration of selective oxytocin receptor antagonist partially obstructs the depolarizing EGABA shift, which is induced by NKCC1 up-regulation, and the consequent emergence of GABAergic excitation in the MNCs of rats subjected to chronic hyperosmotic stress [12]. Meanwhile, a recent study reported that, in a rat model of hypertension produced by salt loading, brain-derived neurotrophic factor (BDNF)-tropomyosin-receptor-kinase B (TrkB) activation causes the down-regulation of KCC2 and the depolarizing shift of EGABA in MNCs [39].
Thus, it is possible that the somato-dendritic release of oxytocin and the BDNF-TrkB activation are the mechanisms underlying the NKCC1 up-regulation and KCC2 down-regulation in the MNCs of multiparous rats.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
## Conclusion
*In the **Conclusion** section:*
In the mammal, the secretion of oxytocin and AVP from MNCs is increased during lactation. In the current study we found that the EGABA’s of MNCs depolarize during the period of lactation such that the GABAergic inhibition of these cells was weakened in primiparous rats and mostly converted into excitation in multiparous females. Furthermore, we obtained evidence that the inhibitory-to-excitatory switch in GABAergic transmission, which is driven by a combination of NKCC1 up-regulation and KCC2 down-regulation, is reversed after the cessation of lactation. We conclude that the GABAergic responses of rat MNCs are modulated reversibly during lactation, perhaps to enhance the secretion of AVP and oxytocin.
## Materials and methods
*In the **Materials and methods** section:*
## Animal care
*In the **Animal care** section:*
Female Sprague–Dawley rats (250–400 g) from Orient Bio Co (Sungnam, Korea) were used in the current study. They were housed in a temperature-controlled vivarium (22-24 °C) with a 12/12-h light/dark cycle. The experimental procedures described below were approved by the Institutional Animal Care and Use Committee and conformed to the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All possible efforts were made to minimize the number of animals used as well as their suffering.
## Animal groups
*In the **Animal groups** section:*
Four groups of rats were used in this study: virgin rats (Virgin), primiparous lactating rats (Lac1) or rats in third lactation (Lac3) and rats in dry period after giving birth 3 times (Dry3). Virgin and Dry3 groups of rats were killed for hypothalamic slices in ≥7 days after their arrival at the vivarium, while Lac1 and Lac3 groups of rats were sacrificed after giving births to and then nursing their pups (n = 8-12) for 3–14 days in the vivarium. Dry3 rats had not been pregnant for 1–3 months after the end of last lactation.
## Hypothalamic slice preparation
*In the **Hypothalamic slice preparation** section:*
Hypothalamic slices were prepared as previously described [40]. In brief, the rat was anesthetized with urethane (1.25 g/kg, i.p.), and the brain was quickly excised from the skull and submerged in ice-cold artificial cerebrospinal fluid (ACSF; composition in mM: 124 NaCl, 1.3 MgSO4, 3 KCl, 1.25 NaH2PO4, 26 NaHCO3, 2.4 CaCl2, and 10 glucose). After being chilled for 1–2 min, the brain was trimmed to a block containing the hypothalamus. Using a vibroslicer (Campden Instruments, Loughborough, United Kingdom), coronal slices (350 μm) containing the SON were cut from the tissue block in ice-cold ACSF. The slices were transferred to a gas-interface type recording chamber, which was perfused with warm (34-35 °C) aerated (95 % O2/5 % CO2) ACSF at a rate of 0.5-1 mL/min by a peristaltic pump-driven or gravity-fed bath perfusion system [9].
Warm (34-35 °C) air humidified by 95 % O2/5 % CO2 gas mixture was continuously blown over the slices to further ensure adequate oxygenation of cells in the tissue.[](https://www.ncbi.nlm.nih.gov/mesh/D014520)
## Intracellular electrophysiological recording
*In the **Intracellular electrophysiological recording** section:*
Current or voltage clamp recordings were obtained from neurons in the SON of hypothalamic slices equilibrated for 1–8 h in the recording chamber as described previously [12]. In brief, the SON was identified as a translucent region right next to the optic chiasm. Micropipettes (tip diameter, 1.5-2.0 μm; 3–5 MΩ) pulled from borosilicate glass capillaries (P-97; Sutter Instrument Co, Novato, CA) and filled with gramicidin (50 μg/mL)-containing solution (composition in mM: 143 K-gluconate, 2 KCl, 10 HEPES, and 0.5 EGTA; pH 7.2-7.3) were used for recording in a perforated configuration. Stable perforated recording condition was usually achieved 10–25 min after seal was formed. Those recordings having steady series resistances (range 30–50 MΩ) and action potential amplitudes of >45 mV (measured from action potential threshold) were the only ones included in the data pool.
The voltage errors resulting from the series resistance were compensated offline for voltage clamp recordings and online for current clamp recordings by using the bridge circuit. We corrected the liquid junction potential before the experiments; we set the pipette potential to −9 mV just before the formation of patch configuration, knowing that the liquid junction potential was 15.8 mV (at 34.5 °C) while the perforated patch potential arising from gramicidin perforation was −6.8 mV. We assumed that the change in resting membrane potential detected when the recording mode was transformed from perforated to whole-cell configuration represented the perforated patch potential. The signals from neurons amplified by Axoclamp-2B amplifier (bandwidth filter set at 10 Hz) were digitized and sampled at 50 μs intervals (Digidata1320, pClamp 8.0; Molecular Devices, Sunnyvale, CA).[](https://www.ncbi.nlm.nih.gov/mesh/D017640)
## Drugs
*In the **Drugs** section:*
We purchased all the drugs and chemicals used in the current study from Sigma-Aldrich, except for muscimol (GABAA receptor agonist; Ascent Scientific, Cambridge, MA) and VU0463271 (KCC2 blocker; gift from Prof. Craig Lindsley, Vanderbilt University in Nashville, TN, USA). We prepared the solutions of muscimol and AP-5 (NMDA receptor antagonist) by dissolving these drugs in ACSF, the standard slice perfusion medium, and DNQX (non-NMDA receptor antagonist) solution by diluting its dimethylsulfoxide-based stock solution with ACSF (final concentrations of dimethylsulfoxide, 0.05 %). The solution of bumetanide (NKCC blocker) and VU0463271 was prepared by dissolving this agent in AP-5 and DNQX containing ACSF. Muscimol solution was applied focally by “Y-tube” method [41],
while other drug solutions by bath perfusion.[](https://www.ncbi.nlm.nih.gov/mesh/D009118)
## IHC identification of recorded neurons
*In the **IHC identification of recorded neurons** section:*
For post-hoc IHC identification of SON neurons recorded in slices from the Lac3 rats, the cells were infused with biocytin; biocytin (1.5 mg/mL) contained in the recording pipette was allowed to get into the cell by holding the cell in whole-cell mode for 10–15 min after perforated patch recordings. The SON slices were fixed in 4 % paraformaldehyde-containing phosphate buffered saline (PBS, 0.1 M) at pH 7.4 for 24 h, transferred to 30 % sucrose-PBS and kept in this solution for 24–48 h. Then, they were cut into 25–35 μm-thick sections. These sections were incubated for 24 h at 4 °C in a solution containing a rabbit polyclonal antibody against AVP-neurophysin (1:200; Abcam, Cambridge, MA) and a mouse monoclonal antibody against oxytocin-neurophysin (1:5000; Abcam). After being washed with PBS three times, the sections were reacted with DyLight 488-conjugated goat anti-rabbit and DyLight 594-conjugated goat anti-mouse secondary antibodies (1:200 dilution each; Jacson ImmunoResearch, West Grove, PA) for 24 h at 4 °C, and then with avidinAMCA (1:500 dilution; Vector Labs, Burlingame, CA) for 1 h at room temperature.
The secondary antibodies and avidinAMCA were dissolved in 0.1 M PBS containing 0.3 % Triton X-100 and 2 % normal goat serum. The sections were examined under a confocal fluorescence microscope for the presence of AVP-neurophysin and oxytocin-neurophysin immunoreactivity and biocytin labeling.[](https://www.ncbi.nlm.nih.gov/mesh/C013411)
## Statistical analysis
*In the **Statistical analysis** section:*
Numerical data are expressed as the mean ± SEM. Student’s t test was used for the comparison of two independent datasets with normal distribution. One-way analysis of variance (ANOVA) and one-way repeated measures ANOVA were performed to compare multiple independent and dependent datasets with normal distributions, respectively. The pairwise comparisons following the ANOVA’s were done with Holm-Sidak method. Chi-square test was performed to determine whether there is a significant difference between the observed frequencies in one or more categories.
P < 0.05 was considered significant. This significance level, however, was reduced with the Bonferroni correction when the problem of multiple comparisons arises.
# Abbreviations
*In the **Abbreviations** section:*
ACSF
Artificial cerebrospinal fluid
ANOVA
One-way analysis of variance
AP-5[](https://www.ncbi.nlm.nih.gov/mesh/D015763)
DL-2-amino-5-phosphonopentanoic acid[](https://www.ncbi.nlm.nih.gov/mesh/D015763)
AVP
Arginine vasopressin
BDNF
Brain-derived neurotrophic factor
DNQX[](https://www.ncbi.nlm.nih.gov/mesh/C056723)
6,7-dinitroquinoxaline-2,3-dione[](https://www.ncbi.nlm.nih.gov/mesh/C056723)
EGABA
Reversal potential of GABAA receptor-mediated response
IHC
Immunohistochemistry
KCC2
K+-Cl− cotransporter isotype 2
MNCs
Magnocellular neurosecretory cells
NKCC1
Na+-K+-2Cl− cotransporter isotype 1
PBS[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
Phosphate buffered saline[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
PND
Postnatal day
PVN
Paraventricular nucleus
SON
Supraoptic nucleus
TrkB
Tropomyosin-receptor-kinase B
Seung Won Lee and Young-Beom Kim contributed equally to this work.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
S.W.L., H.C.H., Y.-W.C. and Y.I.K. conceived this project. S.W.L., Y.-B.K., J.S.K., W.B.K. and Y.S.K. performed the experiments and analyzed the results. S.W.L., C.S.C., Y.-W.C. and Y.I.K. wrote the manuscript. All authors read and approved the final manuscript.
# References
*In the **References** section:*
|
# Introduction
A novel cancer immunotherapy based on the combination of a synthetic [carbohydrate](https://www.ncbi.nlm.nih.gov/mesh/D002241)-pulsed dendritic cell vaccine and glycoengineered cancer cells
# Abstract
*In the **Abstract** section:*
Immune tolerance to tumor-associated carbohydrate antigens (TACAs) has severely restricted the usefulness of most TACAs. To overcome this problem, we selected a sial[ylated trisaccharide TACA, GM3, as a t](https://www.ncbi.nlm.nih.gov/mesh/D015295)ar[get a](https://www.ncbi.nlm.nih.gov/mesh/D015295)ntigen, and tested a new immunotherapeutic strate[gy by](https://www.ncbi.nlm.nih.gov/mesh/D015295) combining metabolic bioengineering with dendritic ce[ll (DC) vacci](https://www.ncbi.nlm.nih.gov/mesh/D014312)n[atio](https://www.ncbi.nlm.nih.gov/mesh/D015295)n.[ We](https://www.ncbi.nlm.nih.gov/mesh/D005679) engineered cancer cells to express an artificial structure, N-phenylacetyl-D-neuraminic acid, in place of the natural N-acetyl-D-neuraminic acid of GM3 by using N-phenylacetyl-D-mannosamine (ManNPhAc) as a[ biosynthetic precursor. Next, w](https://www.ncbi.nlm.nih.gov/mesh/D009438)e selectively targeted the[ bioengineered cancer cell](https://www.ncbi.nlm.nih.gov/mesh/D019158)s by[ va](https://www.ncbi.nlm.nih.gov/mesh/D005679)ccination [with DCs pulsed with the GM3](https://www.ncbi.nlm.nih.gov/mesh/C510175) N[-phenyla](https://www.ncbi.nlm.nih.gov/mesh/C510175)cetyl derivative. Vaccination with GM3NPhAc-KLH-loaded DCs elicited robust GM3NPhAc-specific T cell-dependent immunity. The results [sho](https://www.ncbi.nlm.nih.gov/mesh/D005679)wed that this strategy could significantly in[hibit FBL3 t](https://www.ncbi.nlm.nih.gov/mesh/C032808)umor growth and prolong the [survival](https://www.ncbi.nlm.nih.gov/mesh/D005732) of tumor-bearing mice; B16F10 lung metastases could also be reduced. These findings lay out a new strategy for overcoming immune tolerance to TACAs, such as GM3, for the development of effective tumor immunotherapies.[](https://www.ncbi.nlm.nih.gov/mesh/D015295)
## INTRODUCTION (cont.)
*In the **INTRODUCTION (cont.)** section:*
Immunotherapy is an attractive approach for the treatment of cancer, especially for patients with tumor metastases [1]. One of the most promising approaches to cancer immunotherapy is the use of antigen presenting cells (APC), such as dendritic cells (DC), loaded with tumor-associated antigens (TAAs) because of the potentially superior efficacy and specificity of this approach [2, 3]. However, a major obstacle in the field is the development of immune tolerance to TAAs, which results in the failure of natural TAAs or vaccines formulations with natural TAAs to induce sufficiently effective immune responses. Traditional approaches to improve the immunogenicity of natural TAAs, such as the conjugation of TAAs with large carrier molecules or the coadministration of potent immunological adjuvants, have achieved limited success.
Even when vaccines made of TAAs provoke immune responses, they usually only elicit B cell-mediated immunity, rather than the more desirable antitumor T cell immune responses [4–6]. Consequently, the development of a method to overcome the problem of immune tolerance to TAAs will be critical for the success of new cancer vaccines or immunotherapies.
GM3, a sialylated trisaccharide tumor-associated carbohydrate antigen (TACA), is overexpressed by many types of tumors, such as melanoma, leukemia, and pulmonary cancer [7–12]. Recently, we synthesized several GM3 derivatives and demonstrated that unnatural GM3 derivatives, in particular N-phenylacetyl GM3 (GM3NPhAc), were more immunogenic than native GM3 and could evoke robust antigen-specific T cell-dependent immunity [13, 14]. For cancer immunotherapy, we exploited a new strategy for selectively targeting the immune response towards cancer cells based on bioengineering. Specifically, we engineered cancer cells to express GM3NPhAc in place of the natural GM3 on the cancer cell surface by using N-phenylacetyl-D-mannosamine (ManNPhAc) as a biosynthetic precursor for GM3NPhAc.
We have shown that several murine and human tumor cell lines can be metabolically glycoengineered to express GM3NPhAc in vitro [15, 16]. Furthermore, in vivo ManNPhAc treatment results in abundant GM3NPhAc expression on tumor tissues, but not on normal tissues in tumor-bearing mice [17]. More importantly, bioengineered cancer cells can be selectively targeted by specific immune reactions evoked by conjugate vaccines containing GM3NPhAc [15–17].[](https://www.ncbi.nlm.nih.gov/mesh/D005679)
In this study, DCs from murine bone marrow loaded with a conjugate of GM3NPhAc and keyhole limpet hemocyanin (GM3NPhAc-KLH) were tested for therapeutic efficacy against cancer in combination with ManNPhAc administration. DCs loaded with GM3NPhAc-KLH could induce robust antigen-specific T cell-dependent immunity. The GM3NPhAc-specific antisera could mediate high cytotoxicity to ManNPhAc-treated B16F10 and FBL3 cells. Lymphocytes isolated from the spleen also showed specific cytotoxicity to the glycoengineered tumor cells. More importantly, the immunity induced by GM3NPhAc-KLH-loaded DCs in combination with ManNPhAc treatment could significantly inhibit tumor growth and metastasis, and also prolong the survival of tumor-bearing mice.[](https://www.ncbi.nlm.nih.gov/mesh/D005732)
## RESULTS
*In the **RESULTS** section:*
## Costimulatory molecule expression and IL-12 production by GM3NPhAc-KLH-pulsed DCs
*In the **Costimulatory molecule expression and IL-12 production by GM3NPhAc-KLH-pulsed DCs** section:*
Day 5 mouse bone marrow-derived DCs displayed typical morphological characteristics. The DCs pulsed for 24 h with GM3NPhAc-KLH or KLH expressed higher levels of the costimulatory molecules CD80 and CD86, and secreted higher amounts of IL-12p70 compared with unpulsed DCs. LPS activated-DCs were used as a positive control (Figure 1A and 1B).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
Costimulatory molecule expression and IL-12 production by murine bone marrow-derived DCs
DCs were cultured for 24 h with 20 μg/mL KLH, 20 μg/mL GM3NPhAc-KLH, 500 ng/mL LPS, or remained untreated (medium alone). (A) Cells were analyzed by flow cytometry. (B) IL-12p70 secreted in supernatants was measured by standard ELISA; n = 3; *P < 0.05 vs untreated controls (medium alone).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## The induction of anti-GM3NPhAc antibodies by GM3NPhAc-KLH-pulsed DCs
*In the **The induction of anti-GM3NPhAc antibodies by GM3NPhAc-KLH-pulsed DCs** section:*
To evaluate the ability of GM3NPhAc-KLH-pulsed DCs to induce GM3NPhAc-specific antibodies, ELISA was used to measure antibody levels in the sera of vaccinated mice. Levels of GM3NPhAc-specific total antibodies (Igκ) and IgG in sera of mice immunized with GM3NPhAc-KLH-DCs increased significantly (P < 0.05) compared with those in KLH-DC-vaccinated mice (Figure 2A). Similar results were also observed in B16F10-bearing mice. Interestingly, ManNPhAc treatment, which was used to metabolically glycoengineer cancer cells to express GM3NPhAc, could further increase levels of GM3NPhAc-specific total antibodies and IgG (Figure 2B).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
GM3NPhAc-specific total antibody (Igκ) and IgG levels in the sera of immunized mice[](https://www.ncbi.nlm.nih.gov/mesh/D005732)
(A) C57BL/6 mice were immunized three times at weekly intervals by s.c. injection of 1 × 106 GM3NPhAc-KLH-DCs or KLH-DCs. Sera were prepared on the day following the final boost. (B) Mice were immunized as in (A) One week after the third immunization, each mouse was injected i.v. with 5 × 105 B16F10 cells, followed by daily i.p. injections of ManNPhAc (50 mg/kg/day) for 7 days. Sera were prepared at the end of experiments, i.e. on day 42. Antibodies were assayed by ELISA as described in the Materials and Methods section. For each data set, the mean ± SD for 10 mice per group is shown.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## ADCC and CDC by immune sera towards metabolically glycoengineered cancer cells
*In the **ADCC and CDC by immune sera towards metabolically glycoengineered cancer cells** section:*
To assess whether the GM3NPhAc-KLH-DC-vaccinated sera could mediate the killing of glycoengineered cancer cells, B16F10 and FBL3 cells were incubated with various concentrations of ManNPhAc for 72 h. ADCC and antibody-mediated CDC were then measured in vitro. Our findings clearly demonstrated that the anti-GM3NPhAc antisera mediated significant cytotoxicity towards ManNPhAc-treated FBL3 cells using either mouse peritoneal macrophages (Figure 3A) or rabbit complement (Figure 3B) as effectors. Similar results were obtained using ManNPhAc-treated B16F10 cells (data not shown). However, anti-GM3NPhAc antisera could not mediate the killing of B16F10 or FBL3 cells (Figure 3C).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
Sera from GM3NPhAc-KLH-DC-vaccinated mice mediated the killing of glycoengineered cancer cells by ADCC or CDC[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
(A, B) FBL3 cells were incubated for 72 h with 0, 0.01, 0.02, 0.04, 0.08, 0.16, or 0.32 mM of ManNPhAc. ADCC assays (A) were performed using peritoneal macrophages from healthy mice as effectors with an effector-to-target cell ratio of 100:1. For antibody-mediated CDC assays (B), rabbit complement was used as an effector. (C) Anti-GM3NPhAc antisera did not mediate the killing of B16F10 or FBL3 cells by CDC. Cell lysis was evaluated by LDH assays. Each data set indicates the mean ± SD of experiments performed in triplicate.[](https://www.ncbi.nlm.nih.gov/mesh/C510175)
## Induction of GM3NPhAc-specific CTL responses by GM3NPhAc-KLH-pulsed DCs
*In the **Induction of GM3NPhAc-specific CTL responses by GM3NPhAc-KLH-pulsed DCs** section:*
To evaluate whether DCs pulsed with GM3NPhAc-KLH were capable of inducing CTL responses, GM3NPhAc-KLH-DCs were injected into naive mice. Then, splenocytes from immunized mice were harvested and restimulated for 24 h with GM3NPhAc-KLH for CTL assays. Splenocytes from GM3NPhAc-KLH-DC-immunized mice displayed significant in vitro cytotoxicity towards metabolically glycoengineered B16F10 and FBL3 cells (Figure 4A).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
GM3NPhAc-KLH-DCs induce CTL responses[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
(A) Mice were immunized as in Figure 2A. Splenocytes from GM3NPhAc-KLH-DC- or KLH-DC-immunized mice were restimulated for 24 h with the corresponding antigens. CTL activity of splenocytes towards glycoengineered B16F10 or FBL3 cells was analyzed by LDH assays at an effector-to-target cell ratio of 100:1; n = 10; *P < 0.05 vs the KLH-DC group. (B) Splenocytes from mice treated as in Figure 2B were tested for CTL activity; n = 10; *P < 0.05 vs the KLH-DC group. (C) ELISPOT assays to determine the number of IFN-γ-secreting lymphocytes from mice treated as in Figure 2B; n = 10; *P < 0.05 vs the KLH-DC group; # P < 0.05 vs the GM3NPhAc-KLH-DC and ManNPhAc group. The figure shows results that are representative of four mice.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
To analyze GM3NPhAc-KLH-DC-induced CTL activity in tumor-bearing mice, splenocytes were harvested from B16F10 tumor-bearing mice treated with GM3NPhAc-KLH-DCs and ManNPhAc and were directly used in CTL assays. ManNPhAc-treated B16F10 cells were selectively targeted and killed by these splenocytes (Figure 4B). IFN-γ ELISPOT assays showed that GM3NPhAc-KLH-DC immunizations could markedly increase the number of IFN-γ-producing splenocytes. ManNPhAc treatment could further increase the number of IFN-γ-producing splenocytes (Figure 4C).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## Induction of immune-mediated protection in mice vaccinated with combined GM3NPhAc-KLH-DCs and ManNPhAc treatment
*In the **Induction of immune-mediated protection in mice vaccinated with combined GM3NPhAc-KLH-DCs and ManNPhAc treatment** section:*
To test the immune-mediated protection generated by GM3NPhAc-KLH-DC vaccination in vivo, we injected mice three times with GM3NPhAc-KLH-DCs at weekly intervals. Mice were inoculated with FBL3 cells by s.c. injection, followed by daily i.p. injections of ManNPhAc (50 mg/kg/day) for 7 days. Mice received GM3NPhAc-KLH-DCs in PBS, or control mice received KLH-DCs. Tumor size was measured using a caliper ruler. Tumors in both groups of control mice grew progressively and developed into palpable tumors (about 1 mm in diameter) 10 days earlier than those in the group treated with GM3NPhAc-KLH-DCs and ManNPhAc (Figure 5A).
The tumor size in the treatment group was significantly smaller than that in both control groups (P < 0.05). Moreover, GM3NPhAc-KLH-DCs and ManNPhAc treatments could significantly prolong the survival time (P < 0.05) of tumor-bearing mice (Figure 5B). All of the mice in the control groups died by day 54. However, 75% of the mice in the treatment group were alive at that time point. The average survival time was 42.6 days for the GM3NPhAc-KLH-DCs and PBS group, and >61.0 days for the GM3NPhAc-KLH-DCs and ManNPhAc group when the experiment was ended on day 66.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
Evaluation of a novel immunotherapy to treat FBL3 cancer
(A) Tumor sizes and growth rates in mice treated with KLH-DCs, GM3NPhAc-KLH-DCs and PBS, or GM3NPhAc-KLH-DCs and ManNPhAc; *P < 0.05, as compared with the other two treatment groups. (B) The survival of tumor-bearing mice treated with KLH-DCs, GM3NPhAc-KLH-DCs and PBS, or GM3NPhAc-KLH-DCs and ManNPhAc.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
We further used the B16F10 lung metastasis model to evaluate the efficacy of the new immunotherapy regimen. In this experiment, mice were treated as described above except that they were also challenged with intravenously (i.v.) injected B16F10 cells. The number of lung nodules was counted 42 days after the initial immunization. Vaccination with GM3NPhAc-KLH-DCs combined with cells metabolically glycoengineered with ManNPhAc led to a dramatic reduction in the number of lung metastases (19.38 ± 7.33 per mouse), which was less than that in both control groups (39.38 ± 13.70 per mouse in the GM3NPhAc-KLH-DCs and PBS group, and 41.89 ± 11.47 per mouse in the KLH-DCs group; Figure 6A and 6B).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
Evaluation of the novel immunotherapy in the B16F10 lung metastasis model
(A) Representative images of lungs 42 days after the initial immunization. (B) B16F10 nodules on the lungs were counted; n = 10; *P < 0.05 vs the GM3NPhAc-KLH-DCs and PBS group.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## DISCUSSION
*In the **DISCUSSION** section:*
TACAs are both the most exposed and often the most abundant TAAs on the surface of cancer cells. Expression of TACAs is closely correlated with tumor progression and metastasis [18, 19]. Therefore, they are ideal molecular targets for immune recognition in cancer immunotherapy [7, 20]. However, a major obstacle in the therapeutic application of TACAs is immune tolerance.
This problem has severely restricted the usefulness of most TACAs. We previously demonstrated that some unnatural TACA derivatives could evoke robust T cell-mediated immune responses [13–17, 21–24]. Additionally, cell glycoengineering has been established as a useful technique for the modification of carbohydrate antigens on tumor cell surfaces in vitro and in vivo [15–17]. These findings render unnatural TACAs derivatives potentially useful for cancer immunotherapy based on the combined application of synthetic vaccines made of unnatural TACA derivatives and cancer cell metabolic glycoengineering.
In this study, we exploited a new method to use ex vivo conditioned DCs as cellular adjuvants for cancer immunotherapy in combination with cell glycoengineering to enhance immunological recognition and generate targeted immune responses.[](https://www.ncbi.nlm.nih.gov/mesh/D015295)
DCs are uniquely potent in their ability to capture and process antigens, and they express high levels of MHC-peptide complexes and costimulatory molecules that allow for efficient activation of T cells that can play critical roles in cancer immunotherapy [3, 25, 26]. Our previous studies have shown that DCs generated from peripheral blood mononuclear cells of multiple myeloma patients and loaded with glycoengineered multiple myeloma antigens evoked strong allogeneic stimulatory activity in mixed lymphocyte reactions and efficiently activated CD4+ and CD8+ T cells. Importantly, these DC-activated T cells were specifically cytotoxic towards the glycoengineered myeloma cells [27]. In our study presented here, we demonstrated that murine bone marrow-DC pulsed with GM3NPhAc-KLH induced a protective immune response against tumor cells expressing GM3NPhAc antigen.
Mice vaccinated with GM3NPhAc-KLH-DCs produced high levels of GM3NPhAc-specific total and IgG antibodies. These results are in accordance with previous studies that showed that GM3NPhAc-KLH immunization induced IgG1, IgG2a, and IgG3 antibodies [17], indicating that a helper T cell response had been primed. The development of GM3NPhAc-specific antibodies in mice that were immunized with GM3NPhAc-KLH-DCs correlated with the levels of ADCC and CDC activities. Thus, antibody-mediated cytotoxicity against tumors may be an important mechanism for tumor protection. CTLs are critical effectors in antitumor immunity.
To investigate the in vivo generation of a GM3NPhAc-specific T cell responses elicited by this DC vaccine, antigen-specific CTL responses were observed in vitro. Effector cells obtained from mice immunized with GM3NPhAc-KLH-pulsed DCs could efficiently lyse GM3NPhAc-expressing murine B16F10 and FBL3 cells. Furthermore, GM3NPhAc-KLH-DC vaccination effectively generated IFN-γ-secreting splenocytes, as indicated by ELISPOT assays. Ultimately, we have demonstrated that GM3NPhAc-KLH-DC immunization in combination with ManNPhAc treatment showed robust effectiveness in vivo and resulted in the elimination of B16F10 and FBL3 tumors. This novel immunotherapeutic strategy could significantly inhibit FBL3 tumor growth and prolong the survival of tumor-bearing mice.
Furthermore, it could prevent B16F10 lung metastases. The results indicate that the use of DC as direct APCs could improve the potency of GM3NPhAc-KLH as a vaccine, which can include both strong T cell-mediated humoral and cellular anti-GM3NPhAc immunity for effective cancer immunotherapy.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
Biochemical engineering of TACAs on cancer cell surfaces is the basis for selectively targeted immunotherapy using vaccines made of unnatural TACA derivatives. The biosynthesis of carbohydrates uses no rigid template and is controlled by a series of specific carbohydrate biosynthetic enzymes, including sialyltransferase, which is highly expressed in cancer cells [18, 28]. The flexible substrate permissibility of the enzymes makes it convenient to engineer cancer cells that express GM3NPhAc on the cell surface by using ManNPhAc instead of the naturally occurring N-acetyl-D-mannosamine (ManNAc) molecule as a biosynthetic precursor. Therefore, this bioengineering strategy for immunotherapy could be generally applicable to different TACAs and tumors that express sialo-TACAs.
This concept is supported by our previous finding that ManNPhAc, the physiological precursor of N-acetyl sialic acid, could be successfully used in vitro and in vivo to glycoengineer numerous tumor cells, but not normal cells, to express GM3NPhAc in place of native GM3. It is theoretically possible that the artificial neoantigen could be expressed on normal cells and cause autoimmune reactions. However, we did not observe GM3NPhAc expression on normal tissues, as shown by immunostaining in tumor-bearing mice [17]. Moreover, no remarkable ManNPhAc treatment-related toxicities that affected the weight or general behavior of the immunized mice were observed in these experiments, perhaps because the concentrations of the modified glycans on normal cells are too low to provoke strong immune responses. The limited negative consequences of ManNPhAc usage in mice are thus important for potential human trials in the future.[](https://www.ncbi.nlm.nih.gov/mesh/D015295)
In summary, we have demonstrated that GM3NPhAc-KLH-DCs can stimulate robust T cell-mediated immunity. This vaccine could significantly inhibit tumor growth and metastasis, and also prolong the survival of cancer-bearing animals when used in combination with ManNPhAc treatment. This work lays out a new strategy to overcome the problem of immune tolerance for the development of effective TACA vaccines. This study also supports the potential benefits of developing new cancer immunotherapies that combine DC vaccination against GM3NPhAc with ManNPhAc treatment for application in humans. We propose that this type of immunotherapy may be applicable to melanoma, leukemia, breast carcinoma, pulmonary cancer, prostatic carcinoma, and other types of tumors that express GM3.
Finally, we chose GM3 as a target antigen to confirm the feasibility of this novel immunotherapeutic strategy. It is worth mentioning that this strategy is theoretically applicable to additional sialo-TACAs expressed by other tumor cells.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## MATERIALS AND METHODS
*In the **MATERIALS AND METHODS** section:*
## Reagents
*In the **Reagents** section:*
GM3NPhAc-KLH and GM3NPhAc-HSA conjugates, as well as ManNPhAc were prepared according to previously reported methods [13, 15]. The homogeneity and purity of gangliosides was >95%, as determined by TLC and densitometry [22]. Lipopolysaccharide (LPS; Escherichia coli, O26:B6) and KLH were obtained from Sigma (St Louis, MO, USA). Recombinant mouse GM-CSF and IL-4 were purchased from R&D Systems (Minneapolis, MN, USA).
FITC- or PE-labeled antibodies (Abs) against murine CD80 and CD86, and isotype control Abs were from BD Pharmingen (San Diego, CA, USA). The alkaline phosphatase-linked goat anti-mouse IgG antibody and Igκ antibody were purchased from Southern Biotechnology (Birmingham, AL, USA). The LDH Assay Kit was purchased from Takara Bio Inc. (Otsu, Japan).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## Animals and cell culture
*In the **Animals and cell culture** section:*
Six- to eight-week-old female C57BL/6 mice were purchased from Shanghai SLAC Laboratory Animal Co. Ltd. (Shanghai, China). All procedures involving animal treatment and care in this study were approved by the animal care committee of the Second Military Medical University in accordance with institutional and Chinese government guidelines for animal experiments.
A murine melanoma cell line, B16F10, and a murine leukemia cell line, FBL3, were obtained from ATCC (Manassas, VA, USA). Cells were cultured in RPMI-1640 supplemented with 10% heat-inactivated FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37°C in a 5% CO2 atmosphere.[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Generation and culture of bone marrow-derived DCs
*In the **Generation and culture of bone marrow-derived DCs** section:*
Murine DCs were prepared and analyzed as previously described [29] with some modifications. Briefly, bone marrow was flushed from the femur and tibia of mice and red blood cells were lysed with 0.84% ammonium chloride. Cells were cultured in RPMI-1640 complete medium for 2 h to allow for adherence. Non-adherent cells were collected and incubated with culture medium supplemented with recombinant murine GM-CSF (10 ng/mL) and recombinant murine IL-4 (10 ng/mL). On day 5, non-adherent cells were harvested as DCs and used for the subsequent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D000643)
## DC pulsed with antigen and phenotypic analysis
*In the **DC pulsed with antigen and phenotypic analysis** section:*
We incubated 2 × 106 DCs for 24 h with 20 μg/mL GM3NPhAc-KLH, 20 μg/mL KLH, or medium alone. The resulting DCs were referred to as GM3NPhAc-KLH-DCs, KLH-DCs, or medium-DCs (unpulsed). DCs were stimulated with 500 ng/mL LPS as a positive control. Expression of the costimulatory markers CD80 and CD86 was quantitated by flow cytometry using FITC- or PE-labeled Abs against murine CD80 and CD86, respectively. IL-12p70 production was measured in supernatants using a murine IL-12p70 ELISA kit (R&D Systems).[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## Preparation of ManNPhAc-glycoengineered tumor cells
*In the **Preparation of ManNPhAc-glycoengineered tumor cells** section:*
FBL3 or B16F10 cells (5 × 105) were cultured in RPMI-1640 complete medium containing different concentrations of ManNPhAc in 6-well plates according to an established protocol [17]. After 72 h incubation, cells were washed twice with PBS and used in subsequent experiments. Unless otherwise indicated, the glycoengineering experiments described in this study were performed using ManNPhAc (1 mmol/L).[](https://www.ncbi.nlm.nih.gov/mesh/C510175)
## Immunizations
*In the **Immunizations** section:*
C57BL/6 mice were vaccinated by three subcutaneous (s.c.) injections of 1 × 106 GM3NPhAc-KLH-DCs or KLH-DCs in 0.1 mL of PBS in the lower right flank at weekly intervals. All mice were bled and spleens were harvested on the day following the final boost. Sera were prepared for antibody assays. Splenocytes were restimulated with 10 μg/mL GM3NPhAc-KLH or KLH for 24 h and these splenocytes were used as effector cells for cytotoxic T lymphocyte (CTL) assays.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## Antibody assays
*In the **Antibody assays** section:*
Specific anti-GM3NPhAc antibody was measured by ELISA as previously described [15, 17, 24]. Briefly, GM3NPhAc-HSA was coated on ELISA plates. Serial dilutions of sera from immunized mice were then added and incubated at 37°C for 2 h. After washing, plates were incubated with 1:1000 dilutions of alkaline phosphatase-linked goat anti-mouse κ, or IgG antibody for 1 h at room temperature. Finally, plates were developed with PNPP substrate.
The absorption (A) value was measured using a plate reader at a 405 nm wavelength. To measure titers, OD values were plotted against dilution numbers, and a best-fit line was obtained. The slope of this line was used to calculate the dilution number at which an OD value of 0.5 was achieved, and this dilution number indicated the antibody titer.[](https://www.ncbi.nlm.nih.gov/mesh/D005732)
## Antibody-dependent cell-mediated cytotoxicity (ADCC) and antibody-mediated complement-dependent cytotoxicity (CDC) assays
*In the **Antibody-dependent cell-mediated cytotoxicity (ADCC) and antibody-mediated complement-dependent cytotoxicity (CDC) assays** section:*
ADCC and antibody-mediated complement-dependent cytotoxicity (CDC) activities were determined as previously described [15, 17] using a LDH cytotoxicity detection kit. Briefly, ManNPhAc-glycoengineered FBL3 or B16F10 tumor cells (1.5 × 104 cells/well) were used as target cells and incubated with antisera derived from GM3NPhAc-KLH-DCs immunized mice (1:20 dilution in DMEM) at 37°C for 2 h. Thereafter, peritoneal macrophages isolated from healthy mice or rabbit complement serum (diluted 1:10 in DMEM) were added to each well as effectors. Plates were incubated at 37°C for another 18 h for ADCC or 1 h for CDC assays, and then cell supernatants were harvested to detect cell lysis in a LDH assay using a LDH cytotoxicity detection kit according to the manufacturer's instructions. Results are expressed as the percentage of cell lysis.[](https://www.ncbi.nlm.nih.gov/mesh/C510175)
## CTL assays
*In the **CTL assays** section:*
CTL assays were performed as previously described. [17] Briefly, splenocytes from immunized mice were cultured for 24 h with or without 2 μg/mL of the corresponding antigen, KLH or GM3NPhAc-KLH. These splenocytes were used as effector cells and were co-cultured with FBL3 cells or ManNPhAc-glycoengineered FBL3 cells (target cells) at an effector-to-target cell ratio of 100:1 in RPMI-1640 with 10% FBS for 24 h. Cell-free supernatants were collected and analyzed using the LDH assay.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## IFN-γ enzyme-linked immunosorbent spot (ELISPOT) assay
*In the **IFN-γ enzyme-linked immunosorbent spot (ELISPOT) assay** section:*
IFN-γ-producing splenocytes were quantified by ELISPOT assay as previously described [17]. In brief, 1 × 106 splenocytes were added to IFN-γ ImmunoSpot plates (Dakewe, Shenzhen, China) and incubated at 37°C for 18 h. After washing, plates were incubated with biotin-conjugated anti-IFN-γ mAb, followed by the addition of streptavidin-alkaline phosphatase. Finally, the activator solution (supplied with the ImmunoSpot kit, Dakewe) was added to each well for spot development. The spots were counted using a microplate reader.[](https://www.ncbi.nlm.nih.gov/mesh/D001710)
## Immune-mediated protection generated by the GM3NPhAc-KLH-DCs vaccine combined with ManNPhAc treatment
*In the **Immune-mediated protection generated by the GM3NPhAc-KLH-DCs vaccine combined with ManNPhAc treatment** section:*
To evaluate the efficacy of the GM3NPhAc-KLH-DCs vaccine to inhibit tumor growth or prolong the survival of tumor-bearing animals, C57BL/6 mice were vaccinated by three s.c. injections of 1 × 106 GM3NPhAc-KLH-DCs in 0.1 mL of PBS administered at weekly intervals. One week after the third immunization, each mouse was inoculated with 5 × 105 FBL3 cells s.c. into the left flank, followed by daily i.p. injections of ManNPhAc (50 mg/kg/day in 0.1 mL of PBS) for 7 days. Mice received immunizations with GM3NPhAc-KLH-DCs plus PBS; mice immunized with KLH-DCs alone were used as a control group. Tumors that developed in immunized mice were monitored and measured using a caliper ruler. Tumor size was calculated using the formula 0.4 × (A2 × B), (where B represents the largest diameter and
A indicates the diameter perpendicular to B). To assess the impact of the immunotherapy on the survival of animals, mice were treated as described above and kept under close observation. Animals were maintained until they died of cancer or, if a tumor reached the size of 700 mm3, the animals were euthanized and similarly scored as ‘died of cancer.’[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
To evaluate the efficacy of the GM3NPhAc-KLH-DCs vaccine to inhibit tumor metastasis, 5 × 105 B16F10 cells were injected into the tail vein of each mouse. GM3NPhAc-KLH-DC immunization and ManNPhAc treatment were carried out using the same protocols as described above. On day 42, all mice were sacrificed. The nodules in the lungs were counted and splenocytes were isolated for both the CTL assay and the enzyme-linked immunosorbent spot (ELISPOT) assay for IFN-γ-secreting lymphocytes.[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
## Statistical analyses
*In the **Statistical analyses** section:*
All experiments were carried out in triplicate, and all data are presented as the mean ± SD. Statistical analyses were performed using the Student's t-test or two-way ANOVA. Differences were considered to be statistically significant when P < 0.05. All statistical analyses were performed using GraphPad Prism 5.0 software (La Jolla, CA, USA).
DISCLOSURE/CONFLICT OF INTEREST
The authors declare no conflcts of interest.
FINANCIAL SUPPORT
This work was supported by the National Natural Science Foundation of China (No.30728032, 21202200).
Authors' contributions
JZ, ZG and LQ designed research; LQ, JL, SC, QW and YL performed research; ZG and QW contributed new reagents; LQ, JL, YL, ZH and JZ analyzed data; LQ and JZ wrote paper. All authors read and approved the final manuscript.
# Abbreviations
*In the **Abbreviations** section:*
TACAs[](https://www.ncbi.nlm.nih.gov/mesh/D015295)
Tumor-associated carbohydrate antigens
KLH[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
Keyhole limpet hemoeyanin[](https://www.ncbi.nlm.nih.gov/mesh/C032808)
HSA[](https://www.ncbi.nlm.nih.gov/mesh/D000075462)
Human serum albumin[](https://www.ncbi.nlm.nih.gov/mesh/D000075462)
CTL
Cytotoxic T lymphocyte
# REFERENCES
*In the **REFERENCES** section:*
|
# Introduction
[Sulphated glycosaminoglycans](https://www.ncbi.nlm.nih.gov/mesh/D006025) and proteoglycans in the developing vertebral column of juvenile Atlantic salmon (Salmo salar)
# Abstract
*In the **Abstract** section:*
In the present study, the distribution of sulphated glycosaminoglycans (GAGs) in the developing vertebral column of Atlantic salmon (Salmo salar) at 700, 900, 1100 an[d 1400 d° was examined by li](https://www.ncbi.nlm.nih.gov/mesh/D006025)gh[t mi](https://www.ncbi.nlm.nih.gov/mesh/D006025)croscopy. The mineralization pattern was outlined by Alizarin red S and soft structures by Alcian blue. The temporal and spatial distribution patterns of different types of GAGs[: chondroitin-](https://www.ncbi.nlm.nih.gov/mesh/C004468)4-sulphate/dermatan sulp[hate, chond](https://www.ncbi.nlm.nih.gov/mesh/D000423)roitin-6-sulphate, chondroitin-0-sulphate and keratan sulphate were add[ress](https://www.ncbi.nlm.nih.gov/mesh/D006025)ed[ by immunohistochemist](https://www.ncbi.nlm.nih.gov/mesh/D002809)r[y using monoclona](https://www.ncbi.nlm.nih.gov/mesh/D003871)l [antibodies against the](https://www.ncbi.nlm.nih.gov/mesh/D002809) d[ifferent GAGs. The spe](https://www.ncbi.nlm.nih.gov/mesh/D002807)cific[ pattern obtaine](https://www.ncbi.nlm.nih.gov/mesh/D007632)d with the different antibodies suggests a unique role of the different GAG types in patte[rn f](https://www.ncbi.nlm.nih.gov/mesh/D006025)ormation and mineralization. In addition, the distribution of the different GAG types in normal and ma[lfo](https://www.ncbi.nlm.nih.gov/mesh/D006025)rmed vertebral columns from 15 g salmon was compared. A changed expression pattern of GAGs was [fou](https://www.ncbi.nlm.nih.gov/mesh/D006025)nd in the malformed vertebrae, indicating the involvement of these molecules during the pathogenesis. The molecu[lar ](https://www.ncbi.nlm.nih.gov/mesh/D006025)size of proteoglycans (PGs) in the vertebrae carrying GAGs was analysed with western blotting, and mRNA transcription of the PGs aggrecan, decorin, biglycan, fibromodulin and lum[ican](https://www.ncbi.nlm.nih.gov/mesh/D006025) by real-time qPCR. Our study reveals the importance of GAGs in development of vertebral column also in Atlantic salmon and indicates that a more comprehensive approach is necessar[y to](https://www.ncbi.nlm.nih.gov/mesh/D006025) completely understand the processes involved.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
The vertebral column is the defining character of all vertebrates. It consists of an alternating pattern of vertebral bodies (centra), providing strength and support, and intervertebral regions (IVR), providing flexibility and resistance to compression. The vertebral column develops from notochord, a flexible rod-like structure derived from the mesoderm. In contrast to the notochord in mammals, where only remnants exist in the intervertebral disc in adulthood, the notochord in many teleosts, including Atlantic salmon (Salmo salar), persists in the entire length of the vertebral column throughout the whole lifespan. During the embryonic and larval stages, the notochord is the main axial support playing essential roles in the development of the vertebral column by serving structural as well as signalling roles for patterning, also of surrounding tissue.
Throughout the early yolk-sac stage, the notochord of Atlantic salmon is unsegmented with a uniform notochord sheath of even thickness surrounded by the sclerotome (Grotmol et al. 2003, 2005). After hatch, when the salmon larvae start to move, a stronger backbone is needed and mineralization of the notochord and sclerotome starts (Nordvik et al. 2005). Mineralization begins already at 650 d° in Atlantic salmon, when matrix in the outer half of the ventral notochordal sheath becomes mineralized (the chordacentrum). During the growth period from 800 to 1400 d°, the main events have occurred, resulting in the architecture of the adult salmon vertebral column (Nordvik et al. 2005).
Many studies have focused on early development of the vertebral column. In different teleost species, a pool of mesenchymal stem cells (MSCs) and neural and haemal arch cartilages surround the external elastic membrane of the notochord at 300 d° are described (Koumans et al. 1990; Koumoundouros et al. 1999; Potthoff et al. 1986; Powell and Tucker 1992). The process where these cells differentiate and develop into specific tissue types and body parts is complicated, highly controlled and orchestrated by a number of signals, including external (e.g. temperature and nutrition) and internal (e.g. hormones and transcription factors). The sequential order of these signals is of great importance for correct and normal development.
The proteoglycans (PGs) represent a ubiquitous family consisting of more than 30 members that are involved in many biological processes, ranging from structural and mechanical foundation characteristics to biochemical involvement in mineralization, growth, differentiation and metabolism. PGs are together with collagen the template for mineralisation, and their structure and temporal presence are of great importance for correct and normal development. In mammals, proteoglycans (PGs), expressed by notochord-like cells in the intervertebral disc, are reported to influence disc integrity and function (Aguiar et al. 1999; Erwin and Inman 2006; Iwasaki et al. 1999). In both mammals and birds, the composition of PGs and the structure of the glycosaminoglycans (GAGs) attached are known to play pivotal roles for skeletal development (Aszodi et al. 2000). PGs in matrix are divided into two major groups, the hyalectans and the small leucine-rich proteoglycans (SLRPs).
The former is able to aggregate with hyaluronic acid and link proteins together, forming huge hydrated space filling polymers. The SLRPs, on the other hand, are smaller extracellular molecules that bind to growth factors, collagens and other matrix molecules.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
The common feature for the PG members is a protein core with one or more covalently attached sulphated carbohydrate side chains (glycosaminoglycans, GAGs). The core protein of the PGs ranges in molecular sizes from around 10 kDa to above 500 kDa, and each protein is encoded by a lonely gene belonging to different gene families (Iozzo 1998). The multifunctionality is dependent on tissue, cell type and/or metabolic context, and where the carbohydrate chains of the proteoglycans with different structurally modifications fine tunes the proteoglycans interactions with a variety of ECM components and cellular ligands, such as chemokines, growth factors, adhesion molecules and collagens (Hardingham and Fosang 1992). Structurally, the GAGs are composed of repeating disaccharide units consisting of a hexosamine and an uronic acid residue, forming long linear chains. Depending on the type of the disaccharide building blocks, the GAGs are divided into three major groups: the chondroitin/dermatan sulphate, (CS/DS), the keratan sulphate (KS) and the heparan sulphate (HS).
The first two groups, CS/DS and KS represent major GAGs of the extracellular matrix (ECM) in mammals and birds. The disaccharide of CS consists of d-glucuronic acid (GlcA) and N-acetyl galactosamine (GalNAc). The GalNAc may be non-sulphated (C-0-S) or sulphated in the 4th (C-4-S) or 6th (C-6-S) position. In DS, the GluA is epimerized to iduronic acid (IdoA). KS is composed of galactose and GluNAc and is the only GAG that does not contain hexuronic acid.
Variations in the amount and position of the sulphate groups in the GAG chains and amount of epimerization of glucuronic acid residues influence the type and strength of interactions with other extracellular components.[](https://www.ncbi.nlm.nih.gov/mesh/D002241)
In the present study, we describe the distribution of sulphated GAGs in the vertebral column of Atlantic salmon during early development by light microscopy methods. Stages in the development (700, 900, 1100 and 1400 d°) were selected to follow the process from notochord to mineralized vertebral column. The different GAG subtypes and sulphate structures were analysed by immunohistochemistry using monoclonal antibodies against, C-0-S, C-4-S, C-6-S and DS after enzymatic treatment (Caterson et al. 1985; Couchman et al. 1984). Monoclonal antibodies against highly sulphated KS were also included. We also compared the distribution of the GAG types in salmon with normal and pathologic vertebral development collected at a later stage of development and approached the number and size of the macromolecules carrying the GAG epitopes by western blotting.
We have previously reported the transcription of proteoglycans in salmon with malformations (Pedersen et al. 2013). In the present study, the transcription of PGs known to carry the C-0-S, C-4-S, C-6-S, DS and KS epitopes was analysed in salmon with healthy development with real-time qPCR. A changing composition of GAGs during spinal segmentation, when tissues are organizing and differentiating, is relevant for the understanding of vertebral development in fish, both under normal conditions and during pathogenesis. Our results reveal new and important distribution patterns of macromolecule components during vertebral development in fish.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
## Materials and methods
*In the **Materials and methods** section:*
## Sampling
*In the **Sampling** section:*
Fish originating from the SalmoBreed strain was collected at the Nofima research station at Sunndalsøra, Norway, in 2007. Eyed salmon eggs were incubated at 8 °C until hatching at approximately 500 d°. Yolk-sac fry were maintained at 8 °C until time of first feeding at 850 d°. At time of first feeding, fry were transferred to circular fibre glass tanks (∅ 0.5 m) supplied with water at a controlled temperature of 12 °C and given commercial feed in excess. Fish tanks were inspected daily.
At sampling, fish were euthanized (tricaine methane sulphonate, Pharmaq, Overhalla, Norway) before dissection. Notochord and surrounding tissue were dissected from salmon at necropsy, at the following developmental stages: 700, 900 (0.2 g), 1100 (0.5–0.6 g) and 1400 d° (1.3–1.4 g), frozen in liquid nitrogen and kept at −80 °C until use. Developmental stages were classified as day degrees (d°), which is defined as the sum of daily ambient water temperatures in degrees for each day of development. Radiography of 2 and 15 g fish was performed at Nofima radiology laboratory at Sunndalsøra, Norway.
The setup was semi-digital, with a mammography X-ray source (MS Giotto, Bologna, Italy) and with the use of reusable image plates with mammography resolution, and exposure at 22 kV/40 mAs. At 2 and 15 g size, 40 and 100 fish per tank, respectively, were X-rayed. Images were read and transferred to the computer in a plate reader (FCR Profect, Fuji Medical Inc., Japan). The images were automatically processed before storage, adjustment of brightness and contrast, equalization of exposure and highlighting of edges (Fuji CR Console software, Fuji Medical Inc., Japan). Based on evaluation of the digital images, fish with no vertebral malformations and fish with severe vertebral malformations were identified and subsequently euthanized and selected for analysis; for histological analyses, samples were stored 24 h in PFA before dehydrated in a graded series of ethanol; for RNA and western blot analysis, samples were snap frozen in liquid nitrogen and stored at −80 °C before further analysis.[](https://www.ncbi.nlm.nih.gov/mesh/C003636)
## Sectioning and mounting
*In the **Sectioning and mounting** section:*
Serial sections of 5 μm of OCT-embedded tissue were cut in the para-sagittal and transversal plane of the notochord and vertebral columns of fish (n = 3) collected at 500, 900, 1100 and 1400 d° of development and of 15 g weight (n = 4) with and without deformities, in a cryostat (Leitz 1720 Digital, Leica Instruments GMBH, Heidelberg, Germany), mounted on poly-l-lysine coated glass slides and kept at −20 °C. All sections were analysed microscopically, using a LEICA DMLB microscope (Leica Microsystems Nussloch GmbH, Germany) and photographed in a spot RT Color Camera (Diagnostic Instruments Inc. Burroughs Sterling Heights, Michigan Heights). Demineralization was not performed in order to avoid destruction of the antigenic epitopes.[](https://www.ncbi.nlm.nih.gov/mesh/D011092)
## Histological staining
*In the **Histological staining** section:*
Sections were stained with haematoxylin (Riedel-de-Haen, Germany) combined with erythrosine B (Aldrich-Chemie, Germany) and saffron to outline general architecture of the vertebrae. In brief, sections were stained in haematoxylin for 5–10 min, washed in water for 20 min, and fixed in 0.25 % hexamine for 3 min before stained in erythrosine B for 4 min. Finally, the sections were dehydrated in ascending concentrations of alcohol before staining with saffron for 3 min. To visualize mineralized regions, sections were stained with Alizarin red S (pH 4.2) for 5 min. To examine the distribution of sulphated GAGs, a solution containing 0.05 % Alcian blue 8GX (Gurr Biological Stains, BDH, Poole, UK) in 0.2 M Na-acetate buffer pH 5.8 added 0.4 M MgCl2, was used.
At a concentration of 0.4 M MgCl2, only negatively charged groups such as sulphated GAGs stain (Scott and Dorling 1965; Scott and Dorling 1965). The sections were immersed in the staining solution at room temperature with gentle shaking overnight, rinsed in running water, dehydrated in absolute ethanol, cleared in xylene, and mounted in Eukitt. All stained sections were microscopically examined with a Leica DMLB microscope (Leica Microsystems Nussloch GmbH, Germany).[](https://www.ncbi.nlm.nih.gov/mesh/D006416)
## Immunohistochemistry
*In the **Immunohistochemistry** section:*
Immunohistochemical identification of the different GAGs CS/DS and KS chains was obtained with the following mAbs: mAb 1B5 for detection of C-0-S, mAb 2B6 for detection of C-4-S and mAb 3B3 for detection of C-6-S (Seikagaku America, MA). The identification of the different GAG types was performed on the same individual and repeated on different fish (n = 3). To generate the antigenic epitopes for detection of CS/DSPGs, the tissue samples were digested with chondroitinase ABC lyase (cABC) from Proteus vulgaris (0.5 units/mL) (EC 4.2.2.4, Sigma-Aldrich Chemie Gmbh, Steinheim, Germany) in 0.1 M Tris–HCl buffer pH 8 for 2 h at 37 °C (Caterson et al. 1982; Yamagata et al. 1968). To detect only CS, sections were treated with chondroitinase AC II, pH 6.0 for 2 h at 37 °C. Digestion with cABC was also performed on the samples to be examined for KS by immunohistochemistry to increase the penetration of the KS antibodies.
The enzyme is inactive with the KS chains (Hamai et al. 1997). After cABC treatment, non-specific binding sites were blocked using 5 % teleost gelatin (Sigma-Aldrich) in PBS-added normal serum from horse (Vectastain Universal Elite ABC kit, Vector Laboratories Inc., Burlingame, USA). The sections were then incubated overnight at 4 °C with the following monoclonal antibody concentrations: 3B3 (1:200), 1B5 (1:400), 2B6 (1:400) and 5D4 (1:200), for detection of C-6-S, C-0-S, 4 sulphated CS/DS and KS, respectively. The mAbs were diluted in PBS-added 5 % teleost gelatin or BSA and 0.005 % Tween-20 (Sigma-Aldrich). Immunostaining was performed using an immunoperoxidase system (Vectastain Universal Elita ABC kit, Vector Laboratories) according to the manufacturer’s recommendations.
Sections were counterstained with haematoxylin. Non-immune serum from the same species as the primary antibody was used in control experiments. Non-specific binding of the secondary antibody was tested by replacing the latter with dilution buffer. Control tests were also performed on sections without digestion with chondroitinase ABC. A Spot RT Color Camera (Diagnostic Instruments Inc.
Burroughs Sterling Heights, Michigan Heights) photographed the sections with a Leica DMLB microscope.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
## RNA isolation, cDNA synthesis and real-time qPCR
*In the **RNA isolation, cDNA synthesis and real-time qPCR** section:*
Primer and probe sequences
Vertebrae dissected from 12 fish of 2 g (n = 12) and 15 g fish (n = 12) size were homogenized in liquid nitrogen before total RNA was isolated using Trizol® (Invitrogen, MD, USA) and reagents from the RNAeasy Micro to Midi Kit® (Qiagen, Hilden, Germany). All samples were DNase I treated (Invitrogen), and RNA quality was assessed by 1 % agarose gel. One microgram of RNA was subjected to reverse transcriptase (RT) reaction by TaqMan Gold RT-PCR kit (Applied Biosystem, CA, USA). The samples were diluted 5× before application of samples (in triplicates) to real-time PCR analysis in an ABI prism 7900HT sequence detection system (Applied Biosystem), using TaqMan 100 rx PCR core master kit (Applied Biosystem). At first, uracil-N-glycosylase (UNG) treatment at 50 °C for 2 min and UNG inactivation at 95 °C for 10 min were performed, followed by amplification of cDNA by 40 two-step cycles (15 s at 95 °C for denaturation of DNA, 1 min at 60 °C for primer annealing and extension).
Cycle threshold (Ct) values were obtained graphically (Applied Biosystem, Sequence Detection System, Software version 2.2). Gene expression was normalized to elongation factor 1 alpha (EF-1α), and ΔCt values were calculated. Comparison of gene expression between samples (15 and 2 g) was derived from subtraction of ΔCt values between the two samples to give a ΔΔCt value, and relative gene expression calculated as (fold difference). The primers and TaqMan probes (Table 1) (5′ labelled-6-FAM and 3′ quencher TAMRA) for real-time PCR amplification of decorin (GenBank accession no. NM001173562.1), biglycan (GenBank accession no. FJ799991.1), lumican (GenBank accession no. NM001140062.1), aggrecan (GenBank accession no. FJ 179677), fibromodulin (GenBank accession no. FJ195618) and internal standard EF-α (Genbank accession no. AF321836) were designed by using Primer Express Program (Applied Biosystem).[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
## Protein extraction and western blot
*In the **Protein extraction and western blot** section:*
Proteins from 15 g vertebrae of five fish (n = 5) were isolated from the residual fractions after the RNA isolation using Trizol® (Invitrogen) according to the manufacturer’s protocol. Protein concentrations were measured with a commercial kit at 750 nm (RC DC Protein Assay, Bio-Rad, USA) in a spectrophotometer (Ultrospec 3000, Pharmacia Biotech) using BSA as a standard. The samples were treated with 0.5 units chondroitinase ABC (P. vulgaris, EC 4.2.2.4, Sigma) in 0.05 M Tris–HCl pH 7.5 for generation of epitopes of C-4-S, DS, C-6-S, C-0-S and KS over night at 37 °C. Equal amounts of chondroitinase ABC-treated samples (20 μg) were subjected to SDS-gel electrophoresis using 4–12 % Bis–Tris gels (Invitrogen), SDS-PAGE NuPAGE® MOPS SDS running buffer (Invitrogen) and Novex Xcell II apparatus (Invitrogen). Transfer of the proteins to nitrocellulose membrane was performed by western blotting in transfer buffer consisting of 10 % methanol with 10 mM 3-cyclohexylamino-1-propanesulphonic acid (ACPS) pH 11 for 1 h at 80 V. Non-specific binding sites were blocked using 5 % teleost gelatin (Sigma) in 0,1 M Tris-saline, pH 7,4.
The nitrocellulose membrane were then probed with monoclonal antibodies against C-4-S, C-6-S, C-0-S and KS (the same antibodies as used in the immunohistochemistry) diluted 1:1000 in TBS-added 0.5 % teleost gelatin. Alkaline phosphatise (AP)-conjugated secondary antibodies [anti-mouse IgG(H + L), Promega, WI, USA] diluted 1:7500 were then applied on the membranes. Immunoreactive bands were revealed with Novex® AP (BCIP/NBT) colour development substrate (Invitrogen), and reaction stopped by distilled water.[](https://www.ncbi.nlm.nih.gov/mesh/C411644)
## Statistical analysis
*In the **Statistical analysis** section:*
Statistical analysis was performed using a two-tailed, unpaired Student’s t test. P values <0.05 were considered statistically significant.
## Results
*In the **Results** section:*
## Histological examination
*In the **Histological examination** section:*
HE-added saffron staining of spinal columns. a At 700 d°, bulging of the ventral part of the notochord begins to emerge (arrow). b At 1100 d°, extensive bulging of the notochord is seen at both ventral and dorsal side and the metameric pattern of the future vertebral column becomes evident. c Transverse section at 1100 d° and d a higher magnification of c, illustrating the lumen of the notochord, the chordoblast and the chordocytes. e The adult spine in 15 g fish with vertebral bodies separated by intervertebral regions. Scale bar 100 µm; nl notochord lumen, nc neural cord, ac arch centra, cb chordoblast, cy chordocytes, ns notochordal sheath, el elastic lamina[](https://www.ncbi.nlm.nih.gov/mesh/D004801)
The architecture of the notochord and the vertebral column in the time span studied was outlined by HE staining-added saffron, here illustrated with longitudinal and transverse sections of salmons obtained at 700 and 1100 d° (Fig. 1a–d). At 700 d°, some bulging of the ventral part of the notochord begins to emerge (arrow, Fig. 1a), and at 1100 d°, extensive bulging of the ventral as well as dorsal side of the notochord is seen (arrows, Fig. 1b), and the metameric pattern of the future vertebral column becomes evident. The transverse section of the fish at 1100 d° (Fig. 1c) outlines the neural chord (nc), the notochord sheath (ns), notochord lumen (nl), and the cartilaginous arches (ac). Figure 1d shows the transverse section at a higher magnification and the characteristic cartilaginous tissue of the arches separated from the sheath by the external elastic lamina (el). In the lumen of the notochord, the chordoblast (cb) layer and the chordocytes (cy) are clearly outlined.
The adult spine, showing the fully developed vertebral bodies and the intervertebral regions (IVRs), is shown in 15 g fish (Fig. 1e).[](https://www.ncbi.nlm.nih.gov/mesh/D004801)
## Mineralization of the developing vertebrae
*In the **Mineralization of the developing vertebrae** section:*
Alizarin red S staining of spinal column. a Longitudinal section at 700 d°, showing staining of the chordacentra in a repetitive pattern on one side of the notochord. b Higher magnification of a where the chordacentra show graded staining, weaker towards the notochord lumen. c Transverse section showing mineralization of the chordacentra on the lateral side of the notochord and the neural arcualia. d Longitudinal section at 900 d° where mineralization of the chordacentra proceeds. This is enlarged in e, showing the curved shape of the chordacentra. f Transverse section showing mineralized tissue surrounding the entire notochord. g, h Longitudinal sections at 1100 d° showing further development of the vertebral column with emerging vertebral endplates and unmineralized IVR. The concave shape of the vertebral bodies is visible. i Transverse section showing a thick mineralized ring with the perichordal bony arcualia. j Longitudinal section at 1400 d°, mineralization of the notochord is completed separated by unmineralized IVRs, enlarged in k. l Staining of the vertebral column from fish at 15 g size, with vertebral bodies containing amphicoel and the trabecular bone, separated by the IVRs. Scale bar 100 µm; nl notochord lumen, nc neural cord, ac arch centra, ns notochordal sheath, tb trabeculae bone, IVR intervertebral regions[](https://www.ncbi.nlm.nih.gov/mesh/C004468)
Alizarin red S staining was used to trace the mineralization process as illustrated in Fig. 2. At 700 d°, the initial mineralization of the chordacentra was revealed in a repetitive way beneath the external elastic lamina in one side of the notochord (Fig. 2a, b), visualized in Fig. 2c on the lateral side of the notochord.
The peripheral part of the chordacentra showed a deeper red colour, fainting towards the notochord lumen (Fig. 2b). The dorsal side of the sheath did not show any mineralization, neither in the longitudinal nor in the transverse sections. But mineralization of the upper, dorsal rim of the neural arcualia had started (Fig. 2c). At 900 d°, the mineralization of the chordacentra proceeded in both caudal and cranial directions in the ventral and now also in the dorsal notochordal sheath (Fig. 2d, e), surrounding the entire notochord (Fig. 2f). The mineralization of the neural and haemal arches also emerged (Fig. 2f).
At 1100 d°, the mineralization of the sheath in the areas of chordacentra was completed in the full thickness of the sheath and the concave shape of the future vertebral bodies appeared (Fig. 2g, h). Hence, the template for the unmineralized intervertebral region (IVR) was established (Fig. 2g, h). The transverse section (Fig. 2i) shows a thicker red-coloured ring with perichordal bony projections, indicative of fully mineralized sheath in the transversal plane.
In addition, the mineralization zones both in the periphery and the central areas of the arcualia were broader (Fig. 2i). At 1400 d°, the metameric segmentation of the notochord was completed, showing the contour of the spine as present in the fully developed salmon with vertebral bodies separated by unmineralized IVRs (Fig. 2j, k), representing the growth zones for vertebral length and thickness. Staining of the vertebral column from a 15 g fish with Alizarin red S showed that the segmentation was completed, with the bone of the vertebral amphicoel and the trabeculae, separated by the IVR (Fig. 2l).[](https://www.ncbi.nlm.nih.gov/mesh/C004468)
## Distribution of sulphated GAGs in the developing vertebrae
*In the **Distribution of sulphated GAGs in the developing vertebrae** section:*
Alcian blue staining of the vertebral column. a Longitudinal section at 700 d°, showing a strong staining of notochordal sheath and the arcualia. Higher magnification in (b), where the metameric staining pattern of the notochord sheath is evident (arrow). c Transverse section showing a faint-stained network in the lumen, indicating the presence of GAGs in chordoblasts and chordocytes at 700 d°, the templates for development of the neural and haemal arches showed strong staining (arrows). d, e Longitudinal sections at 900 d° showing further growth and development of the future IVRs and the cartilaginous arches. f Transverse section showing a continuous staining of the notochord sheath. g Longitudinal section at 1100 d° where the notochord curls along with the formation of the chordacentra, enlarged in h and the transverse section in i showing a layered staining pattern of the notochord sheath. j Longitudinal section at 1400 d°, the IVRs of the notochordal sheath stained dark, enlarged in k where the templates for the notochord endplates emerge. In the transverse section, the layered structures were clearly outlined in the notochordal sheath (triple lines in l). m Longitudinal section of 15 g salmon showing a full-maturated vertebral column with strong staining for GAGs in mineralized and unmineralized compartments. Scale bar 100 µm; nl notochord lumen, nc neural cord, ac arch centra, ns notochordal sheath, tb trabeculae bone, IVR intervertebral regions[](https://www.ncbi.nlm.nih.gov/mesh/D000423)
Alcian blue staining of the consecutive sections showed the distribution of sulphated GAGs in the notochord and surrounding tissue (Fig. 3). Sections collected from fish at 700 d° showed a strong and uniform blue staining of the notochordal sheath, arcualia and a weaker staining of the surrounding tissue (Fig. 3a–c). In the ventral part of the sheath where mineralization of the chordacentra was observed, regularly arranged areas of weaker staining appeared, demarcated on both sides by darker blue colour (arrow, Fig. 3a, b). In the transverse sections of notochord (Fig. 3c), a blue somewhat faint-stained network was evident in the lumen, showing the presence of GAGs also in chordoblasts and chordocytes. Furthermore, the templates for both neural and haemal arches were clearly outlined at 700 d° (arrows, Fig. 3c).
In sections collected from fish at 900 d°, a change in the staining pattern appeared: a darker blue staining was seen in the areas for development of the future IVRs (Fig. 3e). The staining intensity of the chordocyte network of the notochord lumen increased. Although the cartilaginous appearance of the arcualia (ac) became visible at 700 d° (Fig. 3c), the cartilaginous nature of the growing arcualia was more evident at 900 d° (Fig. 3d–f). No major differences were observed in the distribution pattern of GAGs in sections collected at 1100 d° (Fig. 3g–i). At 1400 d°, the IVRs of the sheath with the chordoblasts and chordocytes exhibited a darker blue colour (Fig. 3j–l).
In the IVRs, endplates were surrounded by a darker blue layer (Fig. 3k). In the transverse section at 1400 d°, concentric lamellar structures appeared in the sheath (Fig. 3l). The results after staining longitudinal sections from 15 g salmon with Alcian blue showed strong staining in the various compartments of the vertebral column such as the cartilaginous tissue of the amphicoel, the IVRs with the notochordal sheath and the bony endplates and the chordoblast layer as well as the chordocytes in the lumen (Fig. 3m), revealing an abundance of sulphated GAGs in mineralized as well as unmineralized tissue of the salmon vertebral column in adulthood.[](https://www.ncbi.nlm.nih.gov/mesh/D000423)
## Distribution of GAGs in the developing vertebral column
*In the **Distribution of GAGs in the developing vertebral column** section:*
To trace the contribution of various GAG subtypes to the staining patterns, C-4-S, DS, C-6-S, C-0-S and KS were immunolocalized. Below is the specific pattern obtained by each subtype.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
## Chondroitin 4-sulphate (C-4-S) and dermatan sulphate (DS)
*In the **Chondroitin 4-sulphate (C-4-S) and dermatan sulphate (DS)** section:*
Distribution of chondroitin 4-sulphate (C-4-S) and dermatan sulphate (DS). a Longitudinal section at 700 d° showing strong staining of the notochordal sheath and the perichordal cartilage. The staining of notochordal sheath appeared granular, as clearly outlined in b. c The transverse section showing a more continuous staining of the sheath and staining of the cartilaginous arches. d Longitudinal section at 900 d° showing the staining of the notochordal sheath become weaker in areas of developing chordacentra, enlarged in e. f Transverse section showing stronger staining towards the lumen. g Longitudinal section at 1100 d°, the strong staining persisted in developing IVRs, higher magnification in h, showing a granular staining pattern in the external region of the IVR in contrast to the layered appearance in the central part. i Transverse section showing stronger staining towards the lumen. j, k Longitudinal sections at 1400 d°, showing a strong staining of the IVR in the central region and distinct, labelled wings extending towards the mineralizing zones (arrow), enlarged in k. Some staining was also seen in the chordoblast layer (arrow). l Transverse sections showing strong, lamellar staining of the notochordal sheath. Scale bar 100 µm; nl notochord lumen, nc neural cord, ac arch centra, ns notochordal sheath[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
The mAb 2B6 binds to both C-4-S and DS when the sections are treated with chondroitinase ABC but only to C-4-S when the sections are treated with chondroitinase ACII, as described in materials and methods. Subsequently, C-4-S was identified by direct observations of stained structures after treatment with the latter, whereas DS containing structures were identified as those that were stained with the former and not the latter. A different expression pattern of C-4-S and DS epitopes during development of the salmon spine was detected (Fig. 4). Using mAb 2B6 + cABC lyase for detection of both C-4-S and DS containing structures, strong staining of both the notochordal sheath and the perichordal cartilage was found in sections at 700 d° (Fig. 4a–c). In the longitudinal section (Fig. 4b), staining of the notochordal sheath appeared granular, whereas in the transverse section (Fig. 4c), heavily stained notochordal sheath showed a more uniform appearance, with traces of concentric lamellae.
At 900 d°, the staining of the sheath became weak in the mineralized areas and the granular staining pattern became more uniform (Fig. 4d, e). A strong label persisted in developing IVRs that appeared at 1100 d° in parallel layers (Fig. 4g, h). However, in the outer rims of these regions, more granular staining could be seen (Fig. 4h), thus indicating a different organization pattern of C-4-S/DS (Fig. 4h). A faint staining of the mineralized part of the sheath was also evident and the chondrogenic arches stained strongly at all developmental stages.
At 1400 d°, a well-organized staining pattern appeared in the IVR, showing a strong and uniform staining of the central region with distinct, labelled wings extending towards the mineralizing zones (arrow, Fig. 4j) was visualized. Furthermore, staining was seen in the chordoblast area (arrow, Fig. 4k). Strong staining was seen in the sheath and in the arches in the transverse sections (Fig. 4l).[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
Distribution of chondroitin 4-sulphate (C-4-S) in the salmon spine. In the longitudinal sections a, b, and the transverse section, c collected at 700 d°, staining could hardly be seen. d, e Longitudinal sections at 900 d°, staining in developing IVRs appeared, very strongly in the central part (arrows), enlarged in e. f Transverse section showing that staining appeared in graded levels towards the lumen. g At 1100 d°, staining showed two–three distinct globes in the IVRs, enlarged in h and indicated by arrows. In addition, staining was seen in the layer of IVR towards lumen. i Transverse section showing that C-4-S was present in the notochord. j Longitudinal section at 1400 d°, the pattern of distinct globes and wings became more evident (arrows), enlarged in k. l Transverse section showed the distinct lamellar pattern of the notochordal sheath. No staining was seen in the lumen of the notochord at any stages of development. Scale bar 100 µm; nl notochord lumen[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
Replacing the enzyme cABC lyase with cACII produced a different labelling pattern (Fig. 5). With this enzyme, only C-4-S creates the immunostaining. At 700 d°, staining could hardly be seen in the notochordal sheath and no staining was found in chondrogenic tissue (Fig. 5a–c). This indicates that the dominating GAG in the notochordal sheath at 700 d° was DS.
However, at 900 d°, an increase in the expression of C-4-S was detected, mainly restricted to the developing IVRs of the notochordal sheath, appearing as dark brown circular structures placed centrally in the IVR (Fig. 5d–f). At 1100 d°, the staining clearly showed two distinct spots in the IVRs (arrows, Fig. 5g, h), which persisted at 1400 d° (arrows, Fig. 5j, k).
Furthermore, transverse sections from 1100 and 1400 d° revealed that C-4-S was arranged in the NS in concentric lamellae (Fig. 5i, l). The results indicate that C-4-S was arranged as a gradient, with strongest staining towards lumen and confined to more well-outlined structures, whereas DS showed a more widespread distribution. No staining for C-4-S was seen in the lumen of the notochord at any stages of development and in contrast to DS only weak staining could be detected in the chondrogenic zones.[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
## Chondroitin 6-sulphate (C-6-S) and unsulphated chondroitin (C-0-S)
*In the **Chondroitin 6-sulphate (C-6-S) and unsulphated chondroitin (C-0-S)** section:*
Distribution of C-6-S in the salmon spine. a Longitudinal sections at 700 d° shows a cartilaginous staining, and only weak label in the notochord sheath, enlarged in b with a granular layer close to the lumen. c Transverse section showing uneven staining of the notochordal sheath. d Longitudinal section at 900 d°, showing the same staining pattern as 700 d°, enlarged in e. f In transverse sections, staining was difficult to detect. g Longitudinal sections at 1100 d°, increased staining of the notochord sheath in the IVRs was seen, enlarged in h where a thread-like pattern of staining is evident, indicated by arrows. i Transverse section showing the staining of the arcualia. j Longitudinal section at 1400 d°, the staining remained strong and limited to IVRs, with strongest staining in a layer of the sheath parallel to the chordaepithelium, enlarged in k (indicated by arrows). l Transverse section showing strong staining towards the lumen and in cartilaginous tissue of the arches. Scale bar 100 µm; nl notochord lumen, nc neural cord, ac arch centra[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
The mAb 3B3 + cABC lyase, staining sulphate in the C-6-S position of the galactosamine, resulted in staining in the cartilaginous tissue of the arches at all developmental stages studied (Fig. 6) in mineralizing as well as unmineralized areas (Figs. 2a, d and 6a, d).[](https://www.ncbi.nlm.nih.gov/mesh/D013431)
In contrast to C-4-S/DS, only weak label was detected in the sheath of the notochord at 700 d°, apparent as a granular layer localized above the chordaepithelium (Fig. 6a–c). At 900 d° (Fig. 6d–f), the stain was found mainly in the cartilages. During the segmentation process from 900 to 1100 d°, an increased labelling of the notochord was seen, restricted to the intervertebral ligaments (Fig. 6g–h), and similar circular structures as seen by use of mAb 2B6 after chondroitinase acII treatment appeared (arrows, Fig. 6h). At 1400 d°, the staining of C-6-S remained strongest in a layer of the sheath parallel to the chordaepithelium, limited to IVR, almost disappearing in the outer parts of the sheath (left arrow, Fig. 6k).
In addition, two parallel circular structures, similar to those detected by 2B6 abc/ac in the longitudinal sections at 1100 and 1400 d° (Figs. 4k, 5h, k) appeared (right arrow, Fig. 6k).[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
Distribution of unsulphated GAGs in the salmon spine. a Longitudinal section at 700 d° showing a widespread staining of the notochordal sheath, arcualia and in surrounding tissue. b Transverse section showing that the notochordal sheath stained stronger towards the lumen. Staining of arcualia is also clearly visible. c Longitudinal section at 900 d° showing distinct globule-like structures in the IVRs (arrow) and staining besides these globules in IVR closest towards lumen. d Transverse section at 900 d showing only weak staining. e Longitudinal section at 1100 d°, IVRs was stained and the transverse sections in f showed that the staining was strongest towards the lumen. g This was also evident of longitudinal section at 1400 d° and the transverse section in h show the distinct lamellar ring pattern at this stage. Scale bar 100 µm; nl notochord lumen, nc neural cord, ac arch centra[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
The mAbs 1B5 + cABC lyase, recognizing unsulphated GAGs, were expressed in the same areas as C-6-S and C-4-S/DS, indicating some co-distribution of sulphated and unsulphated GAGs in the arcualia and notochordal sheath (Fig. 7). But 1B5 + cABC lyase resulted in a stronger and more widespread staining of the notochordal sheath and surrounding tissue in sections from salmon collected at 700 d° compared to the mAbs against C-6-S and DS (Fig. 5). In addition, similar structures in the IVR as were outlined by the mAbs against the sulphated epitopes, C-4-S and C-6-S, appeared as early in development as 900 d° (arrow, Fig. 7). Also C-0-S was arranged in concentric lamellae in the transverse sections (Fig. 7h) in a similar way as C-6-S and C-4-S.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
## Keratan sulphate (KS)
*In the **Keratan sulphate (KS)** section:*
Distribution of KS in the salmon spine. a Longitudinal section at 700 d°, KS was detected in the periphery of the lumen of the notochord. In addition, strong staining was seen in the cartilaginous tissue of the arches. The enlarged b shows strong staining beneath the layer of chordaepithelium. c Transverse sections clearly show this pattern. d Longitudinal section at 900 d°, with chordoblasts staining, enlarged in e (indicated by arrow) and cartilaginous staining persisting. f Transverse sections clearly show this pattern. g Longitudinal section showing a highly evident increase in staining intensity, enlarged in h and indicated by arrow. i Transverse sections show highly increased staining at this stage in the lumen of notochord. j Longitudinal section at 1400 d°, showing that the intense staining in the lumen of notochord persisted, enlarged in k, and clearly visible in transverse section l. Scale bar 100 µm; nl notochord lumen, nc neural cord, ac arch centra[](https://www.ncbi.nlm.nih.gov/mesh/D007632)
In contrast to CS and DS epitopes, KS labelling (using mAb 5D4) was detected primarily in the lumen of the notochord, beneath the layer of chordaepithelium (Fig. 8) as early in development as 700 d° (Fig. 8a–c). In addition, strong staining was seen in the cartilaginous tissue of the arcualia at all developmental stages. Both chordoblasts and chordocytes showed strong and consistent staining as seen in sections from fish of age 900 d° (arrow, Fig. 8e, h). An increase in staining intensity was seen during development, most evident beneath the chordaepithelium in the IVR areas (Fig. 8g–l).
Only traces of staining could be seen in the NS (Fig. 8). This staining pattern indicates that another class of proteoglycans carries the KS epitopes.[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
## Distribution of GAGs in normal and malformed vertebrae from 15 g salmon
*In the **Distribution of GAGs in normal and malformed vertebrae from 15 g salmon** section:*
Histology of longitudinal sections of a normal and b malformed vertebrae from 15 g salmon. HE-added saffron staining clearly showing a weaker elastic lamina surrounding malformed vertebrae, as indicated by arrows, and a more elongated appearance containing wavy lamellar structures appearing more stretched in malformed notochordal sheath. Alizarin red S visualized thicker compact bone, flattened endplates and reduced IVRs, between c normal and d malformed. Alcian blue visualized a more homogenous GAG distribution in the notochordal sheath of e normal compared to f malformed. Different layers with weaker staining towards the lumen of the notochord and heavier stained chordoblasts can be observed in the malformed. Scale bar 100 µm; nl notochord lumen, nc neural cord, ac arch centra, ns notochordal sheath, tb trabeculae bone, Norm normal, Malf malformed[](https://www.ncbi.nlm.nih.gov/mesh/D004801)
Immunostaining of different GAG types in longitudinal sections of normal and malformed vertebrae from 15 g salmon. Left panel shows normal vertebrae versus right panel shows malformed vertebrae. a, b KS was found mainly in the lumen of the notochord in normal and malformed, but also appeared strongly in the notochordal sheath in malformed. c, d C-6-S had distinct band with globular structures in the notochordal sheath of normal vertebrae, but these were lost in malformed vertebrae. e, f C-0-S was found closer to the lumen in the notochordal sheath of normal vertebrae in contrast to a more widespread distribution in malformed where this gradient was disturbed. g, h C-4-S/DS showed lamellar structures in the cranio-caudal direction of the notochordal sheath of normal vertebrae, but in malformed, these structures were ruptured and stretched more in the dorsoventral direction. i, j C-4-S showed weak staining in the notochordal sheath of normal vertebrae, but stained more intensively and in a spotted pattern in malformed vertebrae. Scale bar 100 µm; nl notochord lumen, ns notochordal sheath, Norm normal, Malf malformed[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
HE-added saffron staining of normal and malformed spinal columns showed a weaker external elastic lamina surrounding malformed vertebrae, as indicated by arrows in Fig. 9a, b. The notochordal sheath of malformed vertebrae also had a more elongated appearance containing vertical wavy lamellar structures (arrows, Fig. 9a, b). Alizarin red S visualized changes like thicker compact bone, flattened endplates and reduced IVRs, between normal and malformed vertebrae (Fig. 9c, d), as previous described in the literature (Ytteborg et al. 2010a).
Alcian blue visualized a more homogenous GAG distribution in the notochordal sheath of normal vertebrae compared to malformed vertebrae. In the latter, Alcian blue stained the notochordal sheath in different layers, with weaker and intermittent staining in between mainly in the part of the sheath towards the lumen (Fig. 9e, f). Chordoblasts and chordocytes also stained heavier in malformed vertebrae (Fig. 9f), indicating more GAGs in the notochord of these individuals. The distribution of the GAG types and their organization patterns differed markedly between normal and malformed vertebral columns (Fig. 10). The major difference was seen in the distribution of KS epitopes.
KS expression was present mainly in the lumen of the notochord of normal vertebral column (Fig. 10a), but also appeared strongly in the vertebral sheath of malformed samples (Fig. 10b). In addition, a more widespread distribution with changes in the organization pattern of CS epitopes in the intervertebral region was observed in the malformed vertebral column (Fig. 10d, f, h, j) compared to normal (Fig. 10c, e, g, i). The entire notochordal sheath in the malformed vertebral column showed intense staining and the staining pattern changed from parallel arranged lamellae to heavily stained, and more wavy irregular structures appeared. C-6-S had distinct band in the notochordal sheath of normal vertebrae; these were lost in malformed vertebrae (Fig. 10d).[](https://www.ncbi.nlm.nih.gov/mesh/D004801)
## Western blotting
*In the **Western blotting** section:*
Western blotting of proteins isolated from normal and malformed vertebrae of 15 g salmon showing that C-0-S was primarily detected in molecules of higher sizes, whereas C-4-S showed additional bands in the region <60 kDa. Also the main band, appearing by mAb 5D4 against KS, exhibited a molecular size just below the 60-kDa marker. Some broad, weak stain was furthermore seen in the region for higher molecular sizes against KS, most clear in samples from malformed vertebra. mAb against C-6-S was only found this epitope in samples obtained from malformed vertebrae, appearing as a band in the region above the 250-kDa molecular marker. N normal, MF malformed[](https://www.ncbi.nlm.nih.gov/mesh/D002807)
Proteoglycans represent a large family of molecules of different sizes which may carry same GAG chains. The results showed that C-0-S was primarily detected in molecules of higher sizes, whereas C-4-S showed additional bands in the region <60 kDa (Fig. 11). Also the main band, appearing by mAb 5D4 against KS, exhibited a molecular size just below the 60-kDa marker in accordance with the sizes for the small proteoglycans as decorin, biglycan, lumican and fibromodulin. Some broad, weak stain was furthermore seen in the region for higher molecular sizes, most clear in samples from malformed vertebrae. Regarding the mAb against C-6-S, we only found this epitope in samples obtained from malformed vertebrae, appearing as a band in the region above the 250-kDa molecular marker.
The result suggests that this epitope is associated with PG with high molecular size, which changed into lower molecular size able to enter the gel during the pathological development.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
## Q-PCR analyses
*In the **Q-PCR analyses** section:*
Transcription of aggrecan, biglycan, decorin, fibromodulin and lumican in the spinal column of 2 g (n = 12) and 15 g (n = 12) salmon. Biglycan transcription was the only gene that was significantly up-regulated. Levels of transcripts in triplicates are relative to elongation factor 1α (Ef1α), and data are presented as fold difference between 2 and 15 g fish. Asterisk indicates significant difference at p < 0.05 and black bars standard deviation
The results of the qPCR analysis of samples from non-deformed vertebrae collected from fish of 2 and 15 g are shown in Fig. 12. Transcription of the proteoglycans: aggrecan, biglycan, decorin, fibromodulin and lumican were found at both developmental stages. Comparison of the mRNA expression of the proteoglycans present in 2 and 15 g fish showed that biglycan transcription was the only one of the studied PGs which was significantly up-regulated during this period of growth. Decorin was also up-regulated, whereas aggrecan and fibromodulin transcription were down-regulated.
None of these changes were significant. Lumican transcription was similar in samples from 2 and 15 g fish.
## Discussion
*In the **Discussion** section:*
The present study describes the distribution of sulphated GAGs in the vertebral column of Atlantic salmon during development from 700 to 1400 d°, and in 15 g fish. During this period, the notochord evolves from a fire horse-like structure to a fully developed, segmented and mineralized vertebral column. This includes separation of the vertebrae, formation and insertion of ligaments and elaboration of synovial lining to occur simultaneously with continuing growth and mineralization. For all these processes to occur, many signals are required and must be highly coordinated and expressed in distinct patterns for proper differentiation and organization of the tissues involved. Disruption in any of these highly regulated processes may result in pathologic development, such as vertebral malformations.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
Staining with Alcian blue demonstrated the presence of GAGs in the regions of growth and differentiation, in mineralized and unmineralized areas of the notochord, the arch anlagen and surrounding tissues at all developmental stages studied, suggesting a role for sulphated GAGs in vertebral growth and development in Atlantic salmon. The presence of GAGs as judged by the staining intensity of Alcian blue highly correlated with the Alizarin red S staining of calcified tissue, showing the presence of sulphated GAGs in all areas prior to mineralization, making a template for endochondral as well as intramembranous ossification. Intensive staining was also found in regions and tissues that do not mineralize, indicating different roles for the GAGs in growth and development of the vertebral column in salmon. Since Alcian blue with MgCl2 stains all highly anionic GAGs, structural differences in the various subtypes were approached by immunohistochemistry after enzymatic digestion. Our study showed that in the mineralizing areas of the arches, DS, C-0-S, C-6-S and KS were strongly expressed at 700 d° and the staining persisted until adulthood.
In contrast, in the notochord sheath, the different GAG epitopes showed changing temporal and spatial distribution during development. The most striking changes in the composition of the GAGs occurred in the IVR, suggesting a role for GAGs also in the patterning of the vertebral column.[](https://www.ncbi.nlm.nih.gov/mesh/D000423)
At 700 d°, the IVR was part of a uniform-stained sheath of even thickness, consisting primarily of DS. The staining pattern for DS appeared granulated, but was later arranged in IVR as longitudinal layers. Grotmol (Grotmol et al. 2006) has previously described the presence of parallel Collagen 2 fibrils in the notochordal sheath, forming helices around the longitudinal axis of notochord during early life. Thus, the granular staining pattern may be due to DS chains associated with SLRPs, surrounding the collagen fibrils. At later, developmental stages DS were arranged also in parallel layers in the IVRs running from the anterior to the posterior endplates, indicating presence of GAGs in fibres connecting the vertebral bodies.[](https://www.ncbi.nlm.nih.gov/mesh/D003871)
At 900 d°, mineralization of the chordacentra was completed and the composition of the GAGs changed. We observed a folding or bulging of the sheath in the regions were IVR developed and circular structures containing C-4-S, C-6-S and C-0-S epitopes appeared. Using mAb against C-6-S, a worm-like structure appeared in the sections collected at 1100 d°. The structure most likely represents the same circular structures seen in IVR in sections from previous developmental stages, and the observed differences may be a result of minor variations in the para-sagittal plane during sectioning. The functional role of the structures is unknown.
However, they were discussed in Ytteborg et al. (2010a), and a filtering role for these structures, by removing waste products and providing nutrition to and from the notochord, was suggested. They may represent vessels and channels for transportation, in a similar way as the angiogenesis (vasculogenesis) described in the distal femoral chondroepiphysis of rabbit during development and ossification (Doschak et al. 2003) and in human lumbar intervertebral discs during development from foetal to infantile age (Nerlich et al. 2007).[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
The appearance of globes and bulges in the IVR in this experiment occurred simultaneously with completed circumferential mineralization of the chordacentra, thus, at a time where growth, size and structure might demand for a more organized and effective transportation system. During malformation, the epitopes were arranged in dorsoventral waves, ruptures or fissures in between the waves, indicating a higher level of degradation of matrix components associated with the collagenous network in the notochordal sheath. The globular structures containing C-6-S epitopes in the healthy IVR could not be detected in the malformed IVR, suggesting a change in these structures during the pathogenesis, which may possibly alter the nutritional access of the notochord during deformities.[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
An increase in the staining for C-6-S in the IVR seen with development has also been reported in murine vertebrae (Hayes et al. 2001). Staining for C-6-S was found in the regions above the chordaepithelium and in the interface between the soft tissue of the IVR and the bony endplates. The former region has to restrict the pressure from the expanding vacuolated chordocytes (Grotmol et al. 2005). During the period examined, stiffening of the vertebral column occurs due to mineralization of the chordacentra, resulting in greater mechanical challenges on the IVRs, which also have to resist bending movements and compression during swimming. C-6-S is usually a constituent of hyalectans: forming large water binding aggregates and enabling the tissue to resist compression.
Important information on the role of C-6-S in bone development has come from recent study on humans with spondyloepiphyseal dysplasia (Omani type) with a missense in the gene encoding for the enzyme C-6-S T-1 producing C-6-S. Severe progressive scoliosis preceded by diminished intervertebral discs and abnormal vertebral endplates, accompanied by severe arthritic changes with joint dislocations (Thiele et al. 2004). Thus, C-6-S most likely plays a similar role in vertebral development in teleosts as Atlantic salmon. Interestingly, we found increased staining intensity in the malformed IVRs. However, the structure of the sheath seemed to be different from what was found in normal notochordal sheath, indicating a compensatory role of increased protein production in these regions during compression.[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
The major GAG present in the notochord sheath and arches at 700 d° was 4-sulphated DS as only traces of staining from cACII generated epitopes were detectable. DS is a structural isomer of CS, in which some of the glucuronate residues are epimerized to iduronic acid (IdoUA). The epimerization has an important effect on the binding properties of GAGs, as IdoUA can take up more than one ring confirmation, thus increasing the flexibility of the GAG chain (Casu et al. 1988). DS PGs, such as decorin and biglycan, are reported to have a widespread distribution in mammalian tissue such as blood vessel walls, skin, tendon cartilage and undifferentiated mesenchymal tissue (Rosenberg et al. 1985) where they participate in extracellular matrix organization, neurite outgrowth, wound repair, cell adhesion, migration and proliferation, promoting growth factors (Trowbridge and Gallo 2002) (Taylor et al. 2005).
In the present study, we demonstrated an up-regulation of biglycan mRNA during development from 2 to 15 g. Biglycan is reported to control signalling pathways regulating the osteogenic program (Berendsen et al. 2011), and in vitro studies have demonstrated a different influence of biglycan and decorin on mineralization (Mochida et al. 2003). A changed mRNA expression of decorin and biglycan has previous been demonstrated in salmon of 15 g size with vertebral deformities (Pedersen et al. 2013). Thus, the presence of four sulphated DS in the notochord during growth and development of salmon is most likely a prerequisite for proper development, but further studies are needed to explore the consequences of the differences in biglycan and decorin expression in vertebral development. In malformed vertebrae, we found that immunolocalization of CS/DS epitopes changed in the vertebral sheath from parallel arranged lamellae to heavily stained, and more wavy vertical irregular structures appeared.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
In the present study, C-4-S epitopes created by cACII treatment showed a distinct expression pattern, emerging later during development and restricted to the IVR, indicating specific role for this GAG in this particular region of the notochord. It is reported that CS plays important modulatory role on proteinase activities (Georges et al. 2012). Activation of pro-MMP-2 by MMP-16 was significantly enhanced in the presence of excess C-4-S, whereas C-6-S was ineffective. CS-GAGs did also participate in regulating bone resorption through modulation of cathepsin K activity (Li et al. 2002). Moreover, at acidic pH, the collagenase activity of cathepsin K was enhanced in bone and cartilage by CS and KS, whereas DS and HS selectively inhibited its activity (Li et al. 2004).
Hence, the distinct pattern of C-4-S in the IVRs may indicate an inhibitory role of C-4-S in bone formation in the developing notochord of growing Atlantic salmon, ensuring flexibility of the vertebral column by preventing bone and cartilage formation in these regions.[](https://www.ncbi.nlm.nih.gov/mesh/D002809)
KS was hardly detected in the IVR at any stages of development, neither in the globular structures nor in the matrix between. However, KS was present in the lumen of the notochord at all developmental stages studied. The labelling for KS increased in this area during the life span from 700 to 1400 d°, located mainly in the vacuoles of an increasing number of chordocytes. In notochord, the chordoblasts produce the matrix of the sheath and continue to divide throughout life in accordance with sustained notochordal growth (Grotmol et al. 2006). They further maturate into chordocytes, containing large fluid filled vacuoles which functional role is to maintain internal hydrostatic pressure (Adams et al. 1990; Glickman et al. 2003; Nordvik et al. 2005).
KS is known to be the GAG with the highest water binding capacity, and PGs carrying KS chains are considered to regulate water balance of the extracellular matrix. Thus, a role for KS contributing to the hydrostatic pressure of the notochord of Atlantic salmon by vacuole formation is plausible. KS may be a part of aggrecan, a hyalectan, known to form huge aggregates with hyaluronic acid. Aggrecan was identified in the chordocytes of Atlantic salmon notochord (Ytteborg et al. 2010b). Mammalian aggrecan carries a large number of GAG chains with C-6-S epitopes in addition to KS.
In our study, C-6-S did not co-localize with KS in the chordocytes, suggesting the presence of other KS PGs. In a recent study, we immunolocalized lumican in the chordocytes of Atlantic salmon (Pedersen et al. 2013). Lumican is reported to carry KS also in teleosts such as Atlantic cod (Tingbo et al. 2012) and zebrafish (Souza et al. 2007).[](https://www.ncbi.nlm.nih.gov/mesh/D007632)
In contrast to the healthy vertebral column, KS was expressed throughout the entire sheath from the lumen towards the external lamina in malformed vertebral columns, showing a similar wavy distribution pattern as C-4-S/DS and C-6-S. The co-distribution of KS and C-6-S epitopes indicated the presence of aggrecan in the notochord sheath and may reflect a reaction to resist increasing mechanical challenges due to malformation (Witten et al. 2006). Production of KS is also reported in the lack of O2 indicating disturbance in O2 supply. Moreover, SDS and western blot showed more breakdowns of the molecules carrying these epitopes during the pathological process. In contrast to healthy vertebrae, C-6-S epitopes from malformed vertebrae were able to penetrate the gel.
Increased staining pattern of high molecular weight compounds were observed on the blots also for C-0-S and KS epitopes. In a previous study, an up-regulation at mRNA level of metalloproteinases (MMP-9), an enzyme able to degrade PGs (Genovese et al. 2013), was demonstrated in malformed vertebrae from Atlantic salmon (Ytteborg et al. 2010a). In the mammalian IVD, a breakdown of aggregating PG components like aggrecan is related to loss of hydrostatic pressure, tissue dehydration and disc degeneration (Brzoska and Moniuszko-Jakoniuk 2001; Kauppila 1995; Urban and Mcmullin 1985; Yasuma et al. 1993).[](https://www.ncbi.nlm.nih.gov/mesh/D007632)
## Conclusion
*In the **Conclusion** section:*
The present study has outlined the temporal and spatial distribution of different sulphated GAG epitopes, 4-sulphated CS/DS, KS, C-0-S and C-6-S in the Atlantic salmon spinal cord, both in a developmental perspective and in the light of vertebral malformation. The expression of these components was confined to distinct regions of the notochord during the period when the vertebral bodies form and mineralize. The results showed a persistent spatial distribution pattern of the GAGs from 1100 d° to 15 g, indicating similar functions for the different GAGs both in the juvenile and in the mature spine. This pattern was disrupted in vertebral malformations, exhibiting gross alterations in the composition of the ECM accompanied with an increase in matrix degradation during the pathogenesis. Increased lamellar disorganization and fissures in the IVR were features of vertebral malformation.
More GAGs was produced in the matrix of notochord during the initial stages of pathogenesis judged by the stronger staining with Alcian blue. We suggest that GAGs are important for normal sheath integrity and that altered matrix composition is of importance for the pathogenesis.[](https://www.ncbi.nlm.nih.gov/mesh/D006025)
# References
*In the **References** section:*
|
# Introduction
Formation of [Chlorotriophenoxy Radicals](https://www.ncbi.nlm.nih.gov/mesh/C042329) from Complete Series Reactions of [Chlorotriophenols](https://www.ncbi.nlm.nih.gov/mesh/C042983) with [H and OH Radicals](https://www.ncbi.nlm.nih.gov/mesh/D005609)
# Abstract
*In the **Abstract** section:*
The chlorothiophenoxy radicals (CTPRs) are key intermediate species in the formation of polychlorinated dibenzothiophen[es/thianthrenes (PCDT/TAs)](https://www.ncbi.nlm.nih.gov/mesh/C042329). [In th](https://www.ncbi.nlm.nih.gov/mesh/C042329)is work, the formation of CTPRs from the complete s[eries reactions of 19 chlorothiophenol (CTP) c](https://www.ncbi.nlm.nih.gov/mesh/C009449)on[geners w](https://www.ncbi.nlm.nih.gov/mesh/C009449)ith H and OH radicals were investi[gated](https://www.ncbi.nlm.nih.gov/mesh/C042329) theoretically by using the density functi[onal theory (DFT](https://www.ncbi.nlm.nih.gov/mesh/C042983)) [met](https://www.ncbi.nlm.nih.gov/mesh/C042983)hod. The profiles[ of the potential](https://www.ncbi.nlm.nih.gov/mesh/D005609) energy surface were constructed at the MPWB1K/6-311+G(3df,2p)//MPWB1K/6-31+G(d,p) level. The rate constants were evaluated by the canonical variational transition-state (CVT) theory with the small curvature tunneling (SCT) contribution at 600–1200 K. The present study indicates that the structural parameters, thermal data, and rate constants as well as the formation potential of CTPRs from CTPs are strongly dominated by the chlorine substitution at the ortho-position of CTPs. Comparison with [the s](https://www.ncbi.nlm.nih.gov/mesh/C042329)tudy o[f fo](https://www.ncbi.nlm.nih.gov/mesh/C042983)rmation of chlorophenoxy radica[ls (CPRs](https://www.ncbi.nlm.nih.gov/mesh/D002713)) from chlorophenols (CPs) clearly show[s th](https://www.ncbi.nlm.nih.gov/mesh/C042983)at the thiophenoxyl-hydrogen abstraction fro[m CTPs by H is more ef](https://www.ncbi.nlm.nih.gov/mesh/C042329)fi[cien](https://www.ncbi.nlm.nih.gov/mesh/C042329)t than [the phenoxyl-](https://www.ncbi.nlm.nih.gov/mesh/D002733)hy[dro](https://www.ncbi.nlm.nih.gov/mesh/D002733)gen abstraction from CPs [by H, wherea](https://www.ncbi.nlm.nih.gov/mesh/C042983)s[ the thi](https://www.ncbi.nlm.nih.gov/mesh/D006859)ophenoxyl-hydrogen[ abs](https://www.ncbi.nlm.nih.gov/mesh/C042983)trac[t](https://www.ncbi.nlm.nih.gov/mesh/D006859)ion from CTPs by OH is less [impactfu](https://www.ncbi.nlm.nih.gov/mesh/D010636)l[ than th](https://www.ncbi.nlm.nih.gov/mesh/D006859)e phenoxyl-hydroge[n a](https://www.ncbi.nlm.nih.gov/mesh/D002733)bstr[a](https://www.ncbi.nlm.nih.gov/mesh/D006859)ction from CPs[ by OH. Reac](https://www.ncbi.nlm.nih.gov/mesh/C042983)t[ions of ](https://www.ncbi.nlm.nih.gov/mesh/D006859)CTPs with H can oc[cur ](https://www.ncbi.nlm.nih.gov/mesh/C042983)more[ r](https://www.ncbi.nlm.nih.gov/mesh/D017665)eadily than that of CTPs wit[h OH, wh](https://www.ncbi.nlm.nih.gov/mesh/D010636)i[ch is op](https://www.ncbi.nlm.nih.gov/mesh/D006859)posite to the reac[tiv](https://www.ncbi.nlm.nih.gov/mesh/D002733)ity [co](https://www.ncbi.nlm.nih.gov/mesh/D017665)mparison of CPs[ wit](https://www.ncbi.nlm.nih.gov/mesh/C042983)h H an[d](https://www.ncbi.nlm.nih.gov/mesh/D006859) OH.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
## 1. Introduction
*In the **1. Introduction** section:*
Polychlorinated dibenzothiophenes (PCDTs) and polychlorinated thianthrenes (PCTAs) are analogues of dibenzofurans (PCDFs) and polychlorinated dibenzo-p-dioxins (PCDDs), respectively, in which the oxygen atoms are substituted by the sulfur atoms. Therefore, they have similar geochemical behavior, toxicity, and physicochemical properties in the environment. PCDT/TAs have been detected in various environmental samples such as stack gases, incineration of municipal waste, pulp mill effluents, soil and sediments, petroleum refineries, petroleum spills, pine needles, and some aquatic organisms. The long-term adverse environmental effects of PCDT/TAs have been at the forefront of public and regulatory concern, and information about the formation mechanisms of PCDT/TAs in the environment is required.[](https://www.ncbi.nlm.nih.gov/mesh/C016366)
PCDT/TAs were never intentionally synthesized for commercial purposes, but are formed as byproducts from the chemical processes that are similar to those resulting in the formation of PCDD/Fs. The major known sources of PCDT/TAs in environment are combustion, emissions from municipal and hazardous waste incinerators as well as industrial incinerators. High correlation between concentrations of PCDT/TAs and PCDD/Fs in the environmental samples revealed their similar formation mechanism under the pyrolysis or combustion conditions. The most direct route to the formation of PCDT/TAs is the gas-phase reaction of chemical precursors.[](https://www.ncbi.nlm.nih.gov/mesh/C009449)
Chlorophenols (CPs) are structurally similar to PCDD/Fs and the most direct precursors of PCDD/Fs. Similarly, chlorothiphenols (CTPs) are structurally similar to PCDT/TAs and have been demonstrated to be the predominant precursors or key intermediates of PCDT/TA formation. For instance, pentachlorothiophenol, an important additive in the vulcanization process of rubber in the tire industry, represents an important precursor for the formation of octachlorodibenzothiophene (octaCDT), heptachlorothianthrene (heptaCTA), and octachlorothianthrene (octaCTA). CTPs have been widely used in large quantities in various chemical industries, such as in manufacturing of dyes, insecticides, printing inks, pharmaceuticals, and polyvinyl chloride. CTPs are toxic and hazardous to human health and environment due to the presence of sulfur and chlorine.
Variously halogenated derivatives of phenol and thiophenol were subjected to analysis of their inhibitory effect on human cytochrome P450 (CYP) 2E1, which showed that dichlorothiophenols have stronger potent inhibitory activities than dichlorophenols, and the toxicity of CTPs are influenced by chlorine substitution pattern.[](https://www.ncbi.nlm.nih.gov/mesh/D002733)
Similar to the formation of PCDD/Fs from CP precursors, the gas-phase formation of PCDT/TAs from CTP precursors was also proposed involving radical-radical coupling of two CTPRs and radical-molecule recombination of CTPR and CTP. The recent works have shown that radical-radical coupling are more competitive thermodynamically than radical-molecule recombination for the PCDT/TA formation. The dimerization of CTPRs is the major PCDT/TA formation pathway. Thus, the formation of CTPRs is the initial and key step involved in the formation of PCDT/TAs. Under the pyrolysis or combustion conditions, CTPRs can be formed through loss of the triophenoxyl-hydrogen via unimolecular, bimolecular, or possibly other low-energy pathways (including heterogenous reactions).
The unimolecular reaction includes the decomposition of CTPs with the cleavage of the S–H bond. The bimolecular reactions include attack by H, OH, O(3P), or Cl under high-temperature oxidative conditions. As yet, very little work has been done at the high temperatures relevant to these reactions.[](https://www.ncbi.nlm.nih.gov/mesh/C023614)
In recent research from this laboratory, we investigated the formation of chlorophenoxy radicals (CPRs) from the reactions of CPs with H and OH radicals, based on the kinetic model conclusion that PCDD/F yields are most sensitive to the reactions of CPs with H and OH. Thus, as part of our ongoing work in the field, the thiophenoxyl-hydrogen abstraction from CTPs with H and OH are naturally expected to play the most central role in the formation of CTPRs. Here, therefore, we performed a direct density functional theory (DFT) kinetic study on the formation of CTPRs from the complete series reactions of 19 CTP congeners with H and OH radicals. We also studied the reactions of thiophenol with H and OH radicals for comparison. The effect of the chlorine substitution pattern on the structures, energies, and rate constants is discussed.
The formation potential of CTPRs from CTPs with H and OH are compared with that of CPRs from CPs with H and OH, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/C042329)
## 2. Results and Discussion
*In the **2. Results and Discussion** section:*
Due to the different substitution pattern of thiophenol, chlorothiophenols have 19 congeners, including three monochlorothiophenols (2-CTP, 3-CTP and 4-CTP), six dichlorothiophenols (2,3-DCTP, 2,4-DCTP, 2,5-DCTP, 2,6-DCTP, 3,4-DCTP and 3,5-DCTP), six trichlorothiophenols (2,3,4-TCTP, 2,3,5-TCTP, 2,3,6-TCTP, 2,4,5-TCTP, 2,4,6-TCTP and 3,4,5-TCTP), three tetrachlorothiophenols (2,3,4,5-TeCTP, 2,3,4,6-TeCTP and 2,3,5,6-TeCTP), and pentachlorothiophenols (PCTP). Due to the asymmetric chlorine substitution, there are syn and anti-conformers for 2-CTP, 3-CTP, 2,3-DCTP, 2,4-DCTP, 2,5-DCTP, 3,4-DCTP, 2,3,4-TCTP, 2,3,5-TCTP, 2,3,6-TCTP, 2,4,5-TCTP, 2,3,4,5-TeCTP and 2,3,4,6-TeCTP, respectively. The conformer with the sulfydryl-hydrogen facing the closest neighboring Cl is labeled as the syn-conformer and otherwise the anti-conformer (Figure 1). For a given CTP, the syn-conformer is about 0.5 kcal/mol more stable than the corresponding anti form, suggesting a stabilization effect because of intramolecular hydrogen bonding. So, throughout this paper, CTPs denote the syn-conformers.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
syn and anti conformers of CTP.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
The structures of CTPs along with the structure of thiophenol are presented in the Figure S1 of Supplementary Materials. The structural parameters of CTPs are strongly influenced by the ortho-substituted chlorine regardless of the number of chlorine substituents. There exists weak intramolecular hydrogen bonding in the ortho-substituted CTPs. The lengths of the intramolecular hydrogen bonds are from 2.391 to 2.490 Å. No such intramolecular hydrogen bonding forms in the anti-conformers except those with chlorine substitutions at both ortho-positions.
The C–S bonds in CTPs are from 1.749 to 1.761 Å, which are longer than the C–S double bond and shorter than the C–S single bond. The C–S bonds lengths (1.749–1.754 Å) in the ortho-substituted CTPs are consistently shorter than those for all nonortho forms (1.756–1.761 Å). The structures of CTPRs along with the structure of thiophenoxy radical are shown in the Figure S2 of Supplementary Materials.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
Comparison of C–S bond lengths presented in Figure S1 with the C–O bond lengths of CPs in our previous study clearly shows that C–S bond lengths in CTPs are longer than the C–O bond length of CPs (1.331–1.352 Å). Similarly, S–H bond lengths in CTPs (1.332 or 1.333 Å) are longer than the O–H bond length of CPs (0.955–0.960 Å). Table S1 shows the NBO charge of S and H of CTPs (NBOS and NBOH), NBO charge of O and H of CPs (NBOO and NBOH) and HOMO-LOMO gap of CTPs and CPs at MPWB1K/6-31+G(d,p) level. At a given CP and CTP, the NBOO of CP is more negative than the NBOS of CTP, and the NBOH of CP is more positive than that of CTP. This means the O atom in CP have stronger nucleophilicity than the S atom in CTP, i.e., the O–H bond strength in CP is stronger than the S–H bond in CTP.
In addition, the HOMO-LOMO gap of CP is larger than that of CTP, which reconfirms that CP is more stable than CTP. Figure S3 depicts the electron density of 2-CTP/2-CP and 3-CTP/3-CP at MPWB1K/6-311+G(3df,2p) level. The S–H and C–S bond lengths of thiophenol were also studied by Larsen et al. using both experimental investigation and ab initio molecular calculations at B3LYP/aug-cc-pVQZ, MP2(full)/aug-cc-pVTZ and MP2(full)/aug-cc-pVQZ levels. The S–H bond length of 1.333 Å and C–S bond length of 1.761 Å obtained in our study at the MPWB1K/6-31+G(d,p) level are in good agreement with the experimental value of 1.333 Å and 1.773 Å with the discrepancy less than 1.0%. Compared with Larsen’s calculation values of S–H and C–S bond lengths, our MPWB1K/6-31+G(d,p) results are slightly closer to the values at MP2(full)/aug-cc-pVTZ (1.334 Å for S–H bond and 1.763 Å for C–S bond) and MP2(full)/aug-cc-pVQZ levels (1.332 Å for S–H bond and 1.760 Å for C–S bond) than those at B3LYP and aug-cc-pVQZ levels (1.341 Å for S–H bond and 1.779 Å for C–S bond).[](https://www.ncbi.nlm.nih.gov/mesh/D002733)
## 2.1. Reactions of CTPs with H
*In the **2.1. Reactions of CTPs with H** section:*
The formation of CTPRs from the reactions of CTPs with H proceeds via a direct hydrogen abstraction mechanism. The structures of the transition states were located at the MPWB1K/6-31+G(d,p) level and shown in Figure 2. The H–H and C–S bonds in CTPs with ortho-substitution (1.246–1.291 Å for H–H, and 1.752–1.758 Å for C–S) are systematically shorter than those without ortho-substitution (1.295–1.345 Å for H–H, and 1.764–1.770 Å for C–S), respectively. Besides, all the ortho-substituted transition states have relative longer S–H bonds lengths (1.395–1.405 Å) compared to those without ortho-substitution (1.387–1.393 Å). Table 1 gives the potential barriers and reaction heats obtained at the MPWB1K/6-311+G(3df,2p)//MPWB1K/6-31+G(d,p) level.
The formation of CTPRs from the reactions of CTPs with H is strongly exothermic. Table 1 shows that the potential barriers are significantly correlated with the position of the chlorine substitution at the thiophenolicring, but not with the number of chlorine substituents. For example, for dichlorothiophenols, the potential barriers of the thiophenoxyl-hydrogen abstraction from 2,3-DCTP, 2,4-DCTP, 2,5-DCTP and 2,6-DCTP are higher than those from 3,4-DCTP and 3,5-DCTP. For trichlorothiophenols, the potential barriers of the phenoxyl-hydrogen abstraction from 2,3,4-TCTP, 2,3,5-TCTP, 2,3,6-TCTP, 2,4,5-TCTP and 2,4,6-TCTP are higher than that from 3,4,5-TCTP. Obviously, for a given number of chlorine substitutions, the potential barriers for the thiophenoxyl-hydrogen abstraction from the ortho-substituted CTPs are consistently higher than those for other structural conformers.
The chlorine substitution at the ortho-position can lower the barrier heights of thiophenoxyl-hydrogen abstraction from CTPs by H. Intramolecular hydrogen bonding appears to stabilize the CTPs and reduce the reactivity of S–H bonds in CTPs with the ortho-substitution. A similar result was also observed in our previous study of CPs with H.[](https://www.ncbi.nlm.nih.gov/mesh/C042329)
MPWB1K/6-31+G(d,p) optimized geometries for the transition states of the thiophenoxyl-hydrogen abstraction from CTPs by H. Distances are in angstroms. Gray sphere, C; White sphere, H; Yellow sphere, S; Green sphere, Cl. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
Potential barriers ∆E (in kcal/mol), reaction heats ∆H (in kcal/mol, 0 K), imaginary frequencies (in cm−1) of the transition states, and the S–H bond dissociation energies D0 (S–H) (in kcal/mol) for the thiophenoxyl-hydrogen abstraction from CTPs by H.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
In order to further investigate the relative strength of the S–H bonds in CTPs, we also calculated the S–H bond dissociation energies D0 (S–H). The values of D0 (S–H) obtained at the MPWB1K/6-311+G(3df,2p)//MPWB1K/6-31+G(d,p) level are summarized in Table 1. D0 (S–H) of 2-CTP is higher than those of 4-CTP. Similarly, D0 (S–H) of 2,3-DCTP, 2,4-DCTP, 2,5-DCTP and 2,6-DCTP are higher than that of 3,4-DCTP. D0 (S–H) of 2,3,4-TCTP, 2,3,5-TCTP, 2,3,6-TCTP, 2,4,5-TCTP and 2,4,6-TCTP are higher than that of 3,4,5-TCTP.
The chlorine substitution at the ortho-position appears to increase the strength of the S–H bonds in CTPs. However, for a given number of chlorine substitutions, the S–H bond dissociation energies in CTPs with ortho-substitution are not consistently larger than those without ortho-substitution. For example, D0 (S–H) of 2-CTP is smaller than that of 3-CTP. D0 (S–H) of 2,3-DCTP, 2,4-DCTP and 2,6-DCTP are smaller than that of 3,5-DCTP. Chlorine in an aromatic ring is traditionally recognized as an electron-withdrawing group.
The intramolecular hydrogen bonding in the ortho-substituted CTPs as well as the inductive effect of the electron-withdrawing chlorine and steric effect may ultimately be responsible for the relative strength of the S–H bonds in CTPs.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
It is interesting to compare the thiophenoxyl-hydrogen abstraction from CTPs by H with the phenoxyl-hydrogen abstraction from CPs by H. For a given CTP, the potential barrier for the thiophenoxyl-hydrogen abstraction from CTP by H is about 8–11 kcal/mol lower than phenoxyl-hydrogen abstraction from corresponding CP by H. In addition, the thiophenoxyl-hydrogen abstraction by H is more exothermic than the phenoxyl-hydrogen abstraction by H. This indicates that the thiophenoxyl-hydrogen abstraction from CTPs by H can occur more promptly than the phenoxyl-hydrogen abstraction CPs by H.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
## 2.2. Reactions of CTPs with OH
*In the **2.2. Reactions of CTPs with OH** section:*
For thiophenoxyl-hydrogen abstraction from CTPs by OH radical, prereactive intermediates are formed before the transition state. The structures of the prereactive intermediates are presented in Figure 3. As shown in Figure 3, conformations of the intermediates are difference between ortho-substituted structures and nonortho-substituted structures. In the ortho-substituted intermediates, H(1) atom is at the trans-position of O with respect to the O–H(2) bond. In contrast, H(1) atom is at the cis-position of O in the intermediates without ortho-substitution.
In addition, the ortho-substitution also has an effect on other structural parameters, such as the H(1)–O, H(2)–S and C–S bonds. For example, all the ortho-substituted intermediates have relatively shorter C–S bond distances (1.748–1.757 Å) compared to those without ortho-substitution (1.756–1.761Å). The relative energy, ∆EIM, of the intermediate with respect to the total energy of the corresponding CTP and OH is listed in Table 2.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
MPWB1K/6-31+G(d,p) optimized geometries for the prereactive intermediates of the thiophenoxyl-hydrogen abstraction from CTPs by OH. Distances are in angstroms. Gray sphere, C; White sphere, H; Yellow sphere, S; Red sphere, O; Green sphere, Cl. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.).[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
The relative energies of the intermediates ∆EIM (in kcal/mol), potential barriers ∆ETS (in kcal/mol), reaction heats ∆H (in kcal/mol, 0 K), imaginary frequencies (in cm−1) of the transition states for the triophenoxyl-hydrogen abstraction from CTPs by OH.[](https://www.ncbi.nlm.nih.gov/mesh/D010636)
The structures of the transition states are depicted in Figure 4. As shown in Figure 4, H(1) atom is at the trans-position of O with respect to the O–H(2) bond in the transition states with or without ortho-substitution. There are exit weak intramolecular hydrogen bondings in all the structures, which are governed by the chlorine substitution pattern. In the ortho-transition states, the intramolecular hydrogen bondings are between H(1) and ortho Cl atoms. In the transition states without ortho-substitution, the intramolecular hydrogen bondings are between O and H(3) atoms.
The hydrogen bond can lower the energy of the transition state, i.e., lower the reaction potential barrier. Besides, the ortho-substitution also impacts other essential structural parameters of the transition states. Generally, the breaking S–H(2) bonds in the ortho-substituted transition states (1.437–1.456 Å) are longer than those without ortho-substitution (1.414–1.431 Å). The forming O–H(2) bonds in the transition states with ortho-substitution (1.382–1.419 Å) are shorter than those without ortho-substitution (1.420–1.464 Å).[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
MPWB1K/6-31+G(d,p) optimized geometries for the transition states of the thiophenoxyl-hydrogen abstraction from CTPs by OH. Distances are in angstroms. Gray sphere, C; White sphere, H; Yellow sphere, S; Red sphere, O; Green sphere, Cl. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.).[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
The potential barriers and reaction heats calculated at the MPWB1K/6-311+G(3df,2p)//MPWB1K/6-31+G(d,p) level are shown in Table 2. In particular, the potential barrier is the relative energy of the transition state with respect to the total energy of the separated reactants (the corresponding CTP and OH), without considering the very shallow prereactive intermediate. It can be seen from Table 2 that the potential barriers for the thiophenoxyl-hydrogen abstraction from the ortho-substituted CTPs by OH radicals consistently are higher than those from CTPs without ortho-substitution. This reaffirms the conclusion above that the chlorine substitution at the ortho position increases the strength of the S–H bonds and decreases its reactivity.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
It is also necessary to compare the thiophenoxyl-hydrogen abstraction from CTPs by OH with the phenoxyl-hydrogen abstraction from CPs by OH of our previous study. For a given chlorotriophenol, the potential barrier for the thiophenoxyl-hydrogen abstraction from CTPs by OH is about 5–8 kcal/mol higher than phenoxyl-hydrogen abstraction from CPs by OH, which indicates that the thiophenoxyl-hydrogen abstraction from CTPs by OH are more difficult to happen than the phenoxyl-hydrogen abstraction CPs by OH. Compared to the stereo configurations of the transition states from CPs with OH, the transition states from CTPs with OH have the planar structure with all the S, O, H(1), and H(2) atoms almost in the same plane as the benzene ring. This can higher the energy of transition states, i.e., higher the potential energy of CTPs with OH radicals.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
Comparison of the values presented in Table 1 and Table 2 shows that for a given CTP, the potential barrier for the thiophenoxyl-hydrogen abstraction by OH is about 4–6 kcal/mol lower than that of the thiophenoxyl-hydrogen abstraction by H, which indicates that the thiophenoxyl-hydrogen abstraction from CTPs by OH is less efficient than the thiophenoxyl-hydrogen abstraction by H. This is completely on the contrary to the fact of phenoxyl-hydrogen abstraction CPs by OH is more impactfulthan the phenoxyl-hydrogen abstraction by H.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
## 2.3. Rate Constant Calculations
*In the **2.3. Rate Constant Calculations** section:*
Canonical variational transition state theory (CVT) with small-curvature tunneling (SCT) contribution has been successfully performed for formation of CPRs from the complete series reactions of 19 CP congeners with H and OH radicals, and is an efficient method to calculate the rate constants. In this study, we used this method to calculate the rate constants for the formation of CTPRs from the complete series reactions of 19 CTP congeners with H and OH radicals over a wide temperature range of 600–1200 K, as shown in Tables S2 and S3 of Supplementary Materials, respectively. This temperature range covers the possible formation temperature of PCDT/TAs under the pyrolysis or combustion conditions. Due to the absence of the available experimental rate constants, it is difficult to make a direct comparison of the calculated CVT/SCT rate constants with the experimental values for the reactions of CTPs with H and OH. Our previous studies have shown that the CVT/SCT rate constants of phenol + H → phenoxy + H2 and phenol + OH → phenoxy + H2O are in good agreement with the available experimental values.
To be used more effectively, the CVT/SCT rate constants are fitted, and Arrhenius formulas are given in Table 3 for the triophenoxyl-hydrogen abstraction from CTPs by H and in Table 4 for the thiophenoxyl-hydrogen abstraction from CTPs by OH. The pre-exponential factor, the activation energy, and the rate constants can be obtained.[](https://www.ncbi.nlm.nih.gov/mesh/C042329)
Arrhenius formulas (in cm3·molecule−1·s−1) for the thiophenoxyl-hydrogen abstraction from chlorothiophenols and thiophenol by H over the temperature range of 600–1200 K.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
The chlorine substitution pattern of thiophenol strongly affects the CVT/SCT rate constants. At a given temperature, the calculated CVT/SCT rate constants for the thiophenoxyl-hydrogen abstraction from 2-CTP by H or OH radical is smaller than those of the thiophenoxyl-hydrogen abstraction from 3-CTP and 4-CTP by H or OH radical, respectively. The calculated CVT/SCT rate constants for the thiophenoxyl-hydrogen abstraction from 2,3-DCTP, 2,4-DCTP, 2,5-DCTP and 2,6-DCTP by H or OH radical are smaller than those of the thiophenoxyl-hydrogen abstraction from 3,4-DCTP and 3,5-DCTP by H or OH radical, respectively. The CVT/SCT rate constants for the thiophenoxyl-hydrogen abstraction from 2,3,4-TCTP, 2,3,5-TCTP, 2,3,6-TCTP, 2,4,5-TCTP and 2,4,6-TCTP by H or OH are smaller than that of the thiophenoxyl-hydrogen abstraction from 3,4,5-TCTP by H or OH, respectively. For example, at 1000 K, the CVT/SCT rate constants are 2.54 × 10−12, 1.35 × 10−12, 1.49 × 10−12, 1.63 × 10−12, 1.68 × 10−12 cm3·molecule−1·s−1 for reactions of 2,3,4-TCP, 2,3,5-TCP, 2,3,6-TCP, 2,4,5-TCP and 2,4,6-TCP with H, while the value is 3.82 × 10−12 cm3·molecule−1·s−1 for that from 3,4,5-TCP with H. Similarly, at 1000 K, the CVT/SCT rate constants are 2.99 × 10−15, 1.38 × 10−15, 2.76 × 10−15, 2.44 × 10−15, 1.20
× 10−16 cm3·molecule−1·s−1 for reactions of 2,3,4-TCP, 2,3,5-TCP, 2,3,6-TCP, 2,4,5-TCP and 2,4,6-TCP with OH, while the value is 3.44 × 10−14 cm3·molecule−1·s−1 for that from 3,4,5-TCP with OH. This perfectly matches the structural and thermodynamic analysis above that the chlorine substitution at the ortho-position of CTPs increases the strength of the S–H bonds and decreases its reactivity.[](https://www.ncbi.nlm.nih.gov/mesh/D002713)
Arrhenius formulas (in cm3·molecule−1·s−1) for the triophenoxyl-hydrogen abstraction from chlorothiophenols and thiophenol by OH over the temperature range of 600–1200 K.[](https://www.ncbi.nlm.nih.gov/mesh/D010636)
For a given thiochlorophenol, the CVT/SCT rate constants for the thiophenoxyl-hydrogen abstraction by H are noticeably larger than those of the thiophenoxyl-hydrogen abstraction by OH over the whole studied temperature range. For example, at 1000 K, the CVT/SCT rate constant of the thiophenoxyl-hydrogen abstraction from 2,3-DCTP by H is 2.66 × 10−12 cm3·molecule·s−1, whereas the value is 2.91 × 10−15 cm3·molecule·s−1 for the thiophenoxyl-hydrogen abstraction from 2,3-DCTP by OH. This is consistent with thermodynamic analysis: Thiophenoxyl-hydrogen abstraction from CTPs by H is more efficient than the thiophenoxyl-hydrogen abstraction by OH.[](https://www.ncbi.nlm.nih.gov/mesh/D002733)
Comparison with the previous studies of phenoxyl-hydrogen abstraction by H and OH from CPs shows that the CVT/SCT rate constant for the reaction of CTP with H is consistently larger than that of corresponding CP with H at a given temperature, whereas the CVT/SCT rate constant for the reaction of CTP with OH is consistently smaller than that of corresponding CP with OH. This reconfirms thermodynamic analysis that the thiophenoxyl-hydrogen abstraction from CTPs by H is more efficient than the thiophenoxyl-hydrogen abstraction by H and the thiophenoxyl-hydrogen abstraction by OH is less efficient than the phenoxyl-hydrogen abstraction by OH.[](https://www.ncbi.nlm.nih.gov/mesh/D010636)
## 3. Experimental Section
*In the **3. Experimental Section** section:*
## 3.1. Density Functional Theory
*In the **3.1. Density Functional Theory** section:*
The Gaussian 09 program was used to perform all the calculations on the geometries, energies, frequencies for stationary points (reactants, prereactive intermediates, transition states, and products). The MPWB1K method is a hybrid density functional theory (HDFT) model with excellent performance in thermochemistry, thermochemical kinetics, hydrogen bonding and weak interactions. This method has been successfully performed for formation of CPRs from the complete series reactions of 19 CP congeners with H and OH radicals. As a serious ongoing work, it is important to use a consistent method for the species involved in the formation of CTPRs from CTPs with H and OH radicals and compare the formation potential of CTPRs and CTPs. Geometry optimizations were optimized at the MPWB1K/6-31+G(d,p) level.
The vibrational frequencies were also calculated at the same level to determine the nature of the stationary points, the zero-point energy (ZPE), and the thermal contributions to the free energy of activation. The intrinsic reaction coordinate (IRC) calculations were further carried out at the MPWB1K/6-31+G(d,p) level to confirm that the transition state connects to the right minima along the reaction path. For a more accurate evaluation of the energy parameters, a more flexible basis set, 6-311+G(3df,2p), was employed to determine the single-point energies of the various species. The profiles of the potential energy surface were constructed at the MPWB1K/6-311+G(3df,2p)//MPWB1K/6-31+G(d,p)level, including ZPE correction.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## 3.2. Kinetic Calculation
*In the **3.2. Kinetic Calculation** section:*
Rate constants in this study over a wide temperature range (600–200 K) were calculated using the canonical variational transition state theory (CVT) with small-curvature tunneling (SCT) correction. To calculate the rate constants, 40 non-stationary points near the transition state along the minimum energy path, 20 points on the reactants side and 20 points on the product side were selected. Rate constant calculations were carried out using the Polyrate 9.7 program.
## 3.3. Accuracy Verification
*In the **3.3. Accuracy Verification** section:*
The optimized geometries of thiophenol and the calculated vibrational frequencies of thiophenol and 4-chlorotriophenol at the MPWB1K/6-31+G(d,p) level are consistent with the available experimental values, and the relative deviation remains within 1.0% for the geometry parameters and 9.0% for the vibrational frequencies. To verify the reliability of the energy parameters, we calculated S–H bond dissociation energy for the reaction of thiophenol → thiophenoxy + H at the MPWB1K/6-311+G(3df,2p)//MPWB1K/6-31+G(d,p) level. The calculated value of 86.51 kcal/mol at 298.15 K and 1.0 atm is in excellent agreement with the corresponding experimental value of 86.5 kcal/mol. From these results, we inferred that accuracy can be expected for the species involved in this study.[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
## 4. Conclusions
*In the **4. Conclusions** section:*
In this study, we investigated the theoretical formation of chlorothiophenoxy radicals (CTPRs) from the complete series reactions of 19 chlorothiophenol (CTP) congeners with H and OH radicals using DFT electronic structure theory and canonical variational transition-state (CVT) theory with the small curvature tunneling (SCT) contribution. Structural parameters were calculated for all the stationary (reactants, prereactive intermediates, transition states, and products). Potential barriers, reaction heats, and rate constants for all the elementary reactions were studied to compare the formation potential of CTPRs from CTPs with H and OH radicals. Comparison of this study with our previous studies of the chlorophenoxy radical (CPR) formation from chlorophenols (CPs) with H and OH radicals were discussed. Three specific conclusions can be drawn:[](https://www.ncbi.nlm.nih.gov/mesh/C042329)
(1) The ortho chlorine increases the strength of the S–H bond in CTPs and decreased its reactivity, i.e., decreases the formation potential of CTPRs from the ortho-substitued CTPs with H and OH radicals.[](https://www.ncbi.nlm.nih.gov/mesh/D002713)
(2) The triophenoxyl-hydrogen abstraction from CTPs by H is more efficient than the phenoxyl-hydrogen abstraction from CPs by H, whereas the thiophenoxyl-hydrogen abstraction from CTPs by OH is less impactful than the phenoxyl-hydrogen abstraction from CPs by OH.[](https://www.ncbi.nlm.nih.gov/mesh/D010636)
(3) Different from reactions of CPs with H and OH, the thiophenoxyl-hydrogen abstraction from CTPs by H can occur more readily than the thiophenoxyl-hydrogen abstraction by OH radical.[](https://www.ncbi.nlm.nih.gov/mesh/D002733)
The obtained results can support the important input parameters for the PCDT/TA control models in the environment, and be used for future estimates of PCDT/TAs emission quantity based on the well estimated PCDT/TA inventory.[](https://www.ncbi.nlm.nih.gov/mesh/C009449)
# Supplementary Materials
*In the **Supplementary Materials** section:*
MPWB1K/6-31+G(d,p) optimized structures of CTPs and CTPRs. Electron density from total SCF density of 2-CTP, 2-CP, 3-CTP and 3-CP at MPWB1K/6-311+G(3df,2p) level. NBO charge of S and H atoms of CTPs, NBO charge O and H atoms of CPs and HOMO-LOMO gap of CTPs and CPs at MPWB1K/6-31+G(d,p) level. CVT/SCT rate constants for the thiophenoxyl-hydrogen abstraction from CTPs by H and OH radicals. Supplementary materials can be found at .[](https://www.ncbi.nlm.nih.gov/mesh/C042983)
# Author Contributions
*In the **Author Contributions** section:*
Fei Xu designed and performed the mechanism calculations, then wrote the manuscript; Fei Xu and Xiangli Shi performed the kinetic calculation. Fei Xu, Xiangli Shi, Qingzhu Zhang and Wenxing Wang all analyzed the data in the manuscript.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
|
# Introduction
Adding [Sarcosine](https://www.ncbi.nlm.nih.gov/mesh/D012521) to Antipsychotic Treatment in Patients with Stable Schizophrenia Changes the Concentrations of Neuronal and Glial Metabolites in the Left Dorsolateral Prefrontal Cortex
# Abstract
*In the **Abstract** section:*
The glutamatergic system is a key point in pathogenesis of schizophrenia. Sarcosine (N-methylglycine) is an exogenous amino acid that acts as a glycine transporter inhibitor. It modulates glutamatergic transmission by increasing glycine concentration around NM[DA (N-met](https://www.ncbi.nlm.nih.gov/mesh/D012521)hy[l-d-aspartate) ](https://www.ncbi.nlm.nih.gov/mesh/D012521)receptors. In pati[ents with ](https://www.ncbi.nlm.nih.gov/mesh/D000596)schizophrenia, the function of the glutamatergic system in the prefrontal cortex is impaired, which m[ay prom](https://www.ncbi.nlm.nih.gov/mesh/D005998)ote negative and cognitive symptoms. Proton nuclear magnetic resonance (1H-NMR) spectroscopy is a non-invasive imaging method enabling the evaluation of brain metabolite concentration, which can be applied to assess pharmacologically induced changes. The aim[ o](https://www.ncbi.nlm.nih.gov/mesh/D006859)f the study was to evaluate the influence of a six-month course of sarcosine therapy on the concentration of metabolites (NAA, N-acetylaspartate; Glx, complex of glutamate, glutamine and γ-aminobutyric acid (GABA); mI, myo-inositol; Cr, creatine; Cho, c[holine) i](https://www.ncbi.nlm.nih.gov/mesh/D012521)n the left dorso-lateral prefrontal cortex (DL[PFC](https://www.ncbi.nlm.nih.gov/mesh/C000179)) [in patients with ](https://www.ncbi.nlm.nih.gov/mesh/C000179)st[abl](https://www.ncbi.nlm.nih.gov/mesh/D005680)e schizophren[ia. Fifty](https://www.ncbi.nlm.nih.gov/mesh/D018698) p[atients w](https://www.ncbi.nlm.nih.gov/mesh/D005973)ith s[chizophrenia, treat](https://www.ncbi.nlm.nih.gov/mesh/D005680)ed[ wit](https://www.ncbi.nlm.nih.gov/mesh/D005680)h c[on](https://www.ncbi.nlm.nih.gov/mesh/D007294)st[ant antipsyc](https://www.ncbi.nlm.nih.gov/mesh/D007294)ho[ti](https://www.ncbi.nlm.nih.gov/mesh/D003401)cs[ doses, ](https://www.ncbi.nlm.nih.gov/mesh/D003401)in[ st](https://www.ncbi.nlm.nih.gov/mesh/D002794)ab[le clin](https://www.ncbi.nlm.nih.gov/mesh/D002794)ical condition were randomly assigned to administration of sarcosine (25 patients) or placebo (25 patients) for six months. Metabolite concentrations in DLPFC were assessed with 1.5 Tesla 1H-NMR spectroscopy. Clinical symptoms were evaluated wit[h the Pos](https://www.ncbi.nlm.nih.gov/mesh/D012521)itive and Negative Syndrome Scale (PANSS). The first spectroscopy revealed no differences in metabolite concentrations b[et](https://www.ncbi.nlm.nih.gov/mesh/D006859)ween groups. After six months, NAA/Cho, mI/Cr and mI/Cho ratios in the left DLPFC were significantly higher in the sarcosine than the placebo group. In the sarcosine group, NAA/Cr, NAA/Cho, mI/Cr, mI/Cho ratios also s[ign](https://www.ncbi.nlm.nih.gov/mesh/C000179)i[fic](https://www.ncbi.nlm.nih.gov/mesh/D002794)an[tl](https://www.ncbi.nlm.nih.gov/mesh/D007294)y[ i](https://www.ncbi.nlm.nih.gov/mesh/D003401)ncrea[se](https://www.ncbi.nlm.nih.gov/mesh/D007294)d[ co](https://www.ncbi.nlm.nih.gov/mesh/D002794)mpared to baseline values. In the placebo group, only the N[AA/Cr rat](https://www.ncbi.nlm.nih.gov/mesh/D012521)io increased. The addition of sa[rcosine t](https://www.ncbi.nlm.nih.gov/mesh/D012521)o antips[ych](https://www.ncbi.nlm.nih.gov/mesh/C000179)otic [the](https://www.ncbi.nlm.nih.gov/mesh/C000179)r[apy](https://www.ncbi.nlm.nih.gov/mesh/D002794) f[or](https://www.ncbi.nlm.nih.gov/mesh/D007294) [si](https://www.ncbi.nlm.nih.gov/mesh/D003401)x [mo](https://www.ncbi.nlm.nih.gov/mesh/D007294)n[ths](https://www.ncbi.nlm.nih.gov/mesh/D002794) increased markers of neurons viability (NAA) and neurogilal activity (mI) with simultaneous impr[ove](https://www.ncbi.nlm.nih.gov/mesh/C000179)m[en](https://www.ncbi.nlm.nih.gov/mesh/D003401)t of clinical symptoms. Sarcosine,[ two gram](https://www.ncbi.nlm.nih.gov/mesh/D012521)s administered daily, seems to be an effective adjuvant in the pharmacotherapy of[ sc](https://www.ncbi.nlm.nih.gov/mesh/C000179)hizophrenia.[](https://www.ncbi.nlm.nih.gov/mesh/D007294)
## 1. Introduction
*In the **1. Introduction** section:*
Schizophrenia is one of the most devastating mental diseases, with lifetime prevalence from 0.30% to 0.66% and incidence between 10.2 and 22.0 per 100,000 person-years [1]. It is considered a heterogeneous group of psychoses, caused by a constellation of genetic and environmental factors, with documented heritability [2,3]. A few regions of the central nervous system (CNS) are known to play an important role in pathogenesis of schizophrenia. Mostly reported is the prefrontal cortex (PFC), with its dorso-lateral (DLPFC) and medial (MPFC) regions [4,5].
DLPFC dysfunction is responsible for the negative symptoms of schizophrenia, called “axial symptoms”: autistic behavior, anhedonia and avolition, emotional flattening and social withdrawal [6]. DLPFC also plays a substantial role in cognition, including executive functions that are particularly important in daily life, such as working memory, abstract thinking, task flexibility, planning, and impulse control [7,8]. Cognitive impairment is better predictor of long-term functional outcome in schizophrenia than severity of positive, negative or affective symptoms [9].
From the neurochemical perspective, negative and cognitive symptoms are associated with impairment of the glutamatergic system, especially in the PFC, where ionotropic NMDA receptors are abundant [10,11,12]. Glycine is a necessary co-agonist of the NMDA receptor, and sarcosine (N-methylglycine) is an exogenous amino acid that acts as an inhibitor of glycine transporter type 1 (GlyT-1) [13]. Thus, sarcosine should improve the inadequate function of NMDA receptors [12]: a hypothesis confirmed by the observed reduction of schizophrenia symptoms (negative and total symptomatology) associated with the augmentation of antipsychotic therapy with sarcosine [14,15,16,17,18] excluding clozapine [19] or treatment with sarcosine alone [20].[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
In schizophrenia, there is no consensus on the association between changes in CNS metabolites and exacerbation of symptoms, phase of the disease, treatment strategy or analyzed brain region [21,22,23,24]. Decreased concentrations of N-acetylaspartate (NAA), a marker of neuron viability and integrity, are commonly observed [25], and reflect neuronal loss and/or mitochondrial dysfunction [26,27]. However, meta-analyses performed by Steen and Brugger [21,22] found that the NAA concentration in the PFC was similar in patients with a first episode of schizophrenia and in the chronic phase of the disease. It was also not affected by the duration of untreated schizophrenia (DUP)
[28] and had already decreased during the pre-psychotic period [29]. Concentrations of NAA, glutamic acid (Glu) and glutamine (Gln) are important in the pathogenesis of schizophrenia, however, many studies have failed to confirm any correlation between metabolite concentration and clinical symptoms [30,31,32,33,34,35,36]. Nevertheless, a few studies have noted an association between negative symptoms and concentration of NAA in the PFC, thalamus and anterior cingulate cortex (ACC) [37,38,39,40].[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
Although the influence of medications was also evaluated spectroscopically, the findings are ambiguous. In some studies, antipsychotics increased levels of NAA after treatment [34,41], while in others, there were no significant changes [28,42,43,44]. It remains unclear if substances modifying glutamatergic transmission cause changes in concentrations of CNS metabolites detectable in spectroscopy.[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
The aim of the study is to evaluate the influence of sarcosine therapy on the concentrations of NAA, Glx (complex of glutamate, glutamine and γ-aminobutyric acid GABA), mI (myo-inositol), Cho (choline-containing compounds) and Cr (creatine plus phosphocreatine) in the DLPFC of the left frontal lobe in patients with schizophrenia. Our experiment can support new data on the pharmacokinetics, pharmacodynamics and psychopharmacological value of sarcosine, as well as glutamatergic agents in general. It can also reveal new aspects of the role played by the glutamatergic system in the pathogenesis of schizophrenia.[](https://www.ncbi.nlm.nih.gov/mesh/D012521)
## 2. Results and Discussion
*In the **2. Results and Discussion** section:*
At baseline, spectroscopy revealed no significant differences in metabolite concentrations between the groups (Table 1).
Comparison of substances concentrations ratios in study groups.
NAA, N-acetylaspartate; Cr, creatine; Cho, choline; mI, myo-inositol; Glx, glutamate, glutamine and GABA.[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
In a second spectroscopy NAA/Cho, mI/Cr and mI/Cho ratios were significantly higher in patients receiving sarcosine. Moreover in experimental group after the therapy NAA/Cr, NAA/Cho, mI/Cr, mI/Cho ratios increased significantly, comparing to baseline values.[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
Only NAA/Cr ratio increased after therapy in the placebo group, although to a lesser extent than in the sarcosine group (4.9% vs. 18%).[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
At the beginning of the study, no significant difference was noted between groups with regard to PANSS score (71.4 ± 14 vs. 73.3 ± 13 points in total score for sarcosine and placebo groups, respectively; p = 0.6736). However, at the end of the experiment, patients treated with sarcosine had significantly lower results (57.7 ± 15 vs. 71.5 ± 13 points for sarcosine and placebo group, respectively; p = 0.00487). Changes in the negative PANSS subscale followed the trends of the total PANSS subscale. At the beginning of the study there was no significant difference between groups (25.4 ± 5.2 vs. 26.1 ± 5 points for sarcosine and placebo groups, respectively; p = 0.45085). While the negative PANSS score decreased significantly in both groups (25.4 ± 5.2 vs. 18.6 ± 6.1 for the sarcosine group, p = 0.0000; and 26.1 ± 5 vs. 25.4 ± 4.7 for the placebo group, p = 0.03031), this decrease was greater in the sarcosine group (18.6 ± 6.1 vs. 25.4 ± 4.7; p = 0.00001).
The difference in metabolite ratios and negative PANSS subscale scores were calculated between the start-point and end-point of the experiment. Correlations between these differences are presented in Table 2 and in Figure 1.[](https://www.ncbi.nlm.nih.gov/mesh/D012521)
At the time of writing, this paper was the first attempt to spectroscopically assess the impact of the glutamatergic system modulators, particularly sarcosine, on metabolite concentrations in the DLPFC in patients with schizophrenia. Significant changes in the spectral characteristics co-occurring with alleviation of symptoms, assessed with the PANSS scale, imply that two grams of sarcosine daily sufficiently penetrates the blood-brain barrier to modify the neuronal activity in patients with schizophrenia. Moreover, significant negative correlations between differences in negative PANSS subscale score and spectroscopic parameters (NAA/Cho and mI/Cho ratios) suggest that these ratios might quantitatively correspond with clinical outcomes of therapeutic intervention.[](https://www.ncbi.nlm.nih.gov/mesh/D012521)
Correlation between the differences in metabolite ratios (A) NAA/Cho; (B) NAA/Cr; (C) mI/Cho; (D) mI/Cr and differences in negative PANSS subscale score.[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
Correlation between differences in the score of the negative PANSS subscale and metabolite ratios assessed at the beginning and at the end of the experiment.
## 2.1. NAA (N-Acetylaspartate)
*In the **2.1. NAA (N-Acetylaspartate)** section:*
N-acetylaspartate is one of the most common amino acid in the human brain. It is synthesized in neuronal mitochondria and its production closely correlates with glucose metabolism. Due to the fact that it is not present in glial cells, it reflects neuronal activity well [45].[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
In our study, both NAA ratios (NAA/Cr and NAA/Cho) in the sarcosine group were significantly higher after six months, indicating an increase of neuronal viability in the DLPFC. In the placebo group, the NAA/Cr ratio was also significantly raised, however, the change was less distinct. Our findings indicate that sarcosine (and probably other GlyT1 inhibitors) might normalize disturbances in brain metabolism and reverse the tendency for NAA levels to decline in schizophrenia. Increased NAA concentrations were also described as an effect of the antipsychotic drugs [34], which may confirm the value of glutamatergic therapy in the management of schizophrenia. Further investigations should assess whether they act synergistically, and if NAA concentration can be used as a marker of clinical outcome.[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
## 2.2. Glx (Complex of Glutamate, Glutamine and GABA)
*In the **2.2. Glx (Complex of Glutamate, Glutamine and GABA)** section:*
An evaluation of Glx level was performed instead of separate glutamine, glutamate and GABA evaluations, as their peaks closely overlap in 1.5 Tesla spectroscopy. Glx is a surrogate of glutamatergic transmission in grey matter as the concentration of glutamate is five times higher than that of glutamine, and 10 times higher than GABA [46].[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
In schizophrenia, hypofunction of the NMDA receptor may involve GABAergic interneurons, which would result in disturbed glutamatergic transmission [47]. Moreover, there is a decrease of glutamate receptor density on GABAergic interneurons [48]. The summary effect is the inadequate inhibition of glutamatergic neurons, observed in electroencephalography as disturbances of coherent neuronal oscillation at a rate below 0.1 Hz
[49] and γ rhythms (25–100 Hz) in PFC [47,50,51,52]. This information noise negatively affects the concentration process and cognitive functions [53,54,55]. Furthermore it can promote hallucinations [56], delusions, and disturbances of the thinking processes and cognitive functions typical of acute psychosis.
One of the causes of these pathological conditions is a dysfunction of default mode network (DMN), the “rest system’ of the brain, which should be switched off when working memory networks such as the external attention system (EAS) are activated [53]. In schizophrenia, DMN deactivation is impaired, increasing information noise, intensifying cognitive dysfunction [53,57] and general functioning problems [58].
There is no consensus on the glutamate concentration in the DLPFC of patients with schizophrenia [59]. Kegeles et al. showed no significant differences in Glx concentrations between healthy volunteers and groups of medicated and unmedicated patients with schizophrenia [60]. Only three studies have assessed effects of antipsychotics on Glx parameters in the DLPFC before and after treatment. Two studies explored the first episode of schizophrenia: Stanley et al. report a decrease in glutamine levels after 14 weeks of antipsychotic therapy
[61], and Goto et al. note decreased Glx levels in patients after six months of treatment with second-generation antipsychotics [62]. Research conducted in a Polish population showed no changes in Glx levels between baseline assessment and after 40 days of antipsychotic treatment in patients with chronic stage of schizophrenia. However, responders had lower Glx levels at baseline when compared to non-responders [46,63].[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
On the other hand, the administration of ketamine, an NMDA receptor antagonist whose effect is opposite to sarcosine, resulted in increased glutamatergic transmission in ACC [64,65].[](https://www.ncbi.nlm.nih.gov/mesh/D007649)
Most studies have failed to find any significant correlation between glutamatergic parameters and PANSS score [46,60,66,67,68]. Kegeles et al. report that PANSS positive symptoms subscale scores significantly correlated with levels of GABA and Glx only in MPFC but not in DLPFC [60].[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
In the present study, a trend was observed towards a decrease of Glx/Cr ratio in both groups. Although it was more expressed in the sarcosine group, the differences were not significant. Further studies using discreet analysis with a stronger magnetic field are required to support more reliable conclusions.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## 2.3. mI (myo-Inositol)
*In the **2.3. mI (myo-Inositol)** section:*
Myoinositol is a precursor in the transmission of phosphatidylinositol, which is a widely accepted glial marker [69]. In neurodegenerative processes, increased mI concentrations co-occur with reduced NAA concentrations.[](https://www.ncbi.nlm.nih.gov/mesh/D007294)
Significant increases of mI/Cr and mI/Cho ratios in the sarcosine group between two spectroscopies, and in comparison with the placebo group, might indicate unfavourable changes. However, some researchers report greater mI concentrations to be associated with treatment [41,70]. Thus, administration of sarcosine may secondarily activate glial cells, mostly astrocytes, because glycine transporters and other glutamatergic system transporters are abundant in their cell membranes [71].[](https://www.ncbi.nlm.nih.gov/mesh/D007294)
## 2.4. Limitations of the Study
*In the **2.4. Limitations of the Study** section:*
Due to the limited number of patients and application of 1.5 Tesla magnetic resonance, conclusions should be formulated moderately, as precise separation of glutamate, glutamine and GABA spectra requires a 3 Tesla magnetic field, or higher. Analysis of GABA concentration could be of special interest, because sarcosine indirectly acts on the NMDA receptors located also on GABAergic interneurons. A few studies have found that GABA concentrations varied depending on the analyzed region, including different parts of the frontal cortex [60,72]. On the other hand, prior research has revealed an absence of abnormalities in glutamate or glutamine concentrations in the DLPFC of unmedicated patients with schizophrenia.
Thus, the absence of schizophrenia-related glutamate abnormalities in this region may limit the ability to detect a treatment-related change in Glx ratios, which could be detectable in other regions where baseline abnormalities were found, such as the MPFC, striatum, hippocampus or thalamus.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Another important limitation of this work is its application of ratios of metabolites concentrations instead of exact concentrations. Despite changes of Cr and Cho concentrations, depending on duration of schizophrenia, it has previously be demonstrated that treatment with either atypical or typical medication does not alter Cr or Cho levels [73]. Thus, ratios might have a good intra-subject validity [73].[](https://www.ncbi.nlm.nih.gov/mesh/D003401)
Finally, it should be noted that applied statistical methods did not protect against Type I errors associated with multiple testing. However, although significant differences in particular metabolites could be obtained by chance, the clinical improvement seems to confirm their relevance.
## 3. Experimental Section
*In the **3. Experimental Section** section:*
Subjects with schizophrenia, aged 18–60 years who were physically, neurologically and endocrinologically healthy and had normal laboratory values (routine blood tests, biochemical tests including thyroid stimulating hormone, lipid profile, liver and kidney parameters) and electrocardiogram were eligible to enter the study. Patients in acute psychosis, on clozapine treatment or declaring suicidal tendencies were excluded from the study. This study is a part of the Polish Sarcosine Study in Schizophrenia (PULSAR); for further details, please see acknowledgments.[](https://www.ncbi.nlm.nih.gov/mesh/D003024)
Fifty right-handed patients diagnosed with schizophrenia (according to Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision (DSM-IV-TR) criteria) with dominant negative symptoms, and who were in a stable clinical condition, were randomly assigned to a sarcosine or placebo group. 1H-NMR spectroscopy was performed according to the protocol described below at the beginning of the study and six months later. Sarcosine or placebo were added to the ongoing antipsychotic treatment in a double-blind manner. Patients in the study group were given plastic capsules containing 2 grams of the amino acid, while subjects in the placebo group (similar age, sex, clinical presentation, duration of schizophrenia and treatment, Table 3) received capsules with microcrystalline cellulose. Subjects in both groups were ordered to drink the dissolved contents of one capsule once a day in the morning.
All patients were treated with stable doses of antipsychotic and other medication for a minimum of three months before the baseline visit. Doses of antipsychotic and antidepressive drugs were calculated for defined daily dose (DDD) developed by the World Health Organization. Antidepressants were used as a supportive therapy [74] in 14 patients from the sarcosine group and 11 from the placebo group. The differences in the numbers of treated patients and doses in each group were not significant (p > 0.05).
The severity of schizophrenia symptoms was assessed with the Positive and Negative Syndrome Scale (PANSS) [75].[](https://www.ncbi.nlm.nih.gov/mesh/D012521)
Subjects were recruited from the outpatient clinic. All patients included in the study have been informed about the aims and methods of the study, and had expressed their written informed consent for participation in this study. The study protocol was approved by the Bioethics Committee of the Medical University of Łódź (permission number and date: RNN/153/08/KE, 15.07.2008). There was no financial involvement from industry.
Group characteristics.
Abbreviations: n, number of patients; DDD, defined daily dose; PANSS, the Positive and Negative Syndrome Scale; SD, standard deviation.
## 3.1. Spectroscopy
*In the **3.1. Spectroscopy** section:*
Imaging was performed using 1.5 Tesla MR scanner (Siemens Avanto 1.5, Siemens AG, Munich, Germany) equipped with a standard head coil.
NMR acquisition:
(1) FLAIR sequences in axial plane with following parameters: Repetition Time (TR), 9000 ms; Echo Time (TE), 105 ms; inversion time (TI), 2500 ms; flip angle, 150°; voxel size 1.4 mm × 1.3 mm × 3 mm.
(2) T2-weighted sequences were obtained in coronal plane with following parameters: TR = 5000 ms; TE = 100 ms; flip angle, 50°; voxel size 0.6 mm × 0.6 mm × 5.0 mm.
(3) T1-weighted sequences in transverse plane with following parameters: TR = 400 ms; TE = 7.8 ms; flip angle, 90°; voxel size 0.9 mm × 0.9 mm × 0.5 mm.
1H-MRS data was acquired by single voxel spectroscopy (SVS) using a point resolved spin echo (PRESS) sequence 128 averages; TR, 3000 ms; TE, 30 ms; voxel size was 15 mm × 15 mm × 15 mm. The region of interest was placed in the left DLPFC by the neuroradiologist (Figure 2). During the second spectroscopy, voxel localization parameters were copied and adjusted to the position of patient. Automated procedures were used to optimize radiofrequency pulse power, field homogeneity, and water suppression, as well as to convert the lines into a Gaussian shape.
Spectroscopy data was processed by means of Avanto Syngo MR Software (Siemens AG, Munich, Germany), Level B15. The processing included: k-space Fourier transformation and a spatial 50 Hz Hanning filter; subtraction of the residual water signal; time domain 1 Hz exponential apodization; zero filling to 2048 points; Fourier transformation of the time domain signals; frequency shift correction, phase correction and baseline correction. The fitting error was automatically computed as a deviation between theoretical and measured spectrum determined using the last squares method. Values less than 0.5 were considered satisfactory, however, in the whole group mean fitting error was 0.36 (SD, standard deviation 0.07). The following metabolites were assessed: NAA, Glx, mI, Cho and Cr.
No absolute concentrations of metabolites were determined, but their ratios to Cr and Cho.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Images showing voxel location in the left DLPFC (dorso-lateral prefrontal cortex) area and an example before (white line) and after (red line) fitting. Peak areas for N-acetylaspartate (NAA); creatine (Cr and Cr2); choline (Cho); and myo-inositol (Ins dd1) are labelled.[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
## 3.2. Statistical Analysis
*In the **3.2. Statistical Analysis** section:*
Continuous variables are expressed as the mean ± standard deviation (SD). The Shapiro-Wilk test was used to determine the normality of the data. As the distribution was skewed in one or both compared groups in all cases, the Mann-Whitney test was employed to compare the ratios of substance concentrations between groups, and the Wilcoxon sign-rank test was used for comparisons within the same group. To evaluate the association between changes in concentrations ratios and differences in PANSS score, the Spearman’s rank correlation test was applied. Statistical analysis was performed using Statistica for Windows (version 12.0, StatSoft, Tulsa, OK, USA).
A p-value of ≤0.05 was considered significant.
## 4. Conclusions
*In the **4. Conclusions** section:*
Our findings demonstrate that addition of sarcosine to antipsychotic treatment can cause increases of NAA and mI in DLPFC. These changes were associated with clinical improvement. It indicates that sarcosine improves neuron viability and integrity, and may activate neuroglial cells in brain regions essential for the pathogenesis of schizophrenia. It highlights the role of glutamatergic transmission in the pathogenesis of schizophrenia and confirms that two grams of sarcosine administered daily may become an effective adjuvant in the management of schizophrenia.[](https://www.ncbi.nlm.nih.gov/mesh/D012521)
# Author Contributions
*In the **Author Contributions** section:*
Dominik Strzelecki, Piotr Grzelak and Ludomir Stefańczyk conceived and designed the experiments; Dominik Strzelecki, Olga Kałużyńska, Magdalena Kotlicka-Antczak, Agnieszka Gmitrowicz, Piotr Grzelak and Michał Podgórski performed the experiments; Dominik Strzelecki and Michał Podgórski analyzed the data and wrote the paper.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
|
# Introduction
The Deoxynucleoside Triphosphate Triphosphohydrolase Activity of SAMHD1 Protein Contributes to the Mitochondrial DNA Depletion Associated with Genetic Deficiency of Deoxyguanosine Kinase*
# Abstract
*In the **Abstract** section:*
Background: Deoxyguanosine kinase (dGK) deficiency causes mtDNA depletion in quiescent cells where SAMHD1 restricts dGTP concentration.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
Results: siRNA silencing of SAMHD1 corrects mt dNTP imbalance boosting mtDNA content in quiescent dGK-mutated human fibroblasts.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
Conclusion: In noncycling cells SAMHD1 activity worsens the phenotype of dGK genetic deficiency, limiting dGTP supply.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
Significance: Nuclear SAMHD1 and mitochondrial dGK functionally interact in the regulation of mt dGTP for mtDNA maintenance.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
The dNTP triphosphohydrolase SAMHD1 is a nuclear antiviral host restriction factor limiting HIV-1 infection in macrophages and a major regulator of dNTP concentrations in human cells. In normal human fibroblasts its expression increases during quiescence, contributing to the small dNTP pool sizes of these cells. Down-regulation of SAMHD1 by siRNA expands all four dNTP pools, with dGTP undergoing the largest relative increase. The deoxyguanosine released by SAMHD1 from dGTP can be phosphorylated inside mitochondria by deoxyguanosine kinase (dGK) or degraded in the cytosol by purine nucleoside phosphorylase. Genetic mutations of dGK cause mitochondrial (mt) DNA depletion in noncycling cells and[ hep](https://www.ncbi.nlm.nih.gov/mesh/D003854)ato-cerebral mtDNA depletion syndrome in humans. We studied if SAMHD1 and dGK interact in the regulation of the dGTP pool during quiescence [empl](https://www.ncbi.nlm.nih.gov/mesh/D003854)oying dGK-mutated skin fibroblasts derived from three unrelated patients. In the presence of SAMHD1 quiescent mutant fibroblasts m[anif](https://www.ncbi.nlm.nih.gov/mesh/D003854)ested mt dNTP pool imbalance and mtDNA depletion. When SAMHD1 was silenced by si[RNA ](https://www.ncbi.nlm.nih.gov/mesh/D003854)transfection [the ](https://www.ncbi.nlm.nih.gov/mesh/C029603)composition of the mt dNTP pool approached that[ of the contro](https://www.ncbi.nlm.nih.gov/mesh/D003849)ls, and mtDNA copy number[ inc](https://www.ncbi.nlm.nih.gov/mesh/C029603)reased, compensating the depletion to various degrees in the different mutant fibroblasts. Chemical inhibition of purine nucleoside phosphorylase did not improve deoxyguanosine recycling by dGK in WT cells. We conclude that the activity of SAMHD1 contributes to the pathological phenotype of dGK deficiency. Our results prove the importance[ of ](https://www.ncbi.nlm.nih.gov/mesh/C029603)SAMHD1 in the regulation of all dNTP pools and suggest that dGK inside mitochondria has the function of recycling the deoxyguanosine derived from endogenous dGTP degraded [by S](https://www.ncbi.nlm.nih.gov/mesh/D003854)AMHD1 in the nucleus.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
In mammalian cells the concentrations of deoxyribonucleoside triphosphates (dNTPs) are regulated by an interplay of synthesis and degradation. Synthesis of dNTPs occurs through two pathways, de novo synthesis of dNDPs in the cytosol by ribonucleotide reductase (RNR)3 followed by phosphorylation to dNTPs by nucleoside diphosphate kinase and salvage of deoxyribonucleosides by two parallel sets of deoxynucleoside and -nucleotide kinases in the cytosol and in mitochondria. The rate-limiting step of the salvage pathway is the phosphorylation of deoxyribonucleosides to their monophosphates, catalyzed by thymidine kinase 1 (TK1) and deoxycytidine kinase (dCK) outside mitochondria and thymidine kinase 2 (TK2) and deoxyguanosine kinase (dGK) inside mitochondria. The substrate specificity of each pair of enzymes permits the phosphorylation of all physiological deoxyribonucleosides in both instances. Degradation of deoxyribonucleotides is carried out by 5′-nucleotidases that dephosphorylate nucleoside monophosphates to deoxyribonucleosides, inverting the kinase reactions.
Deoxyribonucleoside kinases and 5′-nucleotidases generate substrate cycles that regulate the exchange of deoxyribonucleosides across the plasma membrane, thus modulating the overall dNTP pool size. Intracellular deoxyribonucleosides are also degraded by phosphorylases and deaminases that actively participate in the regulation of dNTP pools, as demonstrated by the large pool imbalances and severe pathologies arising from their genetic deficiency. The catabolic enzymes mentioned so far are constitutively expressed but show tissue-specific variations.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
The dNTP triphosphohydrolase SAMHD1 is a new addition to the set of catabolic enzymes and has recently emerged as a main actor in the regulation of dNTP metabolism. The enzyme cleaves dNTPs to deoxyribonucleosides and triphosphate, reversing in a single step the entire salvage pathway, which may interfere with the de novo pathway as dNTPs are allosteric effectors of ribonucleotide reductase. The impact of SAMHD1 activity on dNTP pools is surprisingly strong. Down-regulation of the enzyme by siRNA silencing, ubiquitin-dependent proteolysis, or genetic inactivation increases the concentrations of all dNTPs in human fibroblasts, especially that of dGTP. This effect is particularly marked in quiescent or differentiated cells where the level of SAMHD1 is higher than during proliferation.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
The combination of synthetic and catabolic reactions provides a fine cyclic regulation of dNTP pools in tune with nuclear DNA replication. Pool sizes expand at the onset of S-phase when the request for dNTPs suddenly increases and diminish when DNA replication is completed and ribonucleotide reduction is turned off by degradation of the RNR small subunit R2. In the absence of SAMHD1 the cyclic variation of dNTP pools is lost, and the pools remain high also when the frequency of S-phase cells is low. In normal human fibroblasts silenced for SAMHD1 the accumulation of dNTPs outside S-phase interferes with the normal progression of the cell cycle, delaying cells in G1.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
SAMHD1 is a modular protein. The protein sequence contains an N-terminal nuclear localization signal followed by a sterile α-motif domain and a catalytic HD-domain typical of a large family of phosphohydrolases. A short C-terminal stretch is important for interactions with proteins regulating SAMHD1 stability and contains a conserved threonine residue (Thr-592) phosphorylated in proliferating cells by cyclin A-dependent cyclin-dependent kinases. The effects of Thr-592 phosphorylation on SAMHD1 function are incompletely understood. Crystallographic studies of the isolated HD domain demonstrate that SAMHD1 is a homotetrameric enzyme allosterically regulated by dNTPs, similar to RNR.
A SAMHD1 tetramer contains four catalytic sites specific for deoxyribonucleoside triphosphates and eight allosteric sites, four specific for dGTP or GTP and four binding all four dNTPs, with the lowest affinity for dCTP. The solved structures of SAMHD1 complexed with effectors and substrates illustrate the structural basis for both allosteric control and substrate specificity. Exonuclease activity has also been reported for SAMHD1, but the nature of the preferred substrate is controversial. A recent study showed that the triphosphohydrolase and nuclease activities of SAMHD1 are inactivated independently by genetic mutations occurring in vivo in patients affected by Aicardi Goutières syndrome. So far, nobody has shown how the solved three-dimensional structure of active SAMHD1 can accommodate both enzyme activities.[](https://www.ncbi.nlm.nih.gov/mesh/D013912)
Although DNA synthesis requires roughly equimolar amounts of the four dNTPs, these occur at unequal concentrations in cells. Pyrimidine dNTPs are generally more abundant than purine dNTPs, with dGTP as the smallest pool. The advantages of small dGTP pools may be rationalized. High dGTP concentrations are toxic because dGTP stimulates the reduction of ADP by RNR, and dATP is the main negative effector of the enzyme. Moreover, dGTP is highly mutagenic, is prone to oxidative damage, and in the oxidized form can be incorporated into DNA, reducing DNA polymerase γ replication fidelity with destabilizing consequences.
An exception to the advantages of small dGTP pools is provided by rat tissue mitochondria, reported to contain huge amounts of dGTP. The control of dGTP concentration so far has been considered to result from a balance between the efficiency of dGTP de novo synthesis and the catabolic action of 5′-nucleotidases cdN and cN-II, the two main intracellular 5′-nucleotidases degrading dGMP, and of purine nucleoside phosphorylase (PNP) that degrades the deoxyguanosine produced by the nucleotidases (Fig. 1). The large effects of SAMHD1 ablation on the concentration of dGTP in various types of mammalian cells show that an important piece of the puzzle was missing. SAMHD1 limits the size of the dGTP pool, especially in quiescent cells, which in culture seem to produce more dGTP than they need.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
Interaction between SAMHD1 and deoxyguanosine kinase activities in the regulation of mitochondrial dGTP. Synthesis of dGTP occurs in the cytosol by de novo synthesis catalyzed by RNR and by salvage of GdR catalyzed by dCK. The two synthetic pathways are counteracted by catabolic activities distributed between the nucleus and cytoplasm. SAMHD1 degrades dGTP to GdR in the nucleus, whereas in the cytosol 5′-nucleotidases (cdN and cN-II) dephosphorylate dGMP to GdR, and PNP degrades GdR to free guanine. Due to the permeability of the nuclear envelope to nucleotides and nucleosides, nuclear and cytosolic pools represent a single kinetic compartment. The mitochondrial dGTP pool derives most of its components from the cytosol via nucleotide and nucleoside carriers in the inner mt membrane. In mitochondria, a separate salvage pathway, depending on dGK, recycles GdR, creating a dGTP pool protected from the catabolic activity of SAMHD1.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
Cells contain two separate dGTP pools, a very small mitochondrial pool enclosed within the mt inner membrane, and a 10-fold larger extramitochondrial pool distributed between the cytosol and nucleus. The two pools communicate through nucleotide and nucleoside carriers in the mt membrane such that guanine deoxynucleotides produced de novo in the cytosol are the main source of mt dGTP both during proliferation and quiescence. However, the presence of a mitochondrial salvage pathway that converts deoxyguanosine (GdR) to dGTP suggests that the import of dGTP from the cytoplasm does not suffice for mtDNA synthesis. Indeed, genetic deficiency of dGK causes a mtDNA depletion syndrome with a specific hepato-cerebral phenotype. As is the case with other forms of mtDNA depletion syndrome, the enzyme deficiency targets differentiated cells that have small dNTP pools.
However, the tissue-specific effects typical of dGK deficiency remain unexplained. The mtDNA depletion in dGK-mutated individuals is commonly attributed to an insufficient supply of dGTP, an explanation based on limited direct experimental evidence published before the discovery of SAMHD1. Wishing to understand how SAMHD1 interacts with other enzymes in the regulation of dNTP pools, we set out to investigate its impact on the mt pools of cells with defective mt salvage of purine deoxyribonucleosides. We found that dGK deficiency reduces mt dGTP in human fibroblasts only moderately but causes mt dNTP imbalances and mtDNA depletion during quiescence. SAMHD1 down-regulation by siRNA transfection restores mt pool balance and boosts mtDNA copy number, compensating the mtDNA depletion to various degrees in different mutant lines.
Interestingly, the activity of SAMHD1 in the nucleus affects the mt dNTP pool by reducing the cellular dGTP content. We propose that inside mitochondria dGK is required to recycle the GdR released from extramitochondrial dGTP by SAMHD1.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
## Experimental Procedures
*In the **Experimental Procedures** section:*
## Materials, Cell Lines, and Cell Growth
*In the **Materials, Cell Lines, and Cell Growth** section:*
[5-3H]Deoxycytidine (9000 cpm/pmol) and [8-3H]deoxyguanosine (6000 cpm/pmol) were from Moravek (Brea, CA). Immucillin H was a gift of Dr. Vern Schramm (Yeshiva University, New York).[](https://www.ncbi.nlm.nih.gov/mesh/D003841)
We used three lines of skin fibroblasts derived from patients with inactivating mutations in DGUOK, the gene for deoxyguanosine kinase. Patient 1 carried the heterozygous mutations D255X and E165V, patient 2 (i.e. patient 14 in Table 3 of Ref.) was heterozygous for two different mutations, i.e. K201fs214X; IVS4 splicing site, and patient 3 carried the homozygous mutation L250S. Patient 1 fibroblasts were donated by Dr R. Martì (Hospital Vall d'Hebron, Barcelona, Spain). Fibroblasts from patients 2 and 3 were donated by Dr Massimo Zeviani from the collection of the Neurological Institute C. Besta, Milano (Italy).
Three lines of age-matched control skin fibroblasts were available in our laboratory. All fibroblasts were grown in DMEM with 4.5 g of glucose per liter, 10% (v/v) fetal calf serum, nonessential amino acids, and antibiotics.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
## Transfections with siRNAs, RNA Extraction, Reverse Transcription, and Real-time PCR
*In the **Transfections with siRNAs, RNA Extraction, Reverse Transcription, and Real-time PCR** section:*
All procedures were performed as described in Franzolin et al.. The siRNAs used were from Qiagen: AllStar negative control siRNA (catalog no.1027281), siRNA 3 as in Laguette et al., siRNA 4 (catalog no. SI00710500), and siRNA 7 (catalog no. SI04243673).
## Western Blotting
*In the **Western Blotting** section:*
Samples of 1–2 million cells were collected by centrifugation, washed with PBS, and lysed with radioimmunoprecipitation assay buffer (10 mm Tris·HCl, pH 7.4, 100 mm NaCl, 1% sodium deoxycholate, 0.1% SDS, 1% Nonidet P-40) containing a mixture of protease inhibitors for mammalian cells (Sigma). The extracts were centrifuged at 19,000 × g for 20 min, the protein concentrations of the supernatant solutions were determined by the BCA protein assay (Pierce), and appropriate amounts of the cleared supernatants were loaded on precast gels, 7.5% (Bio-Rad) for R1 and SAMHD1 separation or Any-Kd gels (Bio-Rad) for R2, p53R2, β actin, and GAPDH, and electrophoresed. The proteins were blotted on Hybond-C extra (GE Healthcare) in the case of ribonucleotide reductase subunits R1, R2, p53R2, β actin, and GAPDH and on PVDF (Millipore) for SAMHD1. Both membranes were saturated with 2% ECL Blocking Agent (GE Healthcare) for 1 h at room temperature and incubated overnight at 4 °C with the following primary antibodies: anti-SAMHD1 1A1 (Abcam), dilution 1:4000; anti-R1 T16 (Santa Cruz) 1:1000; anti-R2 N18 (Santa Cruz) 1:2000; anti-p53R2 N16 (Santa Cruz) 1:1000; anti-GAPDH (Millipore) 1:1000; anti-β actin (Sigma) 1:4000. After 3 washings with PBS + 0.05% Tween 20 for 10 min, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (dilution 1:80,000) for 1 h at room temperature.
Then the membranes were washed and developed with a chemiluminescence ECL kit (LiteAblotTurbo, Euroclone). The signals were detected on Kodak films and quantified with ImageJ software.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Quantitation of dNTP Pools
*In the **Quantitation of dNTP Pools** section:*
Cells were washed with ice-cold PBS, scraped from the plates, and homogenized by repeated aspiration through needles of 0.45 × 23 mm. Cytosolic and mt nucleotides were separated by differential centrifugation of cell homogenates, and nucleotide pools were extracted with ice-cold methanol as described. The concentrations of the four dNTPs were determined by an enzymatic assay modified as described in Ferraro et al.. In some experiments, before extracting the nucleotide pools the quiescent cultures were treated for 24 h with 0.5 μm immucillin H to inhibit purine nucleoside phosphorylase.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Enzyme Assays
*In the **Enzyme Assays** section:*
Whole cell extracts were prepared as described earlier, adding a mixture of mammalian protease inhibitors (Sigma) to the lysis buffer (Tris·HCl, pH 7.5, 10 mm 0.5% Triton X, 2 mm EDTA, 1 mm DTT). Protein concentration was measured by the colorimetric procedure of Bradford with bovine serum albumin as the standard. All enzymatic assays were run with two different aliquots of extract to check for proportionality. Citrate synthase was measured as detailed in Shepherd and Garland. Deoxyguanosine and deoxycytidine kinases were tested by incubating the extracts with the radioactive precursor for 60 min in 40 μl of reaction mix and then spotting a 35-μl aliquot on DE 81 paper as described in Leanza et al..
The specific radioactivity of the labeled substrates was diluted 5–10-fold with non-labeled deoxynucleosides. The activity of dGK was measured with 1 μm [3H]deoxyguanosine (1800 cpm/pmol) in the presence of 0.5 mm nonradioactive CdR to inhibit deoxycytidine kinase and 0.5 μm immucillin H to prevent the degradation of deoxyguanosine. The activity of deoxycytidine kinase was determined with 1 μm [3H]CdR (1100 cpm/pmol). We express enzyme activities as pmol of product min−1 mg−1 of protein extract.[](https://www.ncbi.nlm.nih.gov/mesh/D006851)
## mtDNA Quantification
*In the **mtDNA Quantification** section:*
We determined mtDNA copy numbers with the TaqMan probe system and Applied Biosystem 7500 real-time PCR as described in Andreu et al.. Genomic DNA was extracted by Puregene Core kit B (Qiagen). Mitochondrial rRNA 12S TaqMan probe 6FAM-5′-TGCCAGCCACCGCG-3′-MGB (Applied Biosystems) and primers rRNA 12S forward (5′-CCACGGGAAACAGCAGTGATT-3′) and reverse (5′-CTATTGACTTGGGTTAATCGTGTGA-3′) were used to quantify mtDNA. For nuclear DNA, we used RNase P primers and probe VIC mix (Applied Biosystems). To quantify mtDNA and nuclear DNA we used calibration curves generated by serial dilution of a mixture of plasmids carrying the two PCR amplicons. Each DNA sample was analyzed in triplicate.
## Flow-cytometric Determinations of Cell-cycle Distribution and Mitochondrial Mass
*In the **Flow-cytometric Determinations of Cell-cycle Distribution and Mitochondrial Mass** section:*
Cell cycle distribution in fibroblast cultures was determined by flow cytometry after propidium iodide staining of fixed resuspended cells, with a BD Biosciences FACSCanto II Flow cytometer. To measure mitochondrial mass 0.8 × 106 cells were detached, washed once, and incubated under rotation in 400 μl of 0.2 μm nonyl acridine orange (Sigma) at 37 °C for 10 min. Then cells were washed once in 10 ml of PBS and centrifuged. The pellet was resuspended in 700 μl of PBS and analyzed immediately by flow cytometry.[](https://www.ncbi.nlm.nih.gov/mesh/D011419)
## Statistics
*In the **Statistics** section:*
Where appropriate, statistical analyses (two-way analysis of variance with Bonferroni post test) were performed using GraphPad Prism 5 (GraphPad Software, La Jolla, CA).
## Results
*In the **Results** section:*
The concentration of dGTP in human cells is strongly influenced by the activity of SAMHD1, particularly in noncycling cells that contain higher amounts of the protein than cycling cells. When the enzyme is down-regulated, all four dNTP pools expand, with dGTP undergoing the largest increase. The deoxyguanosine released from dGTP by SAMHD1 can be recycled through the salvage pathway via phosphorylation by dGK. To investigate possible functional interactions between the two enzymes, we employed skin fibroblasts derived from three unrelated mtDNA depletion syndrome patients with different inactivating mutations in DGUOK, the gene coding for deoxyguanosine kinase, and compared them with WT human fibroblasts.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
## Characterization of the dGK-mutated Fibroblast Lines
*In the **Characterization of the dGK-mutated Fibroblast Lines** section:*
The two kinases that phosphorylate GdR in the cells, dCK and dGK, undergo opposite variations when normal skin fibroblasts pass from proliferation to quiescence, with dCK activity declining and dGK activity increasing. We measured the activities of the two enzymes with their preferred substrates, CdR and GdR, respectively, in whole cell extracts from proliferating and quiescent WT and dGK-mutated fibroblasts (Table 1). In the dGK assay the addition of nonradioactive CdR inhibited the phosphorylation of GdR by dCK present in the extracts. The activity of dGK was one order of magnitude lower in the mutant fibroblasts than in the WT controls and hardly changed during quiescence, in contrast to that of the controls. The activity of dCK was between 15 and 20 pmol/min/mg protein in all cell lines after 48 h in culture and decreased with time, reaching after 10 days of quiescence values 3-fold lower in the controls and ∼2-fold lower in the mutants.[](https://www.ncbi.nlm.nih.gov/mesh/D003849)
Enzyme activities of deoxycytidine kinase and deoxyguanosine kinase in whole cell extracts from proliferating or quiescent control and dGK-mutated skin fibroblasts
Deoxycytidine kinase was tested with 1 μm [5-3H]CdR and deoxyguanosine kinase with 1 μm [8-3H]GdR. Values are specific activities calculated in duplicate samples from two experiments ± S.E. C, WT controls; P, dGK-mutated patient lines.[](https://www.ncbi.nlm.nih.gov/mesh/D003841)
During proliferation the P2 mutant line showed a 30–40% depletion of mtDNA compared with the controls, whereas in the other two mutants the content of mtDNA was similar to that of WT fibroblasts (Fig. 2A). After 10 days of quiescence all three mutant lines contained significantly less mtDNA than the WT fibroblasts, in agreement with previous reports of mtDNA depletion in cultured dGK-deficient fibroblasts. The low mtDNA content was not accompanied by a reduced number of mitochondria in the mutants. We determined mt mass by flow cytometry in quiescent cells stained with nonyl acridine orange and by measuring the activity of citrate synthase in cell extracts. Both methods showed in mutant cells values of mt mass within the range observed in control fibroblasts (Fig. 2B).
The depletion of mtDNA during quiescence may depend on an insufficient supply of dGTP in mitochondria. We measured the concentrations of the four dNTPs in the cytosolic and mitochondrial pools of quiescent WT and mutant fibroblasts (<5% S-phase cells). Fig. 3 reports the pool sizes of the individual lines, with the dCTP pool representing in all instances the largest pool and dGTP representing the smallest. The size of each cytosolic dNTP pool varied among the different lines, but on average dGTP corresponded to 5.7 ± 0.3% (mean ± S.E.) of the total cytosolic dNTPs in WT fibroblasts and to 5.1 ± 0.5% in the mutants.
The single recurrent difference between WT and mutant cells was a somewhat higher content of dTTP in the latter. We do not have at present an explanation for this difference (Fig. 3A). Quantitative variations of each dNTP were present also in the mt pools both within the controls and the mutant lines (Fig. 3B). The mutant cells contained less mt dGTP than the WT cells.
The fraction of the mt pools corresponding to dGTP was 17.6 ± 2.4% in WT and 10.8 ± 1.4% in mutant cells. To evaluate how this difference reflected on the pool composition of each line, we normalized the individual pool sizes by that of the dGTP pool (Fig. 3, C and D). The composition of cytosolic pools did not differ statistically in mutant and WT fibroblasts (Fig. 3C), but this procedure revealed that the mitochondria of mutant cells contained an excess of dCTP and that dTTP and dATP tended to be more abundant relative to dGTP than in WT cells, although not significantly (Fig. 3D). Thus dGK deficiency, possibly changing the dynamics of the mt dGTP pool through a reduction of GdR phosphorylation, affected mt dNTP pool balance.
The combination of lower dGTP concentration and mt pool imbalance might account for the mtDNA depletion. Much larger DNA precursor asymmetries observed in rat tissue mitochondria were shown to reduce the replication fidelity of human mt DNA polymerase when tested in vitro. Here, in cultured human fibroblasts, a milder mt pool imbalance was associated with mtDNA depletion.[](https://www.ncbi.nlm.nih.gov/mesh/C041691)
mtDNA copy number and mitochondrial mass in WT and dGK-mutated fibroblasts.
A, we measured mtDNA copy number per nuclear genome in proliferating (48 and 72 h) and quiescent cultures of three lines of wt skin fibroblasts (C1, C2, C3) and three dGK-mutated lines (P1, P2, P3). Control values were combined (wt). All data are the means ± S.E. from at least two experiments for each cell line analyzed in triplicate. B, mitochondrial mass was evaluated by measuring citrate synthase activity (units (U) = nmol/min) in whole cell extracts and by normalizing the nonyl acridine orange (NAO) fluorescence of each line determined by flow cytometry by that of control C1. Data are the means ± S.E. from two experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C041691)
Cytosolic and mitochondrial dNTP pool sizes in quiescent skin fibroblasts. Cytosolic (A) and mitochondrial (B) dNTP pools were measured in quiescent cultures of two WT (C1, C2) and three dGK-mutated (P1, P2, P3) lines of skin fibroblasts. To compare the two groups we calculated the mean ratios between the size of each dNTP pool and that of dGTP in the cytosol (C) and mitochondria (D) of WT or dGK-mutated fibroblasts. Data are the means ± S.E. from two experiments for each cell line analyzed in duplicate. Asterisks indicate a significant difference (***, p < 0.001) of the dCTP/dGTP ratio in mutant fibroblasts relative to controls.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
We then compared dGK mutant and control fibroblasts for their expression of ribonucleotide reductase and SAMHD1 during quiescence in the absence of any experimental treatment. In protein extracts from three lines of WT fibroblasts and the three dGK mutant lines, we measured by Western blotting the expression of the three subunits of ribonucleotide reductase and of SAMHD1 (Fig. 4). The RNR small subunit p53R2, which increases in quiescent cells, was expressed in all lines, whereas the R2 subunit, which is S-phase-specific, was undetectable in all quiescent extracts with the exception of the P1 sample that gave a faint signal. The large subunit R1 produced much weaker bands in extracts from quiescent than from proliferating cells, and again the P1 extracts contained more enzyme than the other lines. All lines expressed SAMHD1, with similar variability in WT and mutant cells.
The lowest expression was detected in C3 and P2 fibroblasts. The P1 fibroblasts had medium-high levels of SAMHD1 and higher expression of RNR subunits R2 and R1 than the other two dGK-mutated lines, a combination that may explain their higher mtDNA content (Fig. 2). We decided to test if lowering SAMHD1 activity in the cells with lower ribonucleotide reduction could increase their mtDNA.[](https://www.ncbi.nlm.nih.gov/mesh/D012265)
Expression of ribonucleotide reductase subunits and SAMHD1 in untreated quiescent skin fibroblasts. Ribonucleotide reductase subunits (R2, p53R2, and R1) and SAMHD1 were detected by immunoblotting in extracts from one proliferating WT control culture (C1p) and quiescent cultures of three WT (C1, C2, and C3) and three dGK-mutated (P1, P2, and P3) lines of skin fibroblasts. To compare R2 or R1 in proliferating and quiescent extracts, different amounts of proteins were loaded in the respective lanes of the gels: 4 and 40 μg for R2 and 20 and 40 μg for R1. In the case of p53R2 we used 2 μg of quiescent extracts and 20 μg for SAMHD1. Loading controls for R1, R2, and SAMHD1 were two unspecific bands (asterisks), whereas GAPDH was used as the control for p53R2.
## Silencing of SAMHD1 in dGK-mutated Fibroblasts Prevents mtDNA Depletion during Quiescence
*In the **Silencing of SAMHD1 in dGK-mutated Fibroblasts Prevents mtDNA Depletion during Quiescence** section:*
Confluent cultures of WT and dGK-mutated fibroblasts were transfected with anti-SAMHD1 siRNA or control siRNA in medium with 0.1% serum and kept in the presence of siRNAs for 10 days, with 2 changes of medium. In a preliminary experiment we tested in one mutant (P1) and two WT lines three different anti-SAMHD1 siRNAs for the effects on SAMHD1 mRNA and mtDNA copy number. All three siRNAs silenced SAMHD1 to the same extent, with a residual mRNA level below 10% of the control (not shown). The amount of mtDNA remained the same in the WT fibroblasts when SAMHD1 was silenced (Table 2). On the contrary, in P1 fibroblasts mtDNA increased almost 3-fold to the level of the controls, suggesting that the down-regulation of dGTP catabolism by SAMHD1 prevented the depletion of mtDNA during quiescence.
We also repeated the experiment with the other two mutant lines using siRNA 7 that in all instances decreased the level of SAMHD1 mRNA to ≤5% that of the control. Fig. 5 shows mtDNA copy numbers at the time of confluence, when siRNA transfections started, and after 10 days of transfection with control or anti-SAMHD1 siRNAs. In dGK-mutated fibroblasts transfected with control siRNA, the depletion of mtDNA, already present at confluence, became more marked after 10 days of quiescence. In contrast, SAMHD1 down-regulation led to a recovery of mtDNA that was complete in the case of P1 cells and only partial in the other two lines.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
mtDNA copy number in quiescent control and dGK-mutated skin fibroblasts transfected with one nontargeting siRNA or three different anti-SAMHD1 siRNAs
mtDNA content was measured by quantitative real-time PCR and normalized per nuclear genome. Values are the means of three technical replicates for each condition ± S.E. from six experiments for control siRNA and siRNA 7 and from two experiments for siRNA 3 and siRNA 4. C1 and C2, controls; P1, dGK-mutant fibroblasts. ND, not done.
Effect of SAMHD1 silencing on mtDNA content of quiescent skin fibroblasts. mtDNA copy number normalized per nuclear genome was determined in WT and dGK-mutated fibroblasts at confluence, i.e. before starting the transfections with siRNAs in low serum medium and after 10 days of transfection during quiescence with control siRNA or anti-SAMHD1 siRNA. The values of three WT lines (C1, C2, and C3) were averaged, whereas those of the patient lines (P1, P2, P3) are reported separately. The percent residual level of SAMHD1 mRNA in the silenced cultures is indicated. Data are the means ± S.E. from three to six experiments for each cell line analyzed in triplicate.
Earlier we showed that the expansion of the dNTP pools that results from silencing of SAMHD1 in quiescent fibroblasts depends on the activity of p53R2-dependent ribonucleotide reductase. Here the level of p53R2 was similar in the three mutant lines; thus, the lower level of mtDNA in P2 and P3 fibroblasts may depend on the lower content of R1 compared with P1 fibroblasts that in addition retained traces of R2 (Fig. 4).[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
To understand how the boosting effect on mtDNA copy number developed during SAMHD1 silencing, we followed for 10 days of transfection with siRNA the variations of mtDNA in one line of WT fibroblasts and in the dGK-mutant fibroblasts with the maximum (P1) and minimum (P2) recovery of mtDNA. Transfections with control or anti-SAMHD1 siRNAs started at day 0 when the cultures had reached confluence and continued for 10 days during which we measured the levels of SAMHD1 protein (Fig. 6A) and mtDNA (Fig. 6C), every second day starting from day 4. Between days 4 and 10 of transfection we controlled the degree of silencing by quantitating SAMHD1 mRNA (Fig. 6B). In the cultures transfected with control siRNA, SAMHD1 increased with time as expected, more in WT and P1 cells than in P2 fibroblasts, which as observed above in Fig.
4 had the lowest endogenous expression of SAMHD1. In the cultures transfected with the anti-SAMHD1 siRNA, the protein became undetectable after day 4, reflecting the high degree of mRNA silencing (Fig. 6B). In agreement with the data in Table 2, the WT fibroblasts maintained their content of mtDNA constant during the whole period independently of the presence and concentration of SAMHD1. In the mutant fibroblasts the depletion was already present at day 0 and during transfection with nontargeting siRNA mtDNA decreased further, more markedly in P1 fibroblasts that started from a higher copy number. In P1 cells the down-regulation of SAMHD1 lead to an actual increase of mtDNA, whereas in P2 fibroblasts it only prevented the additional decline of mtDNA, possibly because these cells originally contained less SAMHD1.
A similar stabilization of mtDNA content was obtained previously in a different line of dGK-mutated fibroblasts incubated with purine deoxynucleoside monophosphates.[](https://www.ncbi.nlm.nih.gov/mesh/D011687)
Quantitative variations of mtDNA during siRNA silencing of SAMHD1 in quiescent WT and dGK-mutated fibroblasts.
A, expression of SAMHD1 protein during transfection with control or anti-SAMHD1 siRNA in quiescent cultures of WT (C2) or dGK-mutated (P1, P2) fibroblasts. The arrow indicates the SAMHD1 signal. An unspecific band appeared just above SAMHD1 and remained stable during silencing. B, residual level of SAMHD1 mRNA in the silenced cultures relative to that in the cultures transfected with control siRNA was determined at the indicated days of transfection during quiescence. Red, C2; green, P1; blue, P2. C, mtDNA content per nuclear genome was measured at 1- or 2-day intervals in cultures treated with control siRNA (continuous line) or anti-SAMHD1 siRNA (broken line) during 10 days of quiescence. Color coding is in B.
## Silencing of SAMHD1 Corrects the Imbalance of mt dNTP Pools in dGK-mutated Fibroblasts
*In the **Silencing of SAMHD1 Corrects the Imbalance of mt dNTP Pools in dGK-mutated Fibroblasts** section:*
The improvement of mtDNA copy number in the mutant fibroblasts silenced for SAMHD1 suggests that their mt dNTP pools might be normalized as a consequence of reduced extramitochondrial catabolism of dNTPs. However, we previously found that SAMHD1 down-regulation in WT quiescent fibroblasts leads to larger and imbalanced dNTP pools, and Fig. 3 shows that the mt pools of dGK-mutated fibroblasts are imbalanced. Thus it was important to assess the effects of SAMHD1 down-regulation on these pools. In WT fibroblasts SAMHD1 silencing induced the expected changes in the cytosolic pools, with a larger expansion of purine dNTPs, a doubling of dTTP content, and only a minor increase of dCTP (Fig.
7A). mt pools were minimally affected (Fig. 7B). The average percentage of dGTP relative to the total dNTPs increased in WT cells from 7.4 ± 0.9 to 13.5 ± 0.5 in the cytosol, remaining unchanged in the mitochondria (24 ± 4) (Fig. 7C). The three mutant lines reacted differently to SAMHD1 silencing.
Whereas the pool changes of P2 and P3 fibroblasts were close to those of the WT cells both in the cytosol and in mitochondria (Fig. 7, A and B), P1 cells expanded their purine dNTP pools to a much larger extent, in particular the cytosolic dGTP pool (Fig. 7A). The higher expression of ribonucleotide reductase in these cells during quiescence (Fig. 4) led to larger pools and complete recovery of mtDNA content (Figs. 5 and 6). Nevertheless, in all three mutant lines SAMHD1 down-regulation increased the percentage of dGTP relative to the total dNTPs not only in the cytosol, as in WT cells, but also in mitochondria (Fig. 7C). Despite the different quantitative alterations, in the mutant lines the composition of mt pools was normalized by SAMHD1 silencing, becoming similar to that of the WT cells.
We report in Fig. 7, D and E, the cytosolic and mt pool composition relative to the content of dGTP for each cell line transfected with nontargeting or anti-SAMHD1 siRNA, calculated as in Fig. 3. In the cytosol the increase of dGTP content reduced the disequilibrium with the other three dNTPs similarly in WT and mutant fibroblasts (Fig. 7D). In mitochondria the large difference between mutant and WT cells disappeared, with a clear normalization of the pools in mutant cells (Fig. 7E).
The correction of mt pool imbalance coincided with the observed improvement of mtDNA copy number.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
Effects of SAMHD1 silencing on cytosolic and mitochondrial dNTP pools in quiescent WT and dGK-mutated fibroblasts. Shown is the increase of cytosolic (A) and mitochondrial (B) dNTP pools induced by SAMHD1 silencing in WT and dGK-mutated skin fibroblasts incubated during 10 days of quiescence with anti-SAMHD1 siRNA or control siRNA. The increases measured in three WT lines were averaged as well as those of the P2 and P3 mutant lines. Values of the P1 mutant are reported separately. C, percentage of dGTP relative to the total dNTPs in the cytosolic and mt pools in WT and dGK-mutated fibroblasts transfected with control siRNA (black) or with anti-SAMHD1 siRNA (gray). D and E, composition of cytosolic (D) and mitochondrial (E) dNTP pools in cultures transfected with control siRNA or with anti-SAMHD1 siRNA, expressed as ratios between the sizes of the individual dNTP pools and the corresponding dGTP pool. The ratios of WT cells were averaged. Mutant P1 values are reported separately from those of the other two mutants (P2/P3). Data are the means ± S.E. from two experiments for each cell line analyzed in duplicate. Asterisks indicate that in nonsilenced cultures significant differences exist in the dTTP/dGTP and dCTP/dGTP ratios of dGK-mutated fibroblasts (P1, P2/P3) relative to controls (wt). **, p < 0.01; *, p < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
## Inhibition of Purine Nucleoside Phosphorylase Does Not Change mtDNA Content in dGK-mutated or WT Fibroblasts
*In the **Inhibition of Purine Nucleoside Phosphorylase Does Not Change mtDNA Content in dGK-mutated or WT Fibroblasts** section:*
The GdR derived from dGTP hydrolysis by SAMHD1 can be further degraded by PNP and thus subtracted from recycling via dGK-mediated salvage (Fig. 1). We performed experiments with quiescent WT and dGK-mutated fibroblasts in which during the last 24 h of transfection with control or anti-SAMHD1 siRNA we added to the medium 0.5 μm immucillin H, a potent chemical inhibitor of PNP. The prediction was that if PNP and dGK compete for their common substrate GdR, PNP inactivation may favor GdR incorporation into the mt dGTP pool of dGK-proficient cells but not of the dGK mutants. However, the treatment with immucillin H produced only minimal increases of cytosolic dGTP in the WT fibrobasts and had no effect on mt dGTP independently of SAMHD1 knockdown (data not shown). These negative results indicate that in cultured fibroblasts PNP interferes negligibly with salvage of endogenous GdR by dGK, and its inactivation does not modify the loss of dGTP due to SAMHD1 activity.[](https://www.ncbi.nlm.nih.gov/mesh/D003849)
## Discussion
*In the **Discussion** section:*
Genetic deficiency for either of the two mt deoxynucleoside kinases, TK2 and dGK, is associated with mtDNA depletion and severe mt diseases. The phenotypes differ for the two enzymes; TK2 mutations primarily affect skeletal muscle, and dGK mutations affect liver and brain. As the genetic defects are in each case present in all somatic cells, their specific phenotypic expression is ascribed either to specific energetic requests of the affected tissues that would require fully functional mitochondria or to variable expressions of the affected protein and other competing or cooperating enzymes. One example is the particularly low expression of TK2 in skeletal muscle, which makes this tissue specifically sensitive to any further decline of TK2 activity, but in most cases the reason of the tissue-specific phenotype remains hypothetical. The expression of dGK, for instance, is comparable in liver and brain and in other tissues, which suggested that the specificity of the dGK− phenotype might depend on differential expression of other enzymes functionally interacting with dGK.
TK2 and dGK deficiencies affect non-dividing cells where de novo synthesis of dNTPs is strongly down-regulated and depends on the R1-p53R2 form of ribonucleotide reductase. In cultured human fibroblasts p53R2-dependent ribonucleotide reduction occurs at a roughly 40-fold lower rate than R2-dependent reduction during S-phase, yet the produced dNTPs largely exceed the needs of the cells during quiescence. In fact ≈90% of these dNTPs are dephosphorylated and the resulting deoxynucleosides are released into the medium. Thus in noncycling WT cells de novo synthesis of dNTPs per se does not appear to be insufficient for the maintenance of mtDNA. Nevertheless, the existence of the two mitochondrial deoxynucleoside kinases demonstrates that deoxynucleoside salvage is needed to support mtDNA synthesis under such conditions, but why are two distinct kinases required, and why inside mitochondria?
At the end of S-phase TK1 is degraded, thus the need for a second thymidine kinase is obvious, although it may well reside outside mitochondria. In contrast to TK1, dCK is expressed with some variation in every phase of a cell's life and in all tissues, and it can phosphorylate both deoxyguanosine and deoxyadenosine besides deoxycytidine, its preferred substrate. Yet the cells also produce dGK, a purine-specific kinase enclosed in the mt matrix essential for the maintenance of mt dNTP pool balance during quiescence. In fact, differently from cultured TK2-deficient fibroblasts that do not manifest any mtDNA defect, dGK-deficient fibroblasts readily develop imbalance of mt dNTPs (Fig. 3, see Ref.) and mtDNA depletion in quiescent cultures (Fig. 2A; Refs. and).
Why are noncycling cells so dependent on GdR salvage inside mitochondria? Our present data suggest that the answer may be SAMHD1, whose concentration increases in non-dividing cells. SAMHD1 is a nuclear enzyme, but its influence spreads to the overall cellular dNTP pool due to the permeability of the nuclear envelope to deoxynucleotides. Indeed, the anti-HIV-1 effect of SAMHD1 was attributed to the restriction of cytosolic dNTP pools that fall below the km of the viral reverse transcriptase. By degradation of all dNTPs in cells with low RNR activity, SAMHD1 restricts the source of deoxynucleotides for mtDNA synthesis.
We demonstrated earlier that mt dNTPs largely derive from the cytosolic pool. Two main dNTP transporters are known in mammalian cells, SLC25A33 and SLC25A36, both active with pyrimidine (d)NTPs, the latter also with (d)GTP. The deoxynucleosides produced by SAMHD1 enter mitochondria via nucleoside transporters and can be recycled by TK2 and dGK. In the mt matrix the resulting deoxynucleotides are sheltered from SAMHD1 and can be used for mtDNA replication or re-exported to the cytosol. Thus intramitochondrial salvage may be seen as a cell-autonomous mechanism (primarily) using deoxynucleosides derived from dNTPs synthesized de novo in the cytosol and hydrolyzed by SAMHD1 in the nucleus.
By this means quiescent or differentiated cells would protect themselves from viral infections and at the same time secure balanced amounts of precursors for mtDNA maintenance. For the sake of the present argument we do not consider here the lower amounts of deoxynucleosides produced from nucleoside monophosphates by 5′-nucleotidases. Phosphorylation of deoxynucleosides derived from the extracellular compartment might be a secondary function of mt deoxynucleoside kinases.[](https://www.ncbi.nlm.nih.gov/mesh/D003854)
SAMHD1 is especially effective in restricting the size of the dGTP pool. When SAMHD1 is inactivated dGTP undergoes the highest relative increase compared with the other three dNTPs, implying that recycling of GdR is especially important to provide mitochondria with sufficient dGTP. The low concentration of dGTP in the cytosol of quiescent fibroblasts may reduce the rate of carrier-mediated import of dGTP into mitochondria so that recycling of GdR becomes limiting to sustain a “normal” mt pool.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
With the present experiments we show that SAMHD1 contributes to the phenotype of dGK deficiency in fibroblasts. In vivo the main target cells of this genetic condition are hepatocytes and neurons, where the relative abundance of enzymes in the network outlined in Fig. 1 possibly differs from that in fibroblasts, our experimental model. The expression of cytosolic dCK, for instance, is particularly low in liver and brain, increasing the dependence of the mt dGTP pool on intramitochondrial salvage of GdR. The tissue specificity of dGK deficiency in vivo may be related, at least in part, to SAMHD1 expression.
When the gene coding for the protein now known as SAMHD1 was first described, it was reported to be highly expressed in liver on the basis of multi-organ Northern blotting analysis. However, a recent study of the protein expression profile in humans demonstrated that the intensity of the SAMHD1 signal in an organ mostly depends on the frequency of cells of hematopoietic origin, which express the highest level of SAMHD1, whereas liver cells contain little SAMHD1. On the other hand, SAMHD1 is only one enzyme in the regulatory network, and its influence on the final concentration of dGTP depends also on the relative activities of other enzymes. The importance of ribonucleotide reductase in this respect is highlighted here by the biochemical phenotype of the P1 fibroblasts, which had higher RNR expression than the other two mutant lines (Fig. 4) and could better recover their mt pool balance and mtDNA upon SAMHD1 knockdown.[](https://www.ncbi.nlm.nih.gov/mesh/C029603)
By silencing SAMHD1 during quiescence we managed to abolish or contain the mtDNA depletion caused by dysfunction of dGK. Previous studies have demonstrated the possibility of improving mtDNA copy number in cultured dGK-mutated cells by supplying high concentrations (≥50–100 μm) of purine deoxynucleosides (GdR ± deoxyadenosine) or their monophosphates in the medium of quiescent fibroblast cultures (Ref. and references therein). The two protocols did not substantially differ, as nucleoside monophosphates are dephosphorylated by ectonucleotidases, and only the resulting deoxynucleosides are able to enter the cells. In dGK-deficient cells GdR (and deoxyadenosine) can only be phosphorylated in the cytosol by dCK. However, only at high concentrations did GdR compete with intracellular CdR for phosphorylation by dCK, whose Km for CdR is at least one order of magnitude lower than that for GdR. High concentrations of purine deoxynucleosides are also needed in these experiments to saturate the catabolic activity of adenosine deaminase and purine nucleoside phosphorylase.
As an alternative, the two enzymes can be blocked by specific inhibitors. Here we tried to favor recycling of endogenous GdR produced by SAMHD1 by treating cells with immucillin H, but we did not observe any reproducible change in dGTP concentration. Our conclusion is that in cultured skin fibroblasts the enzyme does not have a measurable influence on the regulation of endogenous dGTP and that the degradation of exogenous GdR observed earlier by us and others is carried out mostly by the PNP present in the culture medium. With the present experiments we have updated the picture of the regulation of mt dGTP in human fibroblasts revealing a functional interaction between SAMHD1 and dGK that changes our perspective on the role of mitochondrial salvage of deoxynucleosides.[](https://www.ncbi.nlm.nih.gov/mesh/D011687)
## Author Contributions
*In the **Author Contributions** section:*
V. B. and C. R. conceived and coordinated the study and wrote the paper. E. F. and C. R. designed and performed the experiments. E. F. analyzed all the experiments and prepared the figures. C. S. contributed to the cytofluorometric analysis. E. F., V. B., and C. R. critically revised the manuscript for important intellectual content.
All authors analyzed the results and approved the final version of the manuscript.
This work was supported by Fondazione Telethon Grant GGP14005 (to V. B.) and Associazione Italiana per la Ricerca sul Cancro Grant IG 2014 15818 (to V. B.). The authors declare that they have no conflicts of interest with the contents of this article.
RNR
ribonucleotide reductase
SAMHD1
sterile α motif/histidine-aspartate domain-containing protein 1
dCK
deoxycytidine kinase
TK1
cytosolic thymidine kinase
TK2
mitochondrial thymidine kinase
dGK
deoxyguanosine kinase
PNP
purine nucleoside phosphorylase
CdR[](https://www.ncbi.nlm.nih.gov/mesh/D003841)
deoxycytidine[](https://www.ncbi.nlm.nih.gov/mesh/D003841)
GdR[](https://www.ncbi.nlm.nih.gov/mesh/D003849)
deoxyguanosine[](https://www.ncbi.nlm.nih.gov/mesh/D003849)
p53R2
p53-inducible R2 subunit of ribonucleotide reductase
R2
ribonucleotide reductase subunit 2
R1
ribonucleotide reductase subunit 1
mt
mitochondria.
The abbreviations used are:
# References
*In the **References** section:*
|
# Introduction
Global [N](https://www.ncbi.nlm.nih.gov/mesh/D009584)-linked Glycosylation is Not Significantly Impaired in Myoblasts in Congenital Myasthenic Syndromes Caused by Defective Glutamine-Fructose-6-Phosphate Transaminase 1 (GFPT1)
# Abstract
*In the **Abstract** section:*
Glutamine-fructose-6-phosphate transaminase 1 (GFPT1) is the first enzyme of the hexosamine biosynthetic pathway. It transfers an amino group from glutamine to fructose-6-phosphate to yield glucosamine-6-phosphate, thus providing the precursor for uridine diphosp[hate N-ace](https://www.ncbi.nlm.nih.gov/mesh/D006595)tylglucosamine (UDP-GlcNAc) synthesis. UDP-GlcNAc is an [essential](https://www.ncbi.nlm.nih.gov/mesh/D005973) sub[strate for all mamma](https://www.ncbi.nlm.nih.gov/mesh/C027618)lian glyco[sylation biosynthetic p](https://www.ncbi.nlm.nih.gov/mesh/C001293)athways and N-glycan branching is e[specially sensitive to alterations in t](https://www.ncbi.nlm.nih.gov/mesh/D014537)he[ concentra](https://www.ncbi.nlm.nih.gov/mesh/D014537)tion of this [sugar nucl](https://www.ncbi.nlm.nih.gov/mesh/D014537)eotide. It has been reported that GFPT1 mutations lead to a distinct sub-class of con[genital ](https://www.ncbi.nlm.nih.gov/mesh/D011134)myasthenic syndromes (CMS) termed “limb-girdle CMS with tubular aggregates”. CM[S are hereditary](https://www.ncbi.nlm.nih.gov/mesh/D000073893) neuromuscular transmission disorders in which neuromuscular junctions are impaired. To investigate whether alterations in protein glycosylation at the neuromuscular junction might be involved in this impairment, we have employed mass spectrometric strategies to study the N-glycomes of myoblasts and myotubes derived from two healthy controls, three GFPT1 patients, and four patients with other muscular diseases, namely CMS caused by mutations in DOK7, [m](https://www.ncbi.nlm.nih.gov/mesh/D009584)yopathy caused by mutations in MTND5, limb girdle muscular dystrophy type 2A (LGMD2A), and Pompe disease. A comparison of the relative abundances of bi-, tri-, and tetra-antennary N-glycans in each of the cell preparations revealed that all samples exhibited broadly similar levels of branching. Moreover, although some differences were observed in the relative [abundance](https://www.ncbi.nlm.nih.gov/mesh/D011134)s of some of the N-glycan constituents, these variations were modest and were not confined to the GFPT1 samples. Therefore, GFPT1 mutations in CMS patients do not appear to compromise global N-glycos[ylation ](https://www.ncbi.nlm.nih.gov/mesh/D011134)in muscle cells.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
## 1. Introduction
*In the **1. Introduction** section:*
Protein glycosylation, the attachment of carbohydrate chains to proteins, is a common post-translational modification occurring ubiquitously in eukaryotic cells. Intriguingly, mutations in genes encoding protein glycosylation enzymes, or enzymes synthesising the building blocks for protein glycosylation, have recently been identified to cause neuromuscular transmission defects called congenital myasthenic syndromes (CMS). The first of these genes to be correlated with CMS was GFPT1 (glutamine-fructose-6-phosphate transaminase 1). Mutations in GFPT1 cause a distinct sub-class of CMS referred to as “limb-girdle CMS with tubular aggregates” [1,2].
Subsequently, mutations in three genes encoding enzymes of the protein N-glycosylation pathway (DPAGT1, ALG2 and ALG14) were also found to cause limb-girdle CMS [3,4]. The GFPT1 gene encodes a homodimeric, cytoplasmic enzyme that catalyses the first step of the hexosamine biosynthetic pathway [1]. Thus GFPT1 converts fructose-6-phosphate and glutamine into glucosamine-6-phosphate and glutamate; the end product of this pathway is uridine diphospho-N-acetylglucosamine (UDP-GlcNAc), which is a basic substrate not only for protein N- and O-linked glycosylation, but also for lipid glycosylation, proteoglycan synthesis, and O-GlcNAc glycosylation [1]. Currently it is not clear which of these pathways is affected most by lower levels of UDP-GlcNAc caused by GFPT1 deficiency.
However, clinical presentation of CMS patients with mutations in DPAGT1, ALG2 and ALG14 is very similar to GFPT1 CMS, suggesting that impaired protein N-glycosylation might be the molecular defect underlying GFPT1 CMS. N-glycosylation is crucial for the assembly and function of a number of key molecules at the neuromuscular junction, like the acetylcholine receptor, the muscle-specific kinase (MuSK), and agrin. However, it is currently not clear why muscles, or more specifically the neuromuscular junctions, are more vulnerable than other tissue[s to GFPT1 d](https://www.ncbi.nlm.nih.gov/mesh/D002241)eficiency.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
We set out to determine whether protein N-glycosylation is impaired or modified in the muscles of CMS patients with GFPT1 mutations. Specifically we employed mass spectrometric glycomic methodologies to rigorously characterise N-glycosylation in primary myoblasts from patients with GFPT1 mutations, as well as myotubes obtained by their in vitro differentiation. As controls for these analyses, we analysed myoblasts from two healthy controls and four patients with muscular diseases that have not previously been linked to glycosylation, namely CMS caused by mutations in DOK7, myopathy caused by mutations in MTND5, limb girdle muscular dystrophy type 2A (LGMD2A) caused by mutations in the CAPN3 gene encoding calpain-3, and Pompe disease caused by mutations in the GAA gene encoding acid maltase responsible for breaking down glycogen in lysosomes. Unexpectedly, no impairment of N-glycosylation was observed in the GFPT1 patients’ myoblasts or myotubes.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
## 2. Results and Discussion
*In the **2. Results and Discussion** section:*
## 2.1. Optimisation of Myoblast Culture Conditions
*In the **2.1. Optimisation of Myoblast Culture Conditions** section:*
Due to the fact that muscle biopsies do not provide sufficient sample for glycomic analysis we first established suitable cell culture conditions for producing the required cell counts (>106) whilst minimising the amount of foetal calf serum (FCS) present in the culture medium. The latter was important because it is known that FCS derived glycans frequently co-purify with cell-derived glycans during glycomic analyses [5]. We found that culturing in media containing 15% FCS was optimal for our glycomics experiments (see Figure S1).[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 2.2. Glycomic Analysis of Patient and Control Myoblasts Reveals no Impairment in N-Glycosylation
*In the **2.2. Glycomic Analysis of Patient and Control Myoblasts Reveals no Impairment in N-Glycosylation** section:*
Myoblasts were cultured from three GFPT1 patients, one DOK7 patient, one MTND5 patient, one LGMD2A patient, one Pompe disease patient, and two healthy controls. MALDI-TOF N-glycomic profiling was performed on duplicate myoblast preparations with high quality data being acquired in all instances although minor glycans were not observed in GFPT1 patient 3 whose myoblasts were difficult to culture. Representative MALDI-TOF spectra from a healthy control, a GFPT1 patient and the DOK7 patient are shown in Figure 1. MALDI data for the other patients and control are reproduced in Supplementary Information (see Figure S2).[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
Figure 1 shows that myoblast N-glycans comprise both high mannose and complex (bi-, tri-, and tetraantennary) structures. Some common characteristics of mammalian cell N-glycomes [6,7] were observed, such as core fucosylated GlcNAc, LacNAc antenna building blocks which, in some cases, are tandemly repeated to produce oligo-LacNAc extensions, and NeuAc-capped antennae. Significantly, the N-glycan profile of healthy control 1 showed a broadly similar pattern to those of GFPT1 patient 1 and the DOK7 patient.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
To investigate whether the GFPT1 patients exhibited impaired N-glycan branching, we categorised the N-glycans into families with differing levels of sialylation and then compared glycan abundances within these families. We first checked that sialylation levels were similar in the different myoblast samples. We did this by calculating abundance ratios of pairs of glycans where one member of the pair had one more sialic acid than the other but otherwise the pair had identical compositions. The results of these calculations are shown in Figure 2 for the various myoblast samples.
The data show that sialylation patterns are broadly similar for all samples.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Annotated MALDI-TOF MS spectra of permethylated N-glycans of myoblasts from healthy control 1 (A), GFPT1 patient 1 (B), and the DOK7 patient (C). In each of A, B and C, the top panel shows the full spectrum of glycans and the bottom panel amplifies the mass range where the majority of tri- and tetra-antennary glycans are found, the starting point and ending point of which have been indicated by red arrows. Profiles were obtained from the 50% acetonitrile fraction from a C18 Sep-Pak column. All ions are [M + Na]+. The number indicated in the spectra is the mass to charge ratio (m/z) of the corresponding glycan ion. Since the ion is monocharged, the value of m/z is equal to the molecular weight value of the glycan. Annotations are based on the molecular weight, N-glycan biosynthetic pathway, and MS/MS data. Glycans at m/z 2966, 3777, and 4587 are clearly annotated, which is due to the fact that their structures are unequivocal because each antenna is capped with a sialic acid and thus they are homogeneous bi-, tri-, and tetraantennary glycans. However, the glycan structure is not always as unequivocal as the glycan at m/z 2966 as biosynthetically non-fully sialylated glycan molecular ion species could be made up of mixtures of structural isoforms. Therefore, for those heterogeneous multiantennary structures with extended LacNAc repeats, the annotations are simplified throughout by using biantennary structures with the extensions and NeuAcs listed outside a bracket. GlcNAc, Man, Gal, Fuc, NeuAc.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Comparison of N-glycan sialylation in the myoblasts. Each point in the graph indicates a ratio which was obtained by comparing the relative intensity (RI) of one glycan to that of another glycan which possesses one fewer NeuAc. The numbers in the brackets correspond to the m/z of the comparing glycans.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
For each family of sialylated glycans we then compared the abundance ratios of pairs of glycans differing in composition by a single LacNAc moiety. For example Figure 3 shows comparative data for mono-sialylated (panel A) and non-sialylated (panel B) glycans containing from two to six LacNAc units. All samples showed a similar profile of relative abundances, albeit there is a two to three-fold divergence of ratios when comparing four LacNAcs with three LacNAcs (for example m/z 3504 and 3055, panel A), with the GFPT1 samples showing somewhat higher levels of the former than observed in the other samples.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
These unexpected results did not support our hypothesis that branching might be impaired in GFPT1 myoblasts. Indeed they suggested that multiantennary glycans were possibly slightly more abundant in GFPT1 patients than in controls. However, it is important to note that increasing numbers of LacNAc units is not necessarily indicative of increased branching because these moieties can be present in extended oligo-LacNAc antennae rather than as additional antennae. Indeed the MALDI data in Figure 1 confirm that myoblasts are capable of extending their antennae because many of the glycans at high mass have more than the four LacNAc moieties that are the basis of a tetra-antennary glycan. Fortunately isomeric glycans differing in branching and oligo-LacNAc extensions can be readily distinguished and their abundances compared by analysing characteristic fragment ions in MS/MS experiments.
For example Figure 4 shows MS/MS spectra obtained from the monosialylated glycan with three lacNAc units (m/z 3055) in the MALDI data (see Figure 1) for healthy control 1, GFPT1 patient 1 and the DOK7 patient. Two antennae arrangements are consistent with this composition: triantennary and/or biantennary with one LacNAc extension. As shown in the annotations on the spectra and the cartoons in Figure 4, the MS/MS spectra are dominated by fragment ions arising from loss of a single terminal LacNAc, with or without sialic acid. These fragment ions can be derived from both the bi-and tri-antennary options. Nevertheless, there are several fragment ions that are diagnostic for the extended biantennary structure.
These are observed at m/z 935, 1781, and 2142. All are minor. Importantly, their abundances relative to the major fragment ions are similar in the three samples. We estimate from abundance comparisons that m/z 3055 is comprised of about 90% tri-antennary and 10% extended bi-antennary structures in all cases.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
MS/MS analyses of other glycans with similarly ambiguous compositions gave results comparable to those described above (see Figure S3). Taken together, the MS and MS/MS data provide convincing evidence that the patterns of branching and/or antennae extensions are very similar amongst the myoblasts from GFPT1 patients, healthy controls, and the other muscle disease patients studied.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 2.3. Sialic Acid Linkages Are Predominantly α2-3 in Myoblasts
*In the **2.3. Sialic Acid Linkages Are Predominantly α2-3 in Myoblasts** section:*
To obtain information on sialic acid linkages we carried out sialidase S digestion on a myoblast preparation from the DOK7 patient. This sialidase is specific for α2-3 linked sialic acid. As shown in Figure 5, digestion with sialidase S resulted in nearly complete desialylation of all of the core fucosylated N-glycans. A handful of minor sialylated glycans were observed at m/z 2431, 2880, and 3242. These α2-6 linked sialylated bi- and tri-antennary glycans are lacking core fucose and are likely to be derived from FCS in the culture medium (see Section 2.1).
We conclude that the sialic acid in the myoblast N-glycans is mainly α2-3 linked.[](https://www.ncbi.nlm.nih.gov/mesh/D012794)
Comparison of the relative intensities of a family of mono-sialylated glycans with different numbers of LacNAc (A) in myoblasts; comparison of the relative intensities of a family of non-sialylated glycans with different numbers of LacNAc (B) in myoblasts. Each point in the graph indicates a ratio which was obtained by comparing the relative intensity (RI) of one glycan to that of the corresponding glycan which possesses one fewer LacNAc moiety. The number under the glycan structure is the m/z value of the glycan; this number is increasing with the addition of LacNAc moiety. In each comparison, the numbers in the brackets correspond to the m/z of the comparing glycans. GlcNAc, Man, Gal, Fuc, NeuAc.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
MALDI-TOF/TOF MS/MS spectra of the permethylated N-glycan at m/z 3055 [M + Na]+ in myoblasts from healthy control 1 (A), GFPT1 patient 1 (B), and the DOK7 patient (C). Assignments of the fragment ions are indicated on the cartoons and on the spectra the horizontal red arrows show antennae losses whilst antennae-derived fragment ions are annotated with their sequences. The number indicated above the peak in the spectra is the m/z value of the fragment ion (resulting ion) that has been detected by the mass spectrometry. GlcNAc, Man, Gal, Fuc, NeuAc.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Annotated MALDI-TOF MS spectra of permethylated N-glycans (A) and sialidase S treated N-glycans (B) of myoblasts from the DOK7 patient. In each of A, B and C, the top panel shows the full spectrum of glycans and the bottom panel amplifies the mass range where the majority of tri- and tetra-antennary glycans are found, the starting point and ending point of which have been indicated by red arrows. Profiles were obtained from the 50% acetonitrile fraction from a C18 Sep-Pak column. All ions are [M + Na]+. The number indicated in the spectra is the mass to charge ratio (m/z) of the corresponding ion. Since the ion is monocharged, the value of m/z is equal to the molecular weight value of the glycan. Annotations are based on the molecular weight, N-glycan biosynthetic pathway and MS/MS data. Glycans at m/z 2966, 3777, and 4587 are clearly annotated, which is due to the fact that their structures are unequivocal because each antenna is capped with a sialic acid and thus they are homogeneous bi-, tri-, and tetraantennary glycans. However, the glycan structure is not always as unequivocal as the glycan at m/z 2966 as biosynthetically non-fully sialylated glycan molecular ion species could be made up of mixtures of structural isoforms. Therefore, for those heterogeneous multiantennary structures with extended LacNAc repeats, the annotations are simplified throughout by using biantennary structures with the extensions and NeuAcs listed outside a bracket. GlcNAc, Man, Gal, Fuc, NeuAc.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 2.4. Glycomic Analysis of Patient and Control Myotubes Reveals no Significant Difference in N-glycosylation
*In the **2.4. Glycomic Analysis of Patient and Control Myotubes Reveals no Significant Difference in N-glycosylation** section:*
Myotubes were obtained from all patients and controls by in vitro differentiation of their myoblasts in culture. They were then subjected to the same glycomic analyses as for myoblasts. Representative MALDI spectra are shown for healthy control 1, GFPT1 patient 1, and the DOK7 patient in Figure 6; spectra from other samples can be found in Supplementary Information (see Figure S4). Most samples gave good quality MALDI data, but, because sample quantities were more limited than for the myoblast preparations, low abundance glycans at high mass were not always observed. Nonetheless, we were able to draw firm conclusions from our data.
Notably no significant differences were observed in the MALDI profiles from normal and GFPT1 samples (Figure 6A,B). Both showed a predominance of biantennary glycans, relatively high levels of tri- and tetraantennary glycans, and only very minor signals for glycans with more than four lacNAc units. The DOK7 myotubes similarly showed only minor signals for extended glycans but, in contrast to the normal and GFPT1 myotubes, their tri- and tetraantennary glycans were more abundant than biantennary structures. Similar glycomic profiling was performed on cells from healthy controls, GFPT1, MTND5, and LGMD2A patients. They all showed broadly similar glycan profiles as those observed in Figure 6A,B (see Figure S4).[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 2.5. Discussion
*In the **2.5. Discussion** section:*
This is the first report of N-glycan profiles of in vitro cultured human myoblasts and human myotubes differentiating from myoblasts in vitro. The main findings of this work are that the N-glycan profiles of myoblasts (Figure 1) and myotubes (Figure 6) from GFPT1 patients and a selection of healthy and disease controls are broadly similar. All have comparable levels of high mannose and complex-type structures. The latter are core fucosylated and capped with α2-3 linked sialic acid. The relative abundances of the bi-, tri- and tetraantennary complex-type glycans are broadly similar in all samples except for the DOK7 myotubes whose multiantennary glycans were found to be somewhat more abundant than the other samples (Figure 6).[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
These findings are unexpected as our initial hypothesis was that cells from GFPT1 patients would have changes in their N-glycan structures and particularly a reduction in N-glycan branching. This was based on data we recently published in which we characterised N-glycosylation of leukocytes from patients with mutations in phosphoglucomutase 3 (PGM3), another key enzyme in the biosynthesis of UDP-GlcNAc. A reduction in levels of tri-antennary and tetra-antennary N-glycans was observed which was rationalized by the fact that if a comparison is made of the GlcNAc transferases that are responsible for N-glycan branching the higher Km values (which in Michaelis-Menten kinetics corresponds to low substrate affinity and thus a low catalytic activity), are associated with GlcNAc transferase IV and GlcNAc transferase V that initiate the third and fourth antennae. Therefore by requiring higher substrate concentrations they will be most impaired by a reduction in UDP-GlcNAc concentration [8].[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Annotated MALDI-TOF MS spectra of permethylated N-glycans of myotubes from healthy control 1 (A); GFPT1 patient 1 (B) and the DOK7 patient (C). In each of A, B and C, the top panel shows the full spectrum of glycans and the bottom panel amplifies the mass range where the majority of tri- and tetra-antennary glycans are found, the starting point and ending point of which have been indicated by red arrows. Profiles were obtained from the 50% acetonitrile fraction from a C18 Sep-Pak column. All ions are [M + Na]+. The number indicated in the spectra is the mass to charge ratio (m/z) of the corresponding ion. Since the ion is monocharged, the value of m/z is equal to the molecular weight value of the glycan. Annotations are based on the molecular weight, N-glycan biosynthetic pathway and MS/MS data. Glycans at m/z 2966, 3777 and 4587 are clearly annotated, which is due to the fact that their structures are unequivocal because each antenna is capped with a sialic acid and thus they are homogeneous bi-, tri- and tetraantennary glycans. However, the glycan structure is not always as unequivocal as the glycan at m/z 2966 as biosynthetically non-fully sialylated glycan molecular ion species could be made up of mixtures of structural isoforms. Therefore, for those heterogeneous multiantennary structures with extended LacNAc repeats, the annotations are simplified throughout by using biantennary structures with the extensions and NeuAcs listed outside a bracket. GlcNAc, Man, Gal, Fuc, NeuAc.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
Our findings that glycomic profiling of cells from GFPT1 patients is not altered does not however rule out that changes in N-glycosylation still plays role in the associated disease pathologies. It could be that occupancy of N-glycosylation sites is altered in the patients. The glycomic methods employed in the current work do not address occupancy. Indeed, assigning N-glycan occupancy is not possible in glycomic experiments on limited quantities of cells. It is also possible that changes in N-glycosylation are restricted to a subset of glycoproteins and that such subtle differences are lost in the general glycomic profiling.
It could also be that the glycosylation in the muscle of the patient is different from that in the cultured cells or the glycosylation might have changed during the use and regeneration of muscle tissue (GFPT1 CMS is later onset than other CMS forms). In addition, the amount of glucose in the medium could influence the glucose flux and subsequently affect the amount of the primary substrate fructose-6-P [9].[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
The next challenge will be to perform detailed glycoproteomic analysis of specific glycoproteins such as acetylcholine receptor, the muscle-specific kinase (MuSK) and agrin. Indeed similarity between CMS due to GFPT1 mutations and CMS due to DPAGT1 mutations would suggest that reduced endplate acetylcholine receptor due to defective N-linked glycosylation is a primary disease mechanism in this disorder [3,10,11].[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
We chose the disease controls, namely CMS caused by mutations in DOK7, myopathy caused by mutations in MTND5, LGMD2A, and Pompe, as exemplars of muscle diseases which had not previously been found to exhibit defects in N-glycosylation. Very recently, however, a study has found that Pompe disease can lead to a Golgi-based glycosylation deficit in human skin fibroblast-derived induced pluripotent stem cells which were differentiated in culture to cardiomyocytes (iPSCCMs) [12]. Using a similar glycomics strategy to ours, Raval et al. have shown that there is a reduced diversity of multi-antennary structures and hyposialylation in Pompe iPSCCMs. It should be noted that even their control cells did not show the same high molecular weight multi-antennary N-glycans that we observed in this study.
It is intriguing that our muscle cells are not showing comparable glycosylation defects. In future work it will be important to investigate whether glycosylation defects in Pompe are cell-specific.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
In conclusion we have performed the first detailed N-glycan glycomic analysis of myoblasts from healthy control, GFPT1 patients, a DOK7 patient, a MTND5 patient, a LGMD2A patient, and a Pompe disease patient. We also performed detailed structural analysis of N-glycans from myotubes derived from a healthy control, a GFPT1 patient, and a DOK7 patient. Whilst greatly expanding our knowledge of glycosylations in these cells we did not observe significant changes in disease and control N-glycan structures.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 3. Materials and Methods
*In the **3. Materials and Methods** section:*
## 3.1. Patients
*In the **3.1. Patients** section:*
Patients with limb-girdle CMS with tubular aggregates have been reported in two papers [1,2]. Detailed clinical data of the patients was previously published [2]. In brief, the patients presented with fatigable weakness restricted mostly to shoulder and pelvic girdle and minimal or no weakness of facial and ocular muscles. Response to treatment with acetylcholinesterase inhibitors was generally positive. The DOK7 patient has DOK7 mutations at c.1124_1127dupTGCC and a deletion of the last 11 bp of exon 4 on cDNA level (second mutation on genomic DNA level not identified), the MTND5 patient has a MTND5 mutation at m.13051G>A, the Pompe disease patient has GAA mutations at c.-32-13T>G and c.52del (compound heterozygous), the limb girdle muscular dystrophy type 2A patient has CAPN3 mutations at c.1322delG and c.1465C>T (compound heterozygous).
Patients and controls were selected at the Newcastle University John Walton Muscular Dystrophy Research Centre, Institute of Genetic Medicine, Newcastle upon Tyne. Their diagnosis was obtained through a combination of clinical and morphological (muscle biopsy) methods and confirmed by genetic testing of the relevant genes. Collection of samples from patients and their use in research have been ethically approved by the NRES Committee North East—Newcastle and North Tyneside 1. All GFPT1 mutations resulted in a reduction of GFPT1 protein amounts in patient muscle. No changes in protein localisation or enzyme activity were observed.
## 3.2. Cell Culture
*In the **3.2. Cell Culture** section:*
Primary human skeletal myoblasts of the following patients were analysed: LGM9.3, (GFPT1 mutations p.V199F and c.*22>A), and LGM5.3 and 5.5 (GFPT1 mutations p.M492T and c.*22C>A); patient numbers correspond to numbering in a previous paper [1]. LGM9.3, LGM5.3, and LGM5.5 are GFPT1 patient 1, 2, and 3 respectively. Six control myoblast lines were used, two from healthy individuals and four from diseases unrelated to GFPT1 (mutations in MTND5, DOK7, CAPN3, and GAA).
Myoblasts from patients and controls were obtained from the MRC Centre for Neuromuscular Diseases Biobank Newcastle, UK, and the Muscle Tissue Culture Collection, Friedrich-Baur-Institute, Munich, Germany. The cells were isolated as previously described [13] and firstly grown in skeletal muscle growth medium (PromoCell, Heidelberg, Germany) supplemented with 5%, 10%, and 15% foetal calf serum. The final concentration of foetal calf serum in the growth medium was 15%. The concentration of glucose in the myoblast growth medium was 1 g/L. Myoblasts were differentiated to myotubes by switching to differentiation medium (DMEM with 2% horse serum) at a cell density of around 80%. The concentration of glucose in the differentiation medium was 4.5 g/L. Cells were grown in differentiation medium for one week.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
All studies were carried out with informed consent of the patients or their parents and were approved by institutional ethics review boards.
## 3.3. Processing of Myoblasts and Myotubes to Acquire N- and O-glycans
*In the **3.3. Processing of Myoblasts and Myotubes to Acquire N- and O-glycans** section:*
All myoblast and myotube samples were treated following a standard protocol [14]. Briefly, cells were suspended in lysis buffer (25 mM TRIS, 150 mM NaCl, 5 mM EDTA and 1% CHAPS (v/v), pH 7.4) before homogenisation and sonication were performed. The homogenates were subsequently dialysed against a 50 mM ammonia bicarbonate buffer, pH 7.5, after which the samples were lyophilized. Extracted glycoproteins were reduced and carboxymethylated and digested with trypsin. The digested glycopeptides were purified using a C18 cartridge (Oasis HLB Plus Waters) prior to the release of protein linked N-glycans by PNGase F (Roche Applied Science, East Sussex, UK) digest and O-linked glycans by reductive elimination.
Released N- and O-glycans were permethylated and then purified using a Sep-pack C18 cartridge (Waters, Milford, MA, USA) prior to MS analysis. Sialidase cleavage was carried out using sialidase S (Prozyme Glyko, Cambridge, UK) in 50 mM sodium acetate, pH 5.5.[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
## 3.4. Mass Spectrometric Glycomic Analysis
*In the **3.4. Mass Spectrometric Glycomic Analysis** section:*
MS data were obtained via a Voyager MALDI-TOF (Applied Biosystems, Foster City, CA, USA) mass spectrometer. Purified permethylated glycans were dissolved in 10 µL methanol and 1 µL of the sample was mixed with 1 µL of matrix, 20 mg/mL 2,5-dihydroxybenzoic acid (DHB) in in 70% (v/v) aqueous methanol and loaded on to a metal target plate. The instrument was run in the reflectron positive ion mode. The accelerating voltage was 20 kV.[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
MS/MS data were acquired using a 4800 MALDI-TOF/TOF mass spectrometer (AB SCIEX). In the MS/MS experiment the dissolved sample was dried and then re-dissolved in 10 µL methanol, 1 µL of the sample was mixed with 1 µL of matrix, 10 mg/mL diaminobenzophenone (DABP) in 70% (v/v) aqueous acetonitrile and loaded on to a metal target plate. The instrument was run in the reflectron positive ion mode. The collision energy was set to 1 kV with argon as the collision gas. The 4700 calibration standard (mass standards kit for the 4700 proteomics analyzer, Applied Biosystems) was used as the external calibrant for the MS and MS/MS modes.[](https://www.ncbi.nlm.nih.gov/mesh/D000432)
## 3.5. Analyses of MALDI Data
*In the **3.5. Analyses of MALDI Data** section:*
The MS and MS/MS data were processed employing Data Explorer Software from Applied Biosystems. The processed spectra were annotated using a glycobioinformatics tool, GlycoWorkBench [15]. Based on known biosynthetic pathways and susceptibility to PNGase F digestion, all N-glycans are presumed to have a Manα1–6(Manα1–3)Manβ1–4GlcNAcβ1–4GlcNAc core structure [16,17]. The symbolic nomenclature used in the spectra annotation is the same as the one used by the Consortium for Functional Glycomics (CFG) () and the Essentials for Glycobiology on-line textbook ().
Reproducibility and comparability were assessed by ANOVA (see Figure S5).[](https://www.ncbi.nlm.nih.gov/mesh/D011134)
## 4. Conclusions
*In the **4. Conclusions** section:*
The glycomic analysis of myoblasts reveals that the patterns of branching and/or antennae extensions are very similar amongst the myoblasts from GFPT1 patients, healthy controls, and the other muscle disease patients studied.
In addition, analysis of myotubes reveals that there is no significant difference in the MALDI profiles from GFPT1 patients, healthy controls, and the other muscle disease patients studied.
# Supplementary Files
*In the **Supplementary Files** section:*
# Author Contributions
*In the **Author Contributions** section:*
Stuart M. Haslam, Hanns Lochmüller and Anne Dell designed the study. Qiushi Chen, Poh-Choo Pang, Juliane S. Müller and Steve H. Laval performed the experiments outlined in the manuscript. All authors analyzed and interpreted the data. All authors discussed and commented on the study.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
|
# Introduction
[Genicunolide A, B and C](https://www.ncbi.nlm.nih.gov/mesh/D014315): three new [triterpenoids](https://www.ncbi.nlm.nih.gov/mesh/D014315) from Euphorbia geniculata
# Abstract
*In the **Abstract** section:*
Summary
Three new triterpenoids, designated as genicunolide A (1), B (2) and C (3), along with friedel[in (4) and fr](https://www.ncbi.nlm.nih.gov/mesh/D014315)iedelinol (5), w[ere isolated from the aerial parts ](https://www.ncbi.nlm.nih.gov/mesh/D014315)of Euphorbia [geniculat](https://www.ncbi.nlm.nih.gov/mesh/C060796)a. They w[ere charact](https://www.ncbi.nlm.nih.gov/mesh/C055670)erized as 1β-acetoxy-3β-hydroxy-11α,12α-oxidotaraxer-14-ene, 1β,3β-diacetoxy-21α-hydroxy-11α,1[2α-oxidotaraxer-14-ene and 3β,9α,20α-trihydroxy-Ψ](https://www.ncbi.nlm.nih.gov/mesh/D014315)-t[araxast-5-ene, respectively, by spectral and chemical m](https://www.ncbi.nlm.nih.gov/mesh/D014315)ethod[s.](https://www.ncbi.nlm.nih.gov/mesh/D014315)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Euphorbia (Euphorbiaceae) is a very large and diverse genus of flowering plants comprising of about 2,000 members and is found all over the world, ranging from short annual plants to well developed tall trees [1].
The plants of the family Euphorbiaceae contain well-known skin irritating and tumor-promoting diterpenoids with tigliane, ingenane and daphnane skeletons [2]. Some of the species are used in folk medicine to cure skin diseases, gonorrhea, migraines, intestinal parasites, and warts [3] and as a purgative [4–6]. Several macrocyclic diterpenoids with antibacterial, anticancer, anti-multidrug-resistant, antifeedant, anti-HIV and analgesic activity have been isolated from different Euphorbia species. They include jatrophane, ingol and myrsinane diterpenoids [7–13].[](https://www.ncbi.nlm.nih.gov/mesh/D004224)
Triterpenoids which have been reported from various species of Euphorbia include β-amyrin [1], β-amyrin acetate [14,18], cycloeucalenol, obtusifoliol, 24-methylenecycloartan-3-β-ol, β-sitosterol, betulin, erythrodiol, oleanolic acid, β-sitosterol glucoside[15], 29-norcycloart-5-ene-5,8-lanostadiene-3β-ol, 3β,24S,25-trihydroxycycloartane, 3β,24(R),25-trihydroxy-cycloartane, 24-methylenecycloartan-3β-ol [16], cycloart-23-ene-3,5-diol [17], lupeol, lupeol acetate, ginnone, ambrein, lupeone [18], 24-methylenecycloartanol
[19], cycloart-25-en-3β,24-diol [20] and cycloart-22-ene-3β,25-diol [21]. In addition, nor-isoprenoids and coumarins have also been reported from few species of Euphorbia [22–24].[](https://www.ncbi.nlm.nih.gov/mesh/D014315)
Euphorbia geniculata Orteg. [25–26], is a wild weed found in the Jammu region of India [27]. The plant is locally used for the treatment of bacterial infections and inflammations. Previous phytochemical investigations have demonstrated that this plant contains flavonoids: kaempferol, quercetin and 3-rhamnosyl quercetin [28] and triterpenes β-amyrin acetate [29] and geniculatin [30].[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
Reinvestigation of chemistry of the plant led to isolation of three new triterpenoids, designated as genicunolide A (1), B (2) and C (3), together with friedelin (4) [31] and friedelinol (5) [32], from the ethyl acetate extract of the aerial parts of the plant. Herein, we report the characterization of the three compounds by spectral and chemical methods.[](https://www.ncbi.nlm.nih.gov/mesh/D014315)
## Results and Discussion
*In the **Results and Discussion** section:*
The compounds 1–3 (Fig. 1) responded positively to the characteristic Liebermann–Burchard [33], TCA [34–35] and TNM tests [35] for unsaturated triterpenoids.[](https://www.ncbi.nlm.nih.gov/mesh/D014315)
Structures of compounds 1, 2, 3, 1a, 2a, 1b, 2b, 3a, 4 and 5.
The compound 1, M+ at m/z 498.0695 (calculated for C32H50O4, 498.0700), possessed eight tertiary methyl groups, an acetoxy functionality [ν max 1736 cm−1, δ 2.03 (s, 3H), δC 170.2, 21.3], a trisubstituted double bond [ν max 1630, 1042, 880 cm−1, δ 5.56 (d, J = 5.2 Hz, 1H, H-15)] [36] and a secondary equatorial hydroxy group [νmax 3500 cm−1] whose carbinylic proton resonated at δ 3.16 (dd, J = 8.1 Hz, 1H, H-3). On acetylation with Ac2O–C5H5N, at room temperature, compound 1 afforded the diacetate 1a, δ 2.06 (s, 6H), and on oxidation with CrO3–C5H5N yielded a ketoacetate 1b, νmax 1738, 1680 cm–1, δC 216.7, 170.8, which responded positively to the characteristic Zimmermann test for 3-keto function [37], thereby placing the hydroxy group in compound 1 at 3β position, δC 77.2
[38].[](https://www.ncbi.nlm.nih.gov/mesh/D014316)
The mass spectrum of compound 1 displayed the characteristic features of the taraxer-14-ene skeleton [39] by exhibiting RDA fragment ion peaks at m/z 374 (rings A/B/C) and 124 (ring E) and the vinylic carbon resonance signals at δC 118.8 (C-15) and 157.0 (C-14) [40].[](https://www.ncbi.nlm.nih.gov/mesh/D014315)
The presence of a cis-oxido functionality in compound 1 was evident from a pair of AB doublets at δ 2.82 and 3.01 (J ae = 4.7 Hz, 1H each, He-12 and Ha-11), in its 1H NMR spectrum, two methine carbon resonance signals at δC 53.4 (C-11) and 58.3 (C-12) [41] and loss of CO and H2O, via rearrangement of hydrogen [42] from the RDA fragment ion at m/z 374 to give abundant ion peaks at m/z 346 and 354, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
The acetoxy functionality in compound 1 was placed at C-1 β-position on the basis of the chemical shift, multiplicity and coupling constant of the carbinylic proton [δ 4.53 (dd, J ae = 7.4 Hz, J aa = 8.5 Hz, 1H, H-1)] together with the identical chemical shift of H-1 and H-3, in the 1H NMR spectrum of 1a, and the comparable chemical shifts of C-1 and C-3 of 1a (δC 80.7 and 80.6, respectively) with that of 1β,3β-diacetoxylupenes [43].[](https://www.ncbi.nlm.nih.gov/mesh/D002244)
The structure of compound 1 was further confirmed by 1H,1H and 1H,13C COSY, HMBC and HSQC experiments which allowed unambiguous fixation of protons to appropriate carbons and also 13C-chemical shifts (Table 1). The long range correlations between the protons at δ 1.96 (Ha-2) and 1.98 (He-2) and carbonyl signal at δC 170.8 confirmed the presence of acetoxy carbonyl at C-1. This was further substantiated by the long range mutual coupling of the carbinylic proton at δ 4.53 with the proton at δ 3.16 as also with the carbon at δC 41.1 ppm (C-4). The correlation cross peaks between H-23 and H-25 and H-2 in the NOESY experiment confirmed that the acetoxy function at C-1 was β-oriented. The proton at δ 2.82 was correlated to the olefinic carbon at δC 157 ppm (C-14), three bonds away and allowed joining of spin systems separated by a methyl-bearing quaternary carbon on one side.
The chemical shifts, multiplicity and coupling constants of the A-ring carbinylic protons and an inspection of the molecular models suggested that 1,3-cis-diequatorial functions in ring A caused flattening of this ring.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
13C NMR data of 1, 2, 3 and their acetates in CDCl3 (δ in ppm, 125 MHz).[](https://www.ncbi.nlm.nih.gov/mesh/C000615229)
The spectral patterns of compound 2, M+ at m/z 556.0739, C34H52O6, resembled closely with those of 1a, except that it was shown to possess an extra secondary hydroxy group [ν max 3350 cm–1, δ 3.17 (dd, J = 11.1 Hz, 1H)]. Its presence was confirmed by acetylation of 2 to 2a [δ 2.05 (s, 9H), δC 170.1, 170.7 and 170.8] and oxidation to diacetoxy ketone 2b [ν max 1736, 1730, 1680 cm–1, δC 170.5, 170.6, 215.2 (C-21)]. The mass spectrum of compound 2 exhibited RDA fragment ions at m/z 416 and 140, placing the hydroxy group in ring
E. The hydroxy group was placed at C-21 α-position (δC 77.2) [38] on the basis of coupling constant of carbinylic proton in the 1H NMR spectrum of 2, the downfield chemical shift of C-30 methyl protons (δ 1.09) in the spectrum of 2b, and 1H,1H, 1H,13C COSY, HMBC and HSQC spectra of 2 which showed long range correlations of the carbinylic proton at δ 3.17 (H-21) and methyl protons at δ 0.87 (H-29) as also carbons at δC 35.7 (C-16) and 48.1 (C-18).[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
The 1H NMR, 13C NMR (Table 1) and DEPT (135°) spectra of compound 3, M+ at m/z 458.1495, C30H50O3, revealed that the compound possesses seven tertiary methyls, one of which resonated downfield at δ 1.53; a secondary methyl [δ 0.85 (d, J = 4.2 Hz, 3H)], a trisubstituted double bond [νmax 1640, 1040, 890 cm−1, δ 5.36 (d, J = 4.7 Hz, 1H)], and a secondary hydroxy group [νmax 3465 cm−1, δ 3.52 (dd, J aa = 7.6 Hz, H-3, 1H)]. On acetylation with Ac2O–C5H5N, at room temperature, it afforded monoacetate 3a [1735, δ 2.05 (s, 3H), δC 170.5] and on oxidation with CrO3–C5H5N, it yielded a ketone which gave a positive Zimmermann test for 3-keto group
[37] confirming the presence of the C-3 equatorial secondary hydroxy group [δC 71.8] in 3. The mass spectrum of compound 3 revealed that the double bond triggered the typical RDA fragmentation of ring B [39] to give densely populated ion peaks at m/z 166 (ring A) and 292 (rings C/D/E) placing the double bond at C-5 [δC 122.0 (C-6), 139.9 (C-5)] [38] and two hydroxy groups in rings C/D/E. Since the monoacetate 3a still retained a hydroxy group and its mass spectrum also showed a RDA fragment ion at m/z 292, compound 3, therefore, carried two tertiary hydroxy groups on rings C/D/E.
The presence of a secondary methyl group together with a pair of doublets at δ 1.56 and 1.85 (d br, J = 10.0 Hz, 1H each, H-18 and H-19) showed that the compound belonged to the Ψ-taraxastane [35] series. The downfield shift of C-30 methyl singlet (δ 1.53) suggested that one of the tertiary hydroxy groups was at C-20 (δC 77.3). Had it been on C-19, the 13C signal would have been observed upfield at δC 73.0–73.2 [44]. The densely populated ion peaks at m/z 221 (rings A/B) and 203 (221 − H2O. +), arising from the fission of 9, 11 and 8, 14 bonds in ring C, together with the downfield carbon signal at δC 89.5 placed the second tertiary hydroxy groups at C-9.
The structure of compound 3 was further confirmed by 1H,1H, 1H,13C COSY and long range 1H,13C COSY experiments. The presence of the C-9 hydroxy group was proved by linking the carbon signal at δC 89.5 to proton signals at δ 1.01 (C-25) and 0.93 (C-26) in the 1H,13C long-range coupled spectrum. Other data for 1D and 2D NMR spectra of 3 were in agreement with the assigned structure.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Comparison of physical characteristics and spectral data of compounds 4 and 5, with those reported in literature [31–32], confirmed them to be friedelin and friedelinol, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/C060796)
## Conclusion
*In the **Conclusion** section:*
The compounds 1–5 were, thus, characterized as 1β-acetoxy-3β-hydroxy-11α,12α-oxido-taraxer-14-ene (1); 1β,3β-diacetoxy-21α-hydroxy-11α,12α-oxido-taraxer-14-ene (2); 3β,9α,20α-trihydroxy-Ψ-taraxast-5-ene (3); friedelin (4) and friedelinol (5); respectively. Compounds 1–3 are new triterpenoids while 4 and 5 appear to have been isolated for the first time from the genus Euphorbia.[](https://www.ncbi.nlm.nih.gov/mesh/D014315)
## Experimental
*In the **Experimental** section:*
## General procedures
*In the **General procedures** section:*
Melting points were determined in centigrade scale in one end open capillaries on a Büchi 570 melting point apparatus and are uncorrected. IR spectra were recorded on a Perkin-Elmer Paragon-1000 spectrophotometer or an Esquire 3000 spectrometer. 1H and 13C NMR spectra were recorded by a Bruker 500 and 125 MHz instrument using TMS as internal standard and CDCl3 as solvent. High-resolution mass spectra were recorded on a Bruker 400 mass spectrometer. Column chromatography was carried out with Merk silica gel (60–120 mesh).
Optical rotation was measured on a Perkin-Elmer polarimeter.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Plant material
*In the **Plant material** section:*
The aerial parts of Euphorbia geniculata (Orteg) were collected from Jammu, (J&K, India) in July 2013. The specimen was identified by Akhtar H. Malik, Curator, Centre for Biodiversity & Taxanomy, University of Kashmir (Specimen deposited under accession No. 1850 – KASH Herbarium).
## Extraction and isolation
*In the **Extraction and isolation** section:*
The shade dried aerial parts of Euphorbia geniculata (3.0 kg) were extracted sequentially with petroleum ether (60–80 °C), ethyl acetate and methanol in a soxhlet apparatus to afford respective extracts which were concentrated under reduced pressure. The ethyl acetate extract (40 g) was subjected to chromatography on silica gel (60–120 mesh, B.D.H.) column using graded solvent systems of petroleum ether–ethyl acetate. The fractions collected with petroleum ether–ethyl acetate (9:1), F-1; (8:2), F-2; (7:3), F-3 and ethyl acetate, F-4, whose components gave green, pink and violet colouration on TLC (silica gel G) plates, after development with cerric ammonium sulfate–H2SO4, were subjected to re-chromatography. The fraction F-1 on re-chromatography and elution with petroleum ether–dichloromethane (8:2) gave 4 (300 mg) and 5 (410 mg). The fraction F-2 on further chromatography and elution with petroleum ether–dichloromethane (7:3) and (8:4) gave two mixtures.
The mixture obtained with petroleum ether–dichloromethane (8:4) was subjected to preparative TLC using petroleum ether–chloroform (19:3) as solvent system to get compound 1 (48 mg). The fraction F-3 on further chromatography and elution with petroleum ether–dichloromethane (8:2) gave compound 2 (45 mg). The fraction F-4 on repeated chromatography using the same sequence of graded solvent systems, as for crude extract, gave compound 3 (38 mg) with petroleum ether–dichloromethane (3:7) and a mixture containing 3 and 2, which was again resolved by preparative TLC using benzene–ethyl acetate (9:1) as solvent system.[](https://www.ncbi.nlm.nih.gov/mesh/C004544)
Genicunolide A (1): Colourless crystals (CHCl3–Me2CO), mp 150 °C; [α]D 25 +20.5° (c 0.50, CHCl3); HRMS: m/z (rel. int.) 498.0695 (18) (M+) (calcd for C32H50O4, 498.0700), 483 (36.2), 480 (21.4), 456 (28.6), 441 (47.1), 374 (71.3) (RDA, rings A/B/C), 346 (57.2), 314 (42.7), 124 (65.8) (RDA, ring E), 108 (100); IR: νmax 3500 (OH), 3030, 2850, 1736 (OAc), 1630, 1456, 1042, 880 cm−1; 1H NMR (500 MHz, CDCl3) δ 0.80 (s, 3H, H-28), 0.86 (s, 6H, H-24, H-29), 0.89 (s, 3H, H-30), 0.90 (s, 3H, H-23), 1.01 (s, 3H, H-25), 1.09 (s, 3H, H-26), 1.25 (s, 3H, H-27), 2.06 (s, 3H, OCOCH3), 2.31 (d, J = 6.9 Hz, 2H, H-16), 2.82 (d, J = 4.7 Hz, 1H, H-12), 3.01 (d, J = 4.7 Hz, 1H, H-11), 3.16 (dd, J = 5.5, 8.1 Hz, 1H, H-3), 4.53 (dd, J = 7.4, 8.5 Hz, 1H, H- 1), 5.56 (d, J = 5.2 Hz, 1H, H-15); 13C NMR: Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D014315)
Genicunolide B (2): Colourless crystals (CHCl3–Me2CO), mp 160 °C, [α]D 25 + 34.2° (c 0.40, CHCl3); HRMS: m/z 556.0739 (M+) (calcd for C34H52O6, 556.0744), 541 (M+ − .CH3), 514 (M+ − CH2CO), 472 (514 − CH2CO), 454 (472 − H2O), 416 (RDA, rings A/B/C), 356 (416 − HOAc), 286 (356 − CO − CH2CO), 140, 124 (RDA, ring E), 108 (124 – H2O)(100); IR: νmax 3550, 3025, 2863, 1736 (OAc), 1730 (OAc), 1625, 1450, 1045, 890 cm−1; 1H NMR (500 MHz, CDCl3) δ 0.80 (s, 3H, H-28), 0.87 (s, 6H, H-24, H-29), 0.97 (s, 3H, H-30), 0.98 (s, 3H, H-23), 1.01 (s, 3H, H-25), 1.09 (s, 3H, H-26), 1.25 (s, 3H, H-27), 2.06 (s, 6H, 2 x OAc), 2.31 (s, 2H, H-16), 2.80 (d, J = 4.7 Hz, 1H, H-12), 3.01 (d, J = 4.8 Hz, 1H, H-11), 3.17 (dd, J = 5.5, 11.1 Hz, 1H, H-21), 4.53 (dd, J = 7.6, 8.5 Hz, 2H, Ha-1, Ha-3), 5.56 (d, J = 5.2 Hz, 1H, H-15); 13C NMR: Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D014315)
Genicunolide C (3): Colourless needles, mp 210–211 °C, [α]D 25 +30.3° (c 0.3, CHCl3); HRMS: m/z 458.1495 (M. +) (calcd for C30H50O3, 458.1500) (M. +), 443 (M+ − CH3), 440 (M+ − H2O), 425 (M+ − CH3 − H2O), 413 (M+ − CH3CH=+OH), 292 (RDA, rings C/D/E), 237 (RDA, rings D/E), 221, 203, 166 (RDA, ring A), 163, 107, 83, 45 (100); IR: ν max 3580, 3465, 1640, 1445, 1040, 1025, 890 cm−1; 1H NMR (500 MHz, CDCl3) δ 0.68 (s, 3H, H-28), 0.79 (s, 3H, H-27), 0.81 (s, 3H, H-23), 0.85 (d, J = 4.2 Hz, 3H, H-29), 0.90 (s, 3H, H-24), 0.93 (s, 3H, H-26), 1.01 (s, 3H, H-25), 1.53 (s, 3H, H-30), 1.56 (d br, J = 10.0 Hz, 1H, H-18), 1.85 (d br, J = 10.0 Hz, 1H, H-19), 2.28 (s, 2H, H-7), 3.52 (dd, J = 4.8, 7.6 Hz, 1H, H-3), 5.36 (d, J = 4.7 Hz, 1H, H-6); 13C NMR: Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D014315)
Acetylation of 1, 2 and 3: Compounds 1, 2 and 3 (20 mg each) were dissolved separately in C5H5N (2 mL) and Ac2O (2 mL) was added. The reaction mixtures were left overnight, diluted with water and extracted with chloroform. The chloroform solutions were washed with 5% HCl–H2O solution (10 mL each time) and dried over anhydrous K2CO3. After removal of the solvent, the crude acetates were purified by column chromatography on silica gel using petroleum ether–benzene (9:1, 7:3 and 1:1 v/v) when 1a, 2a and 3a (18 mg, 17 mg and 19 mg, respectively) were recovered.[](https://www.ncbi.nlm.nih.gov/mesh/C023666)
Oxidation of 1 and 2: Compound 1 (12 mg) and compound 2 (15 mg) were dissolved separately in C5H5N (1 mL) and treated with freshly prepared CrO3–C5H5N complex. The reaction mixtures were left overnight, diluted with water (10 mL) and extracted with chloroform (3 × 20 mL). The chloroform layer was washed with water, 0.1 N HCl, water and dried over anhydrous MgSO4. After removal of the solvent, the residues were purified by column chromatography over silica gel using a petroleum ether–benzene (1:1 v/v) solvent system, and crystallized from CHCl3–Me2CO.[](https://www.ncbi.nlm.nih.gov/mesh/C023666)
## Supporting Information
*In the **Supporting Information** section:*
|
# Introduction
[Ferric carboxymaltose](https://www.ncbi.nlm.nih.gov/mesh/C522335) reduces transfusions and hospital stay in patients with colon cancer and anemia
# Abstract
*In the **Abstract** section:*
Purpose
The purpose of the study was to evaluate the efficacy of preoperative intravenous (IV) ferric carboxymaltose (FCM) administration vs. no-IV iron in colon cancer (CC) anemic patients undergoing elec[tive surgery with cur](https://www.ncbi.nlm.nih.gov/mesh/C522335)at[ive](https://www.ncbi.nlm.nih.gov/mesh/C522335) intention.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
Methods
This was a multicenter, observational study including two cohorts of consecutive CC anemic patients: the no-IV iron treatment group was obtained retrospectively while FCM-treated patients were recorded prospectively.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
Results
A total of 266 patients were included: 111 received FCM (median dose 1000 mg) and 155 were no-IV iron subjects. Both groups were similar in terms of demographic characteristics, tumor location, surgical approach, and intra-operative bleeding severity. The FCM group showed a significant lower need for red blood cell (RBC) transfusion during the study (9.9 vs. 38.7 %; OR: 5.9, p < 0.001). In spite of lower hemoglobin levels at baseline diagnosis and lower transfusion rates in the FCM group, the proportion of responders was significantly higher with respect to the no-IV group both at hospital admission (48.1 vs. 20.0 %, p < 0[.00](https://www.ncbi.nlm.nih.gov/mesh/C522335)01) and at 30 days post-surgery (80.0 vs. [48.9](https://www.ncbi.nlm.nih.gov/mesh/D007501) %, p < 0.0001). The percentage of patients with normalized hemoglobin levels was also higher in the FCM group (40.0 vs. 26.7 % at 30 days, p < 0.05). A lo[wer](https://www.ncbi.nlm.nih.gov/mesh/C522335) number of reinterventions and post-surgery complications were seen in the FCM group (20.7 vs. 26.5 %; p = 0.311). The FCM group presented a significant shorter hospital stay (8.4 ± 6.8 vs. 10.9 ± 12.4 days to discharge; p <[ 0.](https://www.ncbi.nlm.nih.gov/mesh/C522335)001).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
Conclusions
Preoperative ferric carboxymaltose treatment in patients with CC and iron deficiency anemia significantly reduced RBC transfusion requirements and hospital length of stay, reaching higher response rates and percentages of normalized hemoglobin levels both at hospital admission and 30 days post-surgery.[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
A large number of different tumors occur associated with anemia, which ranges from 25–75 % in cancer patients who undergo surgery [1]. Although differing by tumor location, the overall prevalence of anemia in colon cancer was around 48 % with moderate to severe anemia found in more than 20 % of cases [2]. The etiology of anemia in cancer patients is typically multifactorial, mainly caused by systemic inflammation with increased hepcidin levels promoted by the tumor itself as well as by iron deficiency due to gastrointestinal blood loss that occurs in association with tumor ulceration and with malnutrition derived from the disease itself [3, 4]. Besides, blood losses can be stressed during the surgery.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
Preoperative anemia is emerging as a common and important health problem [5]. This condition is extremely common before surgery associated with colorectal cancer (up to 70 % of patients) [6, 7] and has also been demonstrated in association with increased postoperative morbi-mortality and duration of hospitalization as well as with reduced quality of life [5, 7–10]. Anemia might contribute to complications during and post-surgery [6, 7, 9]. In addition, a low preoperative hemoglobin concentration is one of the major risk factors for transfusion in surgery with moderate to high blood losses [5, 6, 10].
Jointly, the perioperative transfusions unfavorably affect patient outcomes and highlight the risk of postoperative infections, surgical reintervention, recurrence-metastasis, and subsequent cancer-related mortality in colorectal cancer surgery [6, 11, 12]. Therefore, in the context of elective surgery, it is advisable as far as possible to detect and evaluate preoperative anemia early enough to start a suitable treatment.
Anemia in this patient population is largely due to absolute or functional iron deficiency. As such, administering iron prior to surgery is considered an appropriate treatment [13, 14]. Most anemic patients are treated with oral iron, although this therapeutic procedure is slow in terms of iron absorption rate and, in the context of a surgical cancer patient, clearly insufficient to restore early enough the hemoglobin levels and for iron deposit repositioning [14, 15]. However, parenteral administration of iron could increase up to five times the erythropoietic response, which is also associated with a lower frequency of adverse effects in comparison with blood transfusions
[13, 16, 17].[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
Previous studies have shown that treatment with intravenous iron administered at least 1 week before surgery increases hemoglobin levels and, consequently, should reduce the need for transfusion of red blood cell (RBC) units during the perioperative period [18–20]. Therefore, the implementation of intravenous iron administration protocols appears to be an effective and safe strategy for the treatment of preoperative anemia and possibly to reduce transfusion requirements and hospitalization in patients scheduled for elective surgery while meeting cost-effectiveness criteria, whenever an early detection and diagnosis of preoperative anemia is achieved with the objective to implement this therapeutic choice [13, 14, 21]. Many unsustained misconceptions constitute a barrier to this approach [15].[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
With this background in mind and considering the lack of studies with a significant number of patients with colon cancer (CC), this study was designed to evaluate the efficacy of a preoperative administration protocol of intravenous (IV) ferric carboxymaltose (FCM) in colon cancer patients with iron deficiency anemia. Evaluation was completed by assessing the relative reduction in RBC transfusion requirements, post-surgery complications (1 month after surgery), and the total length of hospital stay compared with a retrospective cohort of patients that had not received IV iron.[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
## Material and methods
*In the **Material and methods** section:*
This is a non-interventional study conducted in Spain as a multicenter survey involving two cohorts of consecutive patients with colon cancer and iron deficiency anemia. At diagnosis, the comparator group with no-IV iron treatment (no-IV iron group) was obtained retrospectively while patients treated with ferric carboxymaltose (FCM group) (Ferinject®; Vifor Pharma España S.L.) were recorded prospectively.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
Inclusion of prospective patients was completed between February 2012 and September 2012 in a specialized hospital setting at 11 Spanish hospitals. In the same participant centers, the retrospective cohort was obtained in a sequential manner from surgical intervention 2011 registries and independently of outcomes. All subjects gave their informed consent prior to their inclusion in the study. Approval by the appropriate Institutional Review Boards was obtained.
The study population in both groups included patients aged 18 years or over, diagnosed with colon adenocarcinoma located at least 15 cm above the anal margin, with elective surgery programmed under curative purposes. Iron deficiency anemia was defined according to WHO criteria (hemoglobin (Hb) <13 g/dL in men and <12 g/dL in women) [22], serum ferritin <30 ng/mL, and/or transferrin saturation index <20 %. There were no restrictions on the surgical approach (laparoscopy, open surgery, single port, etc.). The study excluded patients who had rectal neoplasms, emergency or palliative surgery, other linked illnesses associated with anemia such as renal failure or hematological syndromes, tumor recurrence, or a clinical history of blood transfusions during the past 30 days.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
The primary outcome of the study was the relative reduction in perioperative and at 30-day postoperative allogenic RBC transfusion requirements. The secondary end points included the reduction of hospital length of stay (or time to discharge), the incidence of postoperative complications registered during the first month after surgery, the evolution of hemoglobin and iron parameters during the study period, and the proportion of patients with normalized Hb levels and response rate.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
The physicians responsible for the patients’ care, surgery, hospital discharge, and follow-up were unaware of study interventions. When the center diagnosed a patient with colon cancer and concurrent iron deficiency anemia, and who met all the selection criteria, the investigator requested the inclusion of the patient in the program of IV iron with ferric carboxymaltose administration. The total dose was obtained using the ferric carboxymaltose product information dosing scheme [16, 17]. Whenever possible, the FCM administration was between 2 and 4 weeks before the scheduled surgery.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
As standardized hospital clinical practice, a complete blood test was also performed at diagnosis, the day of admission for surgery, at discharge, and at 30 days post-surgery. Peri- and postoperative RBC transfusions were based on the pertinent blood tests and as in hospital clinical practice: being always performed in patients with hemoglobin levels under 7 g/dL, under physician criteria between 7 and 9 mg/dL, and not recommended over 9 g/dL. Normalized Hb levels were established as ≥12 g/dL in women and ≥13 g/dL in men [22]. Patient response was considered when Hb increases ≥1.5 g/dL.
Based on the primary end point, the calculated sample size of the study was approximately 111 patients per group. This estimated size allows the detection of at least a 20 % reduction in RBC transfusion requirements with a statistical power of 90 % and using a 95 % confidence interval (CI). Continuous end points were summarized using descriptive statistics: n, mean, standard deviation, minimum, median and maximum, 95 % CI, and number of missing observations. For discrete end points, the frequency and percentage for each response category were calculated as well as 95 % CI and number of missing data. Missing data was excluded when calculating percentages relative to the total sample.
Student t test or Mann-Whitney U test were used depending on distribution of variables.
## Results
*In the **Results** section:*
## Baseline and clinical characteristics
*In the **Baseline and clinical characteristics** section:*
Baseline, clinical, and surgical characteristics of the patients
Baseline characteristics were similar for both cohorts studied (Table 1). In total, 266 patients were included: 111 received FCM (57.3 % males, mean age 72.9 ± 11.1) and 155 were in the no-IV iron group (55.8 % males, mean age 70.8 ± 10.3). Both groups were similar in terms of tumor location, medical history, and surgical approach and procedures, as well as the proportion of patients with moderate or heavy blood loss (≥50 mL) during surgery (Table 1). Anemia at diagnosis was more pronounced in the FCM group.[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
All patients in the retrospective cohort (no-IV iron group) were receiving different doses and formulations of oral iron supplementation at the time of diagnosis. Within the prospective cohort, the median total FCM dose was 1000 mg iron (mean 1275 ± 430.1 mg) and the administration was 28.5 ± 16.7 days before surgery (mean ± SD).[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
## Primary outcome: transfusion requirements
*In the **Primary outcome: transfusion requirements** section:*
Need for allogenic RBC transfusion (percentage of patients) and mean RBC units transfused
A significantly lower percentage of patients in the FCM group required allogenic RBC transfusion during the study: 9.9 vs. 38.7 % (OR: 5.9, 95 % CI: 2.9–11.1, p < 0.001). This statistically and clinically significant difference was also observed in the individual peri- and post-surgery periods until day 30 and independently of the type of surgical approach (laparoscopic or open surgery) (Table 2). Overall, the mean number of RBC units transfused during the study period was statistically lower in patients treated with FCM (0.2 ± 0.5 vs. 0.8 ± 0.4, p < 0.0001) (see Table 2 for peri- and post-surgery periods). When analyzed by surgery, FCM-treated patients received statistically significant lower mean units of RBC as well (Table 2).
Throughout the entire study period, 9 % of patients in the no-IV iron group received 4 or more RBC units vs. 0 % in the FCM group (p = 0.5817).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
## Evolution of hemoglobin and iron metabolism parameters
*In the **Evolution of hemoglobin and iron metabolism parameters** section:*
Mean (±SD) hemoglobin, hematocrit, MCV, and iron parameters at different time points before and after surgery. Groups were not censored for transfusions[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
Evolution of hemoglobin levels (g/dL) at four time points: diagnosis, hospital admission, discharge, and 30 days post-surgery. Groups were not censored for transfusions. Significant differences between groups are marked with an asterisk for Hb and dagger for serum ferritin (*p < 0.05; **/† p < 0.005; ***/‡ p < 0.001). Ferric carboxymaltose (FCM) was administered during diagnosis period[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
All the hematological parameters, except serum ferritin and transferrin saturation index at hospital discharge, presented significant differences between the FCM and no-IV iron groups at admission time point and 1 month after surgery (Table 3, Fig. 1). Importantly, the FCM group received overall less RBC transfusions, and data has not been censored for transfusions.[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
Figure 1 reflects the evolution of hemoglobin levels in both groups at the four study-defined time points. Despite the lower levels of hemoglobin at diagnosis for the FCM group (9.6 ± 1.4 vs. 10.0 ± 1.2 g/dL, p < 0.005), and much lower RBC transfusion rate, significantly higher hemoglobin concentrations were achieved by the FCM group at hospital admission, discharge, and 30 days post-surgery (Table 3; Fig. 1). Mean total hemoglobin increases significantly favor the FCM group between diagnosis and hospital admission (1.5 vs. 0.5 g/dL; p < 0.0001) and between diagnosis and 30 days post-surgery (3.1 vs. 1.5 g/dL; p < 0.0001). This was also the case when doing the analysis for transfused and non-transfused patients: mean total hemoglobin increase between diagnosis and 30 days post-surgery in transfused patients was 3.5 ± 1.9 FCM vs. 1.4 ± 1.6 no-IV (p < 0.05) and in non-transfused patients was 3.1 ± 1.9 FCM vs. 1. 6 ± 1.8 no-IV (p < 0.001).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
A similar percentage of patients with Hb ≤ 10 g/dL was observed at diagnosis (79 %) in both groups. This rate descends by nearly 20 % for the no-IV group and 30 % in the FCM group at time of hospital admission. The percentage of patients with Hb ≤ 10 g/dL was significantly lower in the FCM group at hospital discharge (61.6 % FCM vs. 75.7 % no-IV iron, p < 0.05) and at 30 days post-surgery (12.0 % FCM group vs. 28.9 % no-IV iron, p < 0.05). Furthermore, the percentage of patients with normalized hemoglobin at 30 days post-surgery was significantly higher in the FCM group vs. no-IV iron (40.0 vs. 26.7 %, p < 0.05).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
Figure 1 also shows the evolution of serum ferritin levels. At 30 days after-surgery, the average FCM-treated patient presented no recognizable signs of iron deficiency anemia (being the mean of Hb: 12.6 g/dL [≥12 g/dL]; serum ferritin: 218 ng/mL [≥30 ng/mL]; and saturation transferrin index: 25.1 % [≥20 %]) compared with the no-IV group that did not reach normalized mean values (Table 3).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
Percentage of hemoglobin responders—defined as those with an Hb increase of ≥1.5 g/dL—at hospital admission and 30 days post-surgery with respect to Hb diagnosis levels. Data was not censored for transfusions
Figure 2 illustrates how the percentage of hemoglobin responders (Hb increase of ≥1.5 g/dL) significantly increased in the group treated with FCM compared to the no-IV group: 48.1 vs. 20.0 % between diagnosis and hospital admission (p < 0.0001) and 80.0 vs. 48.9 % between diagnosis and 30 days after surgery (p < 0.0001).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
## Post-surgery complications
*In the **Post-surgery complications** section:*
Incidence of post-surgery complications
A numerically lower total number of reinterventions and complications related to surgery (including suture dehiscence, paralytic ileus, hemoperitoneum, rectal bleeding, thromboembolism, etc.) at 30 days after surgery were seen in the FCM group in comparison with no-IV patients: 20.7 vs. 26.5 % (OR = 1.4; 95 % CI: 0.8–2.4; p = 0.311) (Table 4).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
## Impact on hospital stay
*In the **Impact on hospital stay** section:*
Mean length of hospital stay measured from the day of surgery until the hospital discharge
The length of hospital stay, measured from the day of surgery until the day of discharge, is shown in Fig. 3. The FCM group had a significantly shorter mean length of hospital stay: 8.4 ± 6.8 days compared to the no-IV iron group (10.9 ± 12.4 days) (p < 0.001). Overall, patients that required RBC transfusion had a longer hospital stay than patients without transfusion (FCM 9.2 ± 9.4 days vs. no-IV iron 12.0 ± 13.0 days; p < 0.005).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
## Safety of ferric carboxymaltose
*In the **Safety of ferric carboxymaltose** section:*
From the FCM-treated patients, no deaths, hypersensitivity, or other serious adverse drug reactions were observed.[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
## Discussion
*In the **Discussion** section:*
Preoperative anemia in patients with surgical bleeding risk associated with colorectal cancer has proved very prevalent and is an independent risk factor for the requirements of allogeneic RBC transfusion [12]. Under this global perspective, and considering the lack of controlled studies with a significant number of patients, our prospectively defined and treated population demonstrated a significant clinical benefit for the preoperative administration of an IV iron supplementation using ferric carboxymaltose (FCM) when compared to a retrospective cohort (with similar baseline characteristics). This overall fourfold reduction of transfusions (and hence associated risk factors) was achieved using FCM which was well tolerated in this group of patients and demonstrated a favorable benefit-risk profile. Furthermore, the observed needs for transfusion were five times (pre-and intraoperative) and almost three times higher (postoperative) for the group that had not received IV iron.
Interestingly, the patients in this cohort were all receiving oral iron supplementation at time of diagnosis.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
Similar decreases in transfusion requirements for FCM-treated patients have been reported in a study that compares the pre-operative administration of FCM with the IV iron sucrose administration in a small cohort of 45 patients treated for anemia in colon cancer resection [18]. In this study, Bisbe and collaborators observed a noticeable reduction in the percentage of patients that required RBC transfusion in the FCM group (from 40 % with iron sucrose to 7 % with FCM), as well as fewer iron administration sessions (as FCM may be administered at a single dose of 1000 mg iron per session vs. only 200 mg for iron sucrose). Analogously, similar trends in reducing allogenic RBC transfusion requirements were observed with FCM in a three-cohort retrospective study that included a total number of 154 patients with GI cancer submitted to laparoscopic resection (gastrectomy, right or left colectomy, or rectum resection) [23].
Even though initial levels of Hb were higher in the non-anemic group of patients, the anemic FCM group required similar mean RBC units transfused, and both were significantly lower (p < 0.001) than in the anemic no-IV-treated group (0.5, 0.4, and 2.4, respectively).[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
In contrast, a previous randomized placebo-control clinical trial (n = 60) did not find support for the use of intravenous iron sucrose (600 mg iron in two divided doses, 14 days before surgery) as a preoperative supplementation to reduce the likelihood of allogenic RBC transfusion for patients undergoing resectional surgery for colorectal cancer (19.2 % placebo vs. 5.9 % in iron sucrose; p = 0.335) [24]. However, although not reaching statistical significance, the number of transfusions was still numerically higher in the placebo group, and it might be hypothesized that the administered iron doses were not sufficient to meet the iron deficit in these patients.[](https://www.ncbi.nlm.nih.gov/mesh/D000077605)
Within our study, and even despite significantly fewer patients transfused, at 1 month post-surgery, as overall, the FCM-treated patients did not reflect signs of anemia, as well as no signs of iron deficiency (mean ferritin and transferrin saturation index were at normal values). Our data in a large patient group extends to the 30-day post-surgery period the previously found evidence for improved hemoglobin concentrations and reduced transfusions in gastrointestinal cancer patients with anemia treated with FCM [23]. Moreover, in those anemic patients included in our study who have received FCM, almost one out of two reached normalized hemoglobin levels after 1 month from the colon cancer resection and despite their lower mean hemoglobin levels at diagnosis and their lower rate of transfusion requirement, while in patients with no-IV iron administration, this percentage was only around 26 %. Our data further supports the preoperative treatment with ferric carboxymaltose in anemic colon cancer patients who are planned for surgery, with benefits extending until 30 days post-surgery.[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
In addition to the reduction in transfusion requirements and the improved iron parameters, our study demonstrates for the first time the benefit of a preoperative FCM administration strategy in the peri- and post-surgery periods when treating this kind of patients. Specifically, importance should be given to the observed relation to the mean hospital length of stay which was significantly reduced with the FCM administration by a mean total of 2.5 days. Likewise, the postoperative benefits of intravenous iron administered previously to the surgery have been reported recently in other patient profiles, i.e., those submitted to non-cardiac surgery, gynecological tumor resection, cardiac valve replacement, and orthopedic procedures in terms of reduction of post-surgery complications and reduced hospital length of stay [10, 16, 19, 20, 25]. Finally, it has ruled out any link between the surgical approach and possible differences between groups with regard to complications in the 30-day post-surgery period.[](https://www.ncbi.nlm.nih.gov/mesh/D007501)
As a result of these improvements, pretreatment with FCM in anemic colon cancer patients seems to be a cost-effective intervention. The incremental costs of IV iron treatment are very likely to be offset by the cost savings due to reductions in length of stay and transfusion rates. The extent of this economic advantage will be assessed in a further within-trial analysis. In a previous Spanish cost analysis taking into account both drug acquisition costs of FCM and iron sucrose as well as administration costs in anemic patients undergoing major elective surgery, the final cost benefit of FCM administration per treatment compared with the common IV iron sucrose therapy was demonstrated by providing a mean of €63 savings per patient FCM treatment [18].[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
As strengths of our study, it should be mentioned the number of patients considered for the analysis in the predefined groups as well as their homogenous baseline and clinical characteristics in both groups. It is also important to highlight that the significant trends observed in the measured outcomes were achieved after similar laboratory values at diagnosis and surgical characteristics in both groups, although the FCM group showed a slightly reduced hemoglobin at baseline. The present study has several limitations that warrant acknowledgment. Specifically, the non-randomized design limits the interpretation of the results. However, these patients represent “real-life” clinical practice, and hence, this is also a strength of this research.
Future randomized controlled trial(s) focused on the use of FCM treatment in preoperative anemic colorectal cancer patients vs. a similar active supplement may aid to confirm our findings.[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
In conclusion, our data demonstrate that preoperative ferric carboxymaltose treatment in iron-deficient colon cancer patients with anemia significantly reduced the length of hospitalization, decreased both the perioperative and postoperative allogenic RBC transfusion requirements, showed no signs of iron deficiency anemia at 30 days post-surgery, and was safe and well tolerated. Based on reduction of RBC transfusions and length of hospital stay, it can be assumed that the preoperative administration of ferric carboxymaltose in iron-deficient colon cancer patients with anemia undergoing surgery could result in significant cost savings.[](https://www.ncbi.nlm.nih.gov/mesh/C522335)
The Colon Cancer Study Group also includes
Jesús-Alberto Varela (Hospital Gregorio Marañón, Madrid, Spain), Teresa Broquetas (Hospital del Mar, Barcelona, Spain), Francisco-Javier Esteban (Hospital La Paz, Madrid, Spain), Luis Ferrer (Hospital General Universitario, Valencia, Spain), Laura Sanchis (Hospital General Universitario, Valencia, Spain), Federico Argüelles (Hospital Virgen de la Macarena, Sevilla, Spain), and Montserrat Andreu (Hospital del Mar, Barcelona, Spain).
# Compliance with ethical standards
*In the **Compliance with ethical standards** section:*
## Disclaimer
*In the **Disclaimer** section:*
Partial results of the present study were presented at the Digestive Disease Week 2013 held in Orlando, FL (USA), and at the 16th Annual Meeting of the Spanish Association of Gastroenterology (AEG) 2013 held in Madrid (Spain).
## Funding
*In the **Funding** section:*
The present work was funded by Vifor Pharma España SL. Medical writing support was provided by TFS Develop and was funded by Vifor Pharma España SL.
## Conflict of interest
*In the **Conflict of interest** section:*
Mercedes Cucala is an employee of Vifor Pharma España. The other authors declare no conflict of interest.
## Ethical approval
*In the **Ethical approval** section:*
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards (2013 Fortaleza).
## Informed consent
*In the **Informed consent** section:*
Informed consent was obtained from all individual participants included in the study.
# References
*In the **References** section:*
|
# Introduction
Effects of [β-mercaptoethanol](https://www.ncbi.nlm.nih.gov/mesh/D008623) on in vitro maturation and [glutathione](https://www.ncbi.nlm.nih.gov/mesh/D005978) level of buffalo oocytes
# Abstract
*In the **Abstract** section:*
Aim:
The present study was carried out to evaluate the effect of supplementation of β-mercaptoethanol (β-ME) on in vitro maturation rate and glutathione (GSH) level of buffalo oocyte[s.](https://www.ncbi.nlm.nih.gov/mesh/D008623)ncbi.nlm.nih.gov/mesh/D008623)
Materials and Methods:
Oocytes were recovered from buffalo’s ovaries collected from government approved slaughter house (near Kamela darwaza, Surat) of Surat Municipal Corporation. The obtained oocytes were in vitro matured in maturation media supplemented with 0 μM (117 oocytes), 100 μM (46 oocytes) and 200 μM (42 oocytes) concentration of β-ME. After 24 h of incubation, maturation rate of oocytes and intra-cellular GSH level were determined.[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
Results:
The results showed that the presence of β-ME did not influence (p>0.05) the oocyte maturation rate. However, GSH level increased significantly (p<0.05) in matured oocytes when supplemented with 100 μM and 200 μM β-ME (6.19±0.10 and 6.37±0.20 pmol/oocyte) as compared to control media (4.68±0.26 pmol/oocyte).[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
Conclusion:
It was concluded that β-ME may have a potential to increase the meiotic maturation of in vitro cultured oocytes and protect it from oxidative damage.[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Buffalo (Bubalus bubalis) plays a prominent role in rural livestock production. Problems like late onset of reproductive maturity, seasonality of breeding, late estrus and long calving interval have been attributed to poor reproductive performance of this species [1]. To cope up with these problems, use of modern biotechnologies, such as in vitro fertilization (IVF) and embryo production are required instead of conventional breeding programs [2].
A major factor affecting in vitro mammalian embryo development is increased oxidative stress [3], which is due to high lipid content of buffalo oocytes [4]. Higher amount of reactive oxygen species (ROS) can alter cellular molecules; induce developmental block, apoptosis and fragmentation of embryos [5]. It has been demonstrated that addition of low molecular thiol compound such as β-mercaptoethanol (β-ME) and cysteamine to the maturation medium causes an increase in intracellular glutathione (GSH) synthesis [6,7] and leads to low oxidative stress in many species [8].[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
β-ME and GSH both improve the cell survival by decreasing apoptotic cell death under “redox” state [9]. GSH directly influences cell death, while β-ME has an indirect effect by supporting increase in intracellular GSH level [5]. Buffalo oocytes can synthesize de novo during in vitro maturation (IVM) [7] and β-ME increases cumulus cells expansion which help in GSH synthesis [10].
So far as buffalo oocytes are concerned, meagre studies have been carried to know the effects of β-ME on cumulus expansion of oocyte and intracellular GSH content.[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
Therefore, the present study was carried out to evaluate the effect of supplementation of β-ME on IVM rate and GSH level of buffalo oocytes.[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
## Materials and Methods
*In the **Materials and Methods** section:*
## Reagents and media
*In the **Reagents and media** section:*
All the chemical and media used in the present study were purchased from Sigma (USA).
## Collection of ovaries
*In the **Collection of ovaries** section:*
Ovaries were collected from sexually matured buffaloes immediately after slaughter from nearby government approved slaughter house (near Kamela darwaza, Surat) of Surat municipal corporation and transported to the laboratory in sterile normal saline (NSS:0.85%) solution fortified with antibiotic (50 μl/L Gentamicin) at 38-39°C temperature. At laboratory, ovaries were washed in 70% ethanol for 1 min. to reduce contamination followed by washing in 0.85% NaCl twice for 1 min.[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## Oocyte recovery
*In the **Oocyte recovery** section:*
After final washing, cumulus oocyte complexes (COCs) were aspirated from non-atretic surface follicles (2-8 mm) using 18-guage needle connected to a 5 ml sterile syringe containing oocyte collection media. Further aspirated oocytes were searched and graded as per Khandoker et al. [11]. A, B and C grade oocytes were used for IVM.
## Maturation of oocytes
*In the **Maturation of oocytes** section:*
After final washing with oocyte collection media, oocytes of A, B and C grade were equally distributed in three groups viz: control Group I (117 oocytes) and treatment Groups-II (46 oocytes) and III (42 oocytes). In control group, only basic maturation media (TCM-199 supplemented with 0.2 mM sodium pyruvate, 10% fetal bovine serum, 3 mg/ml bovine serum albumin and 10 IU/ml hCG) was used while in treatment Group-II and III control media supplemented with 100 µM and 200 µM β-ME respectively was used. Before transferring to maturation media, oocytes were washed once with respective maturation media. Each group was individually placed in 50 μl droplet of maturation medium containing 5-10 oocytes covered with mineral oil in a sterile petridish and kept at 38.5°C, 5% CO2 and 95% humidified air in CO2 for 24 h.[](https://www.ncbi.nlm.nih.gov/mesh/D011773)
## GSH content and maturation rate of oocytes
*In the **GSH content and maturation rate of oocytes** section:*
Maturation of oocyte assessed on the basis of their cumulus layer expansion as per Khandoker et al. [11]. GSH level estimation of oocytes was carried out in all groups after 24 h of maturation. Oocytes were carefully denuded by repeated pipetting, washed several times in 1x phosphate buffered saline and 10-12 oocytes from each group were stored at −20°C in eppendorf for further use. On the day of assay, the samples were thawed and 500 µl of ice-cold 5% metaphosphoric acid added to each sample.
Vortexing performed for 3-5 min. After that sonication was performed for 5 min . Samples were then centrifuged (10 min; 3000 × g) in cryo-centrifuge machine at 4°C and 100 µl of supernatant was recovered. Further estimation of GSH was done with the help of GSH Assay Kit (Calbiochem® USA).[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
## Statistical analysis
*In the **Statistical analysis** section:*
Data pertaining to oocyte maturation were analyzed by SPSS software performing Chi-square test and for data pertaining to GSH level by one-way ANOVA among control and treatment groups. A significance level of p<0.05 was used throughout this study.[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
## Results
*In the **Results** section:*
The effects of supplementation of β-ME on IVM rate and GSH level of buffalo oocytes is presented in Table-1. The results showed that the presence of β-ME did not influence the oocyte maturation rate although higher maturation rate was observed in β-ME-100 and 200 μM groups as compared to control. However, intra-cellular GSH level increased significantly (p<0.05) in the presence of 100 and 200 μM β-ME (6.19±0.10 and 6.37±0.20 pmol/oocyte) as compared to control (4.68±0.26 pmol/oocyte).[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
Effects of β-ME on of oocytes and its GSH level in buffalo.[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
βME=βmercaptoethanol, IVM=In vitro maturation, GSH=Glutathione[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
## Discussion
*In the **Discussion** section:*
Antioxidants function as autocrine and paracrine factors that influence growth, differentiation and retardation of developing follicles. Presence of GSH, β-ME is beneficial for follicle development, and there may be an interaction between exogenous antioxidant and developing follicles. Exogenous antioxidants influence follicle growth and nuclear maturation of intra-follicular oocytes. β-ME is a thiol compound, acting as an antioxidant and promotes embryo development [12,13].[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
The results of the present study revealed that the addition of β-ME (100 µM and 200 µM) to the maturation medium did not increase maturation rate, as also reported in porcine oocytes [14]. However, higher numbers of M-II oocytes were found when denuded oocytes were cultured in maturation medium supplemented with 25 µM β-ME [15]. Similarly, supplementation of β-ME positively influence percentage of oocytes from pre-pubertal Boer goats progressing to metaphase II stage during IVM [16]. It has also been reported that supplementation of β-ME in maturation media have positive effect on expansion of COCs and maturation rate of oocytes of bovine as well as pig
[10,17], apart for impact on fertilization rate [18] and improves embryo development rate [19,20].[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
The effect of β-ME may have been mediated through the synthesis of GSH which is known to play an important role in protecting the cell or embryos from oxidative damage. Exogenous β-ME is able to increase GSH synthesis by reducing cystine to cysteine [21] and increased GSH level promotes embryonic development by maintaining intracellular redox state [22]. In the present study also, significantly higher GSH level was observed in groups having maturation medium supplemented with 100 µM and 200 µM β-ME than control groups. Similar results were found in bovine oocytes and embryos [7,15,22], as well as pig oocytes [12]. GSH itself plays a critical role in protecting the cell from oxidative damages [23-26].
These results warrant the supplementation of exogenous β-ME in basic oocyte maturation medium for improvement of IVF.[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
## Conclusion
*In the **Conclusion** section:*
From the present study, it can be concluded that addition of β-ME at different concentration in maturation media helps in the synthesis of GSH that protects the degeneration of oocytes from ROS during IVM and might enhance the process of maturation of oocyte.[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
## Authors’ Contributions
*In the **Authors’ Contributions** section:*
SSC along with GP designed the experiment and PAP conducted the experiment with the help of ABO. GP and VKS helped in analyzing the data and preparing the manuscript. SSC, GP, VKS and PAP reviewed the manuscript. All authors read and approved the final manuscript.
## Competing Interests
*In the **Competing Interests** section:*
The authors declare that there is no conflict of interests regarding the publication of this paper.
# References
*In the **References** section:*
|
# Introduction
Using a Genetically Encoded Sensor to Identify Inhibitors of Toxoplasma gondii [Ca2+](https://www.ncbi.nlm.nih.gov/mesh/D002118) Signaling*
# Abstract
*In the **Abstract** section:*
The life cycles of apicomplexan parasites progress in accordance with fluxes in cytosolic Ca2+. Such fluxes are necessary for events like motility and egress from host cells. We used ge[neti](https://www.ncbi.nlm.nih.gov/mesh/D002118)cally encoded Ca2+ indicators (GCaMPs) to develop a cell-based phenotypic screen for compounds that modulate [Ca2+](https://www.ncbi.nlm.nih.gov/mesh/D002118) signaling in the model apicomplexan Toxoplasma gondii. In doing so, we took advantage of t[he p](https://www.ncbi.nlm.nih.gov/mesh/D002118)hosphodiesterase inhibitor zaprinast, which we show acts in part through cGMP-dependent protein kinase (protein kinase G; [PKG) to r](https://www.ncbi.nlm.nih.gov/mesh/C011145)aise levels of cytosolic Ca2+. We define the pool of Ca2+ regulated by PKG to be a neutral store distinct from the endop[lasm](https://www.ncbi.nlm.nih.gov/mesh/D002118)ic reticulum. Screening [a li](https://www.ncbi.nlm.nih.gov/mesh/D002118)brary of 823 ATP mimetics, we identify both inhibitors and enhancers of Ca2+ signaling. Two such compounds c[ons](https://www.ncbi.nlm.nih.gov/mesh/D000255)titute novel PKG inhibitors and prevent zaprinast from i[ncre](https://www.ncbi.nlm.nih.gov/mesh/D002118)asing cytosolic Ca2+. The enhancers identified are capable of releasing int[racellula](https://www.ncbi.nlm.nih.gov/mesh/C011145)r Ca2+ stores independently[ of ](https://www.ncbi.nlm.nih.gov/mesh/D002118)zaprinast or PKG. One of these enhancers blocks parasite egress an[d in](https://www.ncbi.nlm.nih.gov/mesh/D002118)vasion and shows strong a[ntiparasi](https://www.ncbi.nlm.nih.gov/mesh/C011145)tic activity against T. gondii. The same compound inhibits invasion of the most lethal malaria parasite, Plasmodium falciparum. Inhibition of Ca2+-related phenotypes in these two apicomplexan parasites suggests that depletion of intracel[lula](https://www.ncbi.nlm.nih.gov/mesh/D002118)r Ca2+ stores by the enhancer may be an effective antiparasitic strategy. These results establish[ a p](https://www.ncbi.nlm.nih.gov/mesh/D002118)owerful new strategy for identifying compounds that modulate the essential parasite signaling pathways regulated by Ca2+, underscoring the importance of these pathways and the therapeutic potential of their inhi[biti](https://www.ncbi.nlm.nih.gov/mesh/D002118)on.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Apicomplexan parasites, such as Toxoplasma gondii and Plasmodium spp., the causative agents of toxoplasmosis and malaria, require changes in cytosolic Ca2+ concentrations to egress from host cells and move within the infected organism (1–6). These pathways therefore hold tremendous therapeutic potential, not only due to their importance in parasite biology but because of their divergence from similar pathways in host cells. Compounds targeting Ca2+ signaling in parasites have been shown to be effective antiparasitics (discussed below). However, few of the molecules involved in regulating Ca2+ homeostasis and signaling have been identified in parasites, and their interplay is only evident in live cells. This has created a need for new methods to study Ca2+ signaling pathways in apicomplexan parasites, with the hope of defining the essential components and identifying novel inhibitors.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
The mechanisms for Ca2+ entry into the cytoplasm and the physiologically relevant sources of Ca2+ remain poorly defined in apicomplexan parasites. Ca2+ can be mobilized from the parasite's intracellular stores, or it can be drawn from the environment. Current evidence points toward intracellular stores being sufficient for parasites to move between cells (6–8), although virulence of T. gondii is enhanced by extracellular Ca2+ (9). The best studied of these intracellular stores is the endoplasmic reticulum (ER).2 This organelle is a highly networked, dynamic structure (10) that has been shown to constitute multiple spatially independent Ca2+ stores in some cell types (11). Such compartmentalization has also been hypothesized to occur in T. gondii (4).
Mammalian cells store Ca2+ in endosomes, lysosomes (12), and the Golgi (13), in addition to the ER (14). Some alveolates, like Paramecium, additionally contain a network of alveolar sacs that sequester Ca2+ in an ATP-dependent manner, with physiologically relevant affinities (15, 16). Whether the inner membrane complex, an apicomplexan structure homologous to alveolar sacs, also stores Ca2+ remains to be determined. Like many other eukaryotes, apicomplexans also possess acidic vacuoles known as acidocalcisomes that contain Ca2+ in complex with pyrophosphate and polyphosphates. This Ca2+ can be released pharmacologically, but the function of acidocalcisomes remains unclear in apicomplexans (17).
A final acidic Ca2+ store described in T. gondii is the plantlike vacuole or vacuolar compartment. The plantlike vacuole is an acidic organelle that releases Ca2+ upon treatment with l-phenylalanine-naphthylamide (GPN), which in other systems causes ion leakage from lysosomal compartments (18). Although implicated in ionic homeostasis, these phenotypes have not been linked to the plantlike vacuole's function as a Ca2+ store (18). Due to the lack of characterized regulatable Ca2+ channels, it remains an open question which of these Ca2+ sources are involved in parasite motility and invasion.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Recent evidence suggests that PKG may play a role in regulating parasite Ca2+. In Plasmodium berghei, PKG has been proposed to influence Ca2+ homeostasis, thereby regulating egress from host cells. Activation of PKG leads to change in the levels of the lipid precursors of inositol 1,4,5-triphosphate (IP3). This has been hypothesized to increase IP3, which causes release of Ca2+ from the ER through binding to the IP3 receptor, although such a channel remains to be identified in apicomplexan parasites (19). PKG is also known to regulate egress in T. gondii, although its effect on parasite Ca2+ has not been determined (20).
In both parasites, PKG can be pharmacologically activated using the mammalian phosphodiesterase inhibitor zaprinast (20, 21). In Plasmodium, zaprinast treatment leads to an increase in cyclic GMP levels that presumably activates PKG (22). Further characterization of zaprinast's mechanism of action could therefore shed light on Ca2+ signaling.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Targeting the downstream effectors of Ca2+ signaling has been shown to hold therapeutic value against T. gondii. Compounds targeting Ca2+-dependent protein kinase 1, a regulator of egress and invasion, reduce proliferation in cell culture and cyst burden in the brains of T. gondii-infected mice (23, 24). Drugs targeting Ca2+-related processes are appealing not only because of this historical success but also because many proteins involved in parasite Ca2+ signaling are sufficiently divergent from their mammalian counterparts to enable the design of drugs with minimal off-target effects. For example, protein kinase G (PKG), which regulates egress, invasion, and motility in T. gondii and Plasmodium spp. (19, 20, 25, 26), is sufficiently different from mammalian PKG to be selectively inhibited (27).
Similarly, the Ca2+-dependent protein kinases lack homologues in mammalian cells (28), making them attractive drug targets.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
In this study, we use both chemical and genetic Ca2+ indicators to define the regulatory circuits that mediate Ca2+ release in T. gondii and identify small molecules that modulate this process. We determine the effect of T. gondii PKG on cytosolic Ca2+ following treatment with zaprinast and characterize the source of the Ca2+ released in this process as a neutral store distinct from the ER. Using genetically encoded Ca2+ indicators recently established in T. gondii (29), we develop a cell-based phenotypic screen that allows us to monitor Ca2+ signaling in live cells without the technical challenges of conventional chemical Ca2+ indicators. Using this platform, we have been able to identify, in an unbiased manner, compounds that interfere with Ca2+ signaling. In contrast to enzyme-based assays, this system enables us to probe a broader swath of parasite biology.
Recent analysis indicates that such phenotypic screens are more likely to lead to clinically approved drugs than the far more prevalent molecular target-based approaches (30). Our screen identified two novel PKG inhibitors that abrogate the effect of zaprinast, as well as two compounds that increase cytosolic Ca2+ through an independent pathway. From the latter, one compound blocks invasion of both T. gondii and Plasmodium falciparum. Our results demonstrate the potential of this strategy to explore parasite Ca2+ signaling and identify new compounds with antiparasitic potential against multiple apicomplexan parasites.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Experimental Procedures
*In the **Experimental Procedures** section:*
## Strain Construction and Maintenance
*In the **Strain Construction and Maintenance** section:*
Recombinant human T. gondii strain RH parasites were maintained in human foreskin fibroblasts (HFFs) grown in DMEM supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 10 μg/ml gentamicin. PKG-T and PKG-M alleles were constructed as described previously (31). GCaMP5 was amplified from pCMV-GCaMP5G (32) with primers containing NsiI and PacI restriction sites (forward primer, 5′-gcg atg cat cct ttt tcg aca aaa tgg gtt ctc atc atc atc atc atc; reverse primer, 5′-gcg tta att aat cac ttc gct gtc atc att tg) and cloned directionally, replacing the CAT gene in pSAG1/2-CAT (33) to generate pSAG1-GCaMP5. Recombinant human parasites were co-transfected with pSAG1-GCaMP5 and pSAG1/2-CAT and selected with chloramphenicol (40 μm), and clones were isolated by limiting dilution. The GCaMP6f strains was similarly derived, as described previously (29).
Both GCaMP strains were maintained under selection to prevent loss of the transgene. The GFP-expressing strain was kindly provided by Jeroen P. J. Saeij (34).[](https://www.ncbi.nlm.nih.gov/mesh/D005839)
## Store Activation and Cpd1 Inhibition Experiments with GCaMP6f
*In the **Store Activation and Cpd1 Inhibition Experiments with GCaMP6f** section:*
GCaMP6f-expressing T. gondii were suspended at 2 × 107 parasites/ml in basal Ca2+ buffer (140 mm NaCl, 10 mm potassium gluconate, 2.7 mm MgSO4, 2 mm glucose, 250 μm EGTA, 85 μm CaCl2, 10 mm HEPES, pH 7.3) or extracellular Ca2+ buffer (140 mm NaCl, 10 mm potassium gluconate, 2.7 mm MgSO4, 2 mm glucose, 1 mm CaCl2, 10 mm HEPES, pH 7.3), supplemented with 1% FBS when noted. For the Cpd1 inhibition experiments, parasites were suspended in Ringer's solution (115 mm NaCl, 3 mm KCl, 2 mm CaCl2, 1 mm MgCl2, 3 mm NaH2PO4, 10 mm HEPES, 10 mm glucose, 1% FBS). In all cases, 100 μl of suspended parasites were applied to each well of a Cell Carrier 96-well plate (PerkinElmer Life Sciences). For store activation experiments, parasites were incubated on ice for 5 min before the addition of Enh1 or Enh2 (10 μm final concentration), zaprinast (100 μm final), or DMSO (0.3% final) suspended in the same buffers as the parasites to which they were added. To examine the effect of Cpd1 on zaprinast or Enh1 treatment of GCaMP6-expressing parasites, fluorescence was recorded for 100 s before adding Cpd1 to a concentration of 1.2 μm.
Fluorescence was recorded for an additional 300 s before treatment with 10 μm Enh1 or 100 μm zaprinast. Parasites were incubated on ice for 5 min before the addition of zaprinast to slow down the response and facilitate observation of peak fluorescence. Fluorescence was read with an excitation wavelength of 485 nm and an emission wavelength of 528 nm every 10 s in a BioTek Cytation 3. The assay plate was shaken for 1 s before each read.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Compound Screen
*In the **Compound Screen** section:*
GCaMP5-expressing parasites were suspended in Ringer's solution at 4 × 107 parasites/ml, and 50 μl of parasites were applied to each well of a 96-well plate (Costar, catalog no. 3631). Parasites were treated with compounds from the GSK libraries PKIS and PKIS2 at 13.3 μm or 1.33% DMSO as a vehicle control and then incubated at 37 °C with 5% CO2 for 10 min. Parasites were then treated with 100 μm zaprinast or 0.1% DMSO as a vehicle control and incubated at 37 °C with 5% CO2 for an additional 4 min before measuring fluorescence with excitation and emission wavelengths of 485 and 525 nm, respectively, on a SpectraMax M3 (Molecular Devices). Basal fluorescence, measured from untreated parasites, was subtracted from all values, and results were expressed as -fold change from parasites treated only with zaprinast. Z′ factors were determined from (i) parasites treated with zaprinast versus untreated parasites and (ii) parasites treated with a final concentration of 1 μm Cpd2 or a vehicle control followed by zaprinast, using the formula, Z′ = 1 − 3(σp + σn)/(|μp − μn|), where σp and σn indicate S.D. values of positive and negative controls, and μp and μn indicate the means of positive and negative controls, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
## In Vitro PKG Assays
*In the **In Vitro PKG Assays** section:*
Recombinant T. gondii PKG was expressed using a baculovirus system. Synthetic DNA coding for the protein was amplified and subcloned into the pFBOH-MHL vector, which confers an N-terminal His6 tag with a tobacco etch virus cleavage site. The resulting plasmid was transformed into DH10BacTM Escherichia coli competent cells to produce recombinant viral DNA. P3 viral stocks were used to infect Sf9 insect cells grown in HyQ SFX insect serum-free medium (Thermo Fisher Scientific). The culture was incubated at 27 °C and shaken at 100 rpm.
After 60–72 h, the cells were harvested. The His-tagged T. gondii PKG samples were purified by affinity chromatography and size exclusion chromatography using an ÄKTAxpress system equipped with a SuperdexTM 200 10/300 column (GE Healthcare, Mississauga, Canada). In vitro kinase assays were performed using a PKG assay kit (CycLex) as per the manufacturer's instructions. Inhibitors were tested against 6.76 nm recombinant PKG, which was active within the linear range of the assay.[](https://www.ncbi.nlm.nih.gov/mesh/C033223)
## Structural Analysis
*In the **Structural Analysis** section:*
Structures of various protein kinases in complex with pyrazolopyridazine (PP) and oxindole derivatives were obtained from the RCSB Protein Data Bank (35) and aligned on the basis of their kinase domains. For the PP analysis, the CDK2 structures 3EID and 3EJ1 (36) and the p38 MAPK structure 3GCP (37) were aligned to the ERK2 structure 1WZY (38). For the oxindole analysis, the PDK1 structures 2PE0 and 2PE2 (39) and the Alk5 structure 2X7O were aligned to the Nek2 structure 2JAV (40).[](https://www.ncbi.nlm.nih.gov/mesh/C118531)
## Lactate Dehydrogenase Release Egress Assays
*In the **Lactate Dehydrogenase Release Egress Assays** section:*
Confluent HFF monolayers in 96-well plates were infected with 5 × 104 parasites/well. The HFF monolayer was washed once with Ringer's solution ∼18 h later, and 50 μl of Ringer's solution was applied to each well. Drugs suspended in Ringer's solution were applied at 1.33 times the indicated concentrations. 1.33% DMSO in Ringer's solution was used as a vehicle control.
Cells were incubated at 37 °C with 5% CO2 for 20 min, stimulated with 500 μm zaprinast or 0.5% DMSO (vehicle), and incubated again for an additional 5 min. The cells were centrifuged at 400 × g for 5 min, before collecting 50 μl of each supernatant. The lactate dehydrogenase levels in the supernatant samples were quantified using the CytoTox 96 cytotoxicity assay (Promega) as per the manufacturer's instructions.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
## Plaque Assays
*In the **Plaque Assays** section:*
T. gondii parasites were suspended in growth medium supplemented with Enh1, zaprinast, or DMSO alone at the indicated concentrations. The amount of DMSO in all treatment groups was normalized to 0.1%. Parasites were incubated at 37 °C with 5% CO2 for 20 min before infecting HFF monolayers in 6-well plates using 3 ml carrying 100 parasites/well. Medium containing the highest concentration of each drug was also applied to HFF cells in the absence of parasites to assess the effects of the drugs on host cell viability. The parasites were allowed to plaque for 8 days before fixing with 70% ethanol and staining with 0.1% crystal violet.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
## Lytic Assays
*In the **Lytic Assays** section:*
Drugs were suspended at twice the indicated concentrations in growth medium and mixed with an equal volume of parasites at an initial concentration of 106 parasites/ml. The maximum concentration of DMSO was added as a vehicle control: 1% DMSO in comparisons with zaprinast and 0.013% in comparisons with Enh1. Parasites were preincubated with the compounds for 20 min at 37 °C with 5% CO2, and then 200 μl/well (105 parasites) was added to host cell monolayers in 96-well plates. Assay plates were incubated at 37 °C with 5% CO2 for 3 days and then fixed in 70% ethanol and stained with 0.1% crystal violet. Absorbance at 590 nm was read as a measure of host cell lysis.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
## Cell-wounding Assays
*In the **Cell-wounding Assays** section:*
Parasites suspended in Ringer's solution at 5 × 105 parasites/ml were pretreated with varying concentrations of zaprinast (as indicated) for 20 min at 37 °C with 5% CO2. 105 parasites/well were then applied to confluent host cell monolayers, and plates were centrifuged at 290 × g for 5 min. Following 1 h at 37 °C with 5% CO2, plates were centrifuged again at 500 × g for 5 min. 50 μl of supernatant were collected from each well, and lactate dehydrogenase was quantified using the CytoTox 96 cytotoxicity assay (Promega).[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
## Video Microscopy Egress Assays
*In the **Video Microscopy Egress Assays** section:*
Host cell monolayers in Cell Carrier 96-well plates (PerkinElmer Life Sciences) were infected with 5 × 104 GFP-expressing parasites/well. 18 h postinfection, the medium was exchanged for Ringer's solution, and the intracellular parasites were treated with 10 μm Enh1, 500 μm zaprinast, or 0.5% DMSO. Images were acquired every 10 s for 30 min. To measure the effect of Enh1 on zaprinast and A23187-induced egress, the procedure was repeated, adding either 12.5 μm Enh1 or 0.13% DMSO and imaging for 10 min, before stimulation with 500 μm zaprinast or 1 μm A23187 (Calbiochem) and imaging for an additional 10 min. In all cases, images were acquired with a ×4 objective on a Cytation3 reader (BioTek) using excitation and emission wavelengths of 485 and 528 nm, respectively.
Intact vacuoles were defined as objects of at least 78 μm2 with a circularity of at least 0.5, as determined in Fiji after default thresholding (41).[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
## Video Microscopy
*In the **Video Microscopy** section:*
Parasites expressing GCaMP5 were transfected with p30-DsRed and used to infect host cell monolayers in 3-cm glass bottom dishes (MatTek). Approximately 20 h after infection, medium was exchanged for Ringer's solution, and parasites were imaged on a Nikon Eclipse Ti epifluorescence microscope, equipped with an enclosure heated to 37 °C. Images were acquired every 5 s for 10 min following the addition of zaprinast to a final concentration of 100 μm. To quantify changes in fluorescence, videos were analyzed in Fiji to measure the average fluorescence intensity in specific circular regions of interest 10 μm in diameter. To examine the effects of zaprinast and Enh1 on mammalian cells, 6 × 103 HeLa cells were seeded in each well of a Cell Carrier 96-well plate.
Approximately 24 h later, HeLa cells were transfected with 100 ng of the pCMV-R-GECO (42) using Fugene (Promega) as per the manufacturer's instructions. Cells were washed in Ringer's solution and then treated with 10 μm Enh1, 500 μm zaprinast, or 2 μm A23187. Images were acquired every 12 s for 30 min on a Cytation3 reader (BioTek) using an excitation wavelength of 531 and an emission wavelength of 593. Higher resolution videos were similarly acquired from HeLa cells seeded in 3-cm glass bottom dishes (MatTek) and transfected with 200 ng of pCMV-R-GECO per dish. Cells were treated with either 500 μm zaprinast or 10 μm Enh1 in Ringer's solution and imaged every 250 ms using a Nikon Eclipse Ti microscope.
Videos were analyzed in Fiji using default thresholding for the red channel to determine the mean gray value in each slice for individual cells. Kymographs were constructed from videos of intracellular GCaMP6f-expressing parasites, following the same treatment as for the video microscopy egress assays, except half as many parasites were used to infect the monolayers, and a ×20 objective was used to acquire the images. Regions of interest were defined by outlining parasites in Fiji and measuring the mean fluorescence of each region over time.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
## Fura-2 Recordings
*In the **Fura-2 Recordings** section:*
Tachyzoite loading with Fura-2/AM was done as described previously (17). Briefly, freshly lysed parasites were washed twice with buffer A (116 mm NaCl, 5.4 mm KCl, 0.8 mm MgSO4, 5.5 mm d-glucose, and 50 mm HEPES, pH 7.4) and resuspended to a final density of 1 × l09 parasites/ml in loading buffer (buffer A plus 1.5% sucrose and 5 μm Fura-2/AM). The suspension was incubated for 26 min at 26 °C with mild agitation. Subsequently, the parasites were washed twice with buffer A to remove extracellular dye, resuspended to a final density of 1 × 109 parasites/ ml in buffer A, and kept in ice.
Parasites are viable for a few hours under these conditions. For fluorescence measurements, 2 × 107 parasites/ml were placed in a cuvette with 2.5 ml of Ringer's solution. Fluorescence measurements were done in a thermostatically controlled Hitachi F-7000 spectrofluorometer using the Fura-2 conditions for excitation (340 and 380 nm) and emission (510 nm). The Fura-2 fluorescence response to Ca2+ was calibrated from the ratio of 340/380-nm fluorescence values after subtraction of the background fluorescence of the cells at 340 and 380 nm, as described previously (43). The Ca2+ release rate is the change in Ca2+ concentration during the initial 20 s after the addition of compound.[](https://www.ncbi.nlm.nih.gov/mesh/D016257)
## Chemical Susceptibility of P. falciparum Egress and Erythrocyte Invasion
*In the **Chemical Susceptibility of P. falciparum Egress and Erythrocyte Invasion** section:*
Blood stage P. falciparum parasites of strain 3D7 were obtained from the Walter and Eliza Hall Institute (Melbourne, Australia) and cultured as described previously (44). Parasites were maintained in O+ human erythrocytes (Research Blood Components, Boston, MA), at 2% hematocrit, in RPMI 1640 supplemented with HEPES (25 mm), hypoxanthine (50 mg/liter), sodium bicarbonate (2.42 mm), and Albumax (4.31 mg/ml). Cultures were incubated at 37 °C in microaerophilic atmospheric conditions (1% O2, 5% CO2, 94% N2) within modular incubator chambers.[](https://www.ncbi.nlm.nih.gov/mesh/D010100)
To test the specific effects of Enh1 on egress of parasites from schizonts as well as erythrocyte invasion by liberated merozoites, mature schizont stage parasites were purified from 4 ml of blood stage culture (10–20% hematocrit) by centrifugation (930 × g, 15 min, low acceleration and deceleration) on 4 ml of a 60% Percoll cushion (45). Schizonts were retrieved from the medium supernatant-Percoll interface. After at least three washes in excess volumes of RPMI culture medium, schizonts were diluted with uninfected erythrocytes for a parasitemia of 3–5% and further supplemented with 2 μm Cpd2 to allow schizont maturation up to the point of egress (22). After 1–4 h at standard culture conditions, cultures were washed at least three times in excess RPMI culture medium to remove Cpd2 and added to 1 volume of Enh1 in RPMI or RPMI only (no drug) for a sample volume of 100 μl at 1% hematocrit. We similarly prepared samples of parasites supplemented with heparin (100 units/ml), a specific inhibitor of erythrocyte invasion (46), to assess the background signal for ring stage parasitemia (see below).
We fixed samples (4% paraformaldehyde and 0.0075% glutaraldehyde in PBS (47)) at the outset of the experiment (untreated only) and after 1–2 h of incubation in standard culture conditions (all samples). After extensive washing in PBS and staining with SYBR Green I (1:1000 dilution in PBS; Invitrogen), schizont and ring stage parasitemia were measured by flow cytometry in the FITC channel as described previously (48, 49).[](https://www.ncbi.nlm.nih.gov/mesh/D006493)
## T. gondii Invasion Assays
*In the **T. gondii Invasion Assays** section:*
Invasion assays were performed as described previously (20). Briefly, freshly lysed tachyzoites were preincubated for 10 min in varying concentrations of Enh1, keeping the total concentration of DMSO (vehicle) constant in all samples. HFF monolayers, seeded 48 h earlier, were infected at a multiplicity of infection of ∼10, and invasion was allowed to proceed for 10 min at 37 °C. Following fixation, intracellular parasites were enumerated by immunofluorescence and normalized to the number of host cells in a given field.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
## Results
*In the **Results** section:*
## Zaprinast Increases Cytosolic Ca2+ in a PKG-dependent Manner
*In the **Zaprinast Increases Cytosolic Ca2+ in a PKG-dependent Manner** section:*
Previous work has shown that zaprinast triggers T. gondii and P. falciparum egress in a PKG-dependent manner (20, 22). Because Ca2+ is a necessary second messenger during egress (8), we hypothesized that zaprinast might stimulate egress by increasing cytosolic Ca2+ in the parasite. To test this, we generated a strain that expressed the genetically encoded Ca2+ indicator GCaMP5 (32) in the wild-type T. gondii recombinant human background. These parasites were transfected with a plasmid encoding a constitutively secreted fluorescent fusion protein, p30-DsRed, which accumulates in the parasitophorous vacuole before egress (50). This second sensor enabled us to visualize permeabilization of the parasitophorous vacuole, which has been demonstrated to occur via the Ca2+-regulated secretion of a perforin-like protein (50).
Human fibroblasts infected with recombinant human GCaMP5 were treated with zaprinast and monitored by live video microscopy. Zaprinast elicited a rapid increase in fluorescence compared with the vehicle alone, which was followed by a second, less intense peak of fluorescence (Fig. 1, A and B). This led to permeabilization of the vacuole membrane, as indicated by diffusion of DsRed, and subsequent egress of the parasites from the host cell (supplemental Video S1).[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
Zaprinast raises cytosolic Ca2+ through the activation of PKG. A, video microscopy of intracellular parasites expressing both GCaMP5 and constitutively secreted DsRed, following the addition of zaprinast at 0 s. B, GCaMP5 fluorescence in the region of the parasitophorous vacuole (green) or DsRed fluorescence in adjacent areas of the infected host cell (red) following the addition of zaprinast. Results shown are mean ± S.E. for four experiments. Kymographs for GCaMP5 fluorescence in all four experiments are shown to indicate times of peak fluorescence (white asterisks) relative to the initiation of egress from the region (white vertical lines). C, intracellular Ca2+ concentrations, monitored over time, for wild-type parasites loaded with Fura-2/AM, suspended in buffer containing extracellular (1.8 mm) or basal (100 nm) free Ca2+, and stimulated with 100 μm zaprinast at 400 s. D, similar measurements were performed for PKG-M and PKG-T parasites loaded with Fura-2/AM and suspended in basal Ca2+. Cpd1 or vehicle was added at 100 s, and zaprinast was added at 400 s, as indicated. E–G, intracellular calcium concentrations from parasites loaded with Fura-2/AM and then treated with zaprinast, ionomycin, GPN, or thapsigargin at 100 s (1) and 400 s (2), as indicated. Traces are representative of three independent experiments. Bar graphs report the change in cytosolic calcium over 20 s following the addition of each drug. Results shown are mean ± S.E. (error bars). Z, zaprinast; I, ionomycin; G, GPN; T, thapsigargin; *, p < 0.05; one-tailed t test.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
To determine whether the zaprinast-induced increase in Ca2+ represented release of intracellular stores or Ca2+ entry, we loaded wild-type parasites with the ratiometric Ca2+ indicator Fura-2/AM and stimulated them with zaprinast in buffers containing either a basal Ca2+ concentration (100 nm free Ca2+) or one resembling the extracellular environment (1.8 mm). We found that cytosolic Ca2+ increased in response to zaprinast under both conditions (Fig. 1C), although the response was magnified in the presence of extracellular Ca2+. This indicates that zaprinast-mediated Ca2+ mobilization occurs through release of intracellular stores, which may enhance Ca2+ entry, as has been reported for other agonists capable of releasing intracellular stores (9).[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
Compound 1 (Cpd1), a specific inhibitor of apicomplexan PKG (25), blocks zaprinast-induced egress (20). To determine whether zaprinast's effect on parasite Ca2+ depends on PKG, we measured the response to zaprinast after Cpd1 treatment. Sensitivity of apicomplexan PKG to Cpd1 relies on the identity of a residue at the base of the ATP-binding pocket known as the gatekeeper (25). We therefore used strains engineered to express PKG with either the wild-type threonine or a bulky methionine gatekeeper residue that renders PKG refractory to Cpd1 inhibition (31). These strains were loaded with Fura-2/AM, treated with Cpd1 or vehicle, and then stimulated with zaprinast.
Parasites treated with vehicle responded to the addition of zaprinast with a sharp Ca2+ spike. The addition of Cpd1 did not change the response of the insensitive PKG-M strain. However, the Ca2+ spike was severely diminished by Cpd1 in the sensitive PKG-T strain, indicating the need for PKG in this process (Fig. 1D). We also observed a slight decrease in the cytosolic Ca2+ level of the PKG-T strain after Cpd1 treatment, perhaps indicating that PKG regulates the basal Ca2+ level in addition to enhancing it before egress.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
## Zaprinast Mobilizes a Neutral, SERCA-independent Ca2+ Store
*In the **Zaprinast Mobilizes a Neutral, SERCA-independent Ca2+ Store** section:*
To further characterize the source of zaprinast-mobilized Ca2+, we assessed the involvement of various Ca2+ storage organelles. We loaded parasites with Fura-2/AM and suspended them in a buffer containing basal Ca2+ and then added the Ca2+ ionophore ionomycin. Binding of ionomycin to Ca2+ is pH-dependent and falls to negligible levels below pH 7 (51). Ionomycin has therefore been used to specifically mobilize neutral Ca2+ stores in both mammalian cells (52) and T. gondii (17). We observed a peak in cytosolic Ca2+ levels following ionomycin treatment, indicating that Ca2+ had been mobilized from neutral stores.
Subsequent treatment with zaprinast did not produce an additional Ca2+ peak, indicating that the zaprinast-mobilized store had already been depleted by ionomycin. In contrast, treating parasites with zaprinast followed by ionomycin produced a Ca2+ spike in response to each compound. These results were corroborated by analyzing the rate of Ca2+ release during the 20 s that followed the addition of each compound (Fig. 1E). Taken together, these results indicate that the zaprinast-mobilized store comprises a subset of the ionomycin-mobilized stores and as such is predicted to be neutral.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
To further rule out acidic Ca2+ stores, we treated parasites with GPN before or after the addition of zaprinast. GPN is specifically hydrolyzed in lysosomal compartments, leading to their leakage, as has been shown for the plantlike vacuole in T. gondii (18). As predicted by mobilization with ionomycin, the zaprinast-mobilized store was independent from this acidic compartment (Fig. 1F).[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
The ER is the major neutral Ca2+ store in many organisms (53). Thapsigargin is an inhibitor of the Ca2+ reuptake pump SERCA, which partially localizes to the ER in extracellular T. gondii tachyzoites (54). We treated parasites with thapsigargin before or after the addition of zaprinast and found that zaprinast mobilized Ca2+ with the same efficiency regardless of the treatment order. The independence of the zaprinast- and thapsigargin-mobilized stores was evident in the rate of Ca2+ release following each treatment (Fig. 1F). These results indicate either that zaprinast mobilizes a neutral store that is separate from the ER or that the ER is segmented into SERCA-dependent and SERCA-independent Ca2+ stores, as has been hypothesized (4).[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Identifying Small Molecules That Modulate Parasite Ca2+
*In the **Identifying Small Molecules That Modulate Parasite Ca2+** section:*
Despite its central importance during infection, few compounds have been demonstrated to specifically interfere with the Ca2+ signaling pathways of apicomplexan parasites. Compounds that modulate Ca2+ signaling can have clinical value as well as being useful in research. We predicted that the zaprinast response measured in GCaMP5-expressing parasites could form the basis for a phenotypic screen. Such a screen would benefit from the known importance of Ca2+ signaling in parasite biology along with the reported success rate of cell-based phenotypic assays. We designed a screen wherein extracellular GCaMP5-expressing parasites were pretreated with compounds for 10 min before stimulation with zaprinast, and parasite fluorescence was measured 4 min later.
We expected that compounds that interfered with the zaprinast response would reduce fluorescence. We calculated a Z′-factor for this screen to determine its dynamic range and suitability to high throughput screening. To do so, we compared zaprinast-treated parasites to untreated parasites or those pretreated with the PKG inhibitor Compound 2 (Cpd2) (55). These scenarios yielded Z′-factors of 0.58 and 0.55, respectively, well above the 0.5 Z′-factor considered acceptable for high throughput screening (56). This phenotypic screen is therefore suitable for analyzing large libraries of compounds to identify compounds interfering with parasite Ca2+ signaling (Fig.
2A).[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Compound screen identifies modulators of zaprinast-induced Ca2+ signaling. A, GCaMP5-expressing parasites pretreated with Cpd2 or a vehicle control were stimulated with zaprinast or a vehicle control, and fluorescence was measured to determine a Z′ score for a zaprinast-based screen for Ca2+ modulators. B, GCaMP5-expressing parasites were pretreated with 823 compounds from the PKIS libraries. Fluorescence was measured after zaprinast stimulation. Results are -fold change from zaprinast alone after background subtraction. Compounds that fluoresced independently are indicated (blue) along with selected enhancers (Enh; green) and inhibitors (Inh; red). The mean ± S.D. (error bars) of two experiments is shown, and dashed lines indicate two S.D. values above and below the mean of all compounds. C, structures of Enh1, Enh2, Inh1, and Inh2 as well as the known PKG inhibitors: Cpd1 and Cpd2. D, cumulative frequencies of screen values for all compounds (black), Inh1 and its analogs (red), or Enh1 and its analogs (green).[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
We obtained the compound libraries published kinase inhibitor sets PKIS1 (57) and PKIS2 (unpublished), from GlaxoSmithKline and applied our phenotypic screen to the 823 ATP mimetics represented in these collections (Fig. 2B and supplemental Table S1). The compounds were screened at a single 10 μm dose with two biological replicates. We identified 37 putative inhibitors and 14 putative enhancers, defined as compounds that resulted in fluorescence readings that were more than two S.D. values below or above the mean of all compounds, respectively. Compounds that interfered with accurate measurement of GCaMP5 fluorescence, including two putative enhancers, were identified and excluded from subsequent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
We wondered whether we could identify classes of molecules that inhibit or enhance Ca2+ mobilization. To this end, we examined analogs of the most potent enhancer and inhibitor, which we here refer to as Enh1 (PKIS: GSK260205A) and Inh1 (PKIS: GW827099X) (Fig. 2C). Compounds with the same core structure as Enh1 generally produced a -fold change in fluorescence of >1 in our screen, indicating that parasite fluorescence was greater in the presence of such compounds than with zaprinast alone (Fig. 2D and supplemental Fig.
S1). A Kolmogorov-Smirnov test comparing the values obtained from all compounds with those obtained from the Enh1 analogs revealed a significant difference (p = 0.0042) (Fig. 2D). The combination of a phenyl group at the R1 position and a 3-aminopropyloxy group at the R3 position, as found in Enh1, appeared to produce a particularly effective enhancer; compounds having either group without the other produced only a mild effect. A similar analysis of Inh1 analogs revealed that these compounds tended to inhibit zaprinast-induced Ca2+ mobilization (Fig. 2D and supplemental Fig. S2).
The fluorophenyl group in the R1 position of Inh1 may be responsible for this compound's strong inhibitory effect (as discussed below). This 4-fluorophenyl is also found in Cpd1 and Cpd2. No other R1 side group produced a similarly strong response, although an analogue containing a trifluoromethyl group at the R1 position elicited the second largest response in the group. In summary, although Enh1 and Inh1 are unique within the set of compounds tested for their abilities to augment and repress cytosolic Ca2+, their core structures appear to be generally well suited to these purposes. Indeed, further exploration of these scaffolds could lead to the identification of improved modulators of parasite Ca2+ signaling.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Two Inhibitors Target PKG Using Distinct Chemical Scaffolds
*In the **Two Inhibitors Target PKG Using Distinct Chemical Scaffolds** section:*
Because known PKG inhibitors could suppress zaprinast-induced Ca2+ release (Fig. 1D), we hypothesized that the newly identified compounds could act in a similar manner. We tested the effects of Inh1 and Inh2 (compound GW407323A from the screen) on the in vitro activity of recombinant T. gondii PKG. Both compounds inhibited PKG activity with IC50 values of 580 nm for Inh1 and 670 nm for Inh2 (Fig. 3A), suggesting that PKG inhibition is a plausible explanation for their activity. Furthermore, neither inhibitor showed appreciable activity against human PKG or related AGC kinases (57), demonstrating selectivity for the parasite enzyme.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
PKG inhibitors display distinct modes of inhibition. A, activity of recombinant PKG in the presence of increasing concentrations of Inh1, Inh2, or the PKG inhibitor Cpd2. B, binding of PP derivatives like Inh1 to protein kinases orients the C2 position toward the gatekeeper residue (orange). Left, the PP scaffold's orientation for a disubstituted PP in complex with CDK2. Middle, the structure of Inh1 is highlighted to indicate the PP scaffold (red) and the C2 fluorophenyl group (green). Two other similar compounds from other kinase structures are superimposed over the CDK2 structure in complex with its inhibitor, all showing a similar positioning of the PP scaffold. Right, human p38 MAPK with a trisubstituted monocyclic heterocycle oriented similarly as the PP scaffold and extending a fluorophenyl group in the direction of the gatekeeper. C, oxindole derivatives bind to protein kinases in a manner that orients their C2 and C5 positions away from the gatekeeper. Left, oxindole scaffold of a derivative in complex with NEK2. Right, the oxindole scaffold of Inh2. Two other similar compounds from other kinase structures have been superimposed on the structure of NEK2 with its inhibitor. D, zaprinast-induced egress of parasites carrying Cpd2-sensitive (PKG-T) or resistant (PKG-M) alleles of PKG, following pretreatment with Inh1, Inh2, or Cpd2. Results shown are mean ± S.E. (error bars) for n = 3 independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C118531)
The two inhibitors are derivatives of distinct bicyclic heterocyclic scaffolds with proven activity against protein kinases. Inh1 is a PP with substitutions in the C2 and C3 positions (Fig. 3B). Several available kinase structures display a similar orientation of the PP scaffold in the ATP-binding pocket, with the C3 position engaging the hinge region and the C2 position oriented toward the gatekeeper residue. The structures of two such inhibitors complexed with human CDK2 demonstrate the proximity of the C2 position to the bulky phenylalanine-gatekeeper residue (Fig.
3B, middle). Substitutions at this position decreased the activity of these inhibitors presumably due to steric clash with the gatekeeper (36). The 4-fluorophenyl group in the C2 position of Inh1 is therefore expected to restrict binding of the inhibitor to kinases with relatively small gatekeepers, like the threonine found in T. gondii PKG. This type of binding is evident in the structure of p38 MAPK with a ligand similar to Inh1 (SB203580, a monocyclic heterocycle rather than a bicyclic one) that extends the 4-fluorophenyl functional group into the hydrophobic pocket created by a threonine gatekeeper (37) (Fig. 3B, right).
Furthermore, the related human kinase ERK2, normally resistant to SB203580, can be rendered sensitive to this inhibitor by mutation of its gatekeeper glutamine to either an alanine or a threonine (58). This strongly suggests that a small gatekeeper is required to accommodate heterocycles with substitutions similar to those found on Inh1.[](https://www.ncbi.nlm.nih.gov/mesh/C118531)
Inh2 is an oxindole derivative, with substitutions at the C3 and C5 positions (Fig. 3C). Disubstituted and trisubstituted oxindoles are well studied protein kinase inhibitors. Several structures are available of kinases in complex with such inhibitors. In every instance, the oxindole is positioned such that the 2-oxygen and the nitrogen of the bicyclic scaffold engage the hinge residues (Fig. 3C, left).
This configuration leaves the C3 and C5 positions pointing away from the binding pocket. Although the C6 substituent is inward pointing, it does not approach the direction of the gatekeeper. The binding configuration of oxindole derivatives is highly consistent across several different kinase structures, suggesting that Inh2 (a 3,5-disubstituted oxindole) is likely to be gatekeeper-independent.[](https://www.ncbi.nlm.nih.gov/mesh/C022960)
To determine whether these inhibitors could interfere with other PKG-dependent processes, we next assessed their ability to block zaprinast-induced egress. We directly tested the structural predictions regarding the effect of the gatekeeper on Inh1, but not Inh2, by performing these experiments with the PKG-M and PKG-T strains. We measured zaprinast-induced egress by lactate dehydrogenase release from host cells infected with either strain (Fig. 3D). As expected, PKG-T parasites did not egress after Cpd2 treatment, whereas the refractory PKG-M strain egressed similarly to vehicle-treated parasites. Both Inh1 and Inh2 reliably reduced egress relative to vehicle-treated parasites.
The similar potencies of these compounds in vivo and in vitro are consistent with inhibition of PKG being the primary mechanism through which Inh1 and Inh2 block egress (Fig. 3D). As predicted from our structural analysis, Inh1 worked in a gatekeeper-dependent manner, reducing egress from cells infected with PKG-T parasites with an IC50 of 1.3 μm but leaving PKG-M parasites unaffected. In contrast, Inh2 reduced egress from cells infected with either strain, exhibiting an IC50 of 2.5 μm for PKG-T and 3.5 μm for PKG-M. These results were satisfying because we were able to recover inhibitors with distinct scaffolds and modes of inhibiting an enzyme known to be involved in the zaprinast response. Identification of two novel PKG inhibitors through a screen of only 823 compounds demonstrates the power of our approach to identify compounds with biological effects in pathways critical to parasite survival.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
## Two Enhancers Directly Release Intracellular Ca2+ Stores
*In the **Two Enhancers Directly Release Intracellular Ca2+ Stores** section:*
Having shown that a screen for modulators of zaprinast-induced Ca2+ mobilization reveals novel PKG inhibitors, we turned our attention to Enh1 and another putative enhancer: Enh2, or GSK2188764A (Fig. 2C). During the course of this study, the genetically encoded Ca2+ indicator GCaMP6f became available. Because GCaMP6f has a broader dynamic range and faster kinetics than GCaMP5 (59), we used GCaMP6f throughout much of this study. We tested whether the enhancers could increase parasite Ca2+ independently of zaprinast. Indeed both enhancers could mobilize Ca2+ on their own, although they did so with significantly slower kinetics than zaprinast (Fig.
4A).[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
Enh1 and Enh2 mobilize intracellular Ca2+ stores. A, GCaMP6-expressing parasites suspended in buffer supplemented with either extracellular (1 mm) or basal (100 nm) Ca2+ concentrations, with or without 1% FBS, and treated with 10 μm Enh1 or Enh2 at time 0. B, GCaMP6-expressing parasites treated with 1 μm Cpd1 or vehicle at 100 s and then with 10 μm Enh1 or 100 μm zaprinast at 400 or 750 s, respectively. A gap indicates incubation on ice before the addition of zaprinast, in order to capture the peak of the response. Measurements represent fluorescence after the subtraction of background obtained from samples treated with Cpd1 or vehicle, as indicated. Results shown are mean ± S.E. (error bars) for n = 3 independent experiments. C and D, intensity of R-GECO-expressing HeLa cells treated with Enh1, zaprinast, or A23187 over 10 min (C) or acquired at a faster rate for 1 min (D).[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
To determine the source of the Ca2+ mobilized by the enhancers, we suspended parasites in buffers containing extracellular (1 mm) or basal (100 nm) levels of free Ca2+, treated these parasites with 10 μm Enh1 or Enh2, and monitored GCaMP6f fluorescence (Fig. 4A). Enh1 increased fluorescence under both conditions, indicating that it mobilizes intracellular stores. Surprisingly, we observed only weak mobilization of Ca2+ in response to Enh2 under both conditions. Because the original screen was performed in buffer containing 1% FBS, we tested the effect of this component. The addition of 1% FBS increased Enh2-mediated Ca2+ mobilization.
We interpreted this to mean that Enh2 requires a cofactor found in FBS. Because the activity of Enh2 is relatively weak and depends on specific conditions, we focused subsequent experiments on Enh1.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
We wondered whether, like zaprinast, Enh1 functions through PKG to increase parasite Ca2+. Significant overlap in the emission spectra of Enh1 and Fura-2 prevented us from using this ratiometric indicator. However, we were able to use GCaMP6f to make semiquantitative measurements. We recorded fluorescence from GCaMP6f-expressing parasites treated with Cpd1 or vehicle control, followed by stimulation with Enh1 or zaprinast. To record the peak fluorescence induced by zaprinast, we had to incubate parasites on ice before stimulation so as to slow down the response.
In contrast to the zaprinast response, Cpd1-treated parasites responded to Enh1 in the same manner as vehicle-treated parasites, indicating that Enh1 does not act through PKG. Although Cpd1 accelerated the return to basal Ca2+ levels following zaprinast treatment, it only partially decreased the initial Ca2+ spike induced by zaprinast (Fig. 4B). This is reminiscent of results obtained from Fura-2-loaded parasites, where the increase in Ca2+ following zaprinast could never be fully suppressed by PKG inhibition (Fig. 1D).[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
We wondered whether zaprinast or Enh1 could induce Ca2+ fluxes in mammalian cells. We transfected HeLa cells with the plasmid CMV-R-GECO (42), which encodes a variation of the GCaMP Ca2+ indicators under the mammalian CMV promoter. We treated these transfected cells with zaprinast, Enh1, or, as a positive control, the Ca2+ ionophore A23187. Video microscopy revealed that R-GECO-expressing HeLa cells increased fluorescence in response to A23187 but that neither zaprinast nor Enh1 produced a significant change (Fig. 4C). Given that zaprinast is known to inhibit phosphodiesterases in mammalian cells (60), we considered that these drugs could be having a subtle effect that was below our limit of detection.
We therefore repeated the assay with greater spatial and temporal resolution. We observed a slight increase in R-GECO fluorescence in response to zaprinast during the first 10 s of treatment. However, Enh1 had no effect on host cell Ca2+ (Fig. 4D), indicating that the effects of Enh1 are specific to Ca2+ signaling in the parasite.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
## Enh1 Inhibits T. gondii Survival
*In the **Enh1 Inhibits T. gondii Survival** section:*
Zaprinast and related phosphodiesterase inhibitors were recently shown to block T. gondii proliferation in human fibroblasts (61). Altering Ca2+ levels using ionophores has also been shown to affect parasite viability (62). We therefore hypothesized that Enh1 would similarly inhibit survival of T. gondii. We tested the effect of these compounds using plaque assays as an indication of parasite survival following treatment with Enh1 or zaprinast. Plaque formation was inhibited by zaprinast and Enh1 at 100 and 0.5 μm, respectively, whereas 10-fold lower concentrations allowed normal plaque formation similar to vehicle alone (Fig. 5A).
These results for zaprinast agree with previously published work (61). In order to determine the inhibitory concentrations of these compounds more precisely, we performed lytic assays in which host cells were exposed to parasites pretreated with varying concentrations of zaprinast or Enh1, and the degree of host cell lysis was assessed by crystal violet staining of the monolayer, 3 days later. Enh1 showed a steep dose response, with 350 nm completely inhibiting parasite-induced cell death, and an EC50 of 180 nm. In contrast, 500 μm zaprinast was required to cause complete inhibition of parasite-induced cell death, and the IC50 of zaprinast was 200 μm (Fig. 5B).[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
Enhancers of Ca2+ mobilization show antiparasitic activity. A, plaque formation in the presence of the indicated concentrations of Enh1 or zaprinast. The drug concentrations indicated did not affect host cell survival. B, dose-dependent effect of Enh1 and zaprinast on parasite viability, assayed by monolayer disruption, 3 days postinfection, at the indicated drug concentrations. Results shown are mean ± S.E. (error bars) for n = 3 independent experiments. C, host cell lysis following 1 h of infection in the presence of varying zaprinast concentrations.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
We noticed the discrepancy between the amount of zaprinast required to kill T. gondii in plaque assays (100 μm) and the amount needed in lytic assays (500 μm). Because a rise in cytosolic Ca2+ stimulates parasite motility (2), we wondered if the movement of T. gondii in response to concentrations between 100 and 500 μm might mechanically wound and kill host cells, which would be indistinguishable from parasite survival in the lytic assay. We tested this hypothesis by incubating host cell monolayers with parasites pretreated with various concentrations of zaprinast and then measuring lactate dehydrogenase release from host cells. We found that treatment with 64–510 μm zaprinast resulted in host cell lysis, with a maximal effect at 130 μm. Host cell lysis was not observed with zaprinast in the absence of T. gondii (Fig. 5C).
Due to the lower multiplicity of infection used in plaque assays compared with lytic assays, host cells in plaque assays would be unlikely to experience observable cell wounding. We therefore expect the plaque assays to provide a more accurate measure of zaprinast's antiparasitic activity.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
## Enh1 Induces Asynchronous Ca2+ Fluxes
*In the **Enh1 Induces Asynchronous Ca2+ Fluxes** section:*
We thought it likely that the plaquing defect caused by Enh1 was related to the Ca2+ disregulation induced by this compound. We therefore characterized the Ca2+ response to Enh1 in more detail. We recorded videos of intracellular GCaMP6f-expressing parasites treated with Enh1 and compared the result with zaprinast treatment, which we have shown induces egress. We observed GCaMP6f activity in both cases, indicating that Enh1 can act on intracellular parasites. However, the profiles of zaprinast and Enh1-mediated Ca2+ mobilization were remarkably different.
Whereas zaprinast induced a fast, strong Ca2+ peak (Fig. 6A and supplemental Video S2), Enh1 elicited a slow, asynchronous effect in which some parasites appeared to experience multiple Ca2+ fluxes of similar magnitudes over the course of several minutes (Fig. 6A and supplemental Video S3). The response to these compounds was similar in extracellular parasites (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Enh1 elicits asynchronous cytosolic Ca2+ fluxes and blocks zaprinast-induced egress. A, video microscopy of GCaMP6f-expressing parasites treated with Enh1 or zaprinast. Time after the addition of the compound is indicated. Different times were used to capture the fast and slow responses of zaprinast and Enh1, respectively. B, kymographs illustrate average fluorescence intensities of individual parasites, per row, during the course of the treatment indicated. Black indicates that parasites egressed from vacuoles. C, change in fluorescence of the parasites illustrated in B over the 40 s following the addition of zaprinast. Measurements from each biological replicate are colored separately. Mean change for each group is indicated with a horizontal line. ****, p < 0.0001, two-tailed t test.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
We generated kymographs illustrating the responses of individual GCaMP6f parasites to Enh1. Wondering whether pretreatment with Enh1 would affect the ability of parasites to respond to other Ca2+ agonists, we treated some of these parasites with zaprinast as well (Fig. 6B). The sporadic flashes of Ca2+ triggered by Enh1 were recapitulated, and all vehicle-treated parasites experienced an increase in Ca2+ immediately after the addition of zaprinast. However, Enh1-treated parasites varied in their responses to zaprinast, exhibiting asynchronous Ca2+ fluxes and inconsistent magnitudes in their Ca2+ increases. In rare cases, Ca2+ concentrations in Enh1-treated parasites even dropped upon zaprinast treatment.
We quantified the changes in fluorescence of individual parasites over the 40 s following zaprinast addition and observed a more varied and overall diminished response to zaprinast in Enh1-treated parasites (Fig. 6C). Taken together, these results suggest that Enh1 may partially deplete the Ca2+ stores mobilized by zaprinast.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Enh1 Blocks Ca2+-related Phenotypes in T. gondii and P. falciparum
*In the **Enh1 Blocks Ca2+-related Phenotypes in T. gondii and P. falciparum** section:*
While examining Enh1-treated parasites, we noticed that this compound did not induce egress and in fact suppressed egress in response to zaprinast (Fig. 6C). To confirm these effects, we treated intracellular GFP-expressing parasites with Enh1 and quantified the number of intact vacuoles at various times after treatment, using fluorescence to monitor the infection using automated image analysis. This assay is capable of directly measuring egress for hundreds of vacuoles per sample. Surprisingly, despite its effects on extracellular parasites, Enh1 failed to stimulate egress beyond the spontaneous egress observed in the vehicle control. In contrast, zaprinast treatment resulted in ∼90% of vacuoles egressing within 30 min (Fig.
7A), as described previously (20). We then examined the effect of a 10-min Enh1 pretreatment and found that it robustly blocked zaprinast-induced egress. Analysis of the dose-dependent inhibition of egress by Enh1 revealed an EC50 of 290 nm, within the range of concentrations that inhibit plaque formation (Fig. 7B). These results suggest that the antiparasitic activity of Enh1 is mediated by its inhibition of parasite egress.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
Enh1 blocks egress Ca2+-related phenotypes in T. gondii and P. falciparum. A, egress of intracellular parasites treated with zaprinast, Enh1, or a vehicle control. The number of intact vacuoles was monitored by live microscopy over 30 min. Representative images before and after treatment are shown. B, dose-dependent inhibition of zaprinast-induced egress following pretreatment with Enh1 or vehicle. C, Enh1 inhibition of egress induced by either zaprinast or A23187. Results shown are mean ± S.E. (error bars) for n = 3 independent experiments. ***, p < 0.001; **, p < 0.01. D–E, schizonts were released from Cpd2 arrest immediately preceding the addition of Enh1. D, after 1–2 h, the remaining schizonts (mean ± S.D. for three technical replicates) were counted and normalized to their initial abundance (3.6 and 5.4% in each experiment, respectively). E, Enh1 blocks invasion of erythrocytes by P. falciparum, measured 1–2 h following release from Cpd2, as assessed from the ring stage parasitemia. Mean invasion ± S.E. is expressed as a percentage of invasion without drug (9.5 and 8.1% in each experiment, respectively). Background was assessed using heparin as a specific blocker of invasion and was comparable with the signal observed with saturating concentrations of Enh1. F, dose-dependent inhibition of T. gondii invasion following 10-min pretreatment of parasites before invasion. Results shown are mean ± S.E. for n = 3 independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
We wondered whether we could generalize the Enh1-associated egress defect beyond zaprinast-induced egress. We therefore tested the ability of Enh1 to block egress induced by the Ca2+ ionophore A23187. Enh1 also produced a significant block on A23187-induced egress (Fig. 7C), demonstrating that its effects on parasite Ca2+ signaling extend to a variety of agonists.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
We tested whether Enh1 also affects other apicomplexan parasites that use Ca2+-based signal transduction. Asexual, blood stage Plasmodium parasites undergo Ca2+-dependent egress from infected erythrocytes, followed by Ca2+-dependent invasion into new erythrocytes (6, 63). To assess the effect of Enh1 on these processes at the relevant developmental stage, we allowed purified schizonts to complete their intracellular maturation while blocked with Cpd2 from rupturing and egressing (22). We administered Enh1 to parasites immediately following washout of Cpd2. After allowing 1–2 h of incubation with the compound, we measured egress and reinvasion using a flow cytometry-based assay that distinguishes schizonts from recently invaded ring stage parasites (49).
Whereas Enh1 does not reduce parasite egress at concentrations up to at least 10 μm (Fig. 7D), the compound completely blocks invasion within the tested range (IC50 = 3.2 μm) (Fig. 7E). Wondering whether Enh1 could also inhibit invasion in T. gondii, we incubated tachyzoites with varying doses of Enh1 for 20 min and tested their ability to infect host cells. As in P. falciparum, Enh1 strongly inhibited T. gondii invasion (Fig. 7F). Furthermore, the IC50 of Enh1 in the invasion assay was 180 μm, similar to that in the egress assay, arguing for a common mechanism for the effect of Enh1 on both Ca2+-related phenotypes. In summary, Enh1 modulates Ca2+-dependent processes in diverse apicomplexan parasites.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Discussion
*In the **Discussion** section:*
Ca2+ signaling plays a central role in apicomplexan biology, yet few of the components that regulate Ca2+ uptake and release have been identified. In this study, we extend our understanding of these processes by demonstrating that PKG activity is needed for the robust Ca2+ response elicited by phosphodiesterase inhibitors. Furthermore, we determine the source of Ca2+ to be distinct from other previously described stores. We use this phenomenon as the basis of a phenotypic screen that allowed us to identify several novel inhibitors and enhancers of Ca2+ signaling. Two of the inhibitors could be shown to interfere with Ca2+ signaling by specifically targeting parasite PKG.
In contrast, the enhancers could be shown to increase Ca2+ independently from zaprinast and in fact prevented parasite egress by apparently depleting intracellular Ca2+ stores. This compound displayed antiparasitic properties against both T. gondii and P. falciparum, establishing a new mechanism for interfering with apicomplexan parasitism.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
The signaling events that trigger parasite egress remain poorly defined. Ca2+ ionophores have long been known to stimulate egress in T. gondii (64). However, more recently, phosphodiesterase inhibitors were shown to have similar effects (20, 61). Our experiments revealed a strong and rapid release of Ca2+ in response to zaprinast treatment through the use of both the established Ca2+ indicator Fura-2 and the newly adapted genetically encoded sensors GCaMP5 and GCaMP6f (29). We provide conclusive evidence linking PKG to the zaprinast-induced increase in cytosolic Ca2+, using a chemical-genetic approach to demonstrate specific inhibition of PKG by Cpd1 and Cpd2.
Sensitivity to both inhibitors depends on the relatively small gatekeeper of apicomplexan PKGs, which we mutated to a methionine that preserves kinase activity but renders PKG refractory to inhibition (31). The changes in the zaprinast response caused by Cpd1 can therefore be fully attributed to PKG because no such changes were observed upon treatment of the resistant strain (PKG-M). However, comparing the response to zaprinast in parasites pretreated with Cpd1 with those pretreated with vehicle shows that the initial sharp Ca2+ peak induced by zaprinast is incompletely suppressed by Cpd1, in contrast to its complete inhibition of zaprinast-induced egress (20). Suppression of this peak appeared greater when assayed using Fura-2 (Fig. 1D) than when using GCaMP6 (Fig. 4B).
This may result from the semiquantitative nature of GCaMP6 and the potentially non-linear relationship between fluorescence and Ca2+ concentration. Additionally, differences in the subcellular distribution of the two indicators may influence their responses to different sources of Ca2+. Ratiometric measurements with Fura-2 also revealed that inhibition of PKG by Cpd1 decreased basal Ca2+ concentrations in extracellular parasites, suggesting that PKG might also be necessary to maintain resting levels of Ca2+ in extracellular parasites. These changes in basal Ca2+ concentrations might not be evident with GCaMP6f due to its higher Kd (375 nm) (59) compared with that of Fura-2 (135 nm) (43).[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
We characterized the zaprinast-mobilized store as neutral because it can be depleted by ionomycin. However, we found that this store is independent of the thapsigargin-mobilizable store. This is surprising, given that the ER is the only neutral Ca2+ store that has been characterized in T. gondii. SERCA, the target of thapsigargin, localizes to the ER in intracellular parasites but redistributes so as to only partially colocalize with the ER in extracellular parasites (54). Because our experiments were done in extracellular parasites, it is possible that zaprinast mobilizes Ca2+ from a section of the ER lacking SERCA under these conditions, as previously suggested for the ethanol-mobilized Ca2+ stores (65).
In P. falciparum, zaprinast has been shown to work through P. falciparum PKG to trigger changes in the levels of various precursors of the second messenger IP3. Presumably, this increases IP3, which then interacts with the IP3 receptor to stimulate release of Ca2+ from intracellular stores (19). An IP3 receptor has not been identified in apicomplexans, but treating parasites with ethanol raises levels of IP3 and stimulates Ca2+ release, providing evidence for the presence of such a channel (66). Our results therefore indicate that zaprinast functions, at least in part, through T. gondii PKG and probably mobilizes a neutral, SERCA-independent, IP3 receptor-gated store.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
Genetically encoded calcium indicators provide excellent reproducibility and circumvent many problems associated with loading and compartmentalization of chemical probes. Here we demonstrate that such indicators can be used to identify compounds that alter apicomplexan Ca2+ signaling and its dependent processes, such as egress and invasion. The simplicity and robustness of this cell-based phenotypic screen, with a Z′-factor >0.5, makes it compatible with high throughput screening efforts. As proof of concept, we screened a library of 823 ATP mimetics from GlaxoSmithKline for compounds that could alter the zaprinast-induced increase in cytosolic Ca2+. Identification of two PKG inhibitors, with distinct chemical scaffolds and mechanisms of inhibition, validated our screen and demonstrates the power of the approach.
Unexpectedly, our screen also revealed two compounds that, when used in combination with zaprinast, augmented Ca2+ levels. One such compound, Enh1, elicited repeated cycles of Ca2+ increase and decrease with overall cytosolic Ca2+ building at a population level. These repeated cycles are reminiscent of what others have seen when Fluo-4/AM-loaded parasites were treated with thapsigargin (54), perhaps indicating that Enh1 also inhibits Ca2+ uptake pathways.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
In contrast to zaprinast and Ca2+ ionophores, Enh1 failed to stimulate egress despite raising cytosolic Ca2+ levels. In fact, Enh1 blocked the ability of these agonists to stimulate parasite egress. Furthermore, treatment with Enh1 blocked tachyzoite invasion and plaque formation at concentrations similar to those required to block egress. These results suggest that Enh1 depletes essential intracellular Ca2+ stores, which have been previously suggested to mediate invasion and egress. Consistent with this view, we observed diminished changes in GCaMP fluorescence in response to zaprinast, following Enh1 treatment (Fig. 6C).
However, without a complete understanding of Enh1 function, we cannot rule out the possibility that inhibition of egress may be independent of its ability to modulate Ca2+ because (i) some parasites in which Enh1 did not elicit Ca2+ fluctuations still failed to egress in response to zaprinast, and (ii) A23187, which should equilibrate Ca2+ across membranes, failed to overcome Enh1 inhibition. Compounds similar to Enh1 have been shown to inhibit mammalian kinases belonging to the AGC family (67). Having already established that T. gondii PKG, a member of the AGC kinase family, mediates Ca2+ release, it is an intriguing possibility that a related kinase might oppose its activity.[](https://www.ncbi.nlm.nih.gov/mesh/C011145)
Enh1 robustly inhibited the ability of P. falciparum to invade erythrocytes, although it did not affect egress. Despite the many parallels between egress of P. falciparum and T. gondii, substantial differences have also been uncovered. Plasmodium egress is a fast, highly synchronized process, dependent on a cascade of proteolytic activity (68). Genetic evidence supports this distinction with mutants in the calcium-responsive protein DOC2.1 blocking both egress and invasion in T. gondii but only invasion in P. falciparum (69), mirroring the effects of Enh1. The intracellular Ca2+ stores of Plasmodium have not been investigated as thoroughly as those of T. gondii, but there is some evidence that it is not solely dependent on its intracellular Ca2+ stores and in fact utilizes Ca2+ found within the parasitophorous vacuole (70).
In light of this, the differential responses of P. falciparum and T. gondii to Enh1 are perhaps not surprising. That Enh1 blocks invasion of P. falciparum demonstrates this compound's ability to perturb Ca2+-dependent processes across multiple members of the Apicomplexa.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
Resistance to front line antimalarials is increasing (71, 72), and new treatment options are needed. Rational design strategies can identify drugs with minimal off-target effects but focus on a limited repertoire of signaling pathways. The Toxoplasma Ca2+ signaling network shows evidence of conservation within apicomplexans while being dissimilar to human signaling pathways. By screening for modulation of this pathway, we have prioritized compounds that are likely to interfere selectively with infection without focusing on a single parasite protein. This screen can be performed with commonly available equipment and should be easily scalable to larger collections of compounds.
Our success in identifying new modulators of Ca2+ signaling with effects that extend to multiple apicomplexans highlights the power of such an approach. In particular, Enh1 inhibits parasite viability with an EC50 in the nanomolar range and provides a good lead for the development of antiparasitic compounds. Further characterization of these compounds may identify novel components of apicomplexan Ca2+ signaling pathways and improve our ability to target these pathways specifically.[](https://www.ncbi.nlm.nih.gov/mesh/D002118)
## Author Contributions
*In the **Author Contributions** section:*
S. M. S., M. A. H. T., and A. S. P. designed and conducted the experiments and analyzed the data. S. M. S. wrote the majority of the manuscript, with specific sections contributed by M. A. H. T. and A. S. P. C. G. H. constructed the T. gondii strains with different PKG alleles. M. E. B., R. H., and W. J. Z. provided key reagents and advice. F. T. and N. J. W. synthesized Cpd2. M. T. D., S. N. J. M., and S. L. supervised the work in their respective laboratories and contributed to the analysis of experiments and writing of the manuscript.
## Supplementary Material
*In the **Supplementary Material** section:*
This work was supported in part by National Institutes of Health Grants AI-110027 and AI-096836 (to S. N. J. M.) and 1DP5OD017892 (to S. L.). The authors declare that they have no conflicts of interest with the contents of this article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
This article contains supplemental Table S1, Figs. S1 and S2, and Videos S1–S3.
ER
endoplasmic reticulum
GPN[](https://www.ncbi.nlm.nih.gov/mesh/C079641)
l-phenylalanine-naphthylamide[](https://www.ncbi.nlm.nih.gov/mesh/C079641)
IP3[](https://www.ncbi.nlm.nih.gov/mesh/D015544)
inositol 1,4,5-triphosphate[](https://www.ncbi.nlm.nih.gov/mesh/D015544)
PKG
protein kinase G
HFF
human foreskin fibroblast
PP[](https://www.ncbi.nlm.nih.gov/mesh/C118531)
pyrazolopyridazine[](https://www.ncbi.nlm.nih.gov/mesh/C118531)
Cpd1 and Cpd2
compound 1 and 2, respectively
Enh1 and Enh2
enhancer 1 and 2, respectively
Inh1 and Inh2
inhibitor 1 and 2, respectively.
The abbreviations used are:
# References
*In the **References** section:*
|
# Introduction
Positive Allosteric Modulation of the Glucagon-like Peptide-1 Receptor by Diverse Electrophiles*
# Abstract
*In the **Abstract** section:*
Therapeutic intervention to activate the glucagon-like peptide-1 receptor (GLP-1R) enhances glucose-dependent insulin secretion and improves energy balance in patients with type 2 diabetes [mellitu](https://www.ncbi.nlm.nih.gov/mesh/D005947)s. Studies investigating mechanisms whereby peptide ligands activate GLP-1R have utilized mutagenesis, receptor chimeras, photo-affinity labeling, hydrogen-deuterium exchange, and crystallography of the ligand-binding ectodomain to establish re[ceptor h](https://www.ncbi.nlm.nih.gov/mesh/D006859)o[mology mo](https://www.ncbi.nlm.nih.gov/mesh/D003903)dels. However, this has not enabled the design or discovery of drug-like non-peptide GLP-1R activators. Recently, studies investigating 4-(3-benzyloxyphenyl)-2-ethylsulfinyl-6-(trifluoromethyl)pyrimidine (BETP), a GLP-1R-positive all[osteric modulator, determined that Cys-347 in the GLP-1R is require](https://www.ncbi.nlm.nih.gov/mesh/C570819)d [for ](https://www.ncbi.nlm.nih.gov/mesh/C570819)positive allosteric modulator activity via covalent modific[ati](https://www.ncbi.nlm.nih.gov/mesh/D003545)on. To advance small molecule activation of the GLP-1R, we characterized the insulinotropic mechanism of BETP. In guanosine 5′-3-O-(thio)triphosphate binding and INS1 832-3 insulinoma cell cAMP assays, [BETP](https://www.ncbi.nlm.nih.gov/mesh/C570819) enha[nced GLP-1(9–36)-NH2-stimulated cAM](https://www.ncbi.nlm.nih.gov/mesh/D016244)P signaling. Using isolated pancreatic i[slet](https://www.ncbi.nlm.nih.gov/mesh/D000242)s, BETP p[oten](https://www.ncbi.nlm.nih.gov/mesh/C570819)tiated insulin secretion in a glucose[-dep](https://www.ncbi.nlm.nih.gov/mesh/D000242)endent manner that requires both the peptide l[igan](https://www.ncbi.nlm.nih.gov/mesh/C570819)d and GLP-1R. In studies of the cova[lent me](https://www.ncbi.nlm.nih.gov/mesh/D005947)chanism, PAGE fluorography showed labeling of GLP-1R in immunoprecipitation experiments from GLP-1R-expressing cells incubated with [3H]BETP. Furthermore, we investigated whether other reported GLP-1R activators and compounds iden[ti](https://www.ncbi.nlm.nih.gov/mesh/D014316)f[ied ](https://www.ncbi.nlm.nih.gov/mesh/C570819)from screening campaigns modulate GLP-1R by covalent modification. Similar to BETP, several molecules were found to enhance GLP-1R signaling in a Cys-347-dependent manner. The[se c](https://www.ncbi.nlm.nih.gov/mesh/C570819)hemotypes are electrophiles that react with GSH, and LC/MS deter[min](https://www.ncbi.nlm.nih.gov/mesh/D003545)ed the cysteine adducts formed upon conjugation. Together, our results sug[ges](https://www.ncbi.nlm.nih.gov/mesh/D005978)t covalent modification may[ be used](https://www.ncbi.nlm.nih.gov/mesh/D003545) to stabilize the GLP-1R in an active conformation. Moreover, the findings provide pharmacological guidance for the discovery and characterization of small molecule GLP-1R ligands as possible therapeutics.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
The secretin family of peptide hormone-binding class B G protein-coupled receptors (GPCRs)2 is composed of 15 members that help regulate various physiological systems, including growth, the stress response, and glucose metabolism. These receptors possess a long amino-terminal ectodomain (ECD) that is structurally stabilized by three evolutionarily conserved disulfide bonds (1–3). This characteristic ECD is a globular structure hypothesized to be critical for the initial ligand recognition event. Subsequent interactions between the docked peptide ligand and the helical bundle of the receptor core occur to induce signaling (4). Several reports provide data supporting the initial receptor-ligand recognition mechanism, and among the most informative are the crystallography studies using the purified ECD regions in complex with their cognate ligands (5, 6) and a recent report utilizing electron microscopy and hydrogen-deuterium exchange study of peptide-ligated glucagon receptor (7).
These studies clearly define many intermolecular processes utilized by various ECDs to bind with great specificity to the orthosteric peptide. However, crystal structures of ligands bound to full-length class B receptors are needed to fully determine critical interactions with the helical bundle that induce an active receptor conformation.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
Although more than 20 class A GPCR crystal structures have been resolved to date, the resolution of class B GPCRs has been elusive. Recently, high resolution structures of the transmembrane domains of two class B family members, the glucagon receptor (8) and corticotropin-releasing factor receptor-1 (9), have been solved. Although the structures were determined using receptor constructs that did not include the ECD sequences, data from both predict very deep and open V-shaped binding crevices for the orthosteric ligands (10). Although several peptide-based agonist therapeutics have been developed for class B GPCRs (11), small molecule agonists have not been developed. One hypothesis for this intractability of class B GPCRs is that the intricate nature of receptor activation renders it challenging to identify low molecular weight non-peptide orthosteric ligands that make sufficient high affinity receptor contacts over large interaction surfaces to mimic the activation mechanism used by the peptide agonists.
This may prove especially difficult if crystal structures of other receptor core domains are solved and show this deep crevice-like binding is general to class B GPCRs. Therefore, an alternative approach to enhance receptor activation is to identify small molecule positive allosteric modulators (PAMs) that bind topographically distinct sites and relieve some hindrance of transitioning the receptor to an active state. Several reports have described the identification and development of PAMs for class A, class B, and class C GPCRs (12).
The glucagon-like peptide-1 receptor (GLP-1R) is a class B GPCR that has proved to be an effective therapeutic target as several peptide-based agonists have been developed and registered for treating type 2 diabetes mellitus (13, 14). Reports from preclinical studies have described the discovery of small molecule GLP-1R PAMs (15, 16). Two of the most characterized are compound 2 (17, 18) and compound B/BETP (19); both molecules have been shown to potentiate GLP-1R-stimulated cAMP accumulation by oxyntomodulin (20, 21) and GLP-1(9–36)-NH2 (22, 23). Each compound also enhances the actions of non-natural agonists on the GLP-1R, peptidic and non-peptidic, such as BMS-21, Boc-5, and TT-15 (24). In total, these results support the existence of an allosteric site in the GLP-1R distinct from the orthosteric peptide-binding pocket.
Although structurally distinct, unfortunately both compounds are unstable and display poor pharmacokinetic properties that preclude potential clinical development (16). Not surprisingly, because of its electrophilic nature, BETP has been reported to conjugate with GSH in vitro (25).[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
The finding that nucleophilic GSH displaces the ethyl sulfoxide moiety of BETP led to studies that elegantly demonstrated the allosteric mechanism utilized by BETP to potentiate GLP-1R signaling occurs via irreversible covalent modification of a free cysteine located in the third intracellular loop of the GLP-1R (26). Importantly, substitution of this cysteine (position 347) with alanine does not alter peptide agonism at the mutant receptor but does result in loss of PAM action for both BETP and compound 2 (26). Interestingly, although BETP and compound 2 represent distinct chemical pharmacophores, the compounds share the electrophilicity; this suggests that compound 2 also activates the GLP-1R by covalent modification of Cys-347 (26). Although both molecules act by the same mechanism, modification of the receptor by the compounds results in a differential enhancement of GLP-1R signaling (cAMP accumulation versus intracellular calcium mobilization, β-arrestin recruitment, and ERK phosphorylation) (24). Although BETP and compound 2 are the best characterized GLP-1R PAMs reported to date, other modulators have been described (16), but it is not known whether these molecules interact with the receptor by the same covalent mechanism.
Furthermore, it is unclear whether other electrophiles that can access Cys-347 will modulate GLP-1R activity.[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
For class B GPCRs, the presence of a free cysteine in the GLP-1R at position 347 is unique. Thus, studies evaluating the effect of modifying this site on GLP-1R-mediated cellular function may enable the development of receptor-specific agents. In line with this, the work presented herein shows that BETP potentiates insulin secretion in a glucose- and GLP-1R-dependent manner, therefore bolstering the potential therapeutic utility of allosterically modulating receptor activity. To further interrogate the mechanism whereby covalent modification enhances GLP-1R function, we describe the discovery and characterization of several electrophilic chemotypes that potentiate GLP-1R activity in a Cys-347-dependent manner. Importantly, the results indicate that structurally diverse compounds that have the ability to access the third inner membrane loop of the GLP-1R can enhance receptor signaling of this important therapeutic target.
Because the positive allosteric mechanism likely occurs through structural modification that results in the formation of an intracellular adduct, strategies aimed at targeting this site may facilitate efforts to stabilize the GLP-1R in an active confirmation and possibly enable new ligand identification approaches.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Experimental Procedures
*In the **Experimental Procedures** section:*
## Ligands
*In the **Ligands** section:*
The following compounds were prepared as described previously: BETP and th-BETP (24); 1 and 2 (17, 18); 3 (27); 4 and 5 (28); 6 (29); 7 and 8 (30); 9 (27); 15 (31); 17 (32). Other described compounds were purchased from commercial sources. [3H]BETP was purchased from ViTrax (specific activity, 33.3 Ci/mmol). GLP-1(7–36)-NH2 was synthesized at Lilly Research Laboratories, and GLP-1(9–36)-NH2 was purchased from Bachem (Torrance, CA).[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
## GLP-1R [35S]GTPγS Binding Assays
*In the **GLP-1R [35S]GTPγS Binding Assays** section:*
Preparation of GLP-1R HEK293 cell membranes and measurement of GLP-1R activation via [35S]GTPγS binding to Gαs using an antibody capture scintillation proximity assay were performed as described previously (20). Briefly, reactions contained 5 μg of membrane in 20 mm HEPES, pH 7.4, 50 mm NaCl, 5 mm MgCl2, 40 μg/ml saponin, 0.1% bovine serum albumin, and 500 pm 35S-labeled GTPγS (PerkinElmer Life Sciences). Peptides and compounds were diluted and treated at a final concentration of 1% DMSO. Binding was induced for 30 min at room temperature before solubilization with 0.2% Nonidet P-40 detergent, rabbit anti-Gαs polyclonal antibody, and 0.5 mg of anti-rabbit polyvinyl toluene beads (PerkinElmer Life Sciences). The detection mixtures were developed for 30 min, centrifuged at 80 × g for 10 min, and counted for 1 min/well using a MicroBeta TriLux instrument (PerkinElmer Life Sciences).[](https://www.ncbi.nlm.nih.gov/mesh/D013462)
## INS1 832-3 Cell cAMP Accumulation Assays
*In the **INS1 832-3 Cell cAMP Accumulation Assays** section:*
The INS1-derived 832-3 insulinoma cell line (33, 34) was used to study stimulation of intracellular cAMP by BETP + GLP-1(9–36)-NH2. Cells were maintained by growing adherently in RPMI 1640 medium (HyClone, Pittsburgh, PA) containing 11.2 mm glucose supplemented with 10% fetal bovine serum, 10 mm HEPES, 1 mm sodium pyruvate, 2 mm l-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin, and 50 μm 2-mercaptoethanol at 37 °C in 5% CO2. On the day of the assay, cells were lifted, counted, and resuspended in Earle's balanced salt solution buffer (EBSS) containing 0.1% bovine serum albumin and 11.2 mm glucose. Cells were seeded at a density of 40,000 cells per well into sterile 96-well half-area solid black microplates and incubated in the presence of various treatments supplemented with isobutylmethylxanthine at 37 °C for 30 min. Cells were then assayed for cAMP accumulation using homogeneous time-resolved fluorescence technology (Cisbio Bioassays, Bedford, MA).[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
## Ex Vivo Pancreatic Islet Insulin Secretion Assays
*In the **Ex Vivo Pancreatic Islet Insulin Secretion Assays** section:*
Male wild-type and Glp-1r null mice (35) (both of C57BL/6 background) were maintained in accordance with the Institutional Animal Care and Use Committee of Eli Lilly and Co. and the Guide for the Use and Care of Laboratory Animals by the National Institutes of Health. Mice were housed in microisolator cages on wood chip bedding with food (2014 Teklad Global Diet, Harlan, Indianapolis, IN) and deionized water available ad libitum. Lights were on a 12:12-h light/dark cycle, and temperature and relative humidity were maintained between 21 and 23 °C and 45 and 65%, respectively. For pancreatic islet isolation, mice were euthanized by cervical dislocation. The common bile duct was cannulated with a 27-gauge needle, and the pancreas was distended by infusion of 10 ml of Hanks' balanced salt solution buffer (Sigma) containing 2% bovine serum albumin (Akron Biotech, Boca Raton, FL) and 1 mg/ml collagenase (Vitacyte, Indianapolis, IN).
Pancreatic tissue was then removed and digested in Hanks' balanced salt solution buffer at 37 °C. Islets were purified on a Histopaque (Histopaque-1077/Histopaque-11991 mixture) gradient (Sigma) for 18 min at 750 × g. Islets were then cultured overnight in RPMI 1640 medium containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin (Invitrogen). The next morning, islets were incubated in EBSS containing 2.8 mmol/liter glucose at 37 °C for 30 min. Groups of three size-matched islets were then hand-picked and incubated in 0.3 ml of EBSS containing the indicated concentrations of glucose and ligands at 37 °C for 90 min. Following the static incubation experiments, supernatants were collected and stored at −20 °C until assayed for insulin using an electrochemiluminescence assay (Meso Scale Diagnostics, Rockville, MD). Data were calibrated to external standards and expressed as nanograms/ml insulin in the culture medium.[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Immunoprecipitation, Liquid Scintillation, and Denaturing PAGE Fluorography
*In the **Immunoprecipitation, Liquid Scintillation, and Denaturing PAGE Fluorography** section:*
HEK293 cells were adherently passaged to poly-d-lysine polymer-coated dishes. Transfection was performed using a mixture of FuGENE 6 reagent (Promega, Madison, WI) and pcDNA3.1 expression vector encoding the human GLP-1R-3×FLAG fusion protein. Following 48 h, cells were washed with 1× PBS and allowed to recover in supplemented Gibco 31053 media containing 0.5% fetal bovine serum prior to a 2-h treatment of 500 nm [3H]BETP at 37 °C. Cells were scraped from the dish to pellet material and to wash away unbound radioligand prior to immunoprecipitation.
Cells were disrupted with sonication and removed from insoluble debris via centrifugation in lysis buffer (50 mm Tris-HCl pH 7.5, 150 mm NaCl, 1 mm EDTA pH 8.0, 1% Triton X-100, 1× Complete protease inhibitor (Roche Applied Science)), and 2 mm reduced l-GSH. All lysates were pre-cleared of immunoglobulin (IgG)-binding content for 2 h at 4 °C with mouse IgG-agarose gel (Sigma) prior to affinity purification using anti-FLAG M2-agarose gel (Sigma) and overnight rotation at 4 °C. The bead resins were washed three times following immunoprecipitation with lysis buffer and gravity filtration using empty Bio-Spin chromatography columns (Bio-Rad). Captured proteins were eluted from the M2 affinity gel or IgG control gel using SDS sample buffer at low speed centrifugation.[](https://www.ncbi.nlm.nih.gov/mesh/D011107)
For liquid scintillation, immunopurified protein was added directly to Ultima Gold scintillator mixture (PerkinElmer Life Sciences), and radioactive decay was quantified with a Beckman LS3801 scintillation counter. Standard samples containing known amounts of radioactivity were also quantified to calculate the instrument counting efficiency for conversion to disintegrations/min. For fluorography, immunopurified protein was reduced with 2-mercaptoethanol at room temperature for electrophoretic separation. The 10% polyacrylamide gel containing radiolabeled samples was fixed with acetic acid in 2-propanol, agitated in Amersham Biosciences Amplify fluorographic agent (GE Healthcare), dried under heated vacuum, and exposed to x-ray MP Hyperfilm (PerkinElmer Life Sciences) at −80 °C. The film was developed using an automatic processor following 5–20 days of exposure.[](https://www.ncbi.nlm.nih.gov/mesh/D008623)
## GLP-1R HEK293 Cell cAMP Assays
*In the **GLP-1R HEK293 Cell cAMP Assays** section:*
HEK293 cells expressing either wild-type human GLP-1R (NCBI accession number NP_002053) (20) or mutant C347A human GLP-1R-1×FLAG were used for measurement of cAMP accumulation. Cells were seeded into 96-well half-area black microplates in Gibco 31053 Dulbecco's modified Eagle's medium (Life Technologies, Inc.) containing 0.1% bovine serum albumin fraction. Peptide and compound treatments were diluted into medium containing isobutylmethylxanthine and added to cells for 30 min at 37 °C or 1 h at room temperature with 5% CO2. Cells were then assayed for cAMP accumulation using homogeneous time-resolved fluorescence technology (Cisbio Bioassays, Bedford, MA). Data were analyzed by the ratio method, calibrated to external standards, and reported as concentration of cAMP or expressed as percent cAMP compared with GLP-1(7–36)-NH2.
To enable testing of small molecule compounds in higher throughput format, for some experiments, minor modifications to this protocol were made. Specifically, compound and GLP-1(9–36)-NH2 were dispensed acoustically (Echo, Labcyte, Sunnyvale, CA) into a total volume of 20 μl in Corning 384-well plates. GLP-1(9–36)-NH2 shift assays were performed with a concentration-response curve of GLP-1(9–36)-NH2 and a fixed dose of small molecules. Data were fit to the four-parameter logistic model (36), and ratios of potencies were used to quantify the fold-shift of GLP-1(9–36)-NH2 potency induced by allosteric modulators. Similarly, a concentration-response curve was generated for compounds in the presence of an EC20 of GLP-1(9–36)-NH2, and relative EC50 values were generated using the four-parameter model.
Data were normalized using MIN and MAX controls, as described (36). Data analysis was performed using Screener (Genedata, San Francisco).[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
## Fluorescein-Maleimide Competition Assay
*In the **Fluorescein-Maleimide Competition Assay** section:*
A competitive fluorescence-based assay for assessing the reactivity of compounds was developed and performed as described previously (37) with minor modifications. Briefly, a test compound was added to PBS (calcium/magnesium-free; Hyclone), followed by rapid addition of 100 nm GSH (ThermoFisher, Grand Island, NY), and then 50 nm fluorescein-5-maleimide (AnaSpec, San Jose, CA). The final DMSO concentration in the assay was 3.3% (v/v). Fluorescein emission was measured at 25 °C using a SpectraMax (Molecular Devices, Sunnyvale, CA). Test compounds were run in a concentration-response format, and equilibrium fluorescence was quantified.
To obtain compound inhibitor constants for the reaction with GSH (KGSH), data were fit using PRISM (GraphPad, La Jolla, CA) to Equation 1, where [I] is concentration of test compound; KS is the second order rate constant for reaction of fluorescein maleimide with GSH; B is basal fluorescence, and F is maximum fluorescence observed. KS was determined experimentally (16,000 m−1 s−1) and is concordant with published literature (37). Data were calculated as the harmonic means and standard deviations from three or more experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## GSH Conjugation Followed by LC/MS
*In the **GSH Conjugation Followed by LC/MS** section:*
## Reagents
*In the **Reagents** section:*
Water and acetonitrile (ACN) were HPLC grade from Lab Scand (Dublin, Ireland). Formic acid (FA), ammonium hydrogen carbonate, phosphoric acid, sodium hydroxide, and l-GSH were from Sigma (Steinheim, Germany). For the sample preparation, a 100 mm sodium phosphate buffer, pH 7.4 (stored at 4 °C), 5 mm GSH in water, and stocks of the different target compounds at 5 mm in ACN were prepared.[](https://www.ncbi.nlm.nih.gov/mesh/C032159)
## Sample Preparation
*In the **Sample Preparation** section:*
Test and control samples were prepared and analyzed in parallel. Test samples contained 0.4 mm target compound and 1.4 mm GSH in phosphate buffer/ACN 70:30 in a total volume of 1.4 ml. Control samples contained 0.4 mm target compound in phosphate buffer/ACN 70:/30. Both solutions were incubated at 37 °C for 120 min in the HPLC autosampler and were analyzed at different times (standard times: 0, 30, 60, and 120 min).[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
## LC/MS Methods
*In the **LC/MS Methods** section:*
An Agilent 1200 series rapid resolution LC/MSD SL system equipped with a solvent degasser, binary pump, auto sampler, column compartment, and a diode array detector (Agilent Technologies, Waldbronn, Germany) with XTerra MS column (50 × 2.1 mm, 3.5 μm; Waters Corp.) was used at 60 °C. The method involved gradient elution with water, 0.1% FA (solvent A) and ACN, 0.1% FA (solvent B) at a flow rate of 1.6 ml/min prior to the mass spectrometer, which was split at a ratio of 4:1. The A/B ratio was set at 80:20 as the initial elution condition, changing linearly to 0:100 in 0.8 min, and maintained at 0:100 from 0.8 to 1.0 min. Peaks were detected by absorbance at 214 nm and bandwidth at 16 nm, using the MS detector for identification. Electrospray mass spectrometry measurements were performed on a mass spectrometry detector quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA) interfaced to the HP1200 system, acquired simultaneously in positive and negative ionization modes (fragmentor 120 V, threshold spectral abundance 150, MS peak width 0.09 min) over the mass range of 100–800. Data acquisition and integration for LC-UV and MS detection were collected using ChemStation software (Agilent Technologies).[](https://www.ncbi.nlm.nih.gov/mesh/C030544)
Injections were made at predetermined time points, which were then used to plot the consumption of target compound as a function of time. The MS method allowed for the simultaneous monitoring of conjugate formation.
## Cathepsin L Assay
*In the **Cathepsin L Assay** section:*
Enzyme activity of the cysteine-protease cathepsin L (CatL) was quantified using human CatL (Calbiochem) and the fluorogenic substrate Z-Phe-Arg-7-amido-4-methylcoumarin (Z-Phe-Arg-AMC, Calbiochem). The assay buffer was 50 mm sodium phosphate buffer, pH 6.5, 2.5 mm EDTA, and 0.002 (w/v) Tween 20. Enzyme activity was measured at 25 °C in 384-well plates (Corning) using a Synergy (BioTek) plate reader with 360 nm excitation and 485 nm emission wavelengths. 20 μl of compound dilution series were incubated with 20 μl of CatL enzyme for 30 min.
The assay was initiated by the addition of 20 μl of Z-Phe-Arg-AMC. Data were quantified after 10 min of reaction, and IC50 values were calculated (36). Final assay conditions were 2.6 nm CatL with 200 nm Z-Phe-Arg-AMC in 0.33% (v/v) DMSO. The assay was pharmacologically validated using the covalent CatL inhibitor (2S,3S)-N2-[(1S)-1-benzyl-2-(benzylamino)-2-oxo-ethyl]-N3-[2-(4-hydroxyphenyl)ethyl]oxirane-2,3-dicarboxamide (CAA0225; EMD Millipore) (38).[](https://www.ncbi.nlm.nih.gov/mesh/D003374)
## Results
*In the **Results** section:*
## BETP Potentiates cAMP Signaling and Stimulates Insulin Secretion in a Glucose- and GLP-1R-dependent Manner
*In the **BETP Potentiates cAMP Signaling and Stimulates Insulin Secretion in a Glucose- and GLP-1R-dependent Manner** section:*
Studies were undertaken to determine the mechanism whereby BETP (19, 22) allosterically modulates the GLP-1R to enhance insulin secretion. To develop a proximal measure of receptor activation, a GLP-1R-containing cell membrane GTPγS binding assay was utilized. Here, upon receptor activation, the non-hydrolyzable guanine nucleotide analogue, [35S]GTPγS, binds Gαs. Accumulated radiolabeled Gαs was then captured using anti-Gαs antisera and measured. Consistent with the established ability of GLP-1R agonists to enhance insulin secretion by stimulating Gαs/adenylyl cyclase/cAMP signaling, BETP induced GTP-Gαs binding in a cooperative manner with the inactive metabolite, GLP-1(9–36)-NH2 (Fig. 1A).
Similarly, in INS1 832-3 insulinoma cells, BETP enhanced GLP-1(9–36)-NH2-induced cAMP accumulation in a concentration-dependent manner (Fig. 1B). The results showing BETP enhances GTP binding in the GLP-1R membrane assay and potentiates cAMP signaling in the pancreatic beta cell line extend the original findings from studies using GLP-1R-expressing HEK293 cells that showed BETP potentiates GLP-1(9–36)-NH2 but not GLP-1(7–36)-NH2 activity (22).[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
BETP potentiates GLP-1(9–36)NH2-mediated cAMP signaling and stimulates insulin secretion in a glucose- and GLP-1R-dependent manner. A, in the presence of GLP-1(9–36)-NH2, BETP enhances activation of Gαs in GLP-1R-expressing membranes. Data show responses of varying concentrations of GLP-1(9–36)-NH2 alone (■) or in combination with 10 μm (▿), 1 μm (▵), or 0.1 μm (♢) of BETP. ● = GLP-1(7–36)-NH2. B, BETP potentiates GLP-1(9–36)-NH2-mediated cAMP accumulation in INS1 832-3 insulinoma cells. Intracellular concentrations of cAMP in response to varying concentrations of GLP-1(9–36)-NH2 alone (■) or in combination with 3 μm (▿), 1 μm (▵), or 0.3 μm (♢) of BETP are depicted. ● = GLP-1(7–36)-NH2. C, in cultures of isolated islets from wild-type (black bar) or Glp-1r KO (white bar) mice, BETP stimulates insulin secretion via a mechanism that requires high glucose (11.2 mm), GLP-1(9–36)-NH2, and the presence of the GLP-1R. Insulin levels in the media were measured following incubation of islets with the indicated concentrations of GLP-1(7–36)-NH2, BETP, GLP-1(9–36)-NH2, BETP plus GLP-1(9–36)-NH2, or glucose-dependent insulinotropic polypeptide (GIP)(1–42). D, insulin secretion is not enhanced by BETP in low glucose concentrations (2.8 mm). For each treatment, mean insulin concentrations were determined by measuring insulin in media from six independent wells containing three islets per well. ***, p < 0.001 using one-way ANOVA followed by Dunnett's comparison versus the respective 11.2 mm glucose group in C. For 11.2 mm glucose, #, p < 0.001 compared with the respective 2.8 mm glucose treatment in C and D. Data presented are representative of three to six independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
To investigate the insulin secretory mechanism of BETP, ex vivo cultures of isolated mouse islets were used. As BETP is a reactive molecule that has been shown to covalently modify several proteins (26), studies were undertaken to assess the specificity and glucose dependence of BETP in potentiating insulin secretion. Similar to previous results from rat islets incubated in high glucose concentrations showing BETP potentiated GLP-1(9–36)-NH2-stimulated insulin secretion (22), the combination of BETP and GLP-1(9–36)-NH2 increased insulin levels in static cultures of islets from wild-type mice (Fig. 1C). Importantly, the insulinotropic response did not occur in the presence of low glucose (Fig. 1D), results that are consistent with the established glucose dependence of GLP-1R activation to enhance insulin secretion.
Furthermore, the co-treatment of BETP and GLP-1(9–36)-NH2 did not enhance insulin secretion in either low or high glucose concentrations in islets isolated from Glp-1r null mice (Fig. 1, C and D). As a control, the insulin secretory capacity of GLP-1R-deficient islets was confirmed by demonstrating that glucose-dependent insulinotropic polypeptide enhanced insulin secretion in these cultures (Fig. 1C). Together, the islet studies performed here indicate that the mechanism whereby BETP potentiates insulin secretion requires the presence of high glucose, the GLP-1R, and a peptide ligand.[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
## Electrophilic Mechanism of Action of BETP
*In the **Electrophilic Mechanism of Action of BETP** section:*
Because of the inherent reactivity of BETP (16), studies were performed to assess whether BETP irreversibly modifies the GLP-1R. For these experiments, a radiolabeled analogue of BETP was synthesized ([3H]BETP) and incubated with HEK293 cells expressing human GLP-1R that contains a carboxyl-terminal FLAG epitope (35). Anti-FLAG affinity purification of lysates from cells treated with [3H]BETP showed high disintegrations/min from GLP-1R-FLAG-expressing cells versus purifications using mouse immunoglobulin control beads (Fig. 2A). Furthermore, PAGE fluorography of lysates under denaturing conditions demonstrated bands consistent with the expected migration pattern of glycosylated GLP-1R (Fig. 2A) (35).
The finding that [3H]BETP irreversibly labels the GLP-1R is in concordance with studies showing BETP covalently binds Cys-347 (26).[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
GLP-1R binding and activation by BETP requires its electrophilic nature. A, affinity purification of GLP-1R-3XFLAG from HEK293 cells treated with [3H]BETP shows labeling of the receptor. Liquid scintillation disintegrations/min from immunoprecipitation assays using IgG control (white bar) or anti-FLAG antibody (black bar)-conjugated beads are shown. Under reducing conditions, 10% PAGE fluorography of anti-FLAG immunopurified proteins shows a banding pattern consistent with migration of the GLP-1R. B, in HEK293 cells expressing the GLP-1R, the non-reactive analogue of BETP, th-BETP, does not potentiate GLP-1(9–36)-NH2-stimulated cAMP accumulation. Data show results from assays testing varying concentrations of GLP-1(9–36)-NH2 alone (■) or in combination with 1 μm (▵) or 0.1 μm (♢) BETP or with 10 μm (○) or 1 μm (▿) th-BETP. ● = GLP-1(7–36)-NH2. C, BETP does not potentiate GLP-1(9–36)-NH2 activity in HEK293 cells expressing the mutant C347A GLP-1R-1XFLAG. Accumulation of intracellular cAMP was stimulated by varying concentrations of GLP-1(9–36)-NH2 alone (■) or in combination with 1 μm (▵) or 0.1 μm (♢) BETP or with 10 μm (○) or 1 μm (▿) th-BETP. ● = GLP-1(7–36)-NH2. D, in an antagonist assay format, GLP-1R activation in HEK293 cells by BETP is not blocked by co-incubation of the non-reactive analogue th-BETP. Varying concentrations of the non-reactive th-BETP analogue (□) do not blunt intracellular cAMP accumulation induced by the combination of BETP (1 μm) and GLP-1(9–36)-NH2 (0.1 μm). ● = GLP-1(7–36)-NH2, and ■ = GLP-1(9–36)-NH2. All data are fit to the four-parameter logistic equation, and the EC50 value for BETP in cAMP accumulation assays using GLP-1R-expressing HEK293 cells is presented in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
To evaluate whether covalent interaction of BETP with the GLP-1R is required for allosteric modulation of receptor activity, a non-reactive structurally similar analogue of BETP that contains thioether in place of the sulfoxide moiety was synthesized (th-BETP, Fig. 3A). For these experiments, HEK293 cells overexpressing the human GLP-1R were used, and as opposed to studies of isolated islets or insulinoma cell lines, GLP-1(9–36)-NH2 shows partial stimulation of cAMP accumulation (21–23). Co-treatment of th-BETP and GLP-1(9–36)-NH2 to these cells showed no additional increase in cAMP accumulation compared with peptide alone (Fig. 2B and Table 1). Furthermore, these assays confirmed the previously reported discovery that Cys-347 of the GLP-1R is required for BETP function, as cells expressing the C347A mutant were not potentiated by BETP (Fig.
2C and Table 1). The concentration-response curves shown for the mutant assay were calculated as a percentage of maximum stimulation by GLP-1(7–36)-NH2; the apparent increase in Emax of GLP-1(9–36)-NH2 may therefore reflect subtle differences in the efficacy of the cognate ligand in wild-type versus C347A GLP-1R cells, a phenomenon that was also previously observed (26).[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
A, chemical structures of known GLP-1R-activating chemotypes are depicted. The leaving groups from nucleophilic attack are shaded. EC50 values for each molecule in cAMP accumulation assays using GLP-1R-expressing HEK293 cells are presented in Table 1. B, chemical structures of the electrophilic compounds evaluated for the ability to modulate GLP-1R activity are presented. The leaving groups from nucleophilic attack are shaded and nucleophilic attack to double bonds is represented with an arrow. EC50 values for each molecule are presented in Table 2.[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
Reactivity and pharmacological characterization of known small molecule GLP-1R activators
SR1 = 5-methyl-1,3,4-thiadiazole-2-thiol; SR2 = 1-[1-(1-methyl-4-piperidyl)ethyl]tetrazole-5-thiol. Compound EC50 values are compound potencies determined from cAMP accumulation assays using wild-type (WT) or C347A GLP-1R-1×FLAG expressing HEK293 cells in the presence of 220 nm GLP-1(9–36)-NH2. 10-Point compound concentration-response curves were run from a maximum of 30 μm, with 3-fold dilutions. Data were normalized using min (220 nm GLP-1(9–36)-NH2 + DMSO = 0%) and max (220 nm GLP-1(9–36)-NH2 + 1 μm reference compound = 100%) controls (36). Potency values are represented as geometric means, standard deviations (S.D.), and number of independent replicates (n). Compound efficacy values (Emax) are represented as arithmetic means, standard deviations (S.D.), and number of independent replicates (n). Representative curves are presented in Fig. 5, A and C. NA = not applicable. GLP-1(9–36)-NH2 fold shift values are calculated as the ratio of GLP-1(9–36)-NH2 potencies in the presence of 10 μm test compound versus DMSO vehicle. 10-Point GLP-1(9–36)NH2 compound concentration-response curves were run from a maximum of 75 μm, with 3-fold dilutions. Data were normalized using min (DMSO = 0%) and max (75 μm GLP-1(9–36)-NH2 + 1 μm reference compound = 100%) controls (36). Potency values are represented as geometric means, standard deviations (S.D.), and number of independent replicates (n). GLP-1(9–36)-NH2 efficacy values (Emax) are represented as arithmetic means, standard deviations (S.D.), and number of independent replicates (n).[](https://www.ncbi.nlm.nih.gov/mesh/C049131)
a Parent compound was still detected after 60 min.
b Data could not be calculated as curves were not obtained in the shift assay.
Studies designed to assess the ability of the non-reactive th-BETP analogue to antagonize BETP showed no inhibition of BETP. Here, near maximal stimulation of cAMP production in wild-type GLP-1R HEK293 cells via combination of BETP and GLP-1(9–36)-NH2 was not blunted by co-addition of increasing concentrations of th-BETP (Fig. 2D). The lack of activity of the non-reactive th-BETP analogue in potentiator assays and the inability of this molecule to blunt BETP action support the assertion that there is not a conventional small molecule binding pocket in the local proximity of Cys-347 of the GLP-1R.[](https://www.ncbi.nlm.nih.gov/mesh/C000612088)
## Known Small Molecule GLP-1R Activators Are Electrophilic
*In the **Known Small Molecule GLP-1R Activators Are Electrophilic** section:*
To determine whether other modulators of the GLP-1R activate the receptor by a covalent mechanism, several previously reported compounds that contain potentially electrophilic positions (compounds 1-8, Fig. 3A) were selected. Compound 9 and th-BETP were used as negative controls because these molecules are not predicted to be reactive. Because the reactivity of BETP to nucleophilic cysteine of GSH has been shown previously (25), a GSH-based conjugation assay is likely an appropriate surrogate for evaluating the potential reactivity of compounds toward free cysteines of the GLP-1R. Thus, BETP, th-BETP, and compounds 1–9 were incubated with GSH (39), and the progression of reactions over time (disappearance of parent compound and appearance of the corresponding adduct formed with GSH) was measured by LC/MS. Under the conditions used, BETP disappeared within 60 min (Table 1).
It must be noted that the t½ of BETP conjugation to GSH was previously determined as <0.5 min (25), corresponding to a second order rate constant of >0.5 m−1 s−1. Results for the other molecules show compounds 1–6 are also reactive to cysteine of GSH; interestingly, each parent compound disappeared within 10 min (data not shown) except for compound 3. Compounds 7 and 8 remained unaltered after 2 h and thus are deemed non-reactive. As expected, compound 9 and th-BETP were also chemically stable. The results showing compound 4 was as unstable in these experiments as quinoxalines 1-3 is of interest because compound 4 has been reported to be orally efficacious (27, 40), although its apparent electrophilicity suggests poor in vivo pharmacokinetic properties, as has been shown for BETP (25).
It should be pointed out that the GSH conjugation assay was performed as described previously (39), using buffer containing organic solvent, conditions that are not representative of in vivo physiology. However, it is indicative of the intrinsic reactivity of the compounds to any cysteine within proximity and therefore predictive of short plasma half-lives when conjugation is observed in this assay; results from the in vitro assays performed here confirm the GSH conjugation mechanism for BETP and also indicate that several known modulators of the GLP-1R are as electrophilic as BETP.[](https://www.ncbi.nlm.nih.gov/mesh/C000612088)
An advantage of the GSH conjugation assay is that it utilizes LC/MS, which enabled us to determine the position of the nucleophilic attack, the structure of the leaving group (see shaded atoms in Fig. 3A), and the structure of the new species formed for each reactive compound from the mass of the new adduct (Table 1). Similar to BETP, the adduct formed by incubation of compound 2 with GSH results from addition of GSH to the quinoxaline ring and elimination of the sulfonyl group. This structure is consistent with the mass obtained by LC/MS. A similar mechanism of addition-elimination to electron-poor aromatic rings is presumed for compounds 1 and 3–6, with elimination of a thiol when GSH reacts with compounds 1 and 4, elimination of a sulfone when it reacts with compounds 5 and 6, and elimination of chloride for compound 3.[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
To determine whether these electrophilic GLP-1R modulators activate the receptor via a mechanism similar to BETP, the compounds were tested in cAMP accumulation assays using cells expressing either wild-type or the C347A GLP-1R. Results of these studies are presented in Fig. 5, A and C, and Table 1. We determined the capacity of these molecules to potentiate GLP-1(9–36)-NH2-mediated cAMP accumulation in GLP-1R-expressing cells using two distinct assays. The first assay (Fig. 5 and referred to as compound EC50 in Tables 1 and 2) measures compound potency in the presence of an EC20 amount of GLP-1(9–36)-NH2. Any observed potency and efficacy enhancements in this assay reflect α and β factor modulation of orthosteric and allosteric ligands (41).
The second assay (referred to as GLP-1(9–36)-NH2 shift in Tables 1 and 2) measures the EC50 and maximal efficacy of GLP-1(9–36)-NH2 in combination with a fixed 10 μm concentration of compound. This assay provides an estimate of the maximal β-induced increases in efficacy and α-induced enhancements in affinity of the allosteric modulators for GLP-1(9–36)-NH2 (41). Analogous to BETP, compounds 1–6 potentiated GLP-1(9–36)-NH2 activity on the wild-type receptor but not in cells expressing the mutant GLP-1R that lacks Cys-347. Together with findings from the GSH conjugation studies, the functional data support the required role of this cysteine in the allosteric mechanism used by the various chemotypes. It is important to note that the mutant receptor was functional as compound 9, a non-electrophilic molecule reported to activate the GLP-1R (42), potentiated GLP-1(9–36)-NH2 in both wild-type and C347A assays (Fig.
5, A and C, and Table 1). To our knowledge, this is the first study showing potentiation of GLP-1(9–36)-NH2 by a compound from this series. Furthermore, these results are significant because the data indicate potentiation of the GLP-1R can be achieved without covalent modification.[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
Reactivity and pharmacological characterization of electrophilic compounds that activate GLP-1R in a cysteine 347-dependent manner[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Compound EC50 values are compound potencies determined from cAMP accumulation assays using wild-type (WT) or C347A GLP-1R-1×FLAG expressing HEK293 cells in the presence of 220 nm GLP-1(9–36)-NH2. 10-Point compound concentration-response curves were run from a maximum of 30 μm, with 3-fold dilutions. Data were normalized using min (220 nm GLP-1(9–36)-NH2 + DMSO = 0%) and max (220 nm GLP-1(9–36)-NH2 + 1 μm reference compound = 100%) controls (36). Potency values are represented as geometric means, standard deviations (S.D.), and number of independent replicates (n). Compound efficacy values (Emax) are represented as arithmetic means, standard deviations (S.D.), and number of independent replicates (n). Representative curves are presented in Fig. 5, B and D. NA = not applicable. GLP-1(9–36)-NH2 fold shift values are calculated as the ratio of GLP-1(9–36)-NH2 potencies in the presence of 10 μm test compound versus DMSO vehicle. 10-Point GLP-1(9–36)-NH2 compound concentration-response curves were run from a maximum of 75 μm, with 3-fold dilutions. Data were normalized using min (DMSO = 0%) and max (75 μm of GLP-1(9–36)-NH2 + 1 μm of reference compound = 100%) controls (36). Potency values are represented as geometric means, standard deviations (S.D.), and number of independent replicates (n). GLP-1(9–36)-NH2 efficacy values (Emax) are represented as arithmetic means, standard deviations (S.D.), and number of independent replicates (n).[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
a Parent compound was still detected after 60 min.
b Addition of GSH to the double bond is shown.[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
c Compounds gave quantifiable activity in GLP-1(9–36)-NH2 shift assay but not in compound dose-response assay. Compounds demonstrated activity in WT (but not C347A) cells, but this was weakly potent and refractory to curve fitting.
## Innate Reactivity Influences Compound-induced Potentiation of the GLP-1R
*In the **Innate Reactivity Influences Compound-induced Potentiation of the GLP-1R** section:*
Compounds 1 and 3 require Cys-347 of the GLP-1R for activity and are predicted to generate the same adduct upon reacting with a cysteine and therefore to have similar potentiation activity (Table 1 and Fig. 4A). However, these molecules showed different modulator activities in the wild-type GLP-1R assay (Fig. 5A and Table 1). The different electrophilic character of these compounds appears to account for the difference. In the GSH conjugation assay, adduct 10 was obtained from both compounds; however, the parent species was still detected after 60 min of incubation of GSH with compound 3 but not for compound 1 (parent had disappeared after 30 min, first time point of this experiment) (Fig. 4B), suggesting the rate of adduct formation is slower for compound 3.
To confirm these results using conditions more similar to those of the GLP-1R cellular assay, the molecules were tested in an aqueous assay (phosphate buffered saline, pH 7.4) where compounds compete with fluorescein-5-maleimide for reaction with GSH (37). Consistent with the LC/MS findings, calculation of the rate constant for the reaction of compounds 1 and 3 with GSH (KGSH) demonstrated that compound 1 is more reactive (1 = 13 m−1 s−1 (S.D. = 10, n = 11) than compound 3 = <1 m−1 s−1 (n = 4) (p = 0.0015 Mann-Whitney U test)). BETP (19 m−1 s−1 (S.D. = 10, n = 6)) was used as a positive control in these experiments. Taken together, these data indicate compound 3 may need a longer incubation time in the functional assays to allow more adduct formation. Thus, time course studies using the GLP-1R [35S]GTPγS binding assay were performed and demonstrate that prolonged incubation of compound 3 induced a similar response to compound 1 (Fig.
4C).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Intrinsic chemical reactivity of electrophilic modulators of the GLP-1R influences receptor activation. A, nucleophilic attack of GSH to quinoxalines 1 and 3 shows formation of the same adduct 10. B, compounds 1 and 3 display different rates of reactivity in a GSH conjugation assay. Data show disappearance of 1 over time (□), formation of adduct 10 over time when 1 is incubated with GSH (■), disappearance of 3 over time (○), and formation of adduct 10 when 3 is incubated with GSH (●). C, time course studies show Gαs activation of the GLP-1R by the quinoxalines. In combination with GLP-1(9–36)-NH2 (3 μm), compounds 1 (3 μm; black bar) and 3 (3 μm; white bar) enhance activation of Gαs in GLP-1R-expressing membranes. Data show mean responses over time of the compounds tested in duplicate in combination with GLP-1(9–36)-NH2. ***, p < 0.0001 using one-way ANOVA followed by pairwise Student's t tests to respective durations of treatment. **, p < 0.01; *, p < 0.05. Data presented are representative of three independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
## Discovery of New GLP-1R Allosteric Modulators
*In the **Discovery of New GLP-1R Allosteric Modulators** section:*
Because the covalent binders shown above represent three different chemotypes, and because the common feature shared by these molecules is the presence of an electrophile within each compound, several structurally distinct reactive compounds that had been previously identified in GLP-1R screening campaigns were tested for their ability to modulate GLP-1R activity in a Cys-347-dependent manner. These molecules were discovered as putative GLP-1R potentiators from screens of the Lilly chemical library using GLP-1R-expressing HEK293 cells where compounds were tested in the presence of a submaximal concentration of GLP-1(9–36)-NH2. The chemical structures of these compounds are depicted in Fig. 3B, and the potentiator results are presented in Fig. 5, B and D, and Table 2.
A variety of compounds showed potentiation of GLP-1(9–36)-NH2 at the wild-type receptor. However, in all cases, compound activity was lost in assays testing the mutant GLP-1R. Consistent with the covalent mechanism, most compounds were confirmed to be reactive in the GSH conjugation assay (Table 2). As negative controls, compounds 18 and 19 were unstable in the presence of GSH and did not potentiate the activity of GLP-1(9–36)-NH2 in the wild-type GLP-1R assay. These are the most reactive compounds in our studies (18 = 22,000 m−1 s−1 (S.D. = 9000, n = 5); 19 = 1200 m−1 s−1 (S.D. = 500, n = 6)), and it is not known whether the lack of PAM activity is due to failure of the molecules to covalently bind the GLP-1R or because the adduct formed is insufficient to facilitate structural changes in the receptor that allow its activation by GLP-1(9–36)-NH2. It is also possible that these compounds cannot access the intracellular location of Cys-347.
Interestingly, as expected for a covalent binder, compound 16 showed activity in the Cys-347-containing wild-type receptor but not in the mutant; however, this molecule did not react with GSH in the conjugation assay or in the fluorescein-5-maleimide competition assay (16 = <1 m−1 s−1 (n = 3)). Precipitation of the compound was observed in the conjugation assay followed by LC/MS. Although low solubility could explain the lack of reactivity of this compound with GSH, other hypotheses for this discrepancy can be envisioned as follows: 1) compound 16 activates the receptor by covalent modification, but Cys-347 in the local cellular environment is slightly more nucleophilic than GSH in the conjugation assays; 2) compound 16 is a pro-drug and requires cellular metabolism to produce the active species.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Requirement of Cys-347 of the GLP-1R for positive allosteric modulation by small molecules. GLP-1(9–36)-NH2-induced cAMP accumulation was measured in wild-type (A and B) and C347A (C and D) GLP-1R-expressing HEK293 cells in the presence of small molecule GLP-1R ligands. Cells were treated with serial dilutions of the compounds in the presence of an EC20 concentration of GLP-1(9–36)-NH2. Resultant cAMP was quantified and data were normalized and fit to the four-parameter logistic equation as described under “Experimental Procedures.” Data graphed exemplify single representative experiments. Summarized data and statistics from multiple experiments are presented in Tables 1 and 2.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
In addition to Cys-347 of the GLP-1R, Cys-438 was also shown to be covalently modified by BETP, but neither BETP nor compound 2 showed reduced abilities to potentiate GLP-1(9–36)-NH2 activity in C438A GLP-1R assays (26). Although it is clear that the newly identified reactive GLP-1R potentiators in our report require Cys-347 for activity, it is possible these electrophiles also bind Cys-438. Although the previous studies showed mutation of Cys-438 does not alter PAM activity, even for the ability of BETP and compound 2 to potentiate oxyntomodulin (26), it cannot be absolutely excluded that formation of an adduct at this position could positively or negatively influence Cys-347-required activity.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
To assess the promiscuity of the new PAMs toward proteins containing nucleophilic cysteine residues, we investigated the inhibitory action of BETP and compounds 11–17 on the activity of several enzymes. These studies utilized several kinases known to be irreversibly inhibited by reactive compounds (EGFR (43), CDK7 (44), BTK (45), and ITK (45)) and the cysteine protease CatL. The results are summarized in Table 3. Weak inhibitory activity was observed for compounds 13 and 15–17, although none of the compounds inhibited all of the enzymes. The kinases tested contain non-catalytic cysteine residues in their active sites.
Thus, non-inhibitory covalent modification of cysteines by inactive compounds cannot be ruled out from these experiments. For CatL, the cysteine-thiol is fundamental for catalysis, and therefore, covalent modification of active site CatL will parallel enzyme inhibition. The lack of activity of BETP and compounds 11–14 against CatL shows that these electrophiles do not react with the catalytic cysteine of this enzyme, whereas the covalent inhibitor CAA0225 exhibited potent activity (IC50 = 6 nm (S.D. = 1, n = 4)).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Inhibition of several enzymes by electrophiles
a Active site occupancy was quantified using TR-FRET (LanthaScreen, ThermoFisher) with compound preincubation of 180 min and a maximal compound concentration of 100 μm, essentially as described (64).
b Enzymatic activity was quantified using radiometric filter binding with a compound preincubation of 180 min and a maximal compound concentration of 20 μm, as described previously (65).
c Cathepsin L activity was quantified as described under “Experimental Procedures.” Data are the mean of three experiments.
Based on potency in the GLP-1R-expressing cellular assays, compounds 13 and 15 were selected for further characterization. Compared with the potentiator activity of BETP and compound 9, two chemotypes that have been structurally optimized using GLP-1R activity assays, compounds 13 and 15 showed lower efficacy in assays where a fixed concentration of compound (10 μm) was tested in the presence of various concentrations of GLP-1(9–36)-NH2 (Fig. 6A). Consistent with lower Emax values, these molecules also do not potentiate insulin secretion in cultures of isolated pancreatic islets from normal mice (Fig. 6, B–D) .
Although larger data sets are needed to define the level and kinetics of cAMP induction in GLP-1R HEK293 cells that correlate with the insulinotropic actions of GLP-1R PAMs, it is likely that a minimum threshold of cAMP accumulation is needed.[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
Comparison of compound-stimulated cAMP generation in HEK293 cells expressing GLP-1R versus insulin secretion from isolated pancreatic islets. A, GLP-1(9–36)-NH2-induced cAMP accumulation was measured in GLP-1R-expressing cells in the presence of small molecule GLP-1R ligands. Cells were treated with varying concentrations of GLP-1(9–36)-NH2 in the presence of 10 μm BETP (□), compound 9 (●), compound 13 (■), or compound 15 (▴), and cAMP was quantified. Data were normalized and fit to the four-parameter logistic equation as described under “Experimental Procedures.” Data were graphed as the mean (±S.D.) of three independent experiments. B–D, insulin levels in media from cultures of wild-type (C57/Bl6 mice) islets treated with GLP-1(7–36)-NH2, GLP-1(9–36)-NH2, compound 9, 13, or 15 alone, or GLP-1(9–36)-NH2 plus each compound. Glucose-stimulated insulin secretion is shown for each islet preparation (2.8 versus 11.2 mm concentrations of glucose), and all compounds were tested in the presence of high glucose. Data are from three islet experiments where mean insulin concentrations for each treatment group were determined by measuring insulin in media from six independent wells containing three islets per well. ***, p < 0.001, and *, p < 0.05 using one-way ANOVA followed by Dunnett's comparison compared with the respective 11.2 mm glucose treatment, and # = p < 0.001 compared with the respective 2.8 mm glucose treatment.[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
Following the discovery and characterization of the new reactive GLP-1R PAMs, we examined the general prevalence of Cys-347-preferring electrophiles among GLP-1R ligands by testing a large selection of compounds (identified from our screens) in the GLP-1R C347A assay. 560 molecules, identified as putative hits from the GLP-1(9–36)-NH2 potentiation screen, were tested at 30 μm in the presence of an EC20 concentration of GLP-1(9–36)-NH2 in both wild-type and C347A GLP-1R-expressing HEK293 cells. 416 compounds showed >10% stimulation in wild-type GLP-1R cells, and this was used as a threshold for subsequent analyses. Correlation of these 416 screening hits in wild-type and C347A GLP-1R cells indicates two clear populations of compounds: a population with equivalent activity in wild-type and C347A cells (∼55% of compounds) and those with activity in wild-type GLP-1R cells but minimal activity in C347A cells (Fig. 7A). Further assessment of these data was performed by correlating the ratio of activities (wild-type/C347A) with “medchem demerits”.
Medchem demerits derive from a rules-based computational tool for identifying potentially reactive and promiscuous molecules based on substructure searching (46). Compounds with >100 demerits are considered undesirable. Our analysis indicates compounds whose action is Cys-347-dependent are not robustly predicted by these first-pass computational filters as most of the molecules suspected of covalent interaction with the GLP-1R have <100 demerits (non-covalent GLP-1R PAMs are indicated in the box in Fig. 7B with activity ratios of wild-type/C347A ≈1). Moreover, analysis of the pharmacological selectivity of Cys-347-dependent compounds indicated these compounds were generally selective and did not exhibit a pan-assay active profile as has been described for some known problematic chemotypes (Fig. 7C)
(47, 48).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Comparative analysis of GLP-1R Cys-347-dependent and Cys-347-independent non-peptide ligands. A, 560 putative GLP-1R ligands were tested at 30 μm in the presence of an EC20 (220 nm) of GLP-1(9–36)-NH2 for the mobilization of intracellular cAMP in C347A and wild-type (WT) GLP-1R-expressing HEK293 cells. The 416 compounds with >10% stimulation in WT cells are graphed. Two clear populations of compounds are annotated on the scatter plot as follows: compounds with approximately equivalent WT and C347A GLP-1R PAM activity (compounds for further characterization), and compounds with high activity in WT GLP-1R and close to no activity in C347A GLP-1R assays (suspected covalent modifiers). B, activity ratios were calculated from A as % stimulation in WT cells divided by % stimulation in C347A cells. This activity ratio was plotted versus the calculated medchem demerits as described by Bruns and Watson (46). Non-covalent GLP-1R PAMs box annotates compounds with approximately equivalent activity in WT and C347A assays. C, WT/C347A activity ratios were plotted versus compound promiscuity. For a given compound, promiscuity was defined as the percentage of assay activities from total number of tested assays. Active was defined as either a concentration response value <10 μm and/or a single point test of >90% stimulation or inhibition. Compounds were not included in this analysis unless they had been tested in >20 assays. Non-covalent GLP-1R PAMs box annotates compounds with approximately equivalent activity in WT and C347A assays.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Discussion
*In the **Discussion** section:*
Several therapeutic peptides that mimic the binding and receptor activation mechanism of GLP-1 have been approved and registered for the treatment of type 2 diabetes mellitus. Fundamental insights into the complex ligand binding process at the orthosteric site to induce GLP-1R signaling have been gained by use of various experimental approaches. X-ray crystallography studies of the GLP-1R ECD in complex with peptide ligand have revealed how the carboxyl-terminal α-helical region of GLP-1 positions within the binding cleft of the ECD (49). Furthermore, photo-affinity labeling has enabled determining residues that position in close proximity between the ligand and the full-length receptor (50, 51). Complementary GLP-1R chimera and mutagenesis studies have proposed additional key residues that may further define the core peptide-binding pocket (52, 53).
Together, these data have informed and refined molecular modeling of the peptide-bound GLP-1R.
As an alternative approach, the studies presented here were undertaken to investigate an allosteric mechanism for activation of GLP-1R. Identifying and understanding mechanisms that modulate the GLP-1R may enable strategies to discover and develop new therapeutic agents, especially at receptor sites distinct from the orthosteric pocket that binds non-peptide low molecular weight compounds. BETP is an electrophilic small molecule that enhances GLP-1R signaling (19, 22) by binding covalently to Cys-347 within the third intracellular loop of the receptor (26). In general, proteins containing free nucleophilic amino acids undergo modification upon being in close proximity to reactive compounds, which often results in widespread irreversibly bound small molecules in cells and tissues. A significant finding from the studies performed herein is that the insulinotropic activity of BETP was demonstrated to be specific for the GLP-1R in ex vivo pancreatic islet assays; here, BETP required the presence of the GLP-1R, high glucose, and the peptide ligand to induce insulin secretion. The findings from these islet culture experiments are important because the results provide physiologically relevant evidence that modulating intracellular regions of the GLP-1R, especially within the vicinity of the third intracellular loop, may be a mechanism to regulate receptor signaling.[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
Interestingly, dynamic simulations of a variant GLP-1R using a homology model (54) also supports the potential importance of the third intracellular loop of the GLP-1R. In efforts aimed at identifying coding variants that underlie susceptibility for developing type 2 diabetes mellitus, an A316T substitution in the GLP-1R was found to associate with lower fasting blood glucose concentrations but higher glucose levels 2 h following a glucose challenge (55). Position 316 is located in transmembrane domain (TMD) 5, and the modeling shows the threonine variant disrupts hydrogen bonding between residues in TMD5 and TMD6 (55). These changes predict a shift of TMD5 toward the cytoplasm and movement of TMD6 outward, altering the position of the third intracellular loop (55). The possibility that a subtle shift in loop three by a variant may alter GLP-1R function is consistent with the ability of BETP to form a small adduct with Cys-347 in this region to alter receptor signaling.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
The capacity of BETP to modulate GLP-1R signaling by covalently modifying the GLP-1R suggests this mechanism could be exploited therapeutically. In addition to BETP, this report shows that several structurally diverse compounds can also covalently modify the GLP-1R to enhance activation by GLP-1(9–36)-NH2. The newly discovered compounds that potentiate GLP-1R signaling require Cys-347 for activity, and unfortunately, these molecules are also electrophilic to GSH, which may indicate the critical pharmacological feature for recognizing the GLP-1R is the intrinsic electrophilic nature of the compounds. Although covalent molecules have been developed as therapeutic agents (56), if there is not a canonical binding pocket present around Cys-347 (and only reactive compounds interact with this amino acid), it is unlikely that drug-like molecules can be developed due to pharmacokinetic and safety liabilities (57). Classical approaches for the evaluation of covalent enzyme inhibitors classify compounds as affinity labels (reactivity-driven) or quiescent affinity labels (reversible binding-driven) (58, 59).
Only the latter have drug-like potential. By examining the calculated rate constants for GSH modification of electrophiles from this study, they appear to be in the range of <100 m−1 s−1 values typical of affinity labels. Subsequent studies could be directed toward a quantitative understanding of structure-activity relationships of GLP-1 PAMs and inherent electrophilicity toward GLP-1R. This would require either a sensitive kinetic assay to measure time dependence of compound action or a direct binding assay. At this point, we do not have robust assay systems capable of resolving GLP-1R-specific binding from the midst of high nonspecific binding for [3H]BETP. Generation of an effective allosteric GLP-1R probe with less electrophilicity would be useful to enable these studies.
It remains uncertain whether allosteric modulators of the GLP-1R that function via the covalent mechanism can be identified that are stable in the presence of other nucleophiles.[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
The overall results from these studies highlight the nucleophilic nature of the GLP-1R, which may influence the direction of future work in this area. For example, this characteristic should be considered when pursuing new screening campaigns for identifying GLP-1R activators. An early understanding of the mechanism of action can avoid the prosecution of promiscuous compounds (60). Many electrophiles can be recognized by features of their chemical structure; however, assays to assess the degree of electrophilicity are useful for conclusively determining the reactivity of some chemotypes. Our results suggest though that although various assays can be used to filter out reactive molecules, C347A GLP-1R is a critical assay for identifying compounds that covalently modify the GLP-1R to enhance its activity.
For instance, it is possible that active compounds could contain independent electrophilic and GLP-1R-binding pharmacophores. Such compounds would be excluded from follow-up if triage was exclusively based on chemical reactivity. In our view, it will be important to determine whether several recently reported GLP-1R PAMs require Cys-347 for activity (61–63). Moreover, our data clearly show that the types of electrophiles that modulate the GLP-1R can be somewhat cryptic and are not generally well predicted by computational approaches, prior learning, and general compound promiscuity information. Our data also indicate that a substantial fraction of GLP-1R allosteric modulators identified by screening may act via covalent modification of Cys-347.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
The nucleophilic capacity of the GLP-1R may be exploited experimentally to stabilize the GLP-1R in biophysical studies. It is clear that adding steric bulk around the vicinity of Cys-347 produces changes in the protein conformation that promote receptor signaling. Importantly, some of the newly identified scaffolds that modulate the GLP-1R in a Cys-347-dependent manner may possess physicochemical properties suitable for various crystallography conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Author Contributions
*In the **Author Contributions** section:*
J. F. and O. C. designed, performed, and analyzed the islet insulin secretion studies. A. D. S. and C. S. performed and analyzed the GTPγS studies. A. D. S. and D. B. W. designed, performed, and analyzed the cAMP accumulation studies. A. M. developed and performed the GSH conjugation assays using LC/MS detection. F. S. W. developed, performed, and analyzed the fluorescein-maleimide competition assays.
A. B. B., F. S. W., and K. W. S. conceived, coordinated the study, designed the experiments, analyzed the data, and wrote the paper. All authors reviewed the results and approved the final version of the manuscript.[](https://www.ncbi.nlm.nih.gov/mesh/D016244)
The authors declare that they have no conflicts of interest with the contents of this article.
GPCR
G protein-coupled receptor
GLP-1R
glucagon-like peptide-1 receptor
TMD
transmembrane domain
PAM
positive allosteric modulator
GTPγS[](https://www.ncbi.nlm.nih.gov/mesh/D016244)
guanosine 5′-3-O-(thio)triphosphate[](https://www.ncbi.nlm.nih.gov/mesh/D016244)
ECD
ectodomain
th-BETP[](https://www.ncbi.nlm.nih.gov/mesh/C000612088)
4-(3-benzyloxyphenyl)-2-ethylsulfanyl-6-(trifluoromethyl)pyrimidine[](https://www.ncbi.nlm.nih.gov/mesh/C096834)
ANOVA
analysis of variance
ACN[](https://www.ncbi.nlm.nih.gov/mesh/C032159)
acetonitrile[](https://www.ncbi.nlm.nih.gov/mesh/C032159)
FA[](https://www.ncbi.nlm.nih.gov/mesh/C030544)
Formic acid[](https://www.ncbi.nlm.nih.gov/mesh/C030544)
Z
benzyloxycarbonyl
AMC[](https://www.ncbi.nlm.nih.gov/mesh/D003374)
amido-4-methylcoumarin[](https://www.ncbi.nlm.nih.gov/mesh/D003374)
CatL
cathepsin L
BETP[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
4-(3-benzyloxyphenyl)-2-ethylsulfinyl-6-(trifluoromethyl)pyrimidine[](https://www.ncbi.nlm.nih.gov/mesh/C570819)
EBSS
Earle's balanced salt solution buffer.
The abbreviations used are:
# References
*In the **References** section:*
|
# Introduction
Loss of GltB Inhibits Biofilm Formation and Biocontrol Efficiency of Bacillus subtilis Bs916 by Altering the Production of [γ-Polyglutamate](https://www.ncbi.nlm.nih.gov/mesh/C511775) and Three Lipopeptides
# Abstract
*In the **Abstract** section:*
Aims
This study examined the contribution of GltB on biofilm formation and biocontrol efficiency of B. subtilis Bs916.
Methods and Results
The gltB gene was identified through a biofilm phenotype screen and a bioinformatics analysis of serious biofilm formation defects, and then a gltB single knockout mutant was constructed using homologous recombination. This mutant demonstrated severe deficits in biofilm formation and colonisation along with significantly altered production ofγ-polyglutamate (γ-PGA) and three lipopeptide antibiotics (LPs) as measured by a transcriptional analysis of both the wild type B. subtilis Bs916 and the gltB mutant. Consequently, the mutant strain retained almost no antifungal activity against Rhizoctonia solani and exhibited decreased biocontrol e[fficiency again](https://www.ncbi.nlm.nih.gov/mesh/C511775)st[ rice](https://www.ncbi.nlm.nih.gov/mesh/C511775) sheath blig[ht. Very few gltB mutan](https://www.ncbi.nlm.nih.gov/mesh/D055666)t [cel](https://www.ncbi.nlm.nih.gov/mesh/D055666)ls colonised the rice stem, and they exhibited no significant nutrient chemotaxis compared to the wild type B. subtilis Bs916. The mechanism underlying these deficits in the gltB mutant appears to be decreased significantly in production of γ-PGA and a reduction in the production of both bacillomycin L and fengycin. Biofilm restoration of gltB mutant by additionγ-PGA in the EM medium demonstrated that biofilm formation was able to restore significantly at 20 g/L.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
Conclusions
GltB regulates biofilm formation by altering the production ofγ-PGA, the LPs bacillomycin L and fengcin and influences bacterial colonisation on the rice stem, which consequently leads to poor biocontrol efficiency against rice sheath blight.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
Significance and Impact of Study
This is the first report of a key regulatory protein (GltB) that is involved in biofilm regulation and its regulation mechanism and biocontrol efficiency by B. subtilis.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Bacillus biofilms consist of a highly structured extracellular matrix attached to the surface of cells [1–5] that play an important role in the biological control of multiple pathogens [6–8]; for example, the presence of a biofilm can significantly improve the colonisation of Bacillus amyloliquefaciens SQR9, and its biocontrol efficiency towards other microbes. Efficient rhizosphere colonisation and biofilm formation enable Paenibacillus polymyxa to control crown rot disease [9], while biofilm formation, colonisation, and secretion of surfactin by B. subtilis 6051 act together to protect plants against pathogenic attack [10].[](https://www.ncbi.nlm.nih.gov/mesh/C009365)
Although the presence of a biofilm is critical for biological disease control by Bacillus species, the mechanisms of biofilm formation, regulation and biocontrol remain unclear [6, 11–12]. The biofilm is primarily composed of a basic skeleton of amyloid fibres that consist of TasA protein, an exopolysaccharide (EPS), and BslA protein [10, 13–15]. In addition to TasA and EPS, γ-polyglutamate (γ-PGA) plays an intricate role in biofilm formation by different B. subtilis strains. While γ-PGA enhances biofilm formation by B. subtilis strain RO-FF-1, loss production of γ-PGA had no significant influence on biofilm formation because its function could be substituted by other biofactors [16].
In contrast, a mutant strain of B. amyloliquefaciens C06 defective in γ-PGA production was unable to form biofilms effectively [17]. Interestingly, however, overproduction of γ-PGA in B. amyloliquefaciens C06 could not enhance the biofilm formation, but instead inhibited it [18]. Previous studies also showed that many genes are involved in regulating the biofilms of Bacillus, including spo0A, spo0H, and abrB, which are needed for EPS and surfactin expression during early sporulation [19]. spo0A mutants demonstrate defective cell–cell interactions and cannot form a multicellular biofilm; abrB negatively regulates biofilm formation; sipW and yoaW, which are regulated by AbrB, are also required for normal biofilm formation [20–21]; and degQ regulates the production of fengycins and is necessary for normal biofilm formation
[22].[](https://www.ncbi.nlm.nih.gov/mesh/D011135)
In B. subtilis, lipopeptide antibiotics (LPs) significantly impact biofilm formation [10]. Loss of ArfB impairs the production of the LP arthrofactin in Pseudomonas sp. MIS38, but enhances biofilm production, although the biofilm is unstable and flat [23–24]. As a very effective biosurfactant, surfactin has been reported to be involved in biofilm formation on solid surfaces [9, 25–26]. Surfactin is also required for biofilm formation on liquid cultures by B. subtilis strain A1/3 [27].
Other studies have suggested that surfactin contributes to biofilm formation by secreting a signalling molecule for KinC activation of the Spo0A pathway [28–29]. The LP bacillomycin D contributes to biocontrol activity and to biofilm formation in B. amyloliquefaciens SQR9 [30]. We have previously used high performance liquid chromatography-mass spectroscopy (HPLC-MS) to identify three LPs (bacillomycin L, surfactin, and fengycin) secreted by B. subtilis Bs916 [31–33], and determine that both bacillomycin L and surfactin were necessary for biofilm formation by this strain [34–35]. While fengycin had no influence on biofilm formation, it played an important role in the biocontrol efficiency of B. subtilis Bs916 because of its antifungal properties [32].[](https://www.ncbi.nlm.nih.gov/mesh/D055666)
In this study, we report a critical regulatory protein in B. subtilis Bs916, named GltB, identified in a screen for altered biofilm phenotypes. After mutating the gltB gene in B. subtilis Bs916 by homologous recombination, the resulting biofilm was weak, thin and structurally distinct from that of the wild-type (WT) B. subtilis. Transcriptomic analysis revealed the production of γ-PGA and the three LPs, bacillomycin L, surfactin, and fengycin was closely related to biofilm formation; production of these molecules was also investigated in detail by HPLC-MS. Additionally, the antifungal activity, colonization ability and biocontrol efficiency of the gltB mutant against Rhizoctonia solani were also investigated. GltB markedly affected the production of γ-PGA and altered the secretion of the LPs surfactin, bacillomycin L and fengycin, which resulted in poor biofilm formation, colonisation and biocontrol efficiency by B. subtilis Bs916.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
## Materials and Methods
*In the **Materials and Methods** section:*
## Strains and culture conditions
*In the **Strains and culture conditions** section:*
The strains and plasmids used in this study are listed in Table 1. The B. subtilis Bs916 and the gltB mutant were cultured in Luria–Bertani (LB) broth medium [36] (10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl) at 37°C under shaking at 200 rpm. Escherichia coli DH5α cultured in LB broth medium was used to produce mutant vectors for eventual transformation into B. subtilis Bs916. R. solani was cultured at 28°C on potato dextrose agar (PDA) medium (200 g/L potato infusion, 20 g/L glucose, and 20 g/L agar at pH 7.0). Biofilm formation of both B. subtilis Bs916 and the gltB mutant was assessed using MSgg medium [19] [5 mM potassium phosphate (pH 7), 100 mM Mops (pH 7), 2 mM MgCl2, 700 μM CaCl2, 50 μM MnCl2, 50 μM FeCl3, 1 μM ZnCl2, 2 μM thiamine, 0.5% glycerol, 0.5% glutamate, 50 μg/mL tryptophan, 50 μg/mL phenylalanine] and EM medium [(1L, PH 7.4): 20g L- glutamate, 12g citric acid, 80g glycerine, 7g ammonium chloride, 0.5g magnesium sulfate heptahydrate, 0.5g dipotassium phosphate, 0.15g calcium chloride dehydrate, 40mg ferric chloride hexahydrate, and 148mg manganese sulphate monohydrate]
[37]. If necessary, antibiotics were added at the following concentrations: 100 mg/L spectinomycin, 100 mg/L ampicillin, and 5 mg/L chloramphenicol.[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
Bacterial strains and plasmids used in this study.
*Ampr: ampicillin resistance, Specr: spectinomycin resistance, CGMCC: China General Microbiological Culture Collection Centre, Beijing, China.[](https://www.ncbi.nlm.nih.gov/mesh/D000667)
## B. subtilis Bs916 mutant library construction and identification of the gltB gene
*In the **B. subtilis Bs916 mutant library construction and identification of the gltB gene** section:*
This method has been previously described in Scientia Agricultura Sinica [38]. Briefly, a mutant library of B. subtilis Bs916 was created using the plasmid pMarA carrying the transposon TnYLB-1, which resulted in random insertion at chromosomal loci. Single colonies from this mutant library were inoculated into 4 mL of LB medium containing spectinomycin to select for mutants and shaken for 12 h at 37°C, and single colonies of the control strain B. subtilis Bs916 were also cultured without spectinomycin. A 200 μl aliquot of the culture was inoculated into 4 mL of MSgg medium and was grown in a stationary setting for 24 h at 37°C.
Each sample was repeated three times. A visual inspection revealed whether there was an obvious change in biofilm formation between the mutant and the WT strains. After screening 5000 mutants, we identified the gltB mutant as having significant deficiencies, and we detected only a single insertion site by Southern Blot.[](https://www.ncbi.nlm.nih.gov/mesh/D000198)
## Construction of the ΔgltB mutant and the addition of GFP tags
*In the **Construction of the ΔgltB mutant and the addition of GFP tags** section:*
A spectinomycin resistance gene and its promoter were obtained from plasmid pDG1728 using the primers SpecF (5'-TTTGGATCCCTGCAGCCCTGGCGAATG-3') and SpecR (5'-TTTGAATTCAGATCCCCCTATGCAAGG-3'; BamHI and EcoRI restriction sites, respectively, are underlined). The complete spectinomycin cassette was isolated via digestion with BamHI and EcoRI and was cloned into the plasmid pUC19; the resulting plasmid was named pUCSpec. A 741-bp PCR product of gltB was cloned from the genome of WT B. subtilis 916 using the primers gltBF (5'-TTTAAGCTTTAACTGTCGACTTCGCGCG-3') and gltBR (5'-TTTGGATCCTGATGCCGTCATTTTGTGC-3'; HindIII and BamHI restriction sites, respectively, are underlined). Next, this PCR product was inserted into pUCSpec to create the pUCSpec-gltB plasmid, and the ΔgltB mutant was obtained by transforming the pUCSpec-gltB plasmid into WT B. subtilis 916. Finally, the positive emission green fluorescent protein (GFP) strains ΔBs916-gfp and ΔgltB-gfp were constructed by transforming the plasmid pRp22-gfp into both the WT B. subtilis 916 and the ΔgltB mutant.[](https://www.ncbi.nlm.nih.gov/mesh/D000198)
## Biofilm formation by WT B. subtilis Bs916 and the ΔgltB mutant
*In the **Biofilm formation by WT B. subtilis Bs916 and the ΔgltB mutant** section:*
To assess biofilm formation, a single colony of either WT B. subtilis 916 or the gltB mutant was inoculated into 4 mL of LB medium (containing spectinomycin only for mutants) and shaken overnight at 37°C. Approximately 1 mL of this culture was then inoculated into 50 mL of LB medium and shaken for 12 h at 37°C. Subsequently, 200 μl of this culture was inoculated into 4 mL of MSgg containing 20 μg/mL Congo Red and 10 μg/ml Coomassie brilliant blue in 12-well microtiter plates at 37°C for 24 h [13]. Finally, we transferred the biofilm of each mutant into 2 mL tubes, placed the tubes in a drying oven at 37°C, and weighed each sample.
Each experiment was repeated three times.[](https://www.ncbi.nlm.nih.gov/mesh/D000198)
## Analysis of genomic transcript levels of WT B. subtilis Bs916 and the ΔgltB mutant using Illumina HiseqTM2500
*In the **Analysis of genomic transcript levels of WT B. subtilis Bs916 and the ΔgltB mutant using Illumina HiseqTM2500** section:*
Total biofilm RNA of both WT B. subtilis 916 and the ΔgltB mutant cultured in a static MSgg medium (4 mL in 12-well microtiter plates) at 37°C for 24 h was extracted using an RNAprep pure Cell/Bacteria Kit. Qualified RNA was obtained according to the 2100 Bioanalyzer test, and then 10 μg of total RNA was digested using 5U DNaseI (Takara, Japan) for 30 min at 37°C. The RNA was then purified using an RNeasy MinElute Cleanup Kit (Qiagen, Germany) and eluted with 15 μl RNase-free water. All rRNA was removed using a Ribo-Zero™ Magnetic Kit (Gram-negative or Gram-positive bacteria; Epicentre, USA) in which the total reaction volume was brought to 40 μl by adding Ribo-Zero Reaction Buffer and Ribo-Zero rRNA Removal Solution, and the reaction was carried out at 68°C for 10 min and then at room temperature for 5 min. Treated RNA was added into pre-washed beads, mixed well, and incubated at room temperature for 5 min. Next, it was moved to 50°C for 5 min, immediately transferred to a magnetic rack for 1 min, and brought to a volume of 180 μl with water.
Next, 3 M sodium acetate, glycogen (10 mg/ml), and 600 μl ethanol were added to the supernatant, and the mixture was placed at -20°C for at least one hour. The rRNA-depleted RNA was then isolated via centrifugation, and the precipitate was dissolved in water. A 100-ng aliquot of rRNA-depleted RNA was used to construct a cDNA library with the NEB Next ® Ultra TM Directional RNA Library Prep Kit for Illumina (NEB, USA). The cDNA library was checked via Qubit fluorometric quantitation, 2% agarose gel electrophoresis, and a high-sensitivity DNA chip.
A cluster generation of the cDNA library (10 ng) was executed in cBot using the TruSeq PE Cluster Kit (Illumine, USA) and sequenced using a Illumina HiseqTM2500.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
Linker sequences that contained low quality clean reads were removed by fastx_clipper, and unwanted, low quality (>20 bases) reads from 3’- 5’ were removed using a FASTQ Quality filter (FASTX-Toolkit, v0.0.13, http://hannonlab.cshl.edu/fastx_toolkit/), thereby deleting all clean reads less than 50 bp. The quality of the sequencing data was assessed using Fastqc software (v0.10.0). Differential gene expression analysis was completed by mapping clean reads (Bowtie 2, v2.1.0) and using the MA-plot-based method with the random sampling model (version 1.20.0). The overall expression levels of each gene were calculated both in the WT Bs916 and ΔgltB mutant strains. When differentially expressed genes satisfied several conditions, including a fold change > 2, an FDR(q value) < 0.001, and at least one sample with an RPKM > 20, their expression was considered significantly altered, and these genes were set aside for further analysis.
## Antifungal properties of WT B. subtilis Bs916 and the ΔgltB mutant against R. solani
*In the **Antifungal properties of WT B. subtilis Bs916 and the ΔgltB mutant against R. solani** section:*
Mycelial plugs of R. solani (7 mm diameter) cultured in PDA medium for 3 d were placed in the centre of LB plates, and approximately 2 μl of cultured WT B. subtilis 916 or gltB mutant was dropped onto the surface of the circular filter paper (6 mm diameter) in an even distribution around the mycelial plugs. The plates were sealed with Parafilm and incubated at 28°C. The antibacterial bandwidth was measured after 24–72 h. Each experiment was repeated three times.[](https://www.ncbi.nlm.nih.gov/mesh/D000362)
## Detection of bacterial colonisation in the rice plant via confocal microscopy and determination of its biocontrol efficiency against rice sheath blight
*In the **Detection of bacterial colonisation in the rice plant via confocal microscopy and determination of its biocontrol efficiency against rice sheath blight** section:*
The bacterial cakes (6 mm diameter) of R. solani cultured for 3 d in solid PDA medium were inoculated into 100 mL of liquid PDA medium with 200 matchstick without commelina. The cakes were incubated in a 500-mL flask without shaking at 28°C for one week. Thirty seeds of rice cultivar were soaked in water for 24 h and then sown in the nursery containing sterile organic soil. At 5 d before rice heading, 10 matchsticks of R. solani were inoculated into 2–3 cm rice leaves of each rice plant. Approximately 20 mL of broth (cultured for 48 h) containing WT B. subtilis Bs916-gfp or the ΔgltB-gfp mutant were evenly sprayed into the rice plant marked by a matchstick.
From 0–15 d after the initial inoculation, the rice leaf marked by a matchstick was clipped and captured with a 40× objective microscope using a confocal microscope (Carl Zeiss LSM710, Germany). To determine the biocontrol efficiency against rice sheath blight, the lesion area length of the WT ΔBs916-gfp and the ΔgltB-gfp mutant were scored. Each treatment group contained eight individual pots.[](https://www.ncbi.nlm.nih.gov/mesh/D000362)
## Analysis of bacillomycin L, surfactin, and fengycin levels produced by WT B. subtilis Bs916 and the ΔgltB mutant
*In the **Analysis of bacillomycin L, surfactin, and fengycin levels produced by WT B. subtilis Bs916 and the ΔgltB mutant** section:*
A single colony of WT B. subtilis 916 or the gltB mutant was inoculated into 4 mL of LB medium (containing spectinomycin for mutants) and shaken overnight at 37°C, and then approximately 1 mL of the resulting culture was inoculated into 50 mL of LB medium and shaken for 48 h at 37°C. The supernatant of the WT B. subtilis 916 and the ΔgltB mutant cultures were isolated via centrifugation at 9,000 rpm for 20 min and then precipitated by adding an appropriate amount of hydrochloric acid to pH 2.0. The lipopeptide precipitates were collected via centrifugation at 12,000 rpm for 30 min, dried, and extracted using 2 mL of methanol before being passed through a 0.22 μm filter and then successfully separated using an Agilent 1200 high-performance liquid chromatography system with a photodiode array (HPLC-PDA) (Agilent Technologies, Santa Clara, USA) attached to a reversed-phase C18 column (5 μm, 4 mm × 250 mm; Merck, Frankfurt, Germany). The run was performed at a flow rate of 0.5 mL/min, a detection wavelength of 210 nm, and a gradient of solvents A (60% water containing 0.5% trifluoroacetic acid) and B (40% acetonitrile containing 0.5% trifluoroacetic acid) for bacillomycin L, A (20% water containing 0.5% trifluoroacetic acid) and B (80% acetonitrile containing 0.5% trifluoroacetic acid) for surfactin, and A (50% water containing 0.5% trifluoroacetic acid) and B (50% acetonitrile containing 0.5% trifluoroacetic acid) for fengycin [34].
The lipopeptide precipitates were further analysed using an Agilent 6410 Triple Quadrupole liquid chromatography/mass spectrometer (LC/MS) (Agilent Technologies, Santa Clara, CA, USA) in positive ion electrospray mode.[](https://www.ncbi.nlm.nih.gov/mesh/D000198)
## Determination of γ-polyglutamate (γ-PGA) and glutamate content of WT B. subtilis Bs916 and the ΔgltB mutant in the process of biofilm formation
*In the **Determination of γ-polyglutamate (γ-PGA) and glutamate content of WT B. subtilis Bs916 and the ΔgltB mutant in the process of biofilm formation** section:*
To assess the glutamate content of WT B. subtilis Bs916 and the ΔgltB mutant in the process of biofilm formation in the EM medium, the glutamic acid derivative reagent was obtained containing the following ingredients: 0.1g o-phthaldehyde (OPA), 2.5 mL methanol, 100 μL 2-mercaptoethanol, 22.4 mL 0.4 mol/L sodium borate buffer (PH 9.5). 1 g/L standard L-glutamate was prepared. 1 mL bacteria of biofilm below in 12-well microtiter plates was collected via centrifugation at 12,000 rpm for 5 min. 100 μL standard L-glutamate, supernatant, and EM medium were combined with 100 μL glutamic acid derivative reagent for 90 s respectively. Derivatized samples of the B. subtilis Bs916, ΔgltB mutant, standard L-glutamate, and EM medium were filtered by 0.22 μm membrane. In addition to standard L-glutamate and double dilution to B. subtilis Bs916, all treatments were diluted tenfold for detection by using an Agilent 1200 high-performance liquid chromatography system with a photodiode array (HPLC-PDA)
(Agilent Technologies, Santa Clara, USA) attached to a reversed-phase C18 column (5 μm, 4 mm × 250 mm; Merck, Frankfurt, Germany). The run was performed via the gradient elution (Table 2).[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Program of gradient elution.
Mobile phase A: Methanol, B: 10 mmol/L Phosphate buffer (PH 6.85). Detection wavelength: 338 nm. Column temperature: 30°C. Injection volume: 10μL.[](https://www.ncbi.nlm.nih.gov/mesh/D000432)
Method of acid hydrolysis was used to assess theγ-PGA content of WT B. subtilis Bs916 and the ΔgltB mutant in the process of biofilm formation. 200 μL overnight LB medium of the B. subtilis Bs916 and the ΔgltB mutant was inoculated into 4 mL of EM medium in 12-well microtiter plates at 37°C for 24 h. 4 mL bacteria of biofilm below in 12-well microtiter plates was collected via centrifugation at 12,000 rpm for 5 min. The supernatant was precipitated with three times ethanol for 12 h respectively. The precipitation was collected by 12000 rpm centrifugation for 30 min, and was dissolved with 4 mL distilled water. 2 mL 6N hydrochloric acid was added into 2 mL this dissolving liquid, then was allowed to stand at 110°C overnight without oxygen.
Method of glutamic acid detection was same as above.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
## Biofilm restoration of the ΔgltB mutant by addition γ-PGA in the medium
*In the **Biofilm restoration of the ΔgltB mutant by addition γ-PGA in the medium** section:*
Theγ-PGA added to the EM medium to the final concentration at 10 g/L, 20 g/L, and 30 g/L respectively were used to restore the biofim formation of ΔgltB mutant. 200 μL overnight culture of B. subtilis Bs916 or ΔgltB mutant were inoculated in 4 mL EM medium with or without addition γ-PGA respectively, and stand at 37°C for 24 h to form biofilm. Polyaspartic acid was added as a control. Each treatment has three duplicates and repeated three times. To further test biofilm formation of ΔgltB mutant, grew next to the WT B. subtilis Bs916 for 72 h.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
## Results
*In the **Results** section:*
## Detection of gltB transposon insertion by alterations in biofilm formation
*In the **Detection of gltB transposon insertion by alterations in biofilm formation** section:*
By screening 5000 mutants from the random insertion mutant library of the WT B. subtilis 916 by biofilm phenotype variety, we identified eight mutants that resulted in major changes in biofilm formation. In one mutant that contained the gltB transposon insertion, a significant and powerful defect in biofilm formation was observed compared to the WT.
By disrupting the gltB gene through homologous recombination, we successfully constructed a single knockout (ΔgltB mutant). Even after continuous culture in MSgg medium for 24 h, severe deficits in biofilm formation persisted. The ΔgltB mutant biofilm was thin and fragile and consisted only of a planar structure, as opposed to the clear three-dimensional structure observed in the WT B. subtilis Bs916 biofilm (Fig 1). The dry weight of the ΔgltB mutant was calculated as only one-third of the WT B. subtilis Bs916. The morphology of the mutant colonies also differed, with only a thin planar structure observed in the ΔgltB mutant compared with an obvious red bulge of the WT B. subtilis Bs916 (S1 Fig).
This phenotype remained stable over time, suggesting that the gltB gene plays an important role in the proper formation of a B. subtilis Bs916 biofilm.
Changes in biofilm formation due to the B. subtilis Bs916 and ΔgltB mutants in MSgg culture medium with 20 μg/mL Congo Red and 10 μg/mL Coomassie brilliant blue.[](https://www.ncbi.nlm.nih.gov/mesh/D003224)
(1) B. subtilis Bs916 biofilm was dense and solid with clear lines. In contrast, the ΔgltB mutant formed an uneven biofilm with an irregular shape. (2) The net weight of the ΔgltB mutant biofilm was more than three times less than the net weight of the WT B. subtilis Bs916 biofilm.
## Analysis of differentially expressed genes in WT B. subtilis Bs916 and ΔgltB mutant
*In the **Analysis of differentially expressed genes in WT B. subtilis Bs916 and ΔgltB mutant** section:*
Differentially expressed genes were identified by analysing gene transcription (Table 3), and the levels of bmy and fen were significantly lower in the ΔgltB mutant (approximately 5% and 7% of WT levels, respectively). The bmy and fen genes code for the biosynthesis of LPs bacillomycin L and fengycin, respectively. In contrast, the ΔgltB mutant exhibited approximately a five-fold increase in srfA transcription compared to WT B. subtilis Bs916. Another significant increase was noted in dhb transcription. This gene, which is responsible for the biosynthesis of the non-lipopeptide antibiotic bacillibactin, increased by 121-fold in the ΔgltB mutant and it may be a compensatory mechanism for the lack of bacillomycin L and fengycin; however, this change in expression had no impact on antibacterial activity or biocontrol efficiency against R. solani (Fig 2).
The ykuP-N-O gene cluster is identified as responsible for the flavodoxin synthesis, and mainly use to inhibit bacterial, especially Staphylococcus aureus, Streptococcus, Diphtheria, etc, and no significant effect on fungal pathogens. The ybdZ is generally considered for MbtH-like protein synthesis, such as the peptide siderophore coelichelin and the calcium-dependent peptide antibiotic (CDA) [39], and its function is similar to bacillibactin. The yktc1 encodes 4-aminobutyrate aminotransferase synthesis, its main function is responsible for the conversion between amino acids and keto acid, and may be included in the metabolic pathways of glutamate toγ-PGA. Although they are necessary raw materials for biofilm formation, the levels of TasA and EPS did not significantly change in the ΔgltB mutant.
Another important raw material for the synthesis of a biofilm, the levels of CapB (YwsC) was only one fifth of the WT B. subtilis Bs916, which was considered to be necessary for γ-PGA production [40]. It may be due to a lack of γ-PGA for bad biofilm formation in the ΔgltB mutant. Furthermore, the expression levels of several previously identified negative regulatory genes of biofilm formation, including abrB and sinR, were doubled in the ΔgltB mutant [20, 41–43]. In contrast, the expression of the positive regulatory genes spo0A was somewhat higher in the ΔgltB mutant, but the increase was less than that of the negative regulatory genes. These changes in gene transcription suggested that bmy, srfA, fen, and capB were all regulated by gltB, and their altered expression results in the significant deficit in biofilm formation observed in the mutant version of B. subtilis Bs916.[](https://www.ncbi.nlm.nih.gov/mesh/D055666)
Differences in antibacterial activity against R. solani between WT B. subtilis Bs916 and the ΔgltB mutant.
The ΔgltB mutant completely lost its antibacterial activity compared to the WT B. subtilis Bs916.
Analysis of differentially expressed genes in wild type B. subtilis Bs916 and the ΔgltB mutant.
## WT B. subtilis Bs916 had higher antibacterial activity and greater biocontrol efficiency against R. solani
*In the **WT B. subtilis Bs916 had higher antibacterial activity and greater biocontrol efficiency against R. solani** section:*
Basically no antibacterial activity was displayed by the ΔgltB mutant compared to a 5-mm antibacterial bandwidth of WT B. subtilis Bs916 (Fig 2). Unsurprisingly, the ΔgltB mutant also demonstrated poor biocontrol efficiency against rice sheath blight. WT B. subtilis 916 demonstrated 65.7% biocontrol efficiency on the third day when lesions on the rice stem were just appearing, which dropped to 36% by the fifteenth day. In contrast, the ΔgltB mutant only achieved 11.4% biocontrol efficiency on the third day, which is approximately 17% of the value for the WT. As the onset of rice sheath blight progressed, the ΔgltB mutant could only achieve a biocontrol efficiency of 1.7% by the fifteenth day, which was significantly lower than the 36% of the WT B. subtilis 916 at the same time point (Table 4).
Taken together, these results suggest that the ΔgltB mutant exerts little to no biocontrol effect against R. solani.
Biocontrol of rice sheath blight in pot cultures of WT B. subtilis 916 and the ΔgltB mutant strain.*
*Data are the average of five pots ± the standard deviation of three independent experiments with five pots. a, b, and c mean in the same column with different letters are significantly different (P<0.05) according to Duncan’s multiple range tests.
## Poor colonisation by the ΔgltB mutant
*In the **Poor colonisation by the ΔgltB mutant** section:*
The capacity for colonisation is a key factor in the prevention and treatment of fungal diseases because colonisation on host plants is closely related to biofilm formation in that strong colonisation often leads to good biofilm formation. Previous studies have suggested that mucoid mutants of the biocontrol strain Pseudomonas fluorescens CHA0, which have enhanced biofilm formation, also display significantly enhanced colonisation in carrot roots [44]. Using a GFP-labelling technique in this study, we assessed the colonisation abilities of the ΔgltB mutant and the WT B. subtilis Bs916 on a rice stem that had been inoculated with R. solani (Fig 3). After six days, the ΔgltB mutant demonstrated poor colonisation, not only because there were fewer bacterial cells but also because the mutant cells lacked significant nutrient chemotaxis. This was especially clear by day nine, when the WT B. subtilis Bs916 cells displayed strong green fluorescence and an obvious clustering effect, whereas only small scattering of green fluorescence was observed in the mutant cells via confocal microscopy.
Although the overall number of bacteria began to decline, this clustering effect was still clearly observed in the WT B. subtilis Bs916 through day twelve. Similar to day nine, only a small scattering of green fluorescence was observed in the ΔgltB mutant. By day fifteen, only a few bacterial cells were left in both the WT B. subtilis Bs916 and the ΔgltB mutant. The results suggest that the gltB gene is essential for the regulation of bacterial colonisation in B. subtilis Bs916, which may also be influenced by the strength of the biofilm formation and its γ-PGA production [17].[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
Colonisation of the rice plant against rice sheath blight.
Over time, the number of both WT B. subtilis Bs916 and ΔgltB mutant cells initially increased and then began to decrease. However, although WT B. subtilis Bs916 could produce normal clusters of cells, the ΔgltB mutant also demonstrated this same trend. By the last time point (i.e., 15 d), there were very few cells in either the WT B. subtilis Bs916 or the ΔgltB mutant, and the clustering of WT B. subtilis Bs916’s had disappeared.
## Disrupting gltB gene expression significantly altered the production of lipopeptide antibiotics bacillomycin L, surfactin, and fengycin in WT B. subtilis Bs916
*In the **Disrupting gltB gene expression significantly altered the production of lipopeptide antibiotics bacillomycin L, surfactin, and fengycin in WT B. subtilis Bs916** section:*
LPs play a major role in the control of rice fungal diseases such as rice sheath blight and rice blast. Based on our previous data and a comparison of our MS molecular weight data with other B. subtilis mass data, six peaks of surfactin, three peaks of bacillomycin L, and five peaks of fengycin were identified in the WT B. subtilis Bs916 by HPLC-MS [45–47] (Fig 4). However, neither bacillomycin L nor fengycin were detected when we analysed the ΔgltB mutant supernatants via HPLC. Surprisingly, five times the surfactin was detected in the ΔgltB mutant compared to the WT B. subtilis Bs916 (Fig 4).
Approximately 19.7 mg/L and 22.8 mg/L of bacillomycin L and surfactin, respectively, were produced in B. subtilis Bs916 after 36 h of cultivation in LB medium [34]. Although the production of fengycin in B. subtilis Bs916 was lower (3 mg/L), it still has a strong ability to inhibit fungal growth [32]. Although the surfactin production of the ΔgltB mutant was five times higher than WT B. subtilis Bs916, the increased production of surfactin could not compensate for the reduction in bacillomycin L and fengycin production, and therefore the ability of the ΔgltB mutant to inhibit the growth of R. solani was significantly reduced almost to zero. Taken together, these results suggest that gltB plays a key role in regulating the production of LPs, particularly bacillomycin L and fengycin, which play key roles in controlling rice sheath blight.[](https://www.ncbi.nlm.nih.gov/mesh/D055666)
Antibiotic secretion of bacillomycin L (a), fengycin (c), and surfactin (b). The ΔgltB mutant no longer secreted bacillomycin L or fengycin. In contrast, the ΔgltB mutant produced significantly higher (approximately five times higher) levels of surfactin compared to the WT B. subtilis Bs916.[](https://www.ncbi.nlm.nih.gov/mesh/C015015)
## Glutamate consumption and γ-PGA production of ΔgltB mutant and WT B. subtilis Bs916
*In the **Glutamate consumption and γ-PGA production of ΔgltB mutant and WT B. subtilis Bs916** section:*
As the Fig 5 and Table 5 shown, the WT B. subtilis Bs916 was able to consume glutamate efficiently and decrease the concentration of glutamate from 20 g/L to the 2.68 g/L after 24 h culture. Inversely, the ΔgltB mutant was unable to take advantage of glutamate as efficiently as the WT B. subtilis Bs916 and only decreased to 10.41 g/L. As expect, the production of γ-PGA of B. subtilis Bs916 was able to reach 13.46 g/L after 24 culture. However, the production of γ-PGA of ΔgltB mutant was only reach 2.56 g/L after 24 h culture. These results were consistent with genomic transcript levels and strongly suggested that the mutation of gltB not only impaired the glutamate consumption significantly and but also decreased γ-PGA production of WT B. subtilis Bs916 sharply.
In addition, our results also suggested the glutamate consumption and γ-PGA production showed a positive relationship.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
γ-PGA and glutamate content of WT B. subtilis Bs916 and the ΔgltB mutant in the process of biofilm formation.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
For glutamate detection, in addition to standard L-glutamate and double dilution to B. subtilis Bs916, all treatments were diluted tenfold for detection. For γ-PGA detection, all treatments were also diluted tenfold for detection.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Glutamate consumption and γ-PGA production of ΔgltB mutant and WT B. subtilis Bs916.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
a, b, and c mean in the same column with different letters are significantly different (P<0.05) according to Duncan’s multiple range tests.
## Biofilm restoration in the ΔgltB mutant
*In the **Biofilm restoration in the ΔgltB mutant** section:*
Due to the sharply decrease in the production ofγ-PGA in the ΔgltB mutant, the restoration of biofilm by additionγ-PGA were investigated in detail. The results showed that the biofilm of ΔgltB was restored significantly both at the concentration of 10 g/L and 20 g/L of additionγ-PGA (Fig 6 and Table 6). Interestingly, when the concentration ofγ-PGA over 30 g/L, the biofilm formation of not only the ΔgltB mutant but also WT B. subtilis Bs916 were inhibited. This phenomena was also observed by Liu et al previously [18]. By plate confrontation growth for biofilm restoration in ΔgltB mutant, next to the WT B. subtilis Bs916 for 72 h, biofilm formation of ΔgltB mutant was no significant restoration before 36 h.
But in 72h, biofilm formation had been restored to most parts (Fig 7). The control strain was only ΔgltB mutant, and its biofilm was still very poor. Taken together, we presumed that the loss of GltB inhibited biofilm formation of B. subtilis Bs916 maybe by altering the production of γ-PGA via the ability to consume glutamate.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
Biofilm restoration in the ΔgltB mutant.
A final concentration of 10 g/L, 20 g/L, and 30 g/Lγ-PGA solution were added into the ΔgltB mutant in 4mL EM medium. The final concentration of 10 g/L and 20 g/Lγ-PGA were able to recover biofilm formation in the ΔgltB mutant, however, high final concentration of 30 g/Lγ-PGA inhibited biofilm formation in WT B. subtilis Bs916 and the ΔgltB mutant.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
Biofilm restoration by B. subtilis Bs916’sγ-PGA in the ΔgltB mutant.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
The ΔgltB mutant grew next to B. subtilis Bs916 on solid EM plate with 1.5% agar to observe their biofilm formation. Biofilm formation of ΔgltB mutant was no significant restoration before 36 h. But in 72h, its biofilm formation had been restored to most parts.[](https://www.ncbi.nlm.nih.gov/mesh/D000362)
Biofilm dry weight analysis of Bs916 and ΔgltB mutant by adding γ-PGA and polyaspartic acid.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
## Discussion
*In the **Discussion** section:*
Rice sheath blight is a common fungal disease that affects rice globally, and it has resulted in more than a 15% production yield loss in China’s major rice growing areas [47–48]. In addition to the use of chemical pesticides to prevent the disease, B. subtilis has been widely used as a major biocontrol pesticide to prevent and control rice sheath blight [49–50]. The complete genome of B. subtilis Bs916 has been sequenced [33, 35], and sequence analyses along with detection of secondary metabolites showed that B. subtilis Bs916 was able to produce four LPs: bacillomycin L, surfactin, fengycin and locillomycin [34–35]. Here, we report a new and critical gene gltB which markedly affects biofilm formation, bacterial colonization and the biocontrol efficiency of B. subtilis Bs916 by regulating the production of γ-PGA and LPs.[](https://www.ncbi.nlm.nih.gov/mesh/D055666)
The ΔgltB mutant showed serious flaws in its ability to form biofilms; the resulting biofilms were weak and thin with an irregular shape and a two-dimensional structure that differed greatly from the three-dimensional structure of the biofilms of the WT. To elucidate the regulatory mechanisms of gltB in biofilm formation, transcriptome analysis was used to identify differentially expressed genes between WT B. subtilis 916 and the ΔgltB mutant. The expression of capB (ywsC) was only one-fifth of the level of that in WT B. subtilis 916, and it was considered to control the biosynthesis of γ-PGA. Additionally, expression of the synthetic gene clusters for bacillomycin L, surfactin, and fengycin were also significantly altered. γ-PGA and the LPs are very important to biofilm formation
[16–17, 39]. Therefore, we proposed that gltB regulates biofilm formation via a γ-PGA-dependent pathway and the production of the three LPs.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
Our results strongly suggested that disruption of gltB resulted in impairment of the ability to assimilate glutamate and led to decreased γ-PGA biosynthesis. On the basis that γ-PGA was able to restore biofilm formation by the gltB mutant, we concluded that γ-PGA plays an important role in biofilm formation by B. subtilis 916. As we reported previously, surfactin was able to restore the biofilm formation of a bacillomycin L mutant [35]; therefore we think that the reduction of the bacillomycin L level in the ΔgltB mutant did not have a significant influence on biofilm formation by this strain because its increased production of surfactin would compensate for the decrease in bacillomycin L. In general, we confirmed that disruption of gltB led to deficiency in biofilm formation.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Colonization on rice surfaces and antifungal activity of B. subtilis 916 ensure its biocontrol efficiency against rice diseases. Biofilm formation is a prerequisite for bacterial colonization on plant surfaces and biocontrol efficiency [47, 51]. Bacillomycin L and fengycin are the main contributors to the antifungal activity of B. subtilis 916 [34–35]. Surfactin is main agent against bacterial pathogens such as Pectobacterium carotovorum, Xanthomonas campestris and Podosphaera fusca [48].
In our study, we conclusively proved that production of γ-PGA contributed to biofilm formation and verified that bacillomycin L and fengycin contributed to the antifungal activity of B. subtilis 916. Using confocal microscopy to detect colonisation of ΔgltB on rice plants and pot experiments to determinate its biocontrol efficiency, we showed that the ΔgltB mutant demonstrated poor colonisation, including fewer bacterial cells and a lack of significant nutrient chemotaxis. These results suggest that colonisation by B. subtilis Bs916 was regulated by GltB; it may also be influenced by the strength of the biofilm formation and the level of γ-PGA production [17]. The ΔgltB mutant also demonstrated poor biocontrol efficiency against rice sheath blight. So, we confirmed that biocontrol efficiency of B. subtilis Bs916 was regulated by GltB, mainly via the production of bacillomycin L and fengycin.[](https://www.ncbi.nlm.nih.gov/mesh/C015015)
In conclusion, the production of γ-PGA contributes to biofilm formation. Bacillomycin L and fengycin contribute to the antifungal activity of B. subtilis 916 against R. solani. GltB regulates biofilm formation by altering the production of γ-PGA and the LPs bacillomycin L and fengycin, and influences bacterial colonisation of rice stems, which consequently leads to poor biocontrol efficiency of rice sheath blight.[](https://www.ncbi.nlm.nih.gov/mesh/C511775)
## Supporting Information
*In the **Supporting Information** section:*
# References
*In the **References** section:*
|
# Introduction
Pinpointing a Mechanistic Switch Between Ketoreduction and “Ene” Reduction in Short‐Chain Dehydrogenases/Reductases
# Abstract
*In the **Abstract** section:*
Abstract
Three enzymes of the Mentha essential oil biosynthetic pathway are highly homologous, namely the ketoreductases (−)‐menthone:(−)‐menthol reductase[ and (−)‐menthone:(+](https://www.ncbi.nlm.nih.gov/mesh/D009822))‐neomenthol reductase, and the “ene” reductase isopiperitenone reductase. We identified a rare catalytic residue substitution in the last two, and performed comparative crystal structure analyses and residue‐swapping mutagenesis to investigate whether this determines the reaction outcome. The result was a complete loss of native activity and a switch between ene reduction and ketoreduction. This suggests the importance of a catalytic glutamate vs. tyrosine residue in determining the outcome of the reduction of α,β‐unsaturated alkenes, due to the substrate o[ccupying ](https://www.ncbi.nlm.nih.gov/mesh/D018698)diffe[rent bin](https://www.ncbi.nlm.nih.gov/mesh/D014443)ding conformations, and possibly also to the relative ac[idities of the two resi](https://www.ncbi.nlm.nih.gov/mesh/D000475)dues. This simple switch in mechanism by a single amino acid substitution could potentially generate a large number of de novo ene reductases.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Biosynthetic enzymes involved in the production of menthol oil have been investigated for their biological function and biocatalytic potential because of the high commercial demand for this oil (ca. 31 000 t/US‐$ 373–401 million per year).1 Three salutaridine/menthone reductase like subfamilies of short‐chain dehydrogenases/reductases (SDRs)2 from Mentha piperita were identified, namely (−)‐menthone:(−)‐menthol reductase (MMR), (−)‐menthone:(+)‐neomenthol reductase (MNMR), and isopiperitenone reductase (IPR).1a,1b MMR and MNMR catalyze the ketoreduction of (−)‐menthone 1 a to (1R,2S,5R)‐menthol 2 a and (1S,2S,5R)‐neomenthol 2 b, respectively (Scheme 1).1a Additionally, they act on isomenthone 1 b to produce (1R,2R,5R)‐neoisomenthol 2 c and (1R,2S,5S)‐isomenthol 2 d, respectively. In contrast, IPR catalyzes double bond reduction of isopiperitenone 3 a to cis‐isopulegone 4 a (Scheme 1).1b[](https://www.ncbi.nlm.nih.gov/mesh/D008610)
Reactions catalyzed by MMR, MNMR, and IPR.1a,1b
The enzymes of the SDR superfamily are characterized by large sequence divergences (>15 % homology), yet show highly conserved three‐dimensional structures2 and an active‐site tetrad typically containing Ser, Tyr, Lys, and Asn.2, 3, 4, 5 Interestingly, the three Mentha SDRs have high protein‐sequence identities (63–68 %), so we performed comparative studies of MMR, MNMR, and IPR, to pinpoint the determinants of ketoreductase vs. ene‐reductase activity within SDRs.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
Kinetic studies of MMR, MNMR, and IPR (see the Supporting Information for details; Figures S1–S3; Table S1)1c and biotransformations were performed with a variety of (a)cyclic, aromatic, and monoterpenoid enones, enals, and enols (Table 1; Figure S4). In some cases, where not commercially available, product standards were synthesized to confirm whether ene reduction or ketoreduction had occurred. Double bond reduction by IPR was seen for six α,β‐unsaturated cyclic enones (3 a,b and 5 a–d) to produce the corresponding unsaturated ketones (4 a,b and 6 a–d, respectively; Table 1, entries 1–6; Scheme 1). The highest yields were obtained with the (+/−)‐isopiperitenone mixture 3 a (50 % ee) to produce nearly equivalent amounts of cis/trans‐isopulegone 4 a diastereomers. Isophorone 5 b and ketoisophorone 5 c were also reduced with high yields and enantiopurity (Table 1, entries 4 and 5).
However, the predominant enantiomer of 6 c generated was (S)‐levodione, opposite to (R)‐6 c generation by the Old Yellow Enzyme (OYEs) ene reductases.5, 6 Low activity was detected with (R)‐piperitenone 3 b generating enantiopure (R)‐pulegone 4 b. No ketoreduction was observed with any substrate tested.[](https://www.ncbi.nlm.nih.gov/mesh/D000447)
Biocatalytic reduction of cyclic ketones by three SDRs.[a][](https://www.ncbi.nlm.nih.gov/mesh/D007659)
[a] Reactions (1 mL) were performed in buffer (50 mm KH2PO4 pH 6.0 for IPR; 50 mm Tris pH 7.0 for MMR/MNMR) containing monoterpenoid (1 a,b, 3 a,b, 5 a–d; 5 mm), enzyme (5 μm), NADP+ (10 μm), glucose (15 mm), GDH (10 U), and enzyme (2 μm). The reaction solutions were agitated at 25 °C for 10 h at 130 rpm. Product identification was performed by both comparing retention times with authentic standards and identification by GCMS on a DB‐WAX column (only GCMS identification for product 6 c). MMR and MNMR data were obtained from previously published work.1c [b] Product yield and enantiomeric excess were determined by GC analysis using DB‐WAX and Chirasil‐DEX‐CB columns, respectively. [c] Lacking enantiopure product standards to assign diastereomeric/enantiomeric identity. [d] nd=not determined due to low product yield.[](https://www.ncbi.nlm.nih.gov/mesh/C013216)
Biotransformations with MMR and MNMR showed only ketoreduction products, with no detectable double bond reduction (Figure S4).1a,1b Reactions with 1 a and 1 b generated the menthol isomers 2 a–d (Table 1, entries 7, 8, 10, and 11). The product ee values were medium to high (72 to >99 %). The only other ketoreduction seen was the slow conversion of 5 a to 7 by MNMR (5 % yield; Table 1, entry 9).[](https://www.ncbi.nlm.nih.gov/mesh/D008610)
A sequence alignment of the three ketoreductases MMR, MNMR, and salutaridine reductase (SalR; 45–49 % homology to Mentha enzymes) from Papaver somniferum L with IPR showed each enzyme contained typical SDR‐like motifs, such as those involved in central β‐sheet stabilization, and a TGxxxGhG motif (Figure S5).4b The latter motif in the Mentha enzymes contains the motif TGxxKGIG, predictive of a preference for NADP(H) over NAD(H).7 A key difference in the sequences between the ketoreductases and IPR was a substitution of the highly conserved catalytic Tyr residue for Glu (238 in IPR). An further sequence alignment of over 500 SDRs revealed only four additional enzymes had substitutions of the active‐site Tyr residue (results not shown). One of these enzymes was IPR from a related Mentha sp., which also contained an active‐site Glu. Interestingly, the aldo–keto reductase superfamily contains both ketoreductases (e.g. aldose reductase) and double bond reductases (e.g. Δ4‐3‐ketosteroid 5β‐reductase) with high sequence homologies.8 In this case, a substitution of an active‐site His for a Glu residue discriminated between ketoreduction and double bond reduction.9 Therefore, we investigated the role of the different catalytic acid residues in IPR (Glu 238) and MNMR (Tyr 244) in the reaction mechanism.[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
We determined crystal structures of both MNMR and IPR (Figure 1), the latter in combination with NADP+, alkene 3 a, and β‐cyclocitral (non‐substrate). Crystallographic methodology, data refinement statistics, and detailed structural descriptions can be found in the Supporting Information (Table S2 and associated discussion). The crystal structures of apo‐IPR and the 3 a‐ and β‐cyclocitral‐bound complexes were solved by molecular replacement using the known SalR crystal structure (PDB 3O26; 1.2 and 1.7 Å resolution, respectively; Table S2).10 The presence of clear density in the F0−Fc map for NADP+ (Figure 1 B) suggested IPR had scavenged it from host cells during protein expression. Substrate 3 a was bound to the active with the C=C bond close to, and parallel with, the nicotinamide ring of NADP+, and close (3.19 Å) to the site of hydride transfer (Figure 1 B right and Figure S 6 a).
The carbonyl oxygen atom of 3 a hydrogen bonds with Glu 238 and the highly conserved Ser 182 and sits at an equal distance (3.15 Å) from both residues. A water molecule hydrogen bonds with Glu 238 and the ribose ring of NADP+, suggesting a mechanistic role for this water molecule. Conserved residue Asn 154 is hydrogen‐bonded to Glu 238, while Lys 242 forms hydrogen bonds to the ribose of NADP+ and a water molecule, indicating its role in stabilizing NADP+ and contributing to a proton relay.3 The IPR–β‐cyclocitral co‐crystal structure shows the ligand binds in a non‐active conformation compared to 3 a binding (Figure S 6 b). No other major changes in residue positions were observed in the co‐crystal structures.[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
Crystal structure analyses of IPR and MNMR. A) Overlay of IPR (blue; PDB code 5LCX) and SalR (coral; PDB code 3O26) structures. The flap domains of IPR and SalR are indicated by dotted lines. NADP+ is displayed as ball and stick and colored by atom type. B) Left: overlay of IPR (gray; PDB code 5LDG) and MNMR (yellow; PDB code 5L53) structures. Right: active site showing side chains of some active‐site residues of IPR and MNMR along with 3 a (cyan) and NADP+. The Figure was prepared using CCP4mg.11[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
The MNMR structure was solved by molecular replacement using IPR as the search model (resolution 2.3–2.7 Å; Table S2), and was found to be structurally similar (rmsd 0.97 Å; Figure 1 B left). A coenzyme‐bound MNMR structure was obtained by soaking crystals with NADP+, however no structures were obtained with 1 a,b within the active site. Major structural differences were not observed between apoprotein and NADP+‐bound forms. Additional discussion on the crystal structures of the Mentha enzymes and related proteins is found in the Supporting Information (Figures S7–S9).[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
As expected, the conserved Glu 238 of IPR occupied the position of Tyr 244 in MNMR, with a distance between Cβ of 3 a and the NADPH hydride of 3.18 Å in the co‐crystal structure. Therefore the bulkier MNMR Tyr 244 likely positions substrates in a different conformation compared to that observed for 3 a in IPR (Figure 1 B right) because of the larger side‐chain bulk of tyrosine. This is consistent with the helix of the flap domain (MNMR) being shifted compared to that in IPR (Figure 1 B right), to accommodate binding of 1 a,b. This structural comparison suggests that this rare residue substitution might be responsible for the switch in activity seen for IPR to NADPH‐dependent 1,4 conjugate reduction of the α,β‐unsaturated carbonyl compound 3 a to 4 a.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Based on prior mechanistic studies and our structural studies, we propose mechanisms of action for both ketoreduction (MNMR) and double bond reduction (IPR) in SDRs.3, 4, 7, 12 Ketoreduction follows typically an ordered “bi‐bi” mechanism, where the coenzyme binds first and leaves last. MNMR appears to follow this classical SDR ketoreduction mechanism for 1 a to 2 b and 1 b to 2 d (Scheme 2 A).1a The alcohol product is formed by the transfer of a hydride from NADPH to the carbonyl carbon atom of the substrate with facial selectivity. In the case of SDRs, the 4‐pro‐S hydride is transferred, in contrast to MDRs that catalyze 4‐pro‐R hydride transfer.4a Concurrent with hydride attack, the carbonyl oxygen atom takes a proton from the conserved Tyr 244 residue acting as a catalytic acid. This starts a cascade of proton transfers through the NADP+ coenzyme and Lys 248, terminating with removal of a proton from a water molecule.
The conserved Ser 188 residue likely functions to stabilize the substrate, while Lys 248 hydrogen bonds with the nicotinamide ribose moiety, lowering the pK a of the Tyr 244‐OH to promote proton transfer.3 Residue Asn 160 in SDRs interacts with the conserved Lys 248 and bulk solvent via water molecule(s), forming a protein relay or hydrogen‐bonded solvent network (Scheme 2 A). This likely helps to stabilize the position of Lys 248, thereby assisting the overall ketoreduction mechanism.3[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
Proposed mechanisms of A) ketoreduction by MNMR and B) reduction of an α,β‐unsaturated double bond by IPR. The three‐dimensional nature of the active sites is represented as compounds in the foreground and background shown in black and grey, respectively.
The structure of the IPR–3 a co‐crystal reveals that Glu 238 positions the substrate to allow hydride addition at the C=C bond of 3 a, rather than the carbonyl carbon atom. In the proposed IPR double bond reduction mechanism, hydride transfer from NADPH to the 4‐position of the α,β‐unsaturated carbonyl system of 3 a results in formation of the respective enolate ion (Scheme 2 B), which then accepts a proton from the conserved residue Glu 238 to generate the more stable enol. Residue Glu 238 abstracts a proton from a nearby water molecule that may initiate a similar proton transfer cascade to that seen in MNMR. Formation of cis‐isopulegone 4 a then proceeds by Glu 238 abstracting the proton, previously donated to the substrate, resulting in re‐formation of the carbonyl group. Alternatively a nonenzymatic water‐mediated step may occur.
Concomitantly, the enolate double bond accepts a proton from water, giving the 1,4 conjugate reduction product (Scheme 2 B). This mechanism is possible in IPR as the side chain of Glu 238, unlike the Tyr side chain, readily dissociates to its conjugate base in water.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
To test this hypothesis further, we generated the variants IPR E238Y and MNMR Y244E and performed biotransformation reactions to detect ketoreduction and/or double bond reduction (Table 2). We tested IPR E238Y at pH 6.0, consistent with the preference for lower pH values of the wild‐type enzymes, in addition to reactions at pH 7.0 for comparison with the MNMR Y244E variant. IPR E238Y showed no double bond reduction with any substrate tested (3 a,b and 5 a–d), however it performed minor ketoreduction with substrate 3 a to form the equivalent alcohol products 8 a (Table 2, entries 1 and 2). Additionally it showed MNMR‐like activity towards Mentha compounds 1 a,b, forming primarily 2 b and 2 d, respectively (Table 2, entries 3–6), although the product yields and enantiopurity were lower than with wild‐type MNMR. Interestingly, reactions with 1 b at pH 7.0 generated a slightly higher yield of products, but they were obtained in near racemic form (Table 2, entry 6.
Therefore, replacing of active‐site Glu by Tyr has converted the enzyme from an ene reductase into a ketoreductase, albeit with lower catalytic efficiency and enantiospecificity.[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
Biocatalytic reduction of cyclic ketones by enzyme variants IPR E238Y and MNMR Y244E.[a][](https://www.ncbi.nlm.nih.gov/mesh/D007659)
[a] Reactions (1 mL) were performed in buffer (50 mm KH2PO4 pH 6.0 for IPR; 50 mm Tris pH 7.0 for MNMR and IPR) containing monoterpenoid (1 a,b, 3 a,b, and 5 a–d; 5 mm), enzyme (5 μm or 10 μm for IPR and MNMR, respectively), NADP+ (10 μm), glucose (15 mm), GDH (10 U), and enzyme (2 μm). The reaction solutions were agitated at 25 °C for 24 h at 130 rpm. Product identification was performed by both comparing retention times with authentic standards and identification by GCMS on a DB‐WAX column (only GCMS identification for product 8 a). Figure S10 gives the GCMS spectra traces of the additional products and their respective substrates. [b] Product yield and enantiomeric excess were determined by GC analysis using DB‐WAX and Chirasil‐DEX‐CB columns, respectively. nd=not determined due to low product yield. [c] Other isomer formed (20 % yield) was 2 a. [d] Other isomer formed (2 % yield) was 2 c.[](https://www.ncbi.nlm.nih.gov/mesh/C013216)
In the case of MNMR variant Y244E, ketoreduction was not seen with any substrate tested (1 a,b, 3 a,b, and 5 a–d). Minor double bond reduction was detected with substrate 5 c to form 6 c (Table 2, entry 7). MMR and MNMR are known to have narrower substrate specificities than IPR1a (Table 1 and Figure S4), suggesting further mutations are required to form a more active ene reductase.
Interestingly, studies with mechanistically different enzymes of the class I aldolase family (transaldolase and fructose‐6‐phosphate aldolase) have shown that the change of the nature of the catalytic acid/base can have a significant effect on the reaction mechanism.14, 15 However, the effect of active‐site spacial changes by residue substitution needs to be considered. For example the lack of ketoreduction of wild‐type IPR with 3 a and 3 b may be due to a preference for binding in a conformation consistent with double bond reduction, while the steric bulk of Tyr in IPR variant E238Y may orient the substrate in a position suitable for ketoreduction. Further studies will be needed to determine the relative contribution of catalytic residue type vs. steric constraints in determining the overall mechanism of the catalysis.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
We have pinpointed a simple mechanistic switch between ene‐reductase and ketoreduction activity in the SDR superfamily. This simple mechanistic switch, in addition to other residue substitutions to improve catalytic efficiency, could potentially transform SDR ketoreductases into novel ene reductases and provide attractive routes to novel ene‐reductase catalysts. This would reduce the dependence on traditional FMN‐containing OYEs for the biocatalytic reduction of α,β‐unsaturated alkenes and complications (reaction rates, yields, and product enantiopurity) that arise when OYEs are affected by molecular oxygen.13 Access to a new class of ene reductases would open up the possibility of developing new catalytic specificities typical of the SDR superfamily for the reduction of α,β‐unsaturated alkenes.[](https://www.ncbi.nlm.nih.gov/mesh/D000475)
## Supporting information
*In the **Supporting information** section:*
|
# Introduction
The Study of Dynamic Potentials of Highly Excited Vibrational States of [DCP](https://www.ncbi.nlm.nih.gov/mesh/D003903): From Case Analysis to Comparative Study with [HCP](https://www.ncbi.nlm.nih.gov/mesh/D009943)
# Abstract
*In the **Abstract** section:*
The dynamic potentials of highly excited vibrational states of deuterated phosphaethyne (DCP) in the D–C and C–P stretching coordinates with anharmonicity and Fermi coupling are studied in [this article and the res](https://www.ncbi.nlm.nih.gov/mesh/D003903)ul[ts ](https://www.ncbi.nlm.nih.gov/mesh/D003903)show that the D-C-P bending vibration mode has weak effects on D–C and C–P stretching modes under different Polyad numbers (P number). Furthermore, the dynamic potentials and the corresponding phase space trajectories of DCP are given, as an example, in the case of P = 30. In the end, a comparative study between deuterated phosphaethyne (DCP) an[d p](https://www.ncbi.nlm.nih.gov/mesh/D003903)hosphaethyne (HCP) with dynamic potential is done, and it is elucidated that the uncoupled[ mode makes the original](https://www.ncbi.nlm.nih.gov/mesh/D003903) h[ori](https://www.ncbi.nlm.nih.gov/mesh/D003903)zontal[ reversed sym](https://www.ncbi.nlm.nih.gov/mesh/D009943)me[try](https://www.ncbi.nlm.nih.gov/mesh/D009943) breaking between the dynamic potential of HCP () and DCP (), but has little effect on the vertical reversed symmetry, between the dynamic potential of HCP () and DCP ()[.](https://www.ncbi.nlm.nih.gov/mesh/D009943)(https://www.ncbi.nlm.nih.gov/mesh/D003903)
## 1. Introduction
*In the **1. Introduction** section:*
Resonance coupling between the different vibrational modes of molecules, which typically increases with energy, makes triatomic molecules quite intricate. The ways of studying the resonance coupling effect between the different modes in a triatomic molecule are ab initio calculations and semi-classical methods. In recent years, a new semi-classical method, named dynamics potential, has been proposed and has been applied to study highly excited molecular vibrational states. This method could, not only verify the conclusions given by ab initio calculations, but also show visual physical pictures, including molecular isomerization, chaotic dynamics, dissociation dynamics, and other information.
The internal interaction between D–C stretching and C–P stretching in DCP (deuterated phosphaethyne) has attracted a great deal of attention, since the information involved in the interaction is significant for understanding the mechanisms of chemical reactions. In previous articles, we analyzed the dynamic features of deuterated phosphaethyne (DCP) and phosphaethyne (HCP) using dynamic potentials. Because of the drastic change of atomic masses of DCP compared with HCP, instead of the resonance between C–P stretching and D–C–P bending, a 2:1 D–C stretching and C–P stretching resonance governs the DCP spectrum. It is shown that there is dynamic symmetry between DCP and HCP systems, which is significant to analyze the features of homologous compounds.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
In this work, the dynamic potentials of highly excited vibrational states and phase space trajectories of DCP are studied. The effect of the D–C–P bending vibration mode on the D–C and C–P stretching modes, under different Polyad numbers, are also investigated. Finally, a comparative study between DCP and HCP is done to clarify the symmetry breaking of dynamic potentials in DCP and HCP systems with the effects of uncoupled modes, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
## 2. The Semi-Classical Hamiltonian of the DCP
*In the **2. The Semi-Classical Hamiltonian of the DCP** section:*
The dynamic properties of DCP molecules’ highly-excited vibrational states, in the energy region 1.97 × 104–2.35 × 105 cm−1, are essential, and the corresponding Hamiltonian could be obtained as following:[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
The corresponding coefficients of the DCP Hamiltonian are shown in Table 1, where subscripts 1, 2, and 3, correspond to the D–C stretching vibration mode, D–C–P bending vibration mode, and C–P stretching vibration mode, respectively. We will use n to denote the corresponding vibration mode, which will be indicated with in the position coordinate, and the momentum coordinate indicated with . is the corresponding harmonic vibration coefficient, while Xij, yijm, zijmn denote the nonlinear coupling coefficients of different modes (Xij ~ coefficient of the two nonlinear coupling modes, yijm ~ coefficient of the three nonlinear coupling modes, zijmn ~ coefficient of the four nonlinear coupling modes). k, λ1, λ3, and μ11 represent the Fermi resonance strength coefficient, with regard to the quantum numbers of the three vibrational modes. Besides , there is another conserved action called Polyad number P = 2 + (P number). Equation (1) is used to study the dynamic properties of highly excited vibrational states in the region of ≤ 4, P ≤ 30.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
The coset space SU(2)/U(1) could be used as the representing space of Hamiltonian and it could be rewritten in the coordinates (, ) indicates with semi-classical representations as follows:
With the coordinate (), the Hamiltonian can be written as:
The semi-classical Hamiltonian, mentioned above, could further be used to obtain dynamic potentials, which are necessary for studying the dynamic nature of the DCP’s highly excited vibrational states.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
## 3. The Dynamic Features of the Highly Excited Vibration States of DCP
*In the **3. The Dynamic Features of the Highly Excited Vibration States of DCP** section:*
Two main parts would be addressed in the following: (1) the influences of bending modes to D–C and C–P stretching modes; (2) the phase space trajectories for each energy levels in the dynamic potentials when P = 30 (as a case study); and (3) the comparative study between DCP and HCP in the sense of symmetry of dynamic potential.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
## 3.1. The Dynamic Potentials Corresponding Typical Polyad Number with Different Quantum Number
*In the **3.1. The Dynamic Potentials Corresponding Typical Polyad Number with Different Quantum Number** section:*
The case of small P number will be firstly discussed. We take P = 18 as instance and the dynamic potential is shown as Figure 1 (the rule of marking fixed points is the same with the literature). Figure 1 shows that when = 0,1,2,3, the dynamic potentials of coordinates are simple inverse Morse potential. It is known that corresponding to a certain P, the stability of the lowest energy level in an inverse Morse potential is the worst, while that of the highest one is the best, which is totally different from the concept in general potential that the lower the energy level is, the worse the stability is. Furthermore, the shapes of dynamic potentials of and coordinates under different are almost same, which elucidates that D-C-P bending has no effect on the stability of the highly excited vibrational states in DCP under the small P number.
On the other hand, the dynamic potentials of show that the three modes corresponding to the highest three energy levels are localized and this conclusion is consistent when = 0,1,2,3. The dynamic potentials of and corresponding to different is basically the same and all the fixed points are remained when is different, which indicate that the effect of D-C-P bending mode has weak interaction with the two coupling modes, which are different from former studies of HOCl and HOBr systems.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
Figure 2 shows the dynamic potentialins of and and it is shown that the results are different when P is large. For example, when P = 30, the dynamic potential of becomes much more complex. There are three new fixed points emerging, , and in the dynamic potential and the shape of dynamic potential of becomes the combination of Morse and inverse Morse potentials. It is found that there is a phenomenon of fixed point-splitting in the dynamic potential of . The original (the dynamic potentialins of when P = 18 in Figure 1) becomes and , which are similar with the results of HOBr and HOCl.
On the other hand, though the dynamic potentials of and when P = 30 are much more complex than the case of P = 18, but the shape of dynamic potentials of and remain the same when = 0,1,2,3, respectively, which is consistent with the case of P = 18.[](https://www.ncbi.nlm.nih.gov/mesh/C027664)
Through the above study, it is found that the D–C–P bending mode weakly affects the resonant coupling of D–C and C–P stretching modes, thereby weakly affecting the dynamics features of DCP. It is shown that the geometrical shapes of the dynamic potentials and the corresponding fixed points are not sensitive to the D–C–P bending mode, but are sensitive to the P number, which are different to our previous studies. Though the cases of P = 18 and P = 30 are shown here, these conclusions are also suitable for other cases. The reasons we address the cases of P = 18 and P = 30 are that the connotations of corresponding dynamic potentials are abundant and the shapes of dynamics potentials are typical.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
## 3.2. The Trajectories of Phase Space Study for the Energy Levels under Specific Polyad Number (P = 30)
*In the **3.2. The Trajectories of Phase Space Study for the Energy Levels under Specific Polyad Number (P = 30)** section:*
For further quantitative analyzing the dynamic features of DCP, the representative trajectories of phase space in for each energy level is studied when P = 30. The dynamic potentials and corresponding energy levels when P = 30 and = 0 are shown in Figure 3.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
The trajectories of phase space for different energy levels in dynamic potentials of and are shown in Figure 4 and Figure 5.
Based on previous studies, the envelope area of the trajectory in phase space shows the quantum environment of a series of energy levels. In Figure 4, for L0–L13, it is found that the envelope area of the trajectory of phase space increases with the reduction of energy level because these energy levels lie in inverse Morse potential. In contrast, For L14–L15, the envelope area of the trajectory of phase space increases with the increase of energy level because these energy levels lie in Morse potential. Furthermore, because the L0 (L15) is tangential with the top (bottom) of the dynamic potential, the envelope area of the trajectory is zero. Particularly, the trajectories of L8 and L9 are divided into two separate trajectories, which show that these two energy levels are located in at a double-wells dynamic potential.
The conclusions are similar in Figure 5. Because all energy levels (except L0 and L15) are in Morse potential of so the envelope area of the trajectory in phase space increases with the increase of energy level.
## 3.3. Comparative Study between DCP and HCP with Dynamic Potential
*In the **3.3. Comparative Study between DCP and HCP with Dynamic Potential** section:*
In previous work, it was shown that the dynamic potential of DCP in coordinate is similar to the inverse of that of HCP in coordinate and the dynamic potential of DCP in coordinate is similar to that of HCP in coordinate with transformation, which is called “dynamic symmetry”. This conclusion is available when the quantum number of the uncoupling mode (H–C stretching mode for HCP and D–C–P bending mode for DCP) is equal to 0. However, the dynamic symmetry will be broken when the quantum number of the uncoupling mode (, = for HCP and = for DCP) is not equal to 0.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
From ergodic analysis of the P number, it is found that when > 0, the original horizontal reversed symmetry between the dynamic potentials of HCP () and DCP (), mentioned in Reference, does not exist; however, for large P number (P > 22), the clockwise 180 °C rotation symmetry between the dynamic potentials of HCP () and DCP () emerge. This symmetry is not strict but it still could be recognized from the shapes of dynamic potentials and the above conclusion is consistent for = 1, 2, 3 (as shown in Figure 6, for instance). In contrast, the results of the dynamic potentials of HCP () and DCP () are different. As shown in Figure 7, for instance, the vertical reversed symmetry could remain for a small P number (10 < P < 20) but it is broken when P becomes large.[](https://www.ncbi.nlm.nih.gov/mesh/D009943)
From the above results, it is shown that the uncoupled mode has an effect on the dynamic symmetry. It is obvious that the makes the original horizontal reversed symmetry break between the dynamic potential of HCP () and DCP () but has little effect on the vertical symmetry breaking between the dynamic potential of HCP () and DCP (). It is elucidated that the stability of dynamic symmetry is different under the effect of the uncoupling mode.[](https://www.ncbi.nlm.nih.gov/mesh/D009943)
## 4. Conclusions
*In the **4. Conclusions** section:*
In this study, the dynamic potentials of highly excited vibrational states of DCP with an harmonicity and Fermi coupling are studied. The results show that the D–C–P bending mode has weak effects on D–C and C–P stretching mode under different Polyad numbers. Just like previous studies, it is found that the vibrational energy levels could be classified by the quantum environments. From comparative studies, it shows that the uncoupled modes make the original horizontal reversed symmetry breaking between the dynamic potential of HCP () and DCP (), but has little effect on the vertical symmetry between dynamic potential of HCP () and DCP (). Considering the effect of in DCP and in HCP, the original dynamic similarities in these two systems disappear and the characteristics of symmetry become much more complex.
The above results show that the method which enables us to understand the DCP dynamics simply from those of HCP without repeated elaboration are only available in some special conditions, and, on the other hand, there are some new dynamic symmetries appearing when the conditions are different, which indicate that the homologous compounds are intrinsically similar only if the coupling patterns of two systems are analogous.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
# Author Contributions
*In the **Author Contributions** section:*
Aixing Wang and Yibao Liu performed the calculations of dynamic potential and the corresponding analysis. Chao Fang supervised the work.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
Dynamic potentials of DCP (P = 18) with = 0,1,2,3, and the energy levels included in the dynamic potential are shown by the lines.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
Dynamic potentials of DCP (P = 30) with = 0,1,2,3, and the energy levels included in the dynamic potential are shown by the lines.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
Dynamic potentials of DCP when P = 30 and = 0 and the energy levels included in the dynamic potential are shown by the lines.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
Trajectories of phase space of L0–L15 when P = 30 ( coordinate).
Trajectories of phase space of L0–L15 when P = 30 ( coordinate).
Dynamic potentials of HCP () and DCP () when > 0 (P = 30).[](https://www.ncbi.nlm.nih.gov/mesh/D009943)
Dynamic potentials of HCP () and DCP () when > 0 (P = 12).[](https://www.ncbi.nlm.nih.gov/mesh/D009943)
The coefficients of vibration Hamiltonian of deuterated phosphaethyne (DCP).[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
|
# Introduction
The Relationship between Serum [Bilirubin](https://www.ncbi.nlm.nih.gov/mesh/D001663) and Elevated Fibrotic Indices among HBV Carriers: A Cross-Sectional Study of a Chinese Population
# Abstract
*In the **Abstract** section:*
The study probed the association between bilirubin and hepatitis B virus (HBV) infection and progression. A cross-sectional analysis of 28,500 middle aged and elderly Chinese parti[cipants w](https://www.ncbi.nlm.nih.gov/mesh/D001663)as performed to analyze the differences of bilirubin in terms of hepatitis B surface antigen (HBsAg) positive or negative and the correlation between bilirubin and severity of hepati[c fibrosi](https://www.ncbi.nlm.nih.gov/mesh/D001663)s estimated by non-invasive indices. Bilirubin was significantly higher in the HBsAg (+) group tha[n the HBs](https://www.ncbi.nlm.nih.gov/mesh/D001663)Ag (−) group. Higher bilirubin levels were consistently associated wi[th elevat](https://www.ncbi.nlm.nih.gov/mesh/D001663)ed liver fibrosis indices among HBsAg carriers. Compared with quartile 1 of total [bilirubin](https://www.ncbi.nlm.nih.gov/mesh/D001663) (TBil), the multivariable-adjusted ORs (95% CIs) for elevated fibrosis indices of quartile 4 were 2.24 (95% CIs, 1.57–3.21) estim[ated by f](https://www.ncbi.nlm.nih.gov/mesh/D001663)ib[rosi](https://www.ncbi.nlm.nih.gov/mesh/D001663)s 4 score (FIB-4) and 2.22 (95% CIs, 1.60–3.08) estimated by aspartate transaminase to platelet ratio index (APRI). In addition, direct bilirubin (DBil) had a stronger association with elevated liver fibrosis indices than did indirect bilirubin (IBil). Furthermore, the relat[ionship b](https://www.ncbi.nlm.nih.gov/mesh/D001663)et[ween](https://www.ncbi.nlm.nih.gov/mesh/D001663) DBil and elevated fibrosis indices was more robust among participants who were fema[le, overw](https://www.ncbi.nlm.nih.gov/mesh/D001663)ei[ght ](https://www.ncbi.nlm.nih.gov/mesh/D001663)or had central fat distribution. These fi[ndin](https://www.ncbi.nlm.nih.gov/mesh/D001663)gs suggested that bilirubin levels, especially DBil, were independently associated with an increased risk of increased fibrosis indices.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 1. Introduction
*In the **1. Introduction** section:*
Although the hepatitis B vaccine is available, more than 350 million people are chronically infected with hepatitis B virus (HBV) [1], and about 30% of the world’s population shows serological evidence of current or past infection [2]. HBV infection is a major threat to public health, especially in China [3]. It has been estimated that more than 80% of liver cancer worldwide is attributable to hepatitis B or C virus infections [4]. Patients with HBV infection have a high risk of progressive liver fibrosis which can lead to cirrhosis and hepatocellular carcinoma (HCC).
In addition, inflammatory milieu caused by chronic viral infections might influence hepatic glucose sensitivity and increase insulin resistance [5], which could be determined from the findings that diabetes and prediabetes were prevalent among HBV-infected patients [6]. HBV infection is the tenth leading cause of death worldwide, with about 786,000 related deaths every year [7]. Therefore, HBV infection causes high mortality and creates a social burden.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
Bilirubin, a primary end product of heme catabolism, processes cytoprotective properties because of the antioxidant nature of the bile pigment. In 1995, it was first suggested by Breimer that bilirubin might be implicated in the protection of specific kinds of diseases resulting from oxidative damage [8]. Then, as observed by related reports, bilirubin appeared to have the innate capacity to resist oxidative damage [9,10,11]. Meanwhile, all patterns of serum bilirubin, including total bilirubin (TBil), direct bilirubin (DBil) and indirect bilirubin (IBil), display protective properties in cardiovascular diseases [12]. Several studies clarified that the robust anti-oxidative properties of bilirubin could largely explain its protective effects [13,14,15], and the findings that subjects with higher serum bilirubin had elevated total antioxidant status also confirmed its anti-oxidative property [16].[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
On the other hand, bilirubin has previously been proven to be a marker of liver injury and is incorporated in several prognostic scoring models, such as the Child–Pugh (CP) score and the model of end-stage liver disease (MELD) [17]. In recent years, relevant studies focused on the effect of bilirubin on several hepatic disorders. Recent study suggested that DBil independently reduced non-alcoholic fatty liver disease (NAFLD) risk [18]. Patients who were with liver biopsy-proved non-alcoholic steatohepatitis (NASH) had significantly lower bilirubin levels compared with those without NASH, and there was also an inverse association between bilirubin levels and histological features including fibrosis [19,20].
Serum IBil levels were negatively correlated with the progression of liver fibrosis in chronic hepatitis C (CHC) patients [21]. However, the concentrations of serum bilirubin increased along with the increased severity of fibrosis among CHC patients [22]. High levels of bilirubin or combined prognostic index including bilirubin were able to predict short-term mortality in the patients with acute-on-chronic liver failure [23,24].
Meantime, abnormal bilirubin values were even more strongly associated with poor clinical outcome at baseline and up to five years follow-up in the patients with primary billiary cirrhosis [25].[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
Related studies have illustrated the associations between bilirubin and liver disease. However, the study which is performed on participants with HBV infection is lacking. Secondly, the most of these studies did not investigate the associations between all subtypes of bilirubin and liver disease. Finally, the study that assesses the relationship between bilirubin and HBV-related fibrosis based on large sample sizes might be needed.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
Considering the apparently bewildering complexity of bilirubin’s function in different milieus and the high prevalence of HBV infection, the underlying association between bilirubin and liver fibrosis with HBV infection needs to be warranted. The task for this study is to untangle the intrinsic relationship between bilirubin and the different indices reflecting liver function in health check-ups with HBV infection.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 2. Results
*In the **2. Results** section:*
## 2.1. Characteristics of Participants
*In the **2.1. Characteristics of Participants** section:*
A total of 28,500 participants (26,549 with HBsAg negative, 1951 HBsAg positive) were included. Demographics and laboratory data of the subjects are listed in Table 1. Firstly, individuals with HBsAg positive were younger than control subjects. HBsAg seropositive subjects had a higher prevalence of current smoking and drinking, which might be explained by the fact that the individuals with HBV infection had more males compared to HBsAg negative subjects. Then, HBsAg seropositive individuals had a lower prevalence of traditional cardiovascular risk factors like hypertension, diabetes, coronary heart disease (CHD) and fatty liver.
Next, the mean levels of platelet count, total cholesterol (TC), triglycerides (TG) and low-density lipoprotein cholesterol (LDL-C) were significantly lower in the HBsAg (+) group. As expected, subjects who were HBsAg-positive had higher levels of liver injury markers (aspartate transaminase (AST) and alanine transaminase (ALT)) compared with HBsAg-negative subjects. It was notable that mean levels of TBil, IBil and DBil were significantly higher in the HBsAg (+) group than the control group.[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
## 2.2. Associations between Serum Bilirubin and Demographic, Biochemical Parameters, Non-Invasive Liver Fibrosis Indices among HBsAg (+) Participants
*In the **2.2. Associations between Serum Bilirubin and Demographic, Biochemical Parameters, Non-Invasive Liver Fibrosis Indices among HBsAg (+) Participants** section:*
Among the carriers of HBV, the associations between serum bilirubin and demographic, biochemical parameters, non-invasive liver fibrosis indices are presented in Table 2. TBil showed significant associations with age, waist-to-hip ratios (WHR), AST, hemoglobin, platelets count, TG, TC, LDL-C, aspartate transaminase to platelet ratio index (APRI) and Fibrosis 4 score (FIB-4). Surprisingly, DBil exhibited significant associations with more parameters than IBil, possibly reflecting the differences of the two forms of bilirubin. In both DBil and IDil, we observed positive associations with age, WHR, hemoglobin, APRI and FIB-4, and inverse association with platelets count and TC. The positive correlations between DBil and the liver injury makers (AST) were statistically significant, but the significant associations between IBil and certain markers of liver injury (AST and ALT) were not shown.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 2.3. Serum TBil Levels in Relation to Fibrotic Indices among HBsAg (+) Participants
*In the **2.3. Serum TBil Levels in Relation to Fibrotic Indices among HBsAg (+) Participants** section:*
Table 3 shows that there are sequentially higher odds of elevated liver fibrosis indices with ascending quartiles of TBil in multivariate models. After adjusting for age, sex, body mass index (BMI), WHR, smoking, drinking, education, marriage status and physical activity, TBil level increase was still significantly linked to the risks of elevated APRI and FIB-4. The positive relationship was decreased by additional adjustment for medical history but still statistically significant. The corresponding odds ratios (ORs) (95% confidence intervals (CIs)) for risks of elevated APRI comparing the upper 3 TBil quartiles with the lowest TBil quartile were 1.26 (0.90, 1.78), 1.78 (1.28, 2.47), 2.22 (1.60, 3.08). As for elevated FIB-4, corresponding ORs (95% CIs) comparing the upper 3 TBil quartiles with the lowest TBil quartile were 1.57 (1.14, 2.17), 1.64 (1.18, 2.28), 2.24 (1.57, 3.21).[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 2.4. Associations between Different Forms of Bilirubin (IBil and DBil) and Fibrosis Scores among HBsAg (+) Participants
*In the **2.4. Associations between Different Forms of Bilirubin (IBil and DBil) and Fibrosis Scores among HBsAg (+) Participants** section:*
We further analyzed the associations between different forms of bilirubin (IBil and DBil) and fibrosis score among HBsAg (+) participants. The positive associations were found between IBil with the elevated APRI or FIB-4 (Table S1). Corresponding ORs (95% CIs) for elevated APRI comparing the upper 3 IBil quartiles with the lowest IBil quartile were 1.30 (0.93, 1.82), 1.52 (1.10, 2.11), 2.16 (1.57, 2.98) after multi-adjustment. As for elevated FIB-4, the corresponding ORs (95% CIs) comparing the upper 3 IBil quartiles with the lowest IBil quartile were 1.24 (0.89, 1.72), 1.41 (1.01, 1.96), 1.75 (1.24, 2.49).[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
A similar tendency was exhibited between DBil and the two fibrosis indices (Table S2). Corresponding ORs (95% CIs) for elevated APRI comparing the highest DBil quartile with the lowest IBil quartile was 2.64 (1.89, 3.70) after full adjustment. As for elevated FIB-4, the corresponding ORs (95% CIs) comparing the highest DBil quartile with the lowest DBil quartile was 3.07 (2.10, 4.50). The trend between bilirubin and elevated fibrotic indices was statistically significant for IBil (PFIB-4 < 0.001 and PAPRI < 0.001) and DBil (PFIB-4 < 0.001 and PAPRI < 0.001). Moreover, the fully adjusted ORs (95% CIs) for DBil (Q4 vs. Q1: ORs for FIB-4, 3.07 (95% CIs: 2.10, 4.50); ORs for APRI, 2.64 (95% CIs: 1.89, 3.70)) were larger than those for IBil (Q4 vs. Q1: ORs for FIB-4, 1.75 (95% CIs: 1.24, 2.49); ORs for APRI, 2.16 (95% CIs: 1.57, 2.98)).[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 2.5. Comparisons of TBil, IBil and DBil in HBsAg(+) Participants
*In the **2.5. Comparisons of TBil, IBil and DBil in HBsAg(+) Participants** section:*
The areas under the receiver operating characteristic curve (AUROC) that predicts elevated APRI and FIB-4 for each form of bilirubin is presented in Table 4. The AUROC values of TBil, IBil and DBil were 0.61 (0.58, 0.64), 0.59 (0.56, 0.62) and 0.63 (0.60, 0.66) for elevated APRI, respectively. As for FIB-4, the corresponding AUC values were 0.61 (0.58, 0.64), 0.57 (0.54, 0.60) and 0.65 (0.62, 0.68), respectively (Table 4). The DBil had higher AUROCs than TBil and IBil.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 2.6. Serum DBil Levels in Relation to Fibrotic Features in Subgroups among HBsAg(+) Participants
*In the **2.6. Serum DBil Levels in Relation to Fibrotic Features in Subgroups among HBsAg(+) Participants** section:*
To better understand the effect of DBil levels on the liver fibrotic progression across sex, BMI and WHR, we inquired into the relationship between DBil and fibrosis indices in subgroups. Figure 1 shows the risks of elevated fibrosis indices with each standard deviation (SD) increase in DBil in the different sex, overweight or not, central and peripheral fat distribution subgroups. Positive associations between DBil and elevated FIB-4 or APRI were consistent in different subgroups after full adjustment. Subjects who were female were inclined to have a higher risk of elevated fibrosis indices. In addition, after full adjustment, subjects who were overweight or had central fat distribution were inclined to have a higher risk of elevated fibrosis index reflected by APRI than those without these metabolic disorder features.
The ORs (95% CIs) for elevated APRI per 1 SD increase of DBil across overweight (yes or no) were 2.04 (1.57–2.66) vs. 1.54 (1.28–1.86). At the same time, the ORs (95% CIs) for elevated APRI per 1 SD increase of DBil between central and peripheral fat distribution were 1.82 (1.46–2.26) vs. 1.55 (1.25–1.91).[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 3. Discussion
*In the **3. Discussion** section:*
In this study, we observed the significant differences in serum bilirubin between individuals with or without HBV infection. Moreover, higher bilirubin levels might indicate more advanced liver fibrosis in a large group of retired workers with serum evidence of HBV infection. In addition, such associations were statistically significant when adjusted for multiple parameters, especially in female, overweight or central fat distribution individuals.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
Our observations could probably best explain the relationship between bilirubin and validated non-invasive fibrosis indices among HBV carriers. For a start, a lower prevalence of traditional cardiovascular risk factors like hypertension, diabetes, CHD and fatty liver was observed in the report, which was different from the findings that diabetes and prediabetes were prevalent among HBV-infected patients [6]. One possible reason for this would be the distinct characteristics of the two study groups. Following the phenomenon that the bilirubin levels among HBV carriers were higher than the others, we sought to investigate the relationship between bilirubin and the liver fibrosis among individuals with HBV infection. Unlike the meaningful findings that higher bilirubin concentrations were associated with reduced risk of cardiovascular disease, respiratory illness and mortality in epidemiological studies [13,15], the protective effect of bilirubin was not meaningful in our study.
Similar tendency was observed in patients with hepatitis C virus (HCV) associated fibrosis [22], but the concentrations of bilirubin in this study were far beyond the”normal range”. Conversely, the principal results in our study were different from another HCV-related fibrosis study [21] in which the inverse relationship between bilirubin and liver fibrosis was found. The possible reasons for the inconsistent results of the researches above might be the sample size and diverse study design. The results presented indicated that bilirubin might act as an independent risk factor for significant liver fibrosis.
Is this biologically plausible? The process of liver fibrosis is the excessive accumulation of extracellular matrix proteins including collagen. Most types of chronic liver disease without clinical symptoms have developed into liver fibrosis which could result in cirrhosis, liver failure, and other severe complications. In advanced cirrhosis, glucuronyl conjugation of bilirubin and biliary excretion of DBil are markedly impaired and jaundice appears [26]. Therefore, the concentration of bilirubin in serum may be a good prognostic marker for patients with decompensated liver cirrhosis.
Hepatic fibrosis, as the onset of liver cirrhosis, might disturb the bilirubin’s normal production and excretion in the liver. Although it is difficult to determine the exact mechanisms behind the relationship between bilirubin and fibrotic progression because of complexity of the disease, several hints could be identified. Firstly, related studies proposed that bilirubin was able to induce cytotoxic effects [27,28,29,30], unbalance the redox homeostasis [31], and finally affect the mitochondrial integrity and induce apoptosis
[32]. Second, activated retinoid-storing hepatic stellate cells might contribute more to the elevation of DBil levels in blood [33], and the medicine associated with reversion and prevention of cirrhosis could also reduce the levels of serum bilirubin [34]. Lastly, slightly elevated bilirubin could induce a stress response to the endoplasmic reticulum, resulting in a decreased proliferative and metabolic activity of hepatocytes [35].[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
Three important findings were achieved about evaluating serum bilirubin in individuals with HBV infection. First, our results showed the positive associations between all forms of bilirubin and severity of liver disease. DBil was more correlative with the indices of liver fibrosis. Also, DBil had higher AUROCs than TBil and IBil. To be noticed, there were slight differences between DBil and IBil.
IBil had more potent anti-oxidant capacity than the DBil [36] might explain the weaker association between IBil and the risk of elevated fibrosis indices. Second, our results suggested that females should pay more attention to an increase of DBil level. Related study suggested a protective effect of estrogens on fibrogenesis via the inhibition of stellate cell proliferation [37]. The individuals in the current study were featured with old age and estrogens levels were substantially decreased.
The findings suggested that elderly females should pay more attention to the increase in DBil. In addition, overweight subjects and those with central fat distribution should also be careful of an increase in DBil level. The intricate interplay between HBV infection and metabolic factors might be involved in the positive relationship between bilirubin and fibrosis [38]. The data in the present study confirmed this idea which the participants with higher DBil levels had higher risks of elevated fibrosis scores among subjects with central fat distribution or overweight compared to those without such metabolic disorders. Lifestyle advice should be offered to all HBsAg carriers due to its easy implementation with little risk of side-effects or cost.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
The data displayed here can be explained only in the context of the study design. Firstly, due to the limitations of the cross-sectional study, the potentially important function of bilirubin needed to be further investigated. Second, the severity of liver fibrosis was assessed only by noninvasive indices. Although liver pathology is the gold standard, patients’ discomfort and expense should be considered also. Noninvasive methods had overcome the limitations of liver biopsy and were also used as prognostic indices for subjects with hepatitis B-associated HCC
[39]. Furthermore, the accuracy of FIB-4 and APRI were 78% and 76% [40], suggesting they were suitable for regular monitoring of disease progression [41,42]. Thus, using APRI and FIB-4 to assess liver fibrosis was acceptable in the circumstances of the study. Finally, we failed to acquire information on concentrations of virus titer that were potentially linked to the pathophysiology of CHB. The proportion of HBeAg(+) was 1.8%, and relatively small in the study.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
In summary, our study demonstrates a robust association between bilirubin levels and higher surrogate indices of liver fibrosis among participants with HBV infection. This suggests that bilirubin levels, especially DBil, were independently associated with an increased risk of increased fibrosis indices.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 4. Methods
*In the **4. Methods** section:*
## 4.1. Study Population
*In the **4.1. Study Population** section:*
We conducted a cross-sectional study using the data from the Dongfeng-Tongji cohort study of retired worker as described previously [43]. In 2013, a total of 38,295 individuals were subjected to physical examination, laboratory tests and accomplished semi-structured questionnaires which included socio-demographic information and other concerned information during face-to-face interviews. Among these subjects, 28,500 underwent laboratory testing for HBV infection. HBV infection was defined as the presence of hepatitis B surface antigen (HBsAg) in peripheral blood. Participants were divided into 2 study groups: (a) participants with HBsAg; (b) controls without HBsAg.
## 4.2. Ethics Statement
*In the **4.2. Ethics Statement** section:*
Participants were enrolled after obtaining their written informed consent to the study protocol that was approved by the Medical Ethics Committee of the School of Public Health, Tongji Medical College, Huazhong University of Science and Technology and Dongfeng General Hospital, DMC (approval No. 03, 1 August 2008).
## 4.3. Measurements
*In the **4.3. Measurements** section:*
Participants who underwent an overnight fast were given a physical examination at Dongfeng Central Hospital with trained physicians, nurses and technicians. Body mass index (BMI) was calculated as weight in kilograms/(height in m)2. A reasonable estimation of fat distribution might be made using waist-to-hip ratios (WHR). Subjects were separated into those with central fat distribution (WHR ≥ 0.81 for women and ≥ 0.92 for men) and those with peripheral fat distribution (WHR < 0.81 for women and <0.92 for men), as described in related study [44].
After an overnight fast, all blood specimens were collected to test blood lipids, fasting glucose, hepatic function and renal function at the hospital’s laboratory by the ARCHITECT Ci8200 automatic analyzer (Abbott, Chicago, IL, USA) with corresponding reagent kits. The laboratory also provided a complete blood count and urine routine test. Commercially available enzyme immunoassays were used to determine serum HBsAg, hepatitis B e antigen (HBeAg), antibodies to hepatitis B surface antigen, hepatitis B e antigen and hepatitis B core antigen at the same laboratory using a fully automatic immunoanalyzer, Uranus AE 120 (AIKANG, Shenzhen, China). In addition, abdominal B-type ultrasound was inspected using Aplio XG (TOSHIBA, Tokyo, Japan), by experienced radiologists.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
A history of regular smoking was defined as having smoked at least one cigarette per day for more than six months. Smokers who met the definition of regular smoking were divided into current smokers and former smokers according to whether these subjects quitted smoking at the time of the interview. The participants never smoking were defined as non-smokers. Likewise, drinking status was classified into three groups: never drinking, quit drinking, current drinking. CHD, hypertension and diabetes were self-reported chronic diseases.
The diagnosis of these conditions was used in accordance with well-accepted international standards [45]. The existence of fatty liver was based on the information of abdominal B-type ultrasound.
## 4.4. Indices of Liver Fibrosis
*In the **4.4. Indices of Liver Fibrosis** section:*
We calculated two validated non-invasive indices for liver fibrosis, including Fibrosis 4 score (FIB-4), aspartate transaminase to platelet ratio index (APRI). All of them were obtained according to the published formula, as previously described [46,47,48]. (# where ULN = upper limit of normal for that laboratory, the upper limit of normal for both AST and ALT was 40 U/L).
Elevated FIB-4 and APRI were defined as FIB-4 ≥ 1.45 [40], APRI ≥ 0.5 [40,49], in both male and female.
## 4.5. Statistical Analyses
*In the **4.5. Statistical Analyses** section:*
All statistical analyses displayed in tables were performed using SAS 9.4 software (SAS Institute Inc., Cary, NC, USA). A two-sided p value (<0.05) was considered to be of statistical significance. Continuous variables were represented as mean (SD) and further compared using independent t tests. Categorical variables were represented as percentages. A chi-square test was used to determine the distribution of categorical variables among various groups.
Spearman rank correlation was performed to test the relationship between continuous variables. To identify whether bilirubin levels were associated with the possibility of increased fibrosis-related indices, multivariate logistic regression analysis was conducted to correct important factors. The Cochran–Armitage trend test was used to investigate the trend among binomial proportion of disease progression.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
# Supplementary Materials
*In the **Supplementary Materials** section:*
Supplementary materials can be found at .
# Author Contributions
*In the **Author Contributions** section:*
All authors contributed significantly to this work; Xiaoping Miao, Ping Yao and Min Du conceived and designed the study strategy; Min Du, Yanyan Xu, Peiyi Liu, Lin Xiao, Shanshan Zhang, Sheng Wei and Ping Yao recruited the participants and collected their information and blood samples; Min Du, Shanshan Zhang, Lin Xiao, Yanyan Xu, Peiyi Liu, Yuhan Tang and Mingyou Xing contributed to data collection and statistical analyses; Min Du contributed to the writing of the manuscript and preparing the tables and figures; Yuhan Tang, Ping Yao and Min Du contributed to the critical revision of the article; All authors reviewed the manuscript.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare that they have no conflict of interest.
# Abbreviations
*In the **Abbreviations** section:*
AUROC areas under the receiver operating characteristic curve ALT alanine transaminase APRI aspartate transaminase to platelet ratio index AST aspartate transaminase BMI body mass index CHC chronic hepatitis C CIs confidence intervals CHD coronary heart disease DBil direct bilirubin FIB-4 fibrosis 4 score HBV hepatitis B virus HBsAg hepatitis B surface antigen HCV hepatitis C virus HCC hepatocellular carcinoma IBil indirect bilirubin LDL-C low-density lipoprotein cholesterol ORs odds ratios SD standard deviation TBil total bilirubin TC total cholesterol TG triglycerides WHR waist-to-hip ratios[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
# References
*In the **References** section:*
The risks of elevated APRI (A) and FIB-4 (B) with each 1 SD increase in serum DBil among HBsAg(+) participants. The risks of elevated APRI (A) and FIB-4 (B) with each 1 SD increase in serum DBil concentrations according to subgroups of sex (male vs. female), overweight (BMI < 25.0 vs. ≥25.0 kg/m2), fat distribution (WHR ≥ 0.92 for male and WHR ≥ 0.81 for female vs. WHR < 0.92 for male and WHR < 0.81 for female). Adjusted for age (continuous), sex (male, female), BMI (continuous), WHR (continuous), smoking (never smoking, quit smoking, current smoking), drinking (never drinking, quit drinking, current drinking), education (≤6/7–9/10–12/≥13), marriage status (yes/no), physical activity (yes/no) and medical history (yes/no for hypertension, CHD, diabetes, fatty liver). APRI, aspartate transaminase to platelet ratio index; FIB-4, Fibrosis 4 score; SD, standard deviation; WHR, waist-to-hip ratios; ORs, odds ratios; CIs, confidence intervals.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
The characteristics of 28,500 participants.
Data are expressed as mean (standard deviation) *, median (interquartile range) # or % ^. HBsAg, hepatitis B surface antigen; BMI, body mass index; WHR, waist-to-hip ratios; CHD, coronary heart disease; ALT, alanine transaminase; AST, aspartate transaminase; TG, triglycerides; TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol.[](https://www.ncbi.nlm.nih.gov/mesh/D014280)
Univariate associations between bilirubin and demographic, biochemical parameters and non-invasive liver fibrosis indices in participants with HBV infection.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
TBil, total bilirubin; IBil, indirect bilirubin; DBil, direct bilirubin; BMI, body mass index; WHR, waist-to-hip ratios; ALT, alanine transaminase; AST, aspartate transaminase; TG, triglycerides; TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; APRI, aspartate transaminase to platelet ratio index; FIB-4, Fibrosis 4 score; rho: Spearman’s rank correlation coefficient.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
Odds ratios and 95% CIs for serum TBil levels and the presence of elevated levels of APRI or FIB-4 in HBsAg (+) individuals.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
TBil, total bilirubin; APRI, aspartate transaminase to platelet ratio index; FIB-4, Fibrosis 4 score. a Without adjustment; b Adjusted for age (continuous), sex (male, female); c Adjusted for the same set of variables in model 1 plus BMI (continuous), WHR (continuous), smoking (never smoking, quit smoking, current smoking), drinking (never drinking, quit drinking, current drinking), education (≤6/7–9/10–12/≥13), marriage status (yes/no) and physical activity (yes/no); d Adjusted for the same set of variables in model 2 plus the components of the medical history as dichotomized variables.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
Areas under the receiver operating characteristic curve (AUROC) for elevated fibrotic indices for the TBil, IBil and DBil in HBsAg (+) participants.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
Tbil, total bilirubin; IBil, indirect bilirubin; DBil, direct bilirubin; APRI, aspartate transaminase to platelet ratio index; FIB-4, Fibrosis 4 score; HBsAg, hepatitis B surface antigen.[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
|
# Introduction
The Production and Utilization of [GDP-glucose](https://www.ncbi.nlm.nih.gov/mesh/C100085) in the Biosynthesis of [Trehalose 6-Phosphate](https://www.ncbi.nlm.nih.gov/mesh/C082722) by Streptomyces venezuelae*
# Abstract
*In the **Abstract** section:*
Trehalose-6-phosphate synthase OtsA from streptomycetes is unusual in that it uses GDP-glucose as the donor substrate rather than the more commonly used UDP-glucose. We now confirm that OtsA from Strept[omyces vene](https://www.ncbi.nlm.nih.gov/mesh/C100085)zuelae has such a preference for GDP-glucose and can utiliz[e ADP-gluco](https://www.ncbi.nlm.nih.gov/mesh/D014532)se to some extent too. A crystal structure of the enzyme shows that it shares twin[ Rossmann-l](https://www.ncbi.nlm.nih.gov/mesh/C100085)ike domains with [the UDP-glu](https://www.ncbi.nlm.nih.gov/mesh/D000245)cose-specific OtsA from Escherichia coli. However, it is structurally more similar to Streptomyces hygroscopicus[ VldE, a GD](https://www.ncbi.nlm.nih.gov/mesh/D014532)P-valienol-dependent pseudoglycosyltransferase enzyme. Comparison of the donor binding sites reveals that the amino acids associated with the binding of diphosphoribose are almost all identical in these three enzymes. By contrast[, the amino](https://www.ncbi.nlm.nih.gov/mesh/D000596) acids associated with binding g[uanine in VldE ](https://www.ncbi.nlm.nih.gov/mesh/D000246)(Asn, Thr, and Val) are similar in S. venezuelae OtsA (Asp, Ser, an[d Phe, resp](https://www.ncbi.nlm.nih.gov/mesh/D000596)ectively) but not conserv[ed in E](https://www.ncbi.nlm.nih.gov/mesh/D006147). coli Ots[A (](https://www.ncbi.nlm.nih.gov/mesh/D001216)Hi[s, ](https://www.ncbi.nlm.nih.gov/mesh/D013912)Leu, a[nd ](https://www.ncbi.nlm.nih.gov/mesh/D014633)Asp, respectively), providing a ratio[nal](https://www.ncbi.nlm.nih.gov/mesh/D001224)e [for](https://www.ncbi.nlm.nih.gov/mesh/D012694) the p[uri](https://www.ncbi.nlm.nih.gov/mesh/D010649)ne base specificity of S. venezuelae OtsA. To estab[lis](https://www.ncbi.nlm.nih.gov/mesh/D006639)h [whi](https://www.ncbi.nlm.nih.gov/mesh/D007930)ch don[or ](https://www.ncbi.nlm.nih.gov/mesh/D001224)is used in vivo, we generated an otsA null muta[nt in ](https://www.ncbi.nlm.nih.gov/mesh/D011687)S. venezuelae. The mutant had a cell density-dependent growth phenotype and accumulated galactose 1-phosphate, glucose 1-phosphate, and GDP-glucose when grown on galactose. To determine how the GDP-glucose i[s generated, we chara](https://www.ncbi.nlm.nih.gov/mesh/C029973)ct[erized three candid](https://www.ncbi.nlm.nih.gov/mesh/C031590)ate GD[P-glucose p](https://www.ncbi.nlm.nih.gov/mesh/C100085)yrophosphorylas[es. SVEN_](https://www.ncbi.nlm.nih.gov/mesh/D005690)3027 is a UDP-glucose p[yrophosphor](https://www.ncbi.nlm.nih.gov/mesh/C100085)ylase, SVEN_3972 is an unusual ITP-mannose pyrophosphorylase, and SVEN_2781 is a pyrophosphorylase that is capable of generating GDP-glucose as well as GDP-mannose. We have therefore established how S. venezuelae can make and utilize GDP-glucose in[ the biosyn](https://www.ncbi.nlm.nih.gov/mesh/C100085)thesis of tr[ehalose 6-p](https://www.ncbi.nlm.nih.gov/mesh/D006155)hosphate.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Streptomyces venezuelae is a soil-dwelling bacterium with a developmental life cycle that initiates with the germination of spores (1). Vegetative hyphae then form to generate a substrate mycelium. Finally, aerial hyphae form that differentiate into the next generation of spores. The spores contain trehalose as a carbon and energy source for germination (2). This non-reducing disaccharide (α-d-glucopyranosyl-(1→1)-α-d-glucopyranoside) is also known to provide tolerance to stresses such as desiccation, dehydration, heat, cold, and oxidation (3).
In addition, trehalose is utilized by the GlgE pathway (4–7) in this organism (8) for the transient biosynthesis of a glycogen-like α-glucan (Fig. 1) (9). This polymer is disassembled in streptomycetes by TreY (EC 5.4.99.15, (1→4)-α-d-glucan 1-α-d-glucosylmutase) and TreZ (EC 3.2.1.141, 4-α-d-(1,4-α-d-glucano)trehalose glucanohydrolase (trehalose-producing)) to regenerate trehalose during the onset of sporulation (10–14).[](https://www.ncbi.nlm.nih.gov/mesh/D014199)
The only route for the de novo biosynthesis of trehalose in S. venezuelae is via trehalose 6-phosphate (8) (Fig. 1). OtsA (α,α-trehalose-phosphate synthase) is responsible for the formation of this metabolic intermediate from an NDP-glucose donor and glucose 6-phosphate as the acceptor. The enzyme from Streptomyces hygroscopicus and some other actinomycetes has been reported to exhibit a preference for GDP-glucose as the donor (EC 2.4.1.36, GDP-glucose:d-glucose-6-phosphate 1-α-d-glucosyltransferase, of the GT20 CAZy family) (15–19). This contrasts with OtsA enzymes from other bacteria, insects, yeasts, and fungi that most commonly utilize UDP-glucose as the donor substrate (EC 2.4.1.15, UDP-glucose:d-glucose-6-phosphate 1-α-d-glucosyltransferase). For example, the enzyme from Escherichia coli is UDP-glucose-specific, with crystal structures providing a clear understanding of the structural basis for its donor preference (20, 21).
In all cases, trehalose 6-phosphate is dephosphorylated by OtsB (EC 3.1.3.12, trehalose-6-phosphate phosphohydrolase) to give trehalose.[](https://www.ncbi.nlm.nih.gov/mesh/D014199)
Proposed metabolic pathways connecting galactose with GDP-glucose. The ability of SVEN_2781 to produce GDP-glucose was established in this study. The conversion of galactose 1-phosphate to glucose 1-phosphate probably occurs via the epimerization of an NDP-galactose in a Leloir-type pathway (31). Glc6P, glucose 6-phosphate; Glc1P, glucose 1-phosphate; Gal, galactose; Gal1P, galactose 1-phosphate; GDPGlc, GDP-glucose; T6P, trehalose 6-phosphate; M1P, maltose 1-phosphate.[](https://www.ncbi.nlm.nih.gov/mesh/D005690)
Because OtsA enzymes in streptomycetes use GDP-glucose as the donor, it would be expected that these organisms possess a GDP-glucose pyrophosphorylase (EC 2.7.7.34, GTP:α-d-glucose-1-phosphate guanylyltransferase) capable of forming GDP-glucose from GTP and glucose 1-phosphate. Such enzyme activity has been reported in mammalian cells, plant tissues, and streptomycetes (22–26), but no sequence information is available, and no bacterial enzyme has been characterized to date. We therefore determined the donor preference of OtsA from S. venezuelae both in vitro and in vivo, identified its structural basis, and characterized three candidate enzymes for the production of the preferred donor.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
## Results
*In the **Results** section:*
## GDP-glucose Is the Preferred Donor Substrate of Recombinant S. venezuelae OtsA
*In the **GDP-glucose Is the Preferred Donor Substrate of Recombinant S. venezuelae OtsA** section:*
The enzyme OtsA from S. hygroscopicus and other streptomycetes has been reported to have a preference for the donor GDP-glucose (15–17). To establish whether the enzyme from S. venezuelae shares this donor preference, the recombinant enzyme was produced in E. coli. The theoretical size of the protein with its tag is 51 kDa, but the recombinant protein had an estimated molecular mass of 109 kDa according to size exclusion chromatography, consistent with it forming a homodimer in solution.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
By monitoring the generation of nucleotide sugar diphosphate in a coupled assay, it was clear that the affinity of the enzyme for glucose 6-phosphate was independent of the GDP-glucose or ADP-glucose donor substrate used (Table 1). However, the associated value of kcat was nearly 5-fold greater with GDP-glucose. Furthermore, the kcat and Km for GDP-glucose were more favorable than with ADP-glucose, giving a catalytic efficiency an order of magnitude greater (Table 1). No activity with UDP-glucose, UDP-galactose, or GDP-mannose was detected. In addition, none of these three compounds inhibited enzyme activity when used at the same concentration as either GDP-glucose or ADP-glucose, implying that they do not bind to the enzyme active site.
The preference for the donor substrates was confirmed using 1H NMR spectroscopy to monitor the reactions. Potential allosteric regulators of OtsA were tested (fructose 6-phosphate, glucose 1-phosphate, mannose 1-phosphate, GTP, ATP, pyrophosphate, and orthophosphate), but none showed any effect on enzyme activity with GDP-glucose. This contrasts with the activation of the Mycobacterium tuberculosis enzyme by fructose 6-phosphate (27). Therefore, although able to use another purine diphosphoglucose donor, the enzyme had a preference for GDP-glucose and was not subject to allosteric regulation.[](https://www.ncbi.nlm.nih.gov/mesh/D009702)
Kinetic analysis of recombinant S. venezuelae OtsA
## The Structural Basis for Donor Specificity
*In the **The Structural Basis for Donor Specificity** section:*
To establish the structural basis for the donor specificity of S. venezuelae OtsA, the recombinant enzyme was crystallized. Crystals diffracted to 1.95 Å (Table 2), allowing the protein structure to be solved using molecular replacement. This was done using a Phyre2-generated homology model (28) for OtsA based on the structure of S. hygroscopicus VldE (PDB3 code 3T5T (29)), a protein of known structure with which it shares one of the highest sequence identities (30%). VldE catalyzes the formation of validoxylamine A 7′-phosphate with a non-glycosidic C–N bond from GDP-valienol and validamine 7-phosphate.[](https://www.ncbi.nlm.nih.gov/mesh/C052534)
Summary of X-ray data and model parameters for S. venezuelae OtsA
Values in parentheses are for the outer resolution shell.
a Rmerge = Σhkl Σi|Ii(hkl) − 〈I(hkl)〉|/Σhkl ΣiIi(hkl).
b Rmeas = Σhkl(N/(N − 1))½ × Σi|Ii(hkl) − 〈I(hkl)〉|/Σhkl ΣiIi(hkl), where Ii(hkl) is the ith observation of reflection hkl, 〈I(hkl)〉 is the weighted average intensity for all observations i of reflection hkl, and N is the number of observations of reflection hkl.
c CC½ is the correlation coefficient between symmetry equivalent intensities from random halves of the data set.
d The data set was split into “working” and “free” sets consisting of 95 and 5% of the data, respectively. The free set was not used for refinement.
e The R-factors Rwork and Rfree are calculated as follows: R = Σ(|Fobs − Fcalc|)/Σ|Fobs|, where Fobs and Fcalc are the observed and calculated structure factor amplitudes, respectively.
f As calculated using MolProbity (58).
Four copies of the S. venezuelae OtsA protein were present in the asymmetric unit. Consistent with the size observed in solution, the biological unit appeared to be a dimer (Fig. 2). The two copies of the dimer within the asymmetric unit are essentially identical and had subunit interfaces of 1319 Å2 (30). The subunit interface in VldE (PDB code 4F9F) is topologically equivalent but more extensive at 2080 Å2. Interestingly, E. coli OtsA is also known to form a dimer in solution, but it is topologically different (PDB code 1UQT), giving a subunit interface of 1038 Å2 (21).
However, it can also form a tetramer in solution (20) involving a second subunit interface of 1033 Å2 (PDB code 1GZ5) that is topologically equivalent to that of the S. venezuelae enzyme. Despite differences in quaternary structures, superposition of the A chain of S. venezuelae OtsA with S. hygroscopicus VldE (PDB code 4F9F over 368 residues) and E. coli OtsA (PDB code 1UQT over 393 residues) showed that they each had a common fold with a root mean square deviation value of 2.00 Å. The fold comprised twin Rossman-like β/α/β domains in a GT-B configuration. A molecule of MES buffer is bound in a cleft between the two domains of S. venezuelae OtsA in what is likely to be the active site, by analogy with the E. coli enzyme (20, 21).[](https://www.ncbi.nlm.nih.gov/mesh/C004550)
The conservation of the S. venezuelae OtsA dimer interface. The biological dimers of S. venezuelae OtsA, E. coli OtsA (PDB code 1UQT), and S. hygroscopicus VldE (PDB code 4F9F) (29) are shown with equivalent orientations with regard to the subunit depicted in red. The orientation of the second subunit within each dimer is depicted in blue, clearly showing a different orientation in the E. coli OtsA enzyme (21). However, the existence of a tetrameric form of the E. coli enzyme (PDB code 1GZ5; where the additional two subunits are shown in gray) shows that one subunit can be in the same orientation (20).
Attempts to obtain a structure of the S. venezuelae enzyme with either GDP or GDP-glucose bound were unsuccessful. Therefore, comparisons were made between the non-liganded structure and those of the closest structural homologues with nucleotides bound; E. coli OtsA with UDP-glucose bound (PDB code 1UQU) (21) and S. hygroscopicus VldE with GDP (and trehalose) bound (PDB code 4F96) (29). First, the active sites were superposed on the basis of the α-carbons of Arg-263, Leu-365, and Glu-369 (S. venezuelae numbering), which were conserved in all three proteins. Then, the observed and potential hydrogen-bonding interactions between the ligands and proteins were assessed (Fig. 3). It was immediately apparent that amino acid side chains associated with binding the ribose diphosphate were almost completely conserved.
The only exceptions were an additional Arg-341 side chain interaction to the ribose in E. coli OtsA and an additional Ser-388 side chain interaction with phosphate in VldE.[](https://www.ncbi.nlm.nih.gov/mesh/D006153)
The donor binding site is configured to bind GDP-glucose. Two-dimensional maps show the binding interactions between UDP-glucose and E. coli OtsA (PDB code 1UQU) (21) and between GDP-glucose and S. hygroscopicus VldE (PDB code 4F96) (29). The interactions between GDP and S. venezuelae OtsA were predicted based on structural alignments with the known ligand-bound structures. Key differences between the three structures are highlighted in red, hydrogen bonds in dashed blue lines, and hydrophobic interactions in solid blue curves.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
With the exception of a hydrogen bond between a protein backbone NH and an oxygen of the base, the interactions with the base were quite different (Fig. 3). In the case of E. coli OtsA, there was an additional backbone CO interaction (from Phe-399) with an NH of the uridine base. By contrast, VldE formed hydrogen bonds between the side chains of Asn-361 and Thr-366 and the guanidine base together with a hydrophobic interaction involving a Val-363 side chain. All three of these interactions appear to be possible in S. venezuelae OtsA with its Asp-339, Ser-344, and Phe-341 side chains in equivalent positions to Asn-361, Thr-366, and Val-363 in VldE. Thus, the preference of S. venezuelae OtsA for GDP-glucose is evident from its structural similarity to VldE.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## An S. venezuelae otsA Null Mutant Grows Like Wild Type on Maltose
*In the **An S. venezuelae otsA Null Mutant Grows Like Wild Type on Maltose** section:*
Given the ability of S. venezuelae OtsA to utilize ADP-glucose to some extent as well as GDP-glucose, it was uncertain what the physiological donor would be, so we took a reverse genetic approach to address this. We had already established that blocking the GlgE pathway with a treS null mutation did not affect growth of S. venezuelae on complex or minimal media supplemented with maltose (8). This was because the product of TreS, maltose, bypassed the blockage (Fig. 1). Similarly, when the otsA gene was replaced by an apramycin resistance (apr) cassette, the resulting otsA null mutant was also able to grow on media supplemented with maltose, just like the wild-type strain. Metabolite analysis using NMR spectroscopy showed that trehalose levels were similar to the wild type (8) throughout growth on maltose.
Although the maltose level reached 4% dry cell weight during growth, a little higher than the 2.5% observed in wild-type cells (8), it fell to zero during sporulation as normal. α-Glucan accumulated in pre-spore cells as normal, and spore morphology was like that of the wild-type strain, according to electron microscopy. No other changes in metabolite levels were apparent. Therefore, there the mutant strain did not show any significant phenotype on media supplemented with maltose.[](https://www.ncbi.nlm.nih.gov/mesh/D000245)
## The Mutant Strain Develops Slowly in a Cell Density-dependent Manner When Grown on Galactose
*In the **The Mutant Strain Develops Slowly in a Cell Density-dependent Manner When Grown on Galactose** section:*
We then tested whether a phenotype could be observed when the mutant was grown on a minimal medium supplemented with a carbon source other than glucose, maltose, or any other intermediate of the GlgE pathway. Although no growth phenotype of the mutant was observed when using fructose, there was a developmental delay with galactose (Fig. 4), such that sporulation did not appear to occur. Interestingly, the growth phenotype of the mutant was dependent on cell density, allowing isolated colonies to grow more like the wild-type strain.[](https://www.ncbi.nlm.nih.gov/mesh/D002244)
Development of a S. venezuelae otsA null mutant is delayed in a density-dependent manner when grown on galactose. The constructed ΔotsA::apr null mutant was grown on either minimal medium supplemented with galactose for 7 days or complex medium containing malt extract and supplemented with galactose for 2–4 days, as indicated. Growth and development were delayed in the mutant at high cell densities on both media. The complemented strain (ΔotsA::apr attBΦBT1::otsA) grew like the WT strain.[](https://www.ncbi.nlm.nih.gov/mesh/D005690)
To be able to isolate larger quantities of cells, the mutant was grown on a complex medium that contained both galactose and malt extract, the latter providing some maltose. A delayed growth phenotype was still observed, but sporulation did eventually occur (Fig. 4). With this growth medium, the mutant strain was able to generate some trehalose, most probably via TreY/TreZ or possibly via TreS (Fig. 1). However, the spores of the mutant strain contained less trehalose (5.4 ± 0.1% dry cell weight ± S.E., n = 3) than the wild-type strain (17.9 ± 0.4% dry cell weight ±
S.E., n = 3). This suggested that the limited amount of maltose present in the medium was not sufficient to fully compensate for the lack of OtsA. Despite the lower level of trehalose, spore morphology was not affected, according to electron microscopy. Taken together, the mutant showed a delayed growth phenotype that was both galactose- and cell density-dependent.[](https://www.ncbi.nlm.nih.gov/mesh/D005690)
## The Mutant Strain Accumulates GDP-glucose When Grown on Galactose
*In the **The Mutant Strain Accumulates GDP-glucose When Grown on Galactose** section:*
1H NMR spectroscopy of cell-free extracts of the otsA mutant strain grown in the presence of galactose revealed the presence of a number of metabolites (Fig. 5A) that were also observed with the wild-type strain during growth (8). These included trehalose and maltose (and possibly glucose, given the overlap between the maltose and glucose resonances). There were, however, a number of other species that do not accumulate in the wild-type. For example, there were two sets of double doublets at 5.46 and 5.50 ppm (Fig. 5, A and B) that would be expected to arise from phosphosugars.
Galactose is likely to be assimilated by S. venezuelae using a Leloir-type pathway in which galactose 1-phosphate is converted to glucose 1-phosphate via epimerization of an NDP-galactose (31). Indeed, adding the cell-free extract to authentic galactose 1-phosphate and glucose 1-phosphate enhanced these specific resonances, consistent with these two metabolites accumulating in the mutant.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
The S. venezuelae otsA null mutant accumulates galactose 1-phosphate, glucose 1-phosphate, and GDP-glucose when grown in the presence of galactose. The mutant (ΔotsA::apr) and WT strains were grown for 1.5 days on a complex medium supplemented with galactose, and cell-free extracts were analyzed using 1H NMR spectroscopy. A, the mutant accumulated trehalose, maltose, and possibly glucose. The regions of the spectrum overlapping with those shown in the subsequent panels are indicated. B, when the mutant extract was added to galactose 1-phosphate (Gal1P) and glucose 1-phosphate (Glc1P), the enhanced resonances lined up with those already observed in the extract (with different y axis scaling for clarity). C, additional resonances in the mutant extract compared with WT extract were consistent with the accumulation of GDP-glucose. A spectrum of authentic GDP-glucose is also shown.[](https://www.ncbi.nlm.nih.gov/mesh/C029973)
In addition, another double doublet at 5.60 ppm (Fig. 5, A and C), together with additional resonances at low field, was suggestive of the presence of a nucleotide sugar diphosphate. Indeed, the resonances associated with authentic GDP-glucose were clearly present in the cell-free extract, consistent with the accumulation of this metabolite. The resonances of neither UDP-glucose, UDP-galactose, nor ADP-glucose were detected in the extract. Some additional broad resonances were also observed with the mutant extract, but the dominant ones are associated with glucose 1-phosphate, galactose 1-phosphate, and GDP-glucose.
Therefore, GDP-glucose appears to be the donor of S. venezuelae OtsA in vivo.[](https://www.ncbi.nlm.nih.gov/mesh/D009702)
## S. venezuelae SVEN_2781 Can Generate GDP-glucose Efficiently
*In the **S. venezuelae SVEN_2781 Can Generate GDP-glucose Efficiently** section:*
The bacterium S. venezuelae possesses several genes predicted to be NDP-sugar pyrophosphorylases, but it was not known which one(s) might produce GDP-glucose. With the absence of any known sequences of GDP-glucose pyrophosphorylases, we first identified three homologues of the GTP-dependent enzyme GDP-mannose pyrophosphorylase from M. tuberculosis (ManB Rv3264c) (32) in the genome of S. venezuelae. Then we produced these three enzymes as recombinant proteins in E. coli. First, SVEN_3027 is also a homologue of GalU from other organisms (also known as GtaB), which is normally associated with the formation of UDP-glucose. Enzyme assays that monitor the production of pyrophosphate showed that SVEN_3027 did possess NDP-glucose pyrophosphorylase activity such that the donor substrate preference was UTP > dTTP ≫ CTP > GTP > ATP > ITP.
Therefore, this enzyme does indeed possess primarily GalU/GtaB-type UDP-glucose pyrophosphorylase activity and is not an efficient source of GDP-glucose.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
Second, SVEN_3972 was isolated as a dimeric enzyme, according to size exclusion chromatography, that was able to produce GDP-glucose from glucose 1-phosphate (2 mm) plus GTP (1 mm). It seemed at first to be quite specific because it was inactive with glucose 1-phosphate plus any other donor tested (ATP, UTP, ITP, or CTP). However, it turned out to be most active with mannose 1-phosphate plus either GTP or ITP. The highest catalytic efficiency was with ITP and mannose 1-phosphate (Table 3). By contrast, the efficiency with which it generated GDP-glucose was over 100-fold lower, so its primary role seems to be the formation of IDP-mannose.
Enzyme activity was not affected by the potential allosteric effectors glucose 6-phosphate, fructose 6-phosphate, GDP, GMP, or phosphoenolpyruvate up to 5 mm.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
Kinetic analysis of S. venezuelae SVEN_3972
Third, SVEN_2781 did show relatively high activity with glucose 1-phosphate (1 mm) and GTP (0.25–1 mm) and was inactive with any other donor tested (ITP, ATP, CTP, or UTP). The production of GDP-glucose was confirmed using NMR spectroscopy. However, SVEN_2781 also exhibited a slightly higher activity with mannose 1-phosphate plus GTP (Table 4). This was perhaps to be expected because it shared 59% sequence identity with M. tuberculosis GDP-mannose pyrophosphorylase (32). NMR spectroscopy showed that with equimolar concentrations of donor and acceptor, the reaction went to completion.
This showed that the chemical equilibrium was strongly in favor of the formation of GDP-mannose. The enzyme was present primarily in an active dimeric form, but inactive octameric and higher oligomeric forms were also detected using size exclusion chromatography. Enzyme activity was not affected by any of the potential allosteric effectors described above, in contrast to enzymes such as ADP-glucose pyrophosphorylase from other actinomycetes (27, 33).[](https://www.ncbi.nlm.nih.gov/mesh/C031590)
Kinetic analysis of S. venezuelae SVEN_2781
## Discussion
*In the **Discussion** section:*
We have now shown that S. venezuelae OtsA has a preference for GDP-glucose as the donor (Table 1), as has been reported for OtsA enzymes from other Streptomyces species (15–17) and other actinomycetes, such as Rubrobacter xylanophilus (18). Interestingly, not all examples of OtsA from actinomycetes share this donor specificity because the enzymes from mycobacteria show a preference for ADP-glucose (27, 34, 35). The best characterized OtsA enzyme to date has been that from E. coli, which has a preference for UDP-glucose. Several ligand-bound structures have allowed the residues that define its base specificity to be identified (Fig. 3) (20, 21, 36). Our structure of the enzyme from S. venezuelae shows that these defining residues are not conserved.
By contrast, the equivalent residues that define the donor base specificity of VldE (29), which uses GDP-valienol, are similar in S. venezuelae OtsA. This helps explain why the S. venezuelae enzyme shows specificity for purine bases, with guanine being preferred. Interestingly, both open and closed conformations of E. coli OtsA and VldE have been described, where ligands tend to promote the open form (20, 21, 29, 36). Our structure of S. venezuelae OtsA most closely resembles the open conformations despite not having any ligands bound. In addition, there is a lack of electron density associated with a loop comprising amino acids 18–22 (peptide sequence GEDGE) of the S. venezuelae enzyme, implying disorder in this region.
This loop is topologically in a similar position to a loop known to interact with the acceptor substrate in the E. coli enzyme that can be disordered in crystals lacking an acceptor bound to the active site (21). It is therefore conceivable that when ligands are bound, the S. venezuelae enzyme adopts a different conformation, and the GEDGE loop forms ordered contacts with the acceptor substrate.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
An otsA null mutant of S. venezuelae exhibits no phenotype when grown on maltose because this GlgE pathway intermediate bypasses the need for OtsA to generate trehalose (Fig. 1). However, when grown on galactose, the mutant accumulates GDP-glucose (Fig. 5), showing that OtsA would normally utilize this donor substrate, although it is also capable of using ADP-glucose to some extent (Table 1). Then again, S. venezuelae does not possess an obvious glgC homologue that would code for an ADP-glucose pyrophosphorylase (5, 8), implying that this organism cannot produce ADP-glucose. The accumulation of both glucose 1-phosphate and galactose 1-phosphate is consistent with this organism assimilating galactose via a Leloir-type pathway (31). What was more unexpected was the cell density-dependent growth phenotype of the mutant when grown on galactose (Fig. 4).
The accumulation of maltose 1-phosphate is known to lead to a delayed growth phenotype in S. venezuelae, but a lack of α-glucan has no effect (8). More dramatically, the accumulation of maltose 1-phosphate causes bacterial cell death in M. tuberculosis (4). It is therefore possible that the accumulation of glucose 1-phosphate and galactose 1-phosphate slows growth in S. venezuelae. Interestingly, the accumulation of ADP-glucose appears to be lethal in M. tuberculosis (37), so perhaps it is the accumulation of GDP-glucose that actually slows the growth of S. venezuelae. Either way, it is intriguing that the phenotype is cell density-dependent.
Perhaps these or some other toxic metabolites are exported out of cells. Alternatively, perhaps some nutrients in the solid medium become locally limiting or metabolic fluxes are influenced by quorum sensing. Further studies will be required to establish the basis for the growth phenotype.[](https://www.ncbi.nlm.nih.gov/mesh/D008320)
It is now clear that S. venezuelae OtsA utilizes GDP-glucose both in vitro and in vivo. We therefore sought an enzyme that is capable of generating this donor substrate. With the absence of a known sequence being associated with such an enzyme, we tested three candidate enzymes based on homology with the potentially similar GDP-mannose pyrophosphorylases. SVEN_3027 generated UDP-glucose as expected, given that it was also a homologue of GalU/GtaB from other organisms (33, 38). SVEN_3972, on the other hand, was somewhat unusual because its main activity appeared to be associated with the production of IDP-mannose (Table 3).
Such an activity has rarely been reported (39). The values of kcat/Km with this enzyme were quite low (40). This implies that either the enzyme is inherently not very efficient or its primary role is to generate another nucleotide sugar diphosphate that has not yet been tested.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
SVEN_2781 was able to generate GDP-glucose, even if it was also capable of forming GDP-mannose with slightly higher efficiency (Table 4). In contrast to SVEN_2781, GDP-mannose pyrophosphorylases from other organisms are specific for mannose-1-phosphate and do not utilize glucose-1-phosphate in all cases investigated (41, 42). The values of kcat/Km for SVEN_2781 were modest but well within the normal physiological range (40). Therefore, it seems reasonable to deduce that SVEN_2781 is capable of generating the donor substrate for OtsA in vivo. This enzyme therefore constitutes the first example of a GDP-glucose pyrophosphorylase (EC 2.7.7.34) of known sequence, although such enzyme activity has been documented several times (23–26).
There remains the possibility that there are other pyrophosphorylases capable of generating GDP-glucose in S. venezuelae, and this can be explored in the future with reverse genetics and the characterization of the other pyrophosphorylases predicted to exist in this organism.[](https://www.ncbi.nlm.nih.gov/mesh/C100085)
## Experimental Procedures
*In the **Experimental Procedures** section:*
## Recombinant Proteins
*In the **Recombinant Proteins** section:*
Recombinant proteins were produced as described previously (4, 6). Genes were synthesized with optimum codon usage for expression in E. coli (Genscript Corp., Piscataway, NJ) to give proteins with tobacco etch virus protease-cleavable N-terminal His6 tags. S. venezuelae OtsA (gene locus synonyms SVEN15_3951 and SVEN_4043) was produced in E. coli SoluBL21. Cells were grown in lysogeny broth at 37 °C to an optical density of 0.6 at 600 nm before induction with 0.5 mm isopropyl β-d-thiogalactopyranoside (IPTG). After a further 20 h of incubation at 37 °C, cells were harvested by centrifugation for 10 min at 5000 × g at room temperature and disrupted by sonication (5 s on and 3 s off over 20 min on ice).
The protein was purified using nickel affinity chromatography and stored in 5 mm HEPES, pH 7.0, containing 60 mm MgCl2. The size of OtsA was determined using size exclusion chromatography. A Superdex 200 10/300 GL column (GE Healthcare) was calibrated using blue dextran (2 MDa), β-amylase (200 kDa), alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa), carbonic anhydrase (29 kDa), and cytochrome c (12 kDa), using 20 mm HEPES buffer, pH 7.0, containing 10 mm MgCl2 and 100 mm NaCl.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
The pyrophosphorylases were produced in a similar way with modifications. SVEN_3027 (gene locus synonyms GalU, SVEN15_2964) was produced in E. coli BL21(DE3)pLysS, which was incubated at 28 °C after induction. The protein was stored at −80 °C in 20 mm Tris-HCl, pH 8.0, containing 100 mm NaCl and 10% (v/v) glycerol. SVEN_3972 (gene locus synonym SVEN15_3882) was produced in E. coli SoluBL21, which was incubated for 5 h after induction with 0.5 mm IPTG. After nickel affinity chromatography, the enzyme was applied to a Superdex 200 16/600 column equilibrated with 50 mm MOPS, pH 8.0, containing 0.025 mm EDTA and 5% (w/v) sucrose.
Active fractions were concentrated, and 0.5 mm dithiothreitol was added for storage at −80 °C. SVEN_2781 (gene locus synonym SVEN15_2722) was produced in E. coli BL21(DE3), which was grown for 5 h at 18 °C after induction with 0.5 mm IPTG. The enzyme was stored at −80 °C in 50 mm Tris-HCl, pH 8.0, containing 0.5 mm dithiothreitol, 0.05 mm EDTA, and 5% (w/v) sucrose.[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
## Enzyme Assays
*In the **Enzyme Assays** section:*
Enzyme activity was quantified as described previously. Briefly, OtsA activity was determined by monitoring the production of NDP using a continuous coupled assay with phosphoenolpyruvate, pyruvate kinase, and lactate dehydrogenase, allowing the oxidation of NADH to be followed spectrophotometrically (27). NDP-sugar pyrophosphorylase activity was determined by monitoring the production of pyrophosphate using a stopped coupled assay with inorganic pyrophosphatase, allowing the production of inorganic phosphate to be detected spectrophotometrically with malachite green (43). Kinetic constants were calculated for each enzyme and condition using all replicate data using SigmaPlot version 13.0 with the Michaelis-Menten equation embedded in the Simple Ligand Binding macro. Compounds that could potentially allosterically regulate pyrophosphorylases were assessed at a concentration of 5 mm, and in each case, no impact on enzyme activity was detected.
Enzyme-catalyzed reactions were also followed using 1H NMR spectroscopy. The consumption of the sugar 1-phosphate acceptors was monitored using their anomeric double doublet resonances at ∼5.5 ppm, relative to the signal from sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4. The formation of the NDP-sugar products was monitored using their anomeric double doublet resonances, at ∼5.6 ppm, which were distinct from those of the donors.[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
## Construction and Complementation of an S. venezuelae otsA Null Mutant
*In the **Construction and Complementation of an S. venezuelae otsA Null Mutant** section:*
The bacterium S. venezuelae ATCC10712 was cultured at 28 °C in malt extract/yeast extract/maltose medium with 50% tap water (MYM-TAP) supplemented with 0.4 ml of trace element solution per liter (44). The maltose was substituted with other carbohydrates where indicated. The minimal medium (8, 44) was supplemented with 4–8 g/liter carbohydrate as a carbon source. The S. venezuelae null mutant in otsA (gene locus synonyms SVEN15_3951 and SVEN_4043) was generated using Redirect PCR targeting (45) to replace the coding region with an apramycin resistance (apr) cassette (8). The otsA gene was first replaced in the cosmid 1-H1,4 using the primers 5′-CGTTTGAGCGTTTACGGGACGGGCTAGGTTCGCCACATGATTCCGGGGATCCGTCGACC and 5′-CTGGAGCGGCCCCCACCTCGACAAGGTTCCAGGCGCTCATGTAGGCTGGAGCTGCTTC.
The disruption of the cosmid was confirmed using restriction digestion with NruI. The cosmid was introduced into S. venezuelae by conjugation, and a double cross-over null mutant was selected on the basis of apramycin resistance and kanamycin sensitivity to give S. venezuelae ΔotsA::apr strain FM003. The deletion of chromosomal otsA was confirmed using Southern blotting with either XhoI or ApaLI. To complement the mutation, the otsA gene was amplified using 5′-GACCGGCCCAAGCCCACCC and 5′-TCAGGCGTCGCTCAGCCCC to give the open reading frame plus ∼300 base pairs upstream covering the native promotor. This fragment was cloned into pMS82 (46) to give pFM4, which was introduced into the mutant by conjugation to give S. venezuelae ΔotsA::apr
attBΦBT1::otsA strain FM003-pFM4 (FM013) where the plasmid is integrated at the ΦBT1 attachment site. This complemented strain had a wild-type phenotype. An empty vector control ΔotsA::apr attBΦBT1::pMS82 strain (FM012) was also generated, which exhibited the phenotype of the mutant strain. The morphology of cells and spores was assessed using scanning electron microscopy, and the production of α-glucan was determined using transmission electron microscopy with periodic acid/thiocarbohydrazide/silver proteinate staining, as described previously (8). Spore stocks were standardized, allowing identical numbers of spores to be used to inoculate plates or cultures when assessing phenotypic differences.[](https://www.ncbi.nlm.nih.gov/mesh/D008320)
## Metabolite Analysis
*In the **Metabolite Analysis** section:*
Cells were grown on solid media overlaid with sterile cellophane discs and harvested by scraping (8). The cells were freeze-dried and powdered using a micropestle before being boiled, to denature enzymes, and disrupted by sonication. Cell debris was removed by centrifugation and cell-free extracts were analyzed by 1H NMR spectroscopy using sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4 as an internal standard using assignments as previously described (6, 8).[](https://www.ncbi.nlm.nih.gov/mesh/D002476)
## Protein Crystallography
*In the **Protein Crystallography** section:*
Protein crystallization trials were set up in 96-well MRC plates (Molecular Dimensions) using the following screens: JCSG-plus (Molecular Dimensions), PACT premier (Molecular Dimensions), Structure (Molecular Dimensions), Morpheus (Molecular Dimensions), ammonium sulfate (Qiagen), and PEG suite (Qiagen). Each reservoir was filled with 50 μl of the screen precipitant solution using the Freedom EVO liquid-handling robot (Tecan). A 0.3-μl sample of precipitant was mixed with 0.3 μl of protein (10–15 mg ml−1 in 5 mm HEPES, pH 7.0, containing 60 mm MgCl2) in a sitting drop format using an OryxNano robot (Douglas Instruments Ltd.). The plates were sealed and stored at 20 °C.
Drops were monitored using a SMZ800 microscope (Nikon). After 3 days, rectangular plate crystals were present in the Morpheus condition 2-2 (0.12 m ethylene glycols (di-, tri-, tetra-, and penta-ethylene glycols), 0.1 m imidazole and MES, pH 6.5, 20% (v/v) ethylene glycol, and 10% (w/v) PEG 8000). Crystals were mounted directly from the screen in LithoLoops (Molecular Dimensions). They were then flash-cooled by plunging into liquid N2 and stored in Uni-Puck cassettes (MiTeGen) for transport to the Diamond Light Source (Oxfordshire, UK). Crystals were subsequently transferred robotically to the goniostat on the beamline and maintained at −173 °C with a Cryojet cryocooler (Oxford Instruments).
Native diffraction data were recorded on beamline i04-1 (wavelength = 0.920 Å, 1800 images with 0.2° oscillation) using a Pilatus 2M detector (Dectris), processed using xia2 (47), and scaled using SCALA (48). A 5% subset of the total number of reflections was set aside to determine the free R factor (49) during model building and refinement. All subsequent processing was conducted using the CCP4 suite (50). The resultant data collection statistics are summarized in Table 2. The crystals belonged to space group P21 with approximate cell parameters of a = 41.43, b = 168.40, c = 133.90 Å, β = 97.19°.[](https://www.ncbi.nlm.nih.gov/mesh/D000645)
The structure of OtsA was solved by molecular replacement using programs from the CCP4 suite (50). The search model for molecular replacement was obtained by submitting the sequence to the Phyre2 server (28), which generated a template from the structure of S. hygroscopicus VldE (PDB code 3T5T), with which it shares 30% sequence identity. PHASER (51) was used to locate four copies of the protomer in the asymmetric unit giving an estimated solvent content of 46%. Density modification was carried out using PARROT (52), which benefitted from the use of 4-fold averaging and enabled the starting model to be entirely rebuilt with Buccaneer (53). Model building was then completed using COOT (54), alternating with refinement with REFMAC5 (55).
The final model consisted of 1770 residues in four polypeptide chains, 652 water molecules, and four MES and four ethylene glycol molecules. The final Rwork and Rfree values were 0.197 and 0.238 to 1.95 Å resolution (refinement statistics are summarized in Table 2). MolProbity (56) was used to validate the model before deposition in the PDB. All structural figures were prepared using CCP4MG (57).[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Author Contributions
*In the **Author Contributions** section:*
S. B. coordinated the study and wrote the paper. F. M., C. E. M. S., and D. M. L. carried out the crystallography. M. D. A. D. and A. A. I. designed the kinetic experiments and M. D. A. D. carried them out. F. M. carried out the microbiology. All authors analyzed the data, edited the manuscript, and approved the final version of the manuscript.
This work was supported by United Kingdom Biotechnology and Biological Sciences Research Council Doctoral Training Grant BB/F017294/1 and Institute Strategic Programme Grant BB/J004561/1, the John Innes Foundation, and the European Molecular Biology Organization (short term fellowship grant ASTF537-2014). The work in Argentina was supported by CONICET, UNL, and ANPCyT (PICT'15 1767). The authors declare that they have no conflicts of interest with the contents of this article.
The atomic coordinates and structure factors (code 5LQD) have been deposited in the Protein Data Bank (http://wwpdb.org/).
M. J. Bibb and M. J. Buttner, unpublished data.
PDB
Protein Data Bank
IPTG[](https://www.ncbi.nlm.nih.gov/mesh/D013862)
isopropyl β-d-thiogalactopyranoside.[](https://www.ncbi.nlm.nih.gov/mesh/D013862)
The abbreviations used are:
# References
*In the **References** section:*
|
# Introduction
LC-MS/MS Analysis Unravels Deep Oxidation of Manganese Superoxide Dismutase in Kidney Cancer
# Abstract
*In the **Abstract** section:*
Manganese superoxide dismutase (MNSOD) is one of the major scavengers of reactive oxygen species (ROS) in mitochondria with pivotal regulatory role in ischemic disord[ers, inflammation and c](https://www.ncbi.nlm.nih.gov/mesh/D017382)an[cer](https://www.ncbi.nlm.nih.gov/mesh/D017382). Here we report oxidative modification of MNSOD in human renal cell carcinoma (RCC) by the shotgun method using data-dependent liquid chromatography tandem mass spectrometry (LC-MS/MS). While 5816 and 5571 proteins were identified in cancer and adjacent tissues, respectively, 208 proteins were found to be up- or down-regulated (p < 0.05). Ontological category, interaction network and Western blotting suggested a close correlation between RCC-mediated proteins and oxidoreductases such as MNSOD. Markedly, oxidative modifications of MNSOD were identified at histidine (H54 and H55), tyrosine (Y58), tryptophan (W147, W149, W205 and W210) and asparagin[e (N206 a](https://www.ncbi.nlm.nih.gov/mesh/D006639)nd N209) residue[s additi](https://www.ncbi.nlm.nih.gov/mesh/D014443)onal to [methionine](https://www.ncbi.nlm.nih.gov/mesh/D014364). These oxidative insults were lo[cated at t](https://www.ncbi.nlm.nih.gov/mesh/D001216)hree hotspots near the hydrophobic pocke[t of the m](https://www.ncbi.nlm.nih.gov/mesh/D008715)anganese binding site, of which the oxidation of Y58, W147 and W149 was up-regulated around [three fol](https://www.ncbi.nlm.nih.gov/mesh/D008345)ds and the oxidation of H54 and H55 was detected in the cancer tissues only (p < 0.05). When normalized to MNSOD expression levels, relative MNSOD enzymatic activity was decreased in cancer tissues, suggesting impairment of MNSOD enzymatic activity in kidney cancer due to modifications. Thus, LC-MS/MS analysis revealed multiple oxidative modifications of MNSOD at different amino acid residues that might mediate the regulation of the superoxide radicals, mitochondri[al ROS sca](https://www.ncbi.nlm.nih.gov/mesh/D000596)venging and MNSOD activity in kidney cancer.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
## 1. Introduction
*In the **1. Introduction** section:*
Superoxide dismutases (SODs) are the ubiquitous superfamily of antioxidant metalloenzymes that convert the superoxide anion (O2−) into oxygen and hydrogen peroxide (H2O2) when cells are exposed to oxidation. These enzymes are the major scavengers of reactive oxygen species (ROS), the natural side products during the process of aerobic cells respiration in physiological and pathological conditions [1]. Three forms of superoxide dismutases, namely Cu/ZnSOD (SOD1/3 in humans), Fe/MNSOD (also called MNSOD, SOD2 in humans), and NiSOD (only found in bacteria), are identified [2]. Among them, MNSOD is the primary mitochondrial ROS scavenging enzyme that catalyzes the conversion of superoxide to H2O2, which is subsequently transformed to water by catalase and other peroxidases [1,3].
MNSOD is essential for the survival of all aerobic organisms from bacteria to humans under physiological conditions [1,4].[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
Posttranslational modifications (PTMs) play a major role in regulating MNSOD activities, interactions, and localization [1]. Protein oxidative modification, a major class of PTM, is caused by oxidative or nitrative disruption of amino acid residues. Oxidative protein modification has been described during aging and various pathological conditions, and serves as a useful biomarker for assessing oxidative stress processes in aging and disease conditions [4]. Protein oxidation can lead to hydroxylation of aromatic groups and aliphatic amino acid side chains, nitration of aromatic amino acid residues, sulfoxidation of methionine residues [5]. Some specific enzymes, such as tyrosine hydroxylase (TH), facilitate the oxidation process in vivo [6].
Aromatic amino acid side chains, such as tyrosine (Tyr, Y) and tryptophan (Trp, W), are more susceptible to free radical incursion leading to specific modification of the aromatic ring. For example, tyrosine (Tyr) yields greater 3,4-dihydroxyphenylalanine (DOPA) and less 2,4-isomer [7]. Protein oxidation can lead to diverse functional consequences, such as loss of enzymatic and binding activities, increased susceptibility to aggregation, proteolysis and altered immunogenicity [8]. Overall, it is believed that protein oxidation is central to functional deficit of the target proteins [9].[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
In addition to direct oxidative damage, cells could undergo nitrative damage through reactive nitrogen species (RNS) pathway, which often occurs with the redox-sensitive amino acid residues including tyrosine and tryptophan [10]. Nitration on tyrosine 58 (Y58 or Y34 when the 24 amino acids transit peptide is cleaved) of MNSOD has been detected by crystal structures and mutation analysis [1,11]. Previous data showed that human MNSOD Y58 was exclusively nitrated to 3-nitrotyrosine (3-NT) to inactivate its enzymatic activity, and nitration of Y58 alone was sufficient for inactivation of MNSOD enzymatic activity [1]. Similar to nitration, acetylation of MNSOD was shown to impair its enzymatic activity. Recent evidence suggests the presence of multiple MNSOD acetylation sites on lysine (Lys, K), such as 53, 68, 89, 122, 130 [12].
It was also shown that SIRT3 could deacetylate MNSOD and enhance its enzymatic activity in vitro [1,3,13].[](https://www.ncbi.nlm.nih.gov/mesh/D026361)
Oxidative stress in mitochondria becomes a major source of ROS. Free radicals originating from mitochondria interact with surrounding molecules and initiate a cascade of signaling pathways leading to oxidative modifications of cellular organelles. Oxidative modifications within mitochondrial proteins may contribute to the development of carcinogenesis [14]. The precise mechanisms by which oxidative and nitrative conditions affect MNSOD structure and function remain unclear. While MNSOD is reportedly up-regulated in kidney cancer
[15,16], its enzymatic activity in tumor areas is similar to adjacent kidney tissues [16,17], suggesting that MNSOD enzymatic activity does not necessarily correlate with its protein expression. It was recently shown that oxidative stress-mediated MNSOD modifications compromised its enzymatic activity and led to functional deficit [1,3,18,19]. In this study, we first pinpointed MNSOD as a vital protein in the kidney tissues from clear cell renal cell carcinoma (ccRCC) using liquid chromatography-tandem mass spectrometry (LC-MS/MS), then attempted to identify and quantify crucial sites of oxidative modifications of MNSOD using high resolution LC-MS/MS and Progenesis LC-MS software, respectively. The newly identified (to the best of our knowledge) oxidation sites we report may introduce novel perspectives for the refinement of MNSOD regulatory mechanisms.
Characterization of MNSOD oxidative modifications may be vital for interpretation of previous studies on MNSOD in tumor biology and could enhance our intervening ability in tumorigenesis.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
## 2. Results
*In the **2. Results** section:*
## 2.1. Quantitative Proteomic Analysis Suggests the Involvement of Anti-Oxidative Stress Pathway in ccRCC
*In the **2.1. Quantitative Proteomic Analysis Suggests the Involvement of Anti-Oxidative Stress Pathway in ccRCC** section:*
Three pairs of tumor and adjacent tissues from ccRCC patients were analyzed by LC-MS/MS. Totally, 5571 and 5816 non-redundant proteins with FDR < 0.01 were identified in adjacent and ccRCC tissues, respectively (Table S1). Among the identified proteins, the average spectral counts of three repeats were used to evaluate their expression levels, which revealed 100 up- and 108 down-regulated proteins for at least 1.4 folds in ccRCC vs. adjacent tissues with the criteria of p < 0.05, detected in every replicate, and the average spectral counts of >20 (Table S2). According to Database for Annotation, Visualization, and Integrated Discovery (DAVID) functional annotation, 33 of the 208 dysregulated proteins were related to oxido-reductases (p = 2.87 × 10−18) (Tables S3 and S4).
Heatmap analysis (Figure 1a) denoted the involvement of these oxidoreductases in binding with cofactors, coenzyme, NAD, NADH and/or NAD(P). Notably, MNSOD was among these oxidoreductases, suggesting that mitochondrial MNSOD was also involved in the electron transport chain for ROS removal. Consistent with these observations, ontological category based on biological process using Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) (Table S5) indicated that one of the most significant categories was oxidation-reduction process (37 proteins, p = 6.49 × 10−17) (Table S6). Three main interactive clusters were formed among the 37 interacting proteins with MNSOD as an important node (Figure 1b). Because cancer cells usually demand high ROS concentrations to maintain their high proliferation rate [14], these data suggested that oxido-reductases, particularly MNSOD, played an important role in RCC pathogenesis [16].[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## 2.2. Oxidative Modification of MNSOD
*In the **2.2. Oxidative Modification of MNSOD** section:*
For a deep post-translational modification analysis, MNSOD was excised from SDS-PAGE (Figure 2a) and analyzed by LC-MS/MS. With a standard search using MASCOT and SEQUEST, 18 high confident peptides of MNSOD were identified, which covered 76% of the sequence (Figure 2b).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
Next, the MS/MS data were searched again with an open modification search algorithm [20] to identify the peptides of MNSOD containing oxidation at any possible amino acids with a delta mass of +16 Da (Table 1). Totally 168 of +16 modification events were counted, 81 of which occurred at tryptophan (W), 27 at glycine (G), 24 at tyrosine (Y), 14 at histidine (H), 13 at asparagine (N), 7 at alanine (A), and 2 at valine (V).[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
To confirm whether these +16 modifications were due to oxidation, restricted search with dynamic oxidation (+15.995 Da) was performed using MASCOT and SEQUEST. High confident oxidation modifications were confirmed at tryptophan (W), tyrosine (Y), histidine (H), and asparagine (N) residues of MNSOD (Table 2). Markedly, these oxidation sites were located in three hotspots, including hotspot 1 (H54, H55, Y58), hotspot 2 (W147, W149) and hotspot 3 (W205, N206, N209, W210). Additionally, the known nitration (+44.985 Da) at tyrosine (Y58) was also identified in the kidney tissues (Table 2).[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## 2.3. Tryptophan Oxidation at W147/W149 and W205/W210 of MNSOD
*In the **2.3. Tryptophan Oxidation at W147/W149 and W205/W210 of MNSOD** section:*
Over half of the observed oxidation events of MNSOD occurred at tryptophan. Four potentially oxidized sites, W147/W149 and W205/W210, were identified in two hotspots near the C-terminus. In the hotspot around the residues from 200 to 212 (hotspot 3), two of the oxidized tryptophan residues were found in the peptide AIW205NVINW210ENVTER. The oxidized W210 (Figure S1a) was identified by the unmodified b4–7 ions and the mass shift of the b8–13 ions, supported by the unmodified y2–6 and the modified y7–12 ions, which has been reported before [21]. Similarly, oxidation at W205 (Figure S1b) was identified by the unmodified y3–9,11 ions followed by the modified y12 ion; it was supported by the shifted b5,7–8,10–11 ions.
The other oxidized tryptophan residues were found in the peptide LTAASVGVQGSGW147GW149LGFNK in hotspot 2. The oxidized tryptophan at W147 (Figure S1c) was uncovered by the unmodified y4–5,7 and modified y8–12,14 ions, which was supported by the unmodified b6,8–9,11 and modified b16,19 ions. Double oxidation at W147 and W149 in a different MS/MS profile of the same peptide (Figure S1d) was identified by the unmodified b6,8–9,11 ions and double mass shifts of 15.99 Da in the b15–19 ions. Markedly, these identified tryptophan oxidations at W147, W149, W205 and W210 were located in the hotspots that were reported to be important for the enzymatic activity of MNSOD [2].
Notably, tryptophan hydroxylase (TPH), a well characterized enzyme in mammals, was shown to convert tryptophan to 5-hydroxytryptophan, the precursor for the neurotransmitter serotonin and melatonin [22], although it was unclear whether TPH was involved in tryptophan oxidation in proteins.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
## 2.4. Asparagine Oxidation at N206 and N209 of MNSOD
*In the **2.4. Asparagine Oxidation at N206 and N209 of MNSOD** section:*
The asparagine residue in proteins might be oxidized to form beta-hydroxyl asparagine [23]. We identified two asparagine oxidation sites at N206 and N209 in MNSOD, and the two oxidized asparagine residues were found in the peptide AIWN206VIN209WENVTER, which was located in the C-terminal. Notably, this modification was observed only in combination with other modification: double oxidation at W205N206 and N209W210 in the same peptide (Table 2). The oxidized N206 in the peptide (Figure S1e) was identified by the unmodified y3–10 ions and the double mass shifts of 15.99 Da in the y12 ion, and it was further supported by the double mass shifts of 15.99 Da in the ions after b4. The oxidized N209 (Figure S1f) was identified by the first shift between the b4–6 ions and b7 ion, and the oxidized W210 identified by the second shift of 15.99 Da between the b7 ion and b8–13 ions; the double oxidized asparagine/tryptophan residues were confirmed by the unmodified y3–6 ions and double mass shifts of 15.99 Da in the y8–12 ions.
Therefore, we concluded that the asparagine residues at N206 and N209 were hotspots for oxidation that could be spontaneously oxidized synchronously with tryptophan.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
## 2.5. Histidine Oxidation at H54 and H55 of MNSOD
*In the **2.5. Histidine Oxidation at H54 and H55 of MNSOD** section:*
Histidine oxidations in proteins are commonly found in cells undergoing redox-mediated oxidation of histidine to 2-oxo-histidine. The oxidized product has been suggested as a biological marker for oxidatively modified proteins [24]. In this study, we identified two oxidized histidine residues at H54 and H55 in MNSOD. The oxidized histidine at H54 in the peptide H(54)HAAYVNNLNVTEEK (Figure S1g) was identified by the mass shifts at b2–14 ions, which was supported by the unmodified y2–14 ions followed by the total mass shift.
The double oxidized histidine at H54 and H55 in the peptide H54H55AAYVNNLNVTEEK (Figure S1h) was identified by double mass shifts at b11,13 ions, which was supported by the unmodified y4,6–13 ions followed by 32 Da shift of the total mass. The two oxidative histidine residues were located in the first hotspot for oxidation, which was also reported previously in human medulloblastoma cells [21].[](https://www.ncbi.nlm.nih.gov/mesh/D006639)
## 2.6. Tyrosine Oxidation at Y58 of MNSOD
*In the **2.6. Tyrosine Oxidation at Y58 of MNSOD** section:*
Tyrosine could be oxidized to form 3-hydroxyl tyrosine or 3,4-dihydroxyphenylalanine (L-DOPA) by tyrosine hydroxylase (TH) [6,10]; however, it was unclear whether TH oxidizes the tyrosine residue in proteins. The oxidized tyrosine at Y58 was identified because the mass shift of 15.99 Da was absent in the y4–10 ions, but present in the y11–14 ions in the MS/MS profile of the peptide HHAAY58VNNLNVTEEK (Figure S1i); it was further confirmed by the absence of the extra mass in the b2,4 ions and presence in the b5–14 ions.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
## 2.7. Oxidation at H54, H55, Y58, W147 and W149 of MNSOD Is Up-Regulated in ccRCC
*In the **2.7. Oxidation at H54, H55, Y58, W147 and W149 of MNSOD Is Up-Regulated in ccRCC** section:*
The peak areas were commonly considered proportional to the amounts of the peptide. In this study, the level of an oxidized residue was measured by the areas of all peptide fragments containing the oxidized residue. In comparison with unoxidized, the relative level of oxidized residues at H54, H55, Y58, W147, W149, W205, N206, N209 and W210 were calculated as 0.00012, 0.00012, 0.00073, 0.09530, 0.06171, 0.06938, 0.05298, 0.05298 and 0.06938, respectively, in kidney tissues (Figure 3a, Table S7). Apparently, the basal levels of oxidation in MNSOD might indicate a managed balance between formation and neutralization of ROS in vivo. Note that abundances of W205 and N206 were identical to W210 and N209 because they were from two identical peptides with the same abundance but different pattern.
The abundances of H54 and H55 were similar because they came from a low abundant, single oxidation peptide (H54) and a high abundant, double oxidation peptide (H54 and H55), so that their total abundances were mainly contributed by the high abundant peptide. Markedly, oxidation at H54 and H55 only found in kidney cancer tissues (p < 0.05, n = 4), and oxidation at Y58, W147 and W149 increased 2.63, 3.26 and 3.13-fold (p < 0.05, n = 4) in kidney cancer tissues, respectively, in comparison with those in adjacent non-cancer tissues of the same patient (Figure 3b). While, oxidation at N206 and N209 decreased 1.54-fold (p < 0.05, n = 4) and oxidation at W205 and W210 decreased 1.40-fold (p < 0.05, n = 4) in kidney cancer tissues. Thus, these data suggested that oxidation at certain amino acid positions of MNSOD could be dysregulated in kidney cancer.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
## 2.8. The Relative MNSOD Enzymatic Activity Is Attenuated in ccRCC
*In the **2.8. The Relative MNSOD Enzymatic Activity Is Attenuated in ccRCC** section:*
Western blotting analysis of four individual ccRCC and adjacent tissue lysates showed a significant increase of MNSOD expression in ccRCC (p < 0.05, Figure 4a). Then MNSOD enzymatic activity was measured from ccRCC and adjacent kidney tissue lysates using superoxide dismutase (SOD) activity kit. When normalized by MNSOD expression level, the relative enzymatic activity of MNSOD was decreased in ccRCC in comparison with adjacent tissues (p < 0.05, Figure 4b), although the total activity was similar. This observation suggested that modification of MNSOD might impair its enzymatic activity.
## 3. Discussion
*In the **3. Discussion** section:*
MNSOD is highly expressed in kidney and is reportedly up-regulated in ccRCC [15,16], the most common subtype of RCC in adults. Label-free quantitative proteomics approach, which is based on the number of acquired spectra for each protein or the ion intensities of identical peptides, has become a popular alternative to assess relative amount of peptides or proteins [25]. In this study, label-free quantification was performed to compare a number of proteins in kidney tissues, and 208 proteins were found to meet the criteria for dysregulation between ccRCC and adjacent tissues. Among them, MNSOD, a crucial protein regulating ROS was pinpointed as an important target by STRING and DAVID bioinformatics tools. It was followed by ion intensity-based quantification to compare oxidized peptides of MNSOD between ccRCC and adjacent tissues.
Using MNSOD protein bands from 4 pairs of kidney samples, shotgun proteomics, and open search algorithm, this approach revealed 9 oxidative modification sites at 4 different amino acid residues on endogenous MNSOD protein (Table 2). Modified sites involved two histidine (H54 and H55), one tyrosine (Y58), four tryptophan (W147, W149, W205 and W210) and two asparagine (N206 and N209), and most of the oxidized residues clustered at the C-terminus. Consistent with our findings, it was recently reported that MNSOD could be oxidized at H54, H55 and W210 in medulloblastoma cells [21]. Notably, three hotspots at amino acid residues 54–58, 147–149, and 205–210 were observed and multiple oxidation incidents occurred simultaneously in the nearby positions, such as W147W149 and N209W210. Our data revealed that MNSOD could be oxidized at tryptophan/asparagine residues, while tryptophan oxidation and asparagine hydroxylation were previously reported to alter the structure and/or function of the oxidized protein [26,27].
These observations demonstrated that mitochondrial MNSOD was not only upregulated in kidney cancer, but it was also deeply oxidized at several hotspots.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
Based on three-dimensional structure (1LUV.pdb), the catalytic core (active site) of MNSOD is located between the N-terminal helices and the C-terminal α/β domain, coordinating in a strained trigonal bipyramidal geometry by the side chains of His50, His98, Asp183 and His187 and the Mn ion in the active metal ion (Figure 2b and Figure 4c) [3,28]. The residues near the active site were known to affect the enzymatic activity or the stability of sites involving His54, Tyr58 and Trp147 [28,29,30]. The hydrophobic side chains that surrounded the metal-ligand cluster including His51, His54, His55, Tyr58, Phe101, Trp102, Trp149, Tyr190, and Tyr200 could also be altered [31].
Therefore, oxidation modifications of these residues in and around the MNSOD active site were believed to affect the catalytic activity [32]. In this study, we found that the oxidized residues were close to the active site or located in the hydrophobic side chains, which surrounded the manganese-ligand cluster and formed a hydrophobic pocket on the surface of proteins. These included the residues His54 (H54), His55 (H55), Tyr58 (Y58), Trp147 (W147) and Trp149 (W149) (Figure 4) [2,5,6,9,31]. These observations are supported by previous reports demonstrating that mutations and modifications close to the active site significantly affect the structural stability as well as the catalytic activity of MNSOD
[33]. In addition, the relative enzymatic activity, when normalized by MNSOD expression, was decreased in ccRCC in comparison with the adjacent tissues. We, therefore, speculated that the newly identified oxidative modification may initially affect the structure then compromise the catalytic activity of MNSOD, and thereby contribute to ccRCC formation and development.[](https://www.ncbi.nlm.nih.gov/mesh/D006639)
Using MS, we found that human endogenous mitochondrial MNSOD can undergo oxidative modifications at histidine, tyrosine, tryptophan and asparagine residues. To our knowledge, we report for the first time that human MNSOD can be oxidized at tyrosine and asparagine residues. Although MNSOD oxidation at H54, H55 and W210 was reported in medulloblastoma cells [21], the extent of oxidation and its modification patterns have not been fully elucidated. Asparagine hydroxylation of ankyrin repeat domain (ARD) of ankyrin R by factor-inhibiting hypoxia-inducible factor (FIH) altered the structure and function of the erythrocyte cytoskeletal ankyrin protein [27], and tyrosine oxidation affected mouse bone marrow mesenchymal stem cells (BMMSCs) proliferation and differentiation
[26], but little is known about asparagine and tyrosine oxidation in MNSOD. Label-free quantification demonstrated a significant increase of oxidation at H54, H55, Y58, W147 and W149 in ccRCC, comparing with adjacent tissues (Figure 3b). As protein oxidation usually altered the enzymatic activity [8], we speculated that the oxidation of H54, H55, Y58, W147 and W149 mediated MNSOD activity, and thereby, contributed to ccRCC initiation. It also accounts for the inconsistency of enzymatic activity and MNSOD expression in ccRCC versus adjacent tissues. Cancer cells produce greater amount of ROS than their corresponding normal cells, thus more MNSOD aggregates in ccRCC to scavenge ROS.
However, increased ROS also inhibits MNSOD activity via oxidative mechanisms in ccRCC. Our observations suggest that, in addition to overexpression of MNSOD in cancer cells, MNSOD oxidation might also contribute to ROS regulation by mediation of MNSOD activity [34].[](https://www.ncbi.nlm.nih.gov/mesh/D006639)
## 4. Materials and Methods
*In the **4. Materials and Methods** section:*
## 4.1. Sample Preparation and Protein Extraction
*In the **4.1. Sample Preparation and Protein Extraction** section:*
Kidney cancer tissues (ccRCC) and adjacent morphologically normal kidney cortex (adjacent tissues) were collected from 7 cancer patients with nephrectomy (5 males and 2 females, age from 46 to 75, Fuhrman nuclear grading with 2 G1 + 2 G2 + 2 G3 + 1 G4, TNM staging with 2 T1 + 3 T2 + 2 T3) in full compliance with Institutional Ethics Review Board’s guidance. All procedures were consistent with the National Institutes of Health Guide and approved by the institutional board with patients’ written consent. This study was evaluated and approved by the Ethics Committee of Shandong Provincial Hospital Affiliated to Shandong University. Frozen tissues were homogenized in RIPA lysis buffer (Millipore, Billerica, MA, USA) including the protease inhibitor cocktail (Roche, Basel, Switzerland). Protein concentration was determined using BCA protein assay (Biyuntian, Beijing, China).
Total proteins of the ccRCC and adjacent tissues from 3 patients (TNM staging with 1 T1 + 1 T2 + 1 T3) were subjected to quantitative proteomic analysis, and those from 4 patients (TNM staging with 1 T1 + 2 T2 + 1 T3) were used for PTMs and MNSOD enzymatic activity analyses.
## 4.2. Liquid Chromatography Tandem Mass Spectrometry
*In the **4.2. Liquid Chromatography Tandem Mass Spectrometry** section:*
For quantitative proteomic analysis, ~100 μg of total proteins from the tumor and adjacent tissues were separated on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), roughly divided into 10 slices according to molecular weights, and digested with trypsin as described previously [35]. The extracted peptides from each gel band were desalted by ZipTip (Millipore) and subjected to peptide fractionation using an EASY-nLC II system (Thermo Scientific, Waltham, MA, USA). The gradient-eluted peptides were analyzed by a Velos Pro ion trap mass spectrometer (Thermo Scientific). The liquid chromatography column (150 mm × Ø 0.075 mm) was packed with 3 μm 100 Å PepMap C18 (Thermo Scientific). Samples were analyzed using 120 min linear gradient of 5%–35% acetonitrile in 0.1% formic acid with a flow rate of 300 nL/min (solvent A: 0.1% formic acid in water, solvent B: 0.1% formic acid in acetonitrile) and the mass spectrometer was operated in a data-dependent mode, in which MS/MS fragmentation was performed using the 20 most intense peaks of every full MS scan.
MS/MS spectra were searched against the human protein database (UniProtKB; 88,295 entries) using SEQUEST HT, which is part of the Proteome Discoverer 1.4 data analysis package (Thermo Scientific, San Jose, CA, USA). Trypsin (full cleavage) was specified as cleavage enzyme allowing up to two missing cleavages. MS/MS spectra were searched with a maximum allowed deviation of 1 Da for the precursor mass and 0.8 Da for fragment masses. Methionine oxidation was selected as a dynamic modification, and the false discovery rate (FDR) was 1%.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
For deep PTMs analysis, total proteins from the tumor and adjacent tissues were separated on SDS-PAGE and the ~22 kDa bands corresponding to MNSOD were excised. Proteins in the gel slices were digested with trypsin and peptides were recovered with C18 ZipTip. The extracted and desalted peptides were subjected to peptide fractionation by liquid chromatography on an EASY-nLC 1000 system (Thermo Scientific) equipped with a long C18 column (300 mm × Ø 0.075 mm, 3 μm particles). Samples were fractionated with 120 min linear gradient of 5%–35% acetonitrile/0.1% formic acid at a flow rate of 300 nL/min (solvent A: 0.1% formic acid in water, solvent B: 0.1% formic acid in acetonitrile). The MS and MS/MS spectra were acquired by a LTQ-Orbitrap Elite mass spectrometer (Thermo Scientific) in a data-dependent mode, in which MS/MS fragmentation of the 20 most intense peaks were acquired for every full MS scan.
MS/MS spectra were searched against the human protein database using MASCOT and SEQUEST. Trypsin (full cleavage) was specified as cleavage enzyme allowing up to two missing cleavages. MS/MS spectra were searched with a maximum allowed deviation of 10 ppm for the precursor mass and 0.6 Da for fragment masses. The oxidation of various amino acid residues including methionine was selected as dynamic modification, and the false discovery rate (FDR) was 1%. For nitration, tyrosine was selected as dynamic modification.
The thresholds for the accepted MS/MS spectra (peptides) were Ions Score of 38 for MASCOT or XCorr of 1.22 × charges for SEQUEST [36]. All modification site assignments were confirmed by manual spectrum interpretation.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
For open modification search, a multi-blind spectral alignment algorithm, termed MODification via alignment or MODa [20], was used. Amino acid residues with a delta mass of +16 Da and probability higher than 95% were retained for potential oxidation modifications.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## 4.3. Label-Free Quantification
*In the **4.3. Label-Free Quantification** section:*
To quantitatively compare the protein abundances between ccRCC and adjacent tissues, spectral count was performed [37]. This method compares the number of identified MS/MS spectra from the same protein in each of the multiple data sets. For normalization, the spectral counts of each protein were divided by the total spectral counts of all proteins from the same sample. To estimate the levels of oxidation at each oxidized residue, the intensities of all peptide fragments containing the oxidized residue were measured [38] by Progenesis LC-MS software (version 4.1, Nonlinear Dynamics, Newcastle, UK). For quantification of ratio of the oxidized residues, the abundance of oxidized residue was divided by the abundance of unoxidized residue in every four pairs of tissue samples as shown in Table S7.
## 4.4. Pathway and Network Analyses of Dysregulated Proteins in ccRCC
*In the **4.4. Pathway and Network Analyses of Dysregulated Proteins in ccRCC** section:*
The dysregulated proteins were chosen based on the criteria of p < 0.05, presence in every replicate, average spectral counts of >20 and fold change ratio of 1.4 (up or down) in ccRCC comparing with adjacent tissues. For identifying enriched signaling networks and diseases categories, the dysregulated proteins were subjected to bioinformatics tools: STRING (available on: ) and DAVID functional annotation tool (available on: ) [39]. The STRING database version 10 and medium confidence (0.4000) were used and active interaction sources included all default settings, such as text mining, experiments, database, co-expression, neighborhood, gene fusion, co-occurrence. For DAVID functional annotation, DAVID database version 6.8 was used.
## 4.5. Western Blotting (WB) Analysis
*In the **4.5. Western Blotting (WB) Analysis** section:*
Samples were separated on SDS-PAGE gel. After electrophoresis, the proteins were transferred to nitrocellulose (NC) membranes, blocked then probed with primary antibodies against MNSOD (1:750, Santa Cruz, Dallas, TX, USA) and Tubulin (1:10,000, Sigma-Aldrich, St. Louis, MO, USA). Proteins were detected using fluorescence conjugated secondary antibody. The membranes were scanned with Odyssey infrared imaging system (Li-Cor, Lincoln, NE, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## 4.6. MNSOD Enzymatic Activity Assay
*In the **4.6. MNSOD Enzymatic Activity Assay** section:*
Commercial superoxide dismutase (SOD) activity kit (Jiancheng biotechnology, Nanjing, China) was used to measure MNSOD enzymatic activity from kidney tissue lysates according to manufacturer’s instructions and based on modified method as previously reported [16]. Total SOD activity was assayed by the inhibition of xantine/xantine oxidase mediated reduction of cytochrome c, and MNSOD activity was determined with addition of 3 mM KCN to inhibit Cu/ZnSOD.[](https://www.ncbi.nlm.nih.gov/mesh/D011190)
## 4.7. Data Analysis
*In the **4.7. Data Analysis** section:*
The SPSS 17.0 software (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. p < 0.05 was considered statistically significant. Statistical analyses between two groups were performed using a Student’s t-test, whereas comparisons involving multiple groups were performed with a two-way ANOVA test.
## 5. Conclusions
*In the **5. Conclusions** section:*
Our study suggests that MNSOD could be highly susceptible to oxidative modifications in ccRCC. The newly identified oxidative modification sites, particularly those in the C-terminus hydrophobic pocket, may be developed as biomarkers or molecular targets to study ROS regulatory mechanisms, ease the interpretation of previous MNSOD data in tumor biology, and enhance our intervening ability in tumorigenesis [18,19].[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
# Supplementary Materials
*In the **Supplementary Materials** section:*
Supplementary materials can be found at .
# Author Contributions
*In the **Author Contributions** section:*
Jiaju Lu and Jing-Hua Yang conceived and designed the experiments; Zuohui Zhao, Ruirui Jing and Cuiling Li performed the experiments; Fengqin Wang and Han-Pil Choi analyzed the data; Xin Lu contributed to sample collection; Zuohui Zhao, Jing-Hua Yang, Kazem M. Azadzoi and Han-Pil Choi wrote the manuscript and contributed to the revision of manuscript.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
Quantitative proteomic analysis revealed the importance of anti-oxidative stress pathway in ccRCC. (a) The heatmap showed the 33 oxidoreductases (by DAVID) were involved in binding cofactor, coenzyme, NAD, NADH or NAD(P). 1: NAD binding site; 2: NAD or NADH binding; 3: nucleotide phosphate-binding region; 4: NAD(P)-binding domain; 5: NAD; 6: coenzyme binding; 7: cofactor binding; 8: oxidation reduction; 9: oxidoreductase. Red arrow showed the candidate protein (MNSOD, SOD2). Green area: gene-term association positively reported, light blue area: gene term association not reported yet; (b) visualization of protein–protein interactions of the 37 oxidation-reduction related proteins in ccRCC using STRING analysis (confidence mode). 37 oxidation-reduction related proteins were input into STRING software and they formed three main clusters (only 33 connected proteins were shown and the clusters were divided by dotted lines), among which MNSOD (SOD2, red arrow) were participated in the network and were chosen to be validated later. The solid lines represented interactions between proteins and thickness of the solid lines denoted the confidence level associated with each interactions.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
LC-MS/MS coverage of MNSOD. (a) Whole cell lysates from kidney tissues were separated by SDS-PAGE and stained by Coomassie Blue (arrow indicated MNSOD). The gel image is the representative of 4 pairs of tumor and adjacent tissues used for PTMs analysis. A: adjacent; T: tumor; (b) CID-based sequence coverage of MNSOD. After Coomassie Blue staining, the 22 kDa protein bands corresponding to MNSOD were cut from the gel and digested, and peptides were analyzed by LC-MS/MS on LTQ-Orbitrap mass spectrometer (MS). The underlined amino acids (bold letters) were identified by Proteome Discoverer 1.4 (MASCOT and SEQUEST), which covered 76.13% sequence of MNSOD. Signal: the signal peptide; α: the α-helices; β: the β-sheets, subscript numbers represent original numbers; solid arrows: metal (Mn2+) binding sites.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
Quantification of the oxidized residues of MNSOD. (a) The average ratios of the oxidized to unoxidized residues in tumor + adjacent tissues. Abundances of the modified and unmodified peptide fragments containing the nine oxidized and unoxidized residues (H54, H55, Y58, W147, W149, W205, N206, N209 and W210) were used to calculate the ratios (Table S7). Three hotspots are indicated; (b) comparison of the oxidation modification between tumor and adjacent tissues. The relative ratios of oxidized to unoxidized residues at histidine 54 (H54), 55 (H55) and tyrosine 58 (Y58) were low both in ccRCC tumor tissues and adjacent tissues but, dramatically increased in tumor comparing with adjacent tissues (the left panel, * p < 0.03, n = 4) (Table S7). The relative ratios at tryptophan 147 (W147) and 149 (W149) were also increased significantly in tumor, but the relative ratios at tryptophan 205 (W205), asparagine 206 (N206), asparagine 209 (N209) and tryptophan 210 (W210) were decreased slightly in tumor (the right panel, * p < 0.03, n = 4) (Table S7).[](https://www.ncbi.nlm.nih.gov/mesh/D006639)
MNSOD expression, enzymatic activity and the potential oxidation hotspots of MNSOD. (a) The expression of MNSOD in adjacent (A) and ccRCC (T) tissues. Four individual ccRCC and adjacent tissue lysates were separated by SDS-PAGE and the expression of MNSOD was detected by Western blotting. Tubulin served as the loading control (* p < 0.05, n = 4); (b) MNSOD enzymatic activity in adjacent (A) and ccRCC (T) tissues. Superoxide dismutase (SOD) activity kit was used to measure MNSOD enzymatic activity. The relative MNSOD enzymatic activity was normalized by MNSOD expression (* p < 0.05, n = 4); (c) the potential oxidation hotspots in three dimensional structure of MNSOD. The ribbon structure of MNSOD subunit (1LUV.pdb) is depicted with the two separate polypeptide chains. The active sites (manganese ions, Mn) are depicted as solid spheres and the oxidative modification sites (H54, H55, Y58, W147, W149, W205, N206, N209, W210) are indicated as red dashed arrows. The 3D structure shows the three oxidation hotspots (hotspot 1 (H54, H55, Y58), hotspot 2 (W147, W149) and hotspot 3 (W205, N206, N209, W210)) are distributed in the surface of homotetrameric MNSOD and point to protein internal, which composite a hydrophobic side chain and affect the molecular structure of MNSOD. The colors in the ribbon structure of MNSOD subunit are the automatically displayed colors according to 1LUV.pdb.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
Frequency of +16 modification events of MNSOD at different amino acid residues *.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
* Note: total counts of the peptide with a delta mass of +16 Da at the indicated amino acids.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
The modified (oxidized and nitrated) amino acid residues in MNSOD a.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
a The modified peptides in MNSOD (P04179, SODM_HUMAN), which were identified by MASCOT and SEQUEST. See Figure S1 for mass spectra containing b and y ions. b Precursor mass accuracy <10 ppm. c Modified peptide sequences. ox: oxidation, no: nitration. The peptide sequences with bold letters indicate novel oxidized/nitrated residues. d Ions Score (>38) by MASCOT, XCorr (>1.22 × charges) by SEQUEST. e Asn oxidation combined with Trp oxidation.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
|
# Introduction
Neuropeptide S (NPS) variants modify the signaling and risk effects of NPS Receptor 1 (NPSR1) variants in asthma
# Abstract
*In the **Abstract** section:*
Single nucleotide polymorphisms (SNPs) close to the gain-of-function substitution, Asn(107)Ile (rs324981, A>T), in Neuro[peptide S ](https://www.ncbi.nlm.nih.gov/mesh/D009711)Receptor 1 (NPSR1) have been associated with asthma. Furthermore, [a f](https://www.ncbi.nlm.nih.gov/mesh/D001216)uncti[ona](https://www.ncbi.nlm.nih.gov/mesh/D007532)l SNP (rs4751440, G>C) in Neuropeptide S (NPS) encodes a Val(6)Leu substitution on the mature peptide that results in reduced bioactivity. We sought to examine the effect[s o](https://www.ncbi.nlm.nih.gov/mesh/D014633)f d[iff](https://www.ncbi.nlm.nih.gov/mesh/D007930)erent combinations of these NPS and NPSR1 variants on downstream signaling and genetic risk of asthma. In transfected cells, the magnitude of NPSR1-induced activation of cAMP/PKA signal transduction pathways and downstream gene expression was dependent on the combination of the NPS [and ](https://www.ncbi.nlm.nih.gov/mesh/D000242)NPSR1 variants with NPS-Val(6)/NPSR1-Ile(107) resulting in strongest and NPS-Leu(6)/NPSR1-Asn(107) in weakest effects, respectively. One or two copies of the NPS-Leu(6) (rs4751440) were associated with physician-diagnosed childhood asthma (OR: 0.67, 95%CI 0.49–0.92, p = 0.01) and together with two other linked NPS variants (rs1931704 and rs10830123) formed a protective haplotype (p = 0.008) in the Swedish birth cohort BAMSE (2033 children). NPS rs10830123 showed epistasis with NPSR1 rs324981 encoding Asn(107)Ile (p = 0.009) in BAMSE and with the linked NPSR1 rs17199659 (p = 0.005) in the German MAGIC/ISAAC II co[hor](https://www.ncbi.nlm.nih.gov/mesh/D001216)t (14[54 ](https://www.ncbi.nlm.nih.gov/mesh/D007532)children). In conclusion, NPS variants modify asthma risk and should be considered in genetic association studies of NPSR1 with asthma and other complex diseases.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Neuropeptide S (NPS) affects multiple neuroendocrine, behavioral, and inflammatory responses via its G protein-coupled cell surface receptor NPSR1 (Neuropeptide S Receptor 1) [1–4]. NPSR1 was identified as a susceptibility gene for asthma and related traits by positional cloning and the associations of NPSR1 single nucleotide polymorphisms (SNPs) with asthma have been replicated in ethnically diverse populations [5–12], and marginally supported by a large-scale genome-wide association study (GWAS) [13]. In addition, NPSR1 SNPs have shown genetic associations with other inflammatory phenotypes such as inflammatory bowel disease [14] and rheumatoid arthritis
[15, 16]. However, the NPSR1 locus has shown allelic heterogeneity with different markers (tag SNPs, intronic markers, and haplotypes) showing associations depending on the study design and population being studied. We have previously shown that several susceptibility alleles of low-to-moderate-effects in NPSR1 may modify the asthma risk and show epistasis depending on the carrier status for variants in genes belonging to common biological pathways [17]. These effects may not be uncovered by SNP arrays which are suitable for detection of common polymorphisms but cannot detect the effects of less frequent coding mutations and low frequency functional SNPs.[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
The functional NPSR1 SNP rs324981 (A>T), encoding a substitution of Asn(107)Ile in the putative ligand-binding pocket of NPSR1 [5], has shown associations with neuropsychiatric phenotypes, such as panic disorders [18–20], psychological stress [21], and fear responses [22, 23]. A GWAS on circadian sleep parameters found an association between rs324981 and regular bedtime [24] and another study showed that the same SNP was associated with sleep and rest duration [25]. In cell models, the change of Asn(107) to Ile(107) results in 10-fold increase in NPS-mediated intracellular signaling [26] and changes in genome-wide transcriptional profiles [27].
In addition, the NPSR1 SNP rs324981 has been associated with airway hyperresponsiveness to methacholine in a Chinese population [8]. Non-coding functional SNPs have also been detected in NPSR1, for instance rs2530547 which affects luciferase expression in gene reporter assays and NPSR1 mRNA levels in human leukocytes [27]. By changing the levels of expression in this receptor, they may ultimately also affect signaling, and are hypothesized to affect neuroinflammatory phenotypes [28]. Thereby associations with individual NPSR1 SNPs needs to be evaluated in the context of gene-gene interactions because a combination of functional polymorphisms may ultimately determine receptor properties and/or expression levels
[27].[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
A functional SNP rs4751440 in NPS encodes a Val(6)Leu substitution in the mature NPS peptide that results in lower bioactivity [29]. We hypothesized that the interaction of functional NPS and NPSR1 SNPs might lead to either strong or minimal downstream signaling depending on the genotype combination on both loci. The aims of this study were: 1) To clarify the potential functional crosstalk between NPS and NPSR1 variants; 2) to identify the impact of putative NPS risk alleles on the susceptibility to asthma; and 3) to evaluate potential gene-gene interaction effects (epistasis) between NPS and NPSR1 polymorphisms on asthma.[](https://www.ncbi.nlm.nih.gov/mesh/D007930)
## Results
*In the **Results** section:*
## Biological interaction of NPS and NPSR1 variants can lead to either strong or minimal downstream signaling
*In the **Biological interaction of NPS and NPSR1 variants can lead to either strong or minimal downstream signaling** section:*
NPS is known to activate NPSR1 by increasing cAMP production [30, 31]. To compare the bioactivities of the two NPS variants, we measured the dose-responses of NPSR1-Ile(107) and NPSR1-Asn(107) coding variants to either wildtype NPS-Val(6) or alternative NPS-Leu(6) in a luciferase assay dependent on the activation of a cAMP-response element (CRE). This assay was used since elevation of the intracellular cAMP levels activates cAMP response element binding protein (CREB) via phosphorylation of protein kinase A (PKA). Activated CREB binds to CRE and induces the expression of downstream target genes of the cAMP/PKA signaling pathway.
Because of the low endogenous expression levels of NPSR1 in majority of the available cell lines [26, 29], we first used HEK293 cells with forced NPSR1 expression as a model for ligand-receptor interactions. This approach allowed comparisons between the present and previous studies [26, 29]. HEK293 cells were transiently co-transfected with the two NPSR1 variants along with a CRE-responsive firefly luciferase construct and a construct constitutively expressing Renilla luciferase. Firefly and Renilla luciferase activity was measured 3 h after stimulation with serial dilutions of NPS (10 pM-100 μM). As shown in Fig 1, wildtype NPS—Val(6) peptide stimulated CRE—dependent luciferase activity at 50—fold lower concentrations than the NPS—Leu(6) peptide in NPSR1—Ile(107) cells (pEC 50: 8.4 ± 0.2 and 6.7 ± 0.2, p = 0.004, respectively), and at 15-fold lower concentrations in NPSR1—Asn(107) cells (pEC 50: 6.1 ± 0.3 and 4.9 ± 0.1, p = 0.03, respectively).[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
NPS variants and CRE luciferase activity.
Dose-response curves of NPS-Val(6) and NPS-Leu(6) (10 pM– 100 μM) based on the cAMP response element (CRE) luciferase activity in human embryonic kidney epithelial (HEK293) cells transiently transfected with either NPSR1-Ile(107) or NPSR1-Asn(107) coding variants. One representative experiment is shown. The CRE-driven luciferase activity is expressed in mean arbitrary units ± SEM of triplicates 3 h after NPS stimulation.[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
The maximum of luciferase activity was ~2-fold in cells transfected with the NPSR1-Ile(107) variant in comparison to the cells transfected with the NPSR1-Asn(107). Thus, the magnitude of NPSR1-induced activation of CRE was markedly different depending on the combination of the NPS and the NPSR1 variants resulting in either strong (NPS-Val(6) and NPSR1-Ile(107)) or minimal (NPS-Leu(6) and NPSR1-Asn(107)) downstream signal transduction.
The activation of CRE results in transcriptional changes of downstream target genes. To examine whether the observed differences in the promoter activity also affected gene expression, we measured the effects of NPS variants on known NPSR1 downstream target genes [27, 32, 33]. NR4A1 (nuclear receptor subfamily 4, group A, member 1), FOS (FBJ murine osteosarcoma viral oncogene homolog), and chemokine ligands CXCL2 and CCL20 were selected as they were among the most influenced genes after NPS stimulation and their expression also differed between transcriptome comparisons of HEK293 cells expressing either NPSR1-Ile(107) or NPSR1-Asn(107) [27].
In the present study, HEK293 cells transiently transfected with either NPSR1-Ile(107) or NPSR1-Asn(107) coding variants were stimulated with 100 nM of either NPS-Val(6) or NPS-Leu(6) for 1, 3, 6, and 24 h, and changes in gene expression were analyzed with RT-PCR. In agreement with the results from the luciferase assay, NPS induced stronger downstream signaling in cells expressing NPSR1-Ile(107) than in cells expressing NPSR1-Asn(107), and wildtype NPS-Val(6) peptide was more potent than the NPS-Leu(6) peptide. mRNA expression of NR4A1, FOS, and CXCL2 was induced 1 h after NPS stimulation whereas the expression of CCL20 was up-regulated at 3 h (Fig 2A).
Real time RT-PCR analysis of NPSR1 downstream target genes.
(A) Time-dependent mRNA expression of NR4A1, FOS, CXCL2, and CCL20 in human embryonic kidney epithelial (HEK293) cells transiently transfected with either NPSR1-Ile(107) or NPSR1-Asn(107) coding variants and stimulated with either NPS-Val(6) or NPS-Leu(6) NPS (100 nM) for 1, 3, 6, and 24 h. (B) Dose-response curves of NR4A1, FOS, IER3, and EGR1 in human SH-SY5Y stable cell line over-expressing NPSR1-Ile(107) stimulated with either NPS-Leu(6) or NPS V6 (0.0001–1 μM) for 3 h. The results are presented as fold-changes in comparison to the unstimulated cells. GAPDH was used as the endogenous reference, and data are expressed as mean of triplicate samples. The error bars represent 95% confidence intervals. In all experiments, results were calculated with the comparative Ct method.
Altered neuroinflammatory mechanisms have been implicated in allergic airway inflammation [34]. Because NPSR1 is expressed in neuroendocrine tissues [35–37], we used SH-SY5Y neuroblastoma cells of neuroendocrine origin to validate the changes in gene expression observed in HEK293 cells. SH-SY5Y cells stably overexpressing NPSR1-Ile(107) [17, 36] were stimulated with either NPS-Leu(6) or NPS-Val(6) (0.0001–1 μM) for 3 h and the mRNA expression of known downstream target genes in these cells [36] was detected using RT-PCR. As expected, NPS stimulation increased the expression of NR4A1, FOS, EGR1 (early growth response protein 1), and IER3 (immediate early response 3) in SH-SY5Y cells over-expressing NPSR1-Ile(107) in a dose dependent manner (Fig 2B).
Wildtype NPS-Val(6) peptide was more potent than the NPS-Leu(6) peptide. These observations validated the results detected in HEK293 cells in an independent cell line.
## NPS rs4751440 encoding Leu(6) is protective for childhood asthma
*In the **NPS rs4751440 encoding Leu(6) is protective for childhood asthma** section:*
Because the SNP rs4751440 (G/C) encoding for NPS-Val(6) and NPS-Leu(6) is common in European and American mixed populations [29], and given its functional effects on NPSR1 signaling in cell assays, we hypothesized that this SNP might influence the genetic risk of diverse disease traits including asthma. The rs4751440 Val(6)Leu and two other NPS SNPs (rs1931704 and rs10830123) were analyzed for associations with physician-diagnosed childhood asthma in the prospective Swedish birth-cohort BAMSE (n = 2033) [38, 39]. The SNPs spanned a region of 11 kb on chromosome 10q26.2 and were in linkage disequilibrium (LD) with each other (Fig 3A). The allele frequencies of the three NPS SNPs in cases and controls and the results of the association tests with asthma are presented in S1 Table.
Two NPS SNPs (rs1931704 and rs4751440) in strong linkage disequilibrium (D’ = 0.99, r2 = 0.69) were significantly associated with physician-diagnosed childhood asthma at age 8 years under dominant (p = 0.01) and additive models (p = 0.03). For both NPS SNPs the effect was driven by a higher frequency of the minor allele in the control group (Fig 3B). These results revealed that the allele C in NPS rs4751440 encoding for Leu(6) is protective for asthma. The association between NPS variants and asthma was also significant when comparing the haplotype frequencies between asthmatic children and controls with a protective effect driven by the haplotype AGC (p = 0.008) (Table 1).
There were no significant associations between NPS SNPs and atopic sensitization (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D014633)
Location and block structure of three genotyped SNPs in the NPS gene.
(A) The scale represents the relative positions of rs1931704 (G>A) in the upstream regulatory region of NPS, followed by rs10830123 (G>C) in the intronic region and rs4751440 (G>C) in the coding region. Allele changes are indicated by > and given on the positive chromosomal strand of the Human Reference Assembly. The arrow indicates NPS transcription from the positive strand. (B) Effects of NPS SNPs on the risk of physician-diagnosed asthma in carriers of either one or two copies of the alternative allele. Bars represent 95% confidence intervals (CI). OR: Odds Ratio. Given the allele and genotype frequencies, the dominant model which best fits the data is shown.
Haplotype association of NPS polymorphisms with asthma at 8 years.
Global haplotype association p-value = 0.0077
Allele changes are indicated by > and given on the positive chromosomal strand of the Human Reference Assembly.
## Gene-gene interactions between NPS and NPSR1 variants
*In the **Gene-gene interactions between NPS and NPSR1 variants** section:*
Based on the effects of the NPS SNP rs4751440 on genetic asthma risk and NPSR1 signaling, we first tested genetic association of 29 NPSR1 SNPs with physician-diagnosed childhood asthma in BAMSE (n = 2033), and then assessed epistasis between NPS and NPSR1. We detected significant association between three NPSR1 SNPs (rs887020, rs2531840 and rs11770777) with asthma (p<0.05) and confirmed that the functional rs324981 (T>A) encoding for NPSR1-Ile(107) and NPSR1-Asn(107) had no significant main effects at present resolution (S2 Table). The next aim was to investigate if the effects of NPS SNPs (rs1931704, rs10830123 and rs4751440) on the risk of asthma were modified by the carrier status of alleles in the SNP rs324981 (A>T) encoding for NPSR1-Asn(107) and NPSR1-Ile(107), respectively, or by other variants with previous evidence of putative functional effects in the NPSR1 gene. We detected significant interaction between NPS SNP rs10830123 and NPSR1 rs324981 (interaction p value = 0.009). There were also significant interactions between NPS SNP rs10830123 with two additional NPSR1 SNPs (rs323922 and rs324384) in linkage disequilibrium with NPSR1 rs324981 (Fig 4).
A second region with significant interactions between NPS and NPSR1 was detected between two SNPs in the NPSR1 promoter (rs2125404 and rs2168890) and NPS SNP rs10830123 (Fig 4) and S3 Table.
Gene-gene interactions between NPSR1 and NPS.
Location of six NPSR1 SNPs with significant interactions with NPS rs10830123 in the multiplicative model for epistasis in connection with their p-value for interaction and linkage disequilibrium. The most significant interaction signals in BAMSE (green) and MAGIC/ISAAC (orange) are marked in bold. Statistics on LD between the functional rs324981 and other interacting SNPs are provided as D’ and r2.
We then explored the gene-gene interaction between NPS and NPSR1 SNPs in the MAGIC/ISAAC cohort (n = 1454). NPS and NPSR1 SNPs did not show any significant main effects on asthma risk in this dataset (data not shown). However, there was a significant interaction between NPS rs10830123 and NPSR1 rs17199659 (p = 0.005) (Fig 4 and S3 Table). NPSR1 rs17199659 is in LD (D’ = 0.95, r2 = 0.19) with NPSR1 rs324981 encoding for NPSR1-Asn(107) and NPSR1-Ile(107) and with rs324384, both showing significant gene-gene interactions in BAMSE (Fig 4). These results suggested that the epistatic signal with NPS was driven by genetic variants in the proximities of exon 3 and exon 4 in NPSR1.
## Discussion
*In the **Discussion** section:*
Previous studies have shown that some polymorphic variants in the genes encoding NPS and NPSR1 have functional effects in modifying peptide bioactivity or receptor signaling efficacy [26, 29]. However, no study has yet analyzed the functional outcomes of the different combinations of coding NPS and NPSR1 variants upon ligand/receptor interactions. This is of critical relevance, because coding SNPs such as NPS rs4751440 or NPSR1 rs324981 are commonly observed in humans, and cells in any given individual will carry one out of nine alternatives of genotype combinations at these two loci. We examined the effects of the peptides NPS-Val(6) and NPS-Leu(6) on the signal transduction of NPSR1, expressed as two variants with different signaling efficacy (NPSR1-Ile(107) and NPSR1-Asn(107)).
Our results clearly showed that the magnitude of NPSR1 signaling (as determined by reporter gene assays) and expression levels of downstream genes was dependent on the combination of NPS and NPSR1 variants, with the highest induction obtained upon stimulation of NPSR1-Ile(107) overexpressing cells with NPS-Val(6), and the lowest induction obtained with the combination of NPSR1-Asn(107) and NPS-Leu(6). Genetic association analyses further suggested epistasis between NPS and NPSR1 variants with the most significant interactions being centered to the region of NPSR1 rs324981 encoding for NPSR1-Ile(107) and NPSR1-Asn(107).
A key strength of this study is the use of robust in vitro cell assays that allowed us to dissect the functional effects of coding NPS/NPSR1 variants on receptor signaling and downstream gene expression. Our cell assays confirmed that stimulation of cells transfected with either NPSR1-Asn(107) or NPSR-Ile(107) with the wildtype NPS-Val(6), induced CRE activity as described previously [26, 27]. Reinscheid et al. [26] have previously shown that upon stimulation with NPS-Val(6), stable HEK293 cell lines expressing NPSR1-Asn(107) and CRE-luciferase reporter gene displayed an induction of luciferase activity with a mean EC50 of 45 nM, which was roughly 10-fold lower than in this study (428 nM). For NPSR1-Ile (107), the stimulation with NPS-Val(6) produced an increase in reporter gene expression with an EC50 of 0.6 nM, which was 10-fold lower than what we observed (4.8 nM)
[26]. The differences could be explained by the ∼20-fold lower magnitude of reporter gene induction in transiently transfected cells compared with stable clones [26]. In agreement with our study, the maximum of reporter gene expression in HEK293 cells transiently transfected with NPSR-Ile (107) was ∼2-fold higher than in cells expressing NPSR1-Asn(107), indicating an increase in agonist efficacy. Another previous study observed EC50 values of 37 nM and 586 nM for NPS-Val(6) and NPS-Leu(6), respectively, in CRE-luciferase activity 18 h after stimulation of HEK293 cells transiently transfected with NPSR1-Asn(107)
[29]. Overall, our results were in line with these previous observations, and methodological issues, such as differences in NPS incubation time and transfection efficiency, might well explain the differences observed.
Our genetic analyses in the birth cohort BAMSE uncovered the association between NPS rs4751440 and the presence of asthma by 8 years. Another NPS SNP (rs10830123) showed borderline association with asthma, and formed a protective haplotype with NPS rs1931704 and NPS rs4751440. The protective effect of the minor allele C in NPS rs10830123 became significant only in the presence of two copies of rs324981 (TT) encoding NPSR1-Ile(107). In the replication cohort, NPS rs10830123 showed interaction with NPSR1 rs17199659 which is in linkage disequilibrium with rs324981 and rs324384. Due to the LD, NPS rs10830123 can be used as a proxy for the effects of NPS rs4751440 (Fig 4).
Although the latter had significant main effects on asthma risk and proved to be functional in the cell assays, a lower frequency of informative GC heterozygotes in the patient group could have influenced the power to detect gene-gene interactions with NPSR1. Indeed, 24% of non-asthmatic children were heterozygous for NPS rs4751440 and there was a low frequency of CC carriers in the cohorts with only 2% of children being homozygous (CC) for NPS-Leu(6). This was in line with previous findings suggesting that 22% of Europeans are heterozygous for NPS rs4751440 [29]. Nevertheless, it is consistent in BAMSE and MAGIC/ISAAC that the most significant interaction was centered to the region of NPSR1 rs324981 encoding for NPSR1-Ile(107) and NPSR1-Asn(107).
Altogether, our results supported the role of the NPS/NPSR1 pathway in asthma and the notion of a “combinatory effect”, in which the phenotype is driven by the interaction of several variants of moderate effects that coincide and affect a biological pathway [17].
In BAMSE, the NPS rs10830123 also showed significant interactions with two SNPs located in the upstream region of NPSR1 (rs2125404 and rs2168890). These were in LD with NPSR1 rs2530547 previously found to affect NPSR1 mRNA levels in human leukocytes [27] and with three SNPs affecting NPSR1 mRNA expression in cerebellum (rs2530549, rs2530550, and rs2530566) according to the Genotype-Tissue Expression database (http://gtexportal.org/home/gene/NPSR1). Furthermore, a bioinformatic analysis on the transcription binding sites that are created or destroyed by the five NPSR1 SNPs implicated in the gene-gene interactions with NPS (S3 Table) revealed that the change of cytosine (C) to guanine (G) in NPSR1 rs323922 is predicted to affect the binding site of V$NBRE/NBRE.01 related to the monomers of the nur subfamily of nuclear receptors (NR4A1-3, also known as nur77, nurr1, nor-1, S4 Table).[](https://www.ncbi.nlm.nih.gov/mesh/D003596)
Previous transcriptome studies in HEK293 cells [32, 33] and SH-SY5Y cells [36] overexpressing NPSR1-Ile(107) revealed downstream target genes with NR4A1 and FOS being among the most affected transcription factors. In our cells assays, the magnitude of the mRNA expression of NR4A1 was clearly dependent on the combination of NPS and NPSR1 variants (Fig 2). NR4A1 is expressed in lymphocytes upon stimulation and is anti-inflammatory [40–42]. NR4A1 has been recently described as a key regulator of catecholamine production by macrophages linking sympathetic stress response and inflammation [43].
The hypothesis of a regulatory loop influenced by particular combinations of NPS/NPSR1 variants is further suggested by the up-regulation of the transcription factor FOS in cells expressing NPSR1-Ile(107) upon stimulation with NPS-Val(6) and the fact that the promoter region of NPSR1 is known to contain FOS binding motifs [28]. The upregulation of NR4A1and FOS induced by the combination of NPS-Val(6) and NPSR1-Ile(107) is anticipated not only to interact in a feedback loop with their regulatory elements in NPSR1 but also to regulate a number of other target genes including those involved in immune responses. For instance, the levels of serum pro-inflammatory cytokines could be activated by both central and peripheral administration of NPS in pigs [44], and NPS has shown direct effects on phagocytosis [45]
[46] and chemotaxis [47] [48]. Further insights to immune responses are suggested by the upregulation of CXCL2 and CCL20 encoding chemokine receptors ligands (Fig 2).[](https://www.ncbi.nlm.nih.gov/mesh/D002395)
The findings of significant epistatic signals implicating coding and non-coding variants allowed us to suggest a model in which the divergent biological outcomes driven by coding variants in NPS and NPSR1 do occur at the level of ligand/receptor interactions but also implicate the effects of polymorphic variants in non-coding regions and the cross-talk of regulatory elements in diverse loci (Fig 5). We must emphasize that the sample size analyzed in this study is underpowered to detect epistatic effects. However, the facts that the interaction signals in the two independent cohorts were narrowed to the same gene regions (Fig 4) and a recent meta-analysis on a larger dataset such as GABRIEL [49] confirmed the interaction signals that we originally identified in this population between NPSR1 and RORA [17], suggest that the interactions between NPS and NPSR1 may occur and are detectable at present resolution. Additional studies using larger sample sizes more suitable to test for epistasis are needed to confirm these findings.
A schematic model on the functional crosstalk between NPS and NPSR1.
(A) The non-synonymous NPS-Leu(6) substitution affects NPS bioactivity. Depending on the NPSR1-Asn(107) or NPSR1-Ile(107) genotype, binding of NPS can lead to strong or minimal signaling as well as differential expression of downstream target genes such as the transcription factors FOS and NR4A1. (B) Two hypothetical mechanisms in which the differential expression of FOS and NR4A1 can modulate the outcomes of the NPS/NPSR1 pathway. FOS and NR4A1 bind to their regulatory elements in NPSR1 creating a feed-back loop that affect cell surface expression of NPSR1 and availability, and also, affect a number of other downstream genes that modulate neuroinflammation and ultimately asthma risk.
In conclusion, our results supported the association of the NPS/NPSR1 pathway with asthma and provided a molecular framework on how the different NPS and NPSR1 variants interact in asthma. This study provides the basis for further investigations analyzing the interacting effects of NPS/NPSR1 variants in many other phenotypes and also support that polymorphic variants in NPS should be taken into account when analyzing the effects of NPSR1 SNPs. Such gene-gene interactions as observed here may dilute or dampen association signals in genetic association studies.
## Materials and methods
*In the **Materials and methods** section:*
## Ethics statement
*In the **Ethics statement** section:*
All methods were carried out in accordance with relevant guidelines and regulations and the study followed the ethical principles for medical research stated in the Declaration of Helsinki. All of the experimental protocols were approved by the Swedish Regional Ethics Committee at Karolinska Institutet, Stockholm, Sweden in the BAMSE cohort (Dnr: 01–478 and 02–420) and by the Ethic Committee of the Bavarian Medical Council for MAGIC/ISAAC (approval nr. 01218). Written informed consent was obtained from parents/guardians on behalf of all child participants.
## Cell culture and NPS stimulations
*In the **Cell culture and NPS stimulations** section:*
Human embryonic kidney cells (HEK293) were cultured in MEM+GlutaMAX-1 medium (Gibco/Invitrogen) supplemented with 10% fetal calf serum (FCS) (Biosera), 1% sodium pyruvate, 1% non-essential amino acids and 1% penicillin/streptomycin (Gibco/Invitrogen). Cells were kept at 37°C in a humidified 5% CO2 incubator. The cells were transiently transfected (Lipofectamin™2000 Reagent, Invitrogen, Carlsbad, USA) with equal amounts of pCMV-NPSR1-Ile(107) or NPSR1-Asn(107) [27] in a ratio of 1:2 (DNA:Lipofectamin), and stimulated with NPS 24 h after transfection. The cells were stimulated with either 100 nM NPS-Val(6) (SFRNGVGTGMKKTSFQRAKS) or 100 nM NPS-Leu(6) (SFRNGLGTGMKKTSFQRAKS) (both from Proteogenix, France) for 1, 3, 6, and 24 h. The HEK293 cell line was chosen for ligand-receptor studies because it is a well-established model for addressing numerous questions in basic biology and is easy to transfect and amenable of stringent quantitative assessment.[](https://www.ncbi.nlm.nih.gov/mesh/D011773)
The stable NPSR1-A-GFP expressing SH-SY5Y cells were engineered and cultured as described before [17]. For gene expression studies, the SH-SY5Y cells over-expressing NPSR1 were seeded at 0.5×106 cells/ml to 6-well plates. After 24 h, fresh cell media was added, and the cells were incubated with a dilution series of either NPS-Val(6) or NPS-Leu(6) (0.0001–10 μM) for 3 h. The changes in gene expression were analyzed with real-time PCR (RT-PCR).
## Luciferase assays
*In the **Luciferase assays** section:*
For pharmacokinetic studies, HEK293 cells in 24-well plates were co-transfected with 700 ng of NPSR1 cDNA expression vectors specific for different coding variants and 50 ng each of cAMP response element (CRE) Firefly Luciferase reporter vector and CMV Renilla Luciferase reporter plasmid (purchased as pre-formulated mix from SABiosciences, Frederick, MD, USA). The day after transfection the cells were stimulated for 3 h with NPS at various concentrations (range 10 pM– 100 μM) before preparation of total cell lysates. In all reporter assays, Luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions, in a Microplate Reader Infinite 200 (Tecan, Männedorf, CH).
Experiments were performed three times in triplicates, and Firefly Luciferase activity was expressed in arbitrary units relative to the cells transfected with the vector without NPS stimulation, after normalization for transfection efficiency based on values obtained for Renilla Luciferase. The concentration-response curve values for pEC50, and Emax were calculated using nonlinear regression analysis, and the data are presented as means ± SE. Statistical analysis was performed using one-way ANOVA followed by unpaired t-test for comparison between two groups. For the analyses, GraphPad Prism 5.01 (GraphPad Software, San Diego, CA) was used.[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
## Real-time RT-PCR
*In the **Real-time RT-PCR** section:*
Total cellular RNA was isolated with the RNAeasy Mini Kit (Qiagen, Hilden, Germany). Reverse transcription was performed with TaqMan reverse transcription reagents (Applied Biosystems, Rotkreuz, Switzerland) using random hexamer primers according to the manufacturer's protocol.
The mRNA expression was measured with qRT-PCR using either SYBR Green (for NR4A1, FOS, IER3, and EGR1, the primer sequences are given in S5 Table) or TaqMAN assay (Hs00601975 for CXCL2 Hs00601975, Hs01011368 for CCL20, Hs01036497 for NPSR1 and Pre-Developed TaqMAN Assay Reagents, Part No. 4310884 for GAPDH; Applied Biosystems).
The PCR assays were performed in a total volume of 10 μl, using 7500 Fast Real-Time PCR system (Applied Biosystems) with the following reaction conditions: 50°C for 2 min and 94°C for 10 min; followed by 45 cycles of 92°C for 14 s and 1 min at 60°C. A dissociation stage was added to the SYBR Green reactions to confirm primer specificity. Results are shown as relative expression compared with unstimulated cells and GAPDH (Applied Biosystems) was used as an endogenous control. Samples were normalized to GAPDH (in SH-SY5Y cells) or to GAPDH and NPSR1 in order to control for the efficiency of the transient transfections (in HEK293 cells). The variation in NPSR1 expression (Ct values) between cells transfected with the NPSR1-Asn(107) and NPSR1-Ile(107) was within one cycle.
Expression levels of mRNA in each sample were determined by the comparative CT method of relative quantification, and expressed in fold-changes relative to the chosen reference sample.
## Study populations
*In the **Study populations** section:*
Genetic association analyses were conducted in the prospective Swedish birth-cohort BAMSE (Children Allergy Milieu Stockholm an Epidemiological study), which is an unselected, population-based Swedish birth cohort originally designed to assess risk factors for allergic diseases in childhood. This prospective study of 4,089 children has been described in detail elsewhere [38, 39]. At age 8, all children were invited to clinical testing and blood samples were obtained. After exclusion of samples with too little blood, incomplete questionnaire data, or lack of parental consent to genetic analysis there were 2033 samples with DNA extracted for genetic studies (of which 262 were asthmatics). All samples were analyzed coded.
The study was approved by the Regional Ethical Committee of Karolinska Institutet, Stockholm, Sweden. Written informed consent was obtained from the parents and/or legal guardians.
Replication studies were carried out in 1454 children from the Multicenter Asthma Genetic in Childhood (MAGIC) study (mean age 11 y) and the cross sectional International Study of Asthma and Allergies in Childhood phase II (ISAAC II) study (age 9–11 y) [50–52]. From both populations, all samples which had been part of the first GWAS on asthma [50] and for which we had imputation data available were selected for this study. After quality control (QC), 1,454 samples (763 asthmatics and 683 controls) remained for analysis. All study methods were approved by the local ethics committees and written informed consent was obtained from all parents of children included in this study.
To control for population stratification, all children were of German descent. Considering that previous studies have detected only marginal genetic differences between Swedes and Germans, and that GWAS data supported that within Europe, Germans and Swedes are most closely related [53], it can be assumed that the Swedish BAMSE and the German ISAAC/MAGICS were comparable.
## Phenotype assessment
*In the **Phenotype assessment** section:*
Information on the clinical outcomes was obtained from questionnaires filled out by the parents, except for atopic sensitization which was assessed from plasma samples. In BAMSE, Asthma ever up to 8 years was defined as a physician-diagnosed asthma between 3 mo after birth and up to 8 y of age [54]. Atopic sensitization was considered present if a child had levels of allergen-specific IgE >0.35 kU/L to Phadiatop, a mixture of cat, dog, horse, birch, timothy, mugwort, Dermatophagoides pteronyssinus, and Cladosporium allergens at 8 y of age (ImmunoCAP, Thermo Fisher Scientific, Phadia AB, Uppsala, Sweden).
In MAGIC, the assessed variables included asthma and allergy, which were diagnosed by a pediatric pulmonologist according to clinical guidelines and objective measures such as lung function tests, clinical examination, and allergy testing [51, 55]. In ISAAC II, children were classified as having asthma if parents reported a physician’s diagnosis of asthma at least once, or of spastic or asthmatic bronchitis more than once in self-administered questionnaires [56, 57]. The populations of the Swedish BAMSE and the German ISAAC/MAGICS are assumed to be comparable because in both, the definition of the phenotype”asthma” was based on a physician-diagnosis and the age range of the children was comparable (8 years in BAMSE vs. 10 years in ISAAC/MAGICS) [58].
## SNP selection, genotyping and quality control
*In the **SNP selection, genotyping and quality control** section:*
In BAMSE the NPS and NPSR1 polymorphisms were genotyped by using the iPLEX chemistry on the SEQUENOM platform at the Mutation Analysis Facility (Karolinska Institutet), with the exception of rs4751440 which was genotyped by a TaqMan SNP Genotyping Assay in an ABI Prism 7500 Fast Sequence Detection System (Applied Biosystems) according to manufacturer's instructions. Primers for multiplex PCR and extension reactions were designed by the SpectroDesigner software (Sequenom GmbH, San Diego, CA, USA) and are available on request. Each assay was validated using 24 unrelated Caucasians and 3 CEPH DNA samples as well as 14 trios from the CEU population. Success rate for genotyping was 98.8%. In the MAGIC/ISAAC study, NPS and NPSR1 SNPs were genotyped by the Illumina Sentrix HumanHap300 BeadChip as previously described [50].
The NPS gene spans a region of 3.3 kb on chromosome 10q26.2 whereas the NPSR1 gene spans over ~220 kb of genomic DNA on chromosome 7p14. The SNPs included in this study were selected based on their effects on protein function (e.g. variants encoding for NPS-Val(6)Leu and NPSR1-Asn(107)Ile); previous reports of association with asthma in children from BAMSE [6, 17] and other populations [5–12]; and by their functional [27] and tagging properties [17, 19, 28]. We analyzed 3 NPS SNPs (rs1931704, rs10830123 and rs4751440 Val(6)Leu), 29 NPSR1 SNPs in the BAMSE cohort [17, 19], and 23 NPSR1 SNPs in MAGIC/ISAAC
[50]. The discrepancy in the number of NPSR1 SNPs was due to the fact that some SNPs genotyped in BAMSE were not present in the array. Genotype frequencies for all SNPs agreed with the expectations under Hardy-Weinberg equilibrium.
## Genetic association tests
*In the **Genetic association tests** section:*
Genetic association tests were conducted using Plink (version 1.07, http://pngu.mgh.harvard.edu/purcell/plink) using the commands—assoc and—model. The comparison of allele frequencies between cases and controls were done by χ2 and Fisher's exact test. Full model associations were evaluated using the Cochran-Armitage trend test, and genotypic, dominant and recessive models. The genotypic and dominant/recessive tests were only conducted if there was a minimum number of five observations per cell either in the 2-by-3 or 2-by-2 tables. Since we tested 32 individual SNPs it was expected at least 1.6 SNPs to be significant (p < 0.05) due to chance.
Epistasis was evaluated using the—epistasis command in PLINK. All pairwise SNP combinations were tested in the case-control datasets. Allelic by allelic epistasis between NPS and NPSR1 SNPs were calculated by logistic regression. The allele dosage of each SNP (A and B) was modeled in the form of Y ~ β0 + β1.A + β2.B + β3.AB + e and the test for interaction based on the coefficient β3. In the equation, Y represents the presence of asthma as a function of the intercept (β0) and the coeffients of the different alleles, and e the residual error.
## Supporting information
*In the **Supporting information** section:*
# References
*In the **References** section:*
|
# Introduction
HSP90 inhibitors potentiate [PGF2α](https://www.ncbi.nlm.nih.gov/mesh/D015237)-induced IL-6 synthesis via p38 MAP kinase in osteoblasts
# Abstract
*In the **Abstract** section:*
Heat shock protein 90 (HSP90) that is ubiquitously expressed in various tissues, is recognized to be a major molecular chaperone. We have previously reported that prostaglandin F2α (PGF2α), a potent bone remodeling mediator, stimulates the synthesis of i[nterleukin-6 (IL-](https://www.ncbi.nlm.nih.gov/mesh/D015237)6)[ thro](https://www.ncbi.nlm.nih.gov/mesh/D015237)ugh p44/p42 mitogen-activated protein (MAP) kinase and p38 MAP kinase in osteoblast-like MC3T3-E1 cells, and that Rho-kinase acts at a point upstream of p38 MAP kinase. In the present study, we investigated the involvement of HSP90 in the PGF2α-stimulated IL-6 synthesis and the underlying mechanism in MC3T3-E1 cells. Geldanamyci[n, an](https://www.ncbi.nlm.nih.gov/mesh/D015237) inhibitor of HSP90, significantly amplified both the PGF2α-stimulated IL-6[ release and](https://www.ncbi.nlm.nih.gov/mesh/C001277) the mRNA expression levels. In addition, other HSP90 inhi[bitor](https://www.ncbi.nlm.nih.gov/mesh/D015237)s, 17-allylamino-17demethoxy-geldanamycin (17-AAG) and 17-dimethylamino-ethylamino-17-demethox[y-geldanamycin (17-DMAG) and onalespib](https://www.ncbi.nlm.nih.gov/mesh/C112765), [enhanc](https://www.ncbi.nlm.nih.gov/mesh/C112765)ed the[ PGF2α-stimulated IL-6 release. Geldanamycin, 17-AAG ](https://www.ncbi.nlm.nih.gov/mesh/C448659)an[d onale](https://www.ncbi.nlm.nih.gov/mesh/C448659)spib m[arkedly s](https://www.ncbi.nlm.nih.gov/mesh/C552103)trengthened the[ PGF2](https://www.ncbi.nlm.nih.gov/mesh/D015237)α-induced phosphorylation [of p38 MAP k](https://www.ncbi.nlm.nih.gov/mesh/C001277)in[ase. G](https://www.ncbi.nlm.nih.gov/mesh/C112765)eldan[amycin an](https://www.ncbi.nlm.nih.gov/mesh/C552103)d 17-AAG did not affect the PGF2α-induced phosphorylation of p44/p42 MAP kin[ase and myos](https://www.ncbi.nlm.nih.gov/mesh/C001277)in ph[osphat](https://www.ncbi.nlm.nih.gov/mesh/C112765)ase targeting subuni[t (MY](https://www.ncbi.nlm.nih.gov/mesh/D015237)PT-1), a substrate of Rho-kinase, and the protein levels of RhoA and Rho-kinase. In addition, HSP90-siRNA enhanced the PGF2α-induced phosphorylation of p38 MAP kinase. Furthermore, SB203580, an inhibitor of p38[ MAP ](https://www.ncbi.nlm.nih.gov/mesh/D015237)kinase, significantly suppressed the amplification by gel[danamyci](https://www.ncbi.nlm.nih.gov/mesh/C093642)n, 17-AAG or 17-DMAG of the PGF2α-stimulated IL-6 release. Our results strongly [suggest that](https://www.ncbi.nlm.nih.gov/mesh/C001277) H[SP90 n](https://www.ncbi.nlm.nih.gov/mesh/C112765)egat[ively r](https://www.ncbi.nlm.nih.gov/mesh/C448659)egulates[ the ](https://www.ncbi.nlm.nih.gov/mesh/D015237)PGF2α-stimulated IL-6 synthesis in osteoblasts, and that the effect of HSP90 is exerted thr[ough ](https://www.ncbi.nlm.nih.gov/mesh/D015237)regulating p38 MAP kinase activation.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Heat shock proteins (HSPs) are induced in response to biological stress such as heat stress and chemical stress [1]. HSPs, which are generally recognized as molecular chaperones, facilitate the refolding of nonnative proteins, or assist in their elimination via the chaperone-mediated autophagy or the ubiquitin proteasome system. HSPs have recently been classified into seven families, named HSPH (HSP110), HSPC (HSP90), HSPA (HSP70), HSPD/E (HSP60/HSP10), CCT (TRiC), DNAJ (HSP40) and HSPB (small HSP) [1,2]. Among them, HSP90 (HSPC) abundantly express in a variety type of unstressed cells and represents 1–2% of total cellular proteins, which increases to 4–6% under the stress conditions [2].
HSP90 consists of three domains, such as N-terminal domains, middle domains and C-terminal domains, and acts as an ATP-dependent chaperone [3]. It has been shown that HSP90 is overexpressed in many types of cancers, and that HSP90-dependent client proteins are involved in a variety of oncogenic pathways [4,5]. Therefore, inhibition of HSP90 functions has become as one of the leading strategies for anticancer chemotherapeutics
[4,5]. In our previous study [6], we have demonstrated that HSP90 inhibitors such as geldanamycin [7], 17-allylamino-17demethoxy-geldanamycin (17-AAG) [8] and 17-dimethylamino-ethylamino-17-demethoxy-geldanamycin (17-DMAG) [9], cause epidermal growth factor receptor (EGFR) desensitization in human pancreatic cancer cells, and that the activation of p38 mitogen-activated protein (MAP) kinase induced by HSP90 inhibitors regulates the desensitization of EGFR via its phosphorylation at Ser1046/7. HSP90 inhibitors, by interfering the N-terminal domain ATP binding site of HSP90, cause the destabilization and eventual degradation of HSP90 client proteins, and then lead to inhibit ATP-dependent HSP90 chaperone activity [10].
Regarding the MAP kinase superfamily, it is generally recognized that p44/p42 MAP kinase, p38 MAP kinase and stress-activated protein kinase/c-Jun N-terminal kinase play central roles in a variety of cellular functions, including proliferation, differentiation and survival [11]. Therefore, HSP90 is considered to act as a pivotal modulator of various cellular functions via MAP kinases such as p38 MAP kinase.[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
Bone metabolism is strictly regulated by two types of antagonistic functional cells; osteoblasts and osteoclasts [12]. Bone tissue is continuously regenerated through a process so called bone remodeling [13]. To maintain an adequate bone quality and the quantity, osteoblastic bone formation and osteoclastic bone resorption are tightly coordinated. The disruption of bone remodeling process causes metabolic bone diseases such as osteoporosis or fracture healing distress. With regard to HSP90 inhibitor-effects on bone metabolism, 17-AAG reportedly stimulates osteoclast formation and promotes osteolytic bone metastasis in bone metastasis of a breast cancer cell line [14].
In addition, it has been shown that geldanamycin induces autophagy and apoptosis of osteosarcoma cells [15]. However, the exact roles of HSP90 in bone metabolism have not yet been fully clarified.[](https://www.ncbi.nlm.nih.gov/mesh/C112765)
Interleukin-6 (IL-6) is a multifunctional cytokine which belongs to the glycoprotein 130 (gp130) cytokine family, and has important physiological effects on a variety of cell functions, such as the promotion of B-cell differentiation, the T-cell activation and the induction of acute-phase proteins [16,17]. It has been recognized that IL-6 stimulates bone resorption and induces osteoclast formation [17], and IL-6 reportedly plays a pivotal role in the process of bone fracture repair [18]. Thus, accumulating evidence suggests that IL-6 is an osteotropic modulator, and influence bone formation under the condition of increased bone turnover [19]. On the other hand, prostaglandins (PGs) modulate various bone cell functions as autacoids. Among them, PGF2α, which has been conventionally recognized as a potent bone-resorptive agent
[20], is currently recognized as a bone remodeling mediator [21]. It has been previously reported that PGF2α induced IL-6 production in osteoblast-enriched cultured neonatal mouse calvaria, resulting in bone resorption [20,22]. We have previously shown that PGF2α stimulates the synthesis of IL-6 through p44/p42 MAP kinase and p38 MAP kinase in osteoblast-like MC3T3-E1 cells [23,24]. Thus, it is probable that the PGF2α-induced IL-6 synthesis is not specific for osteoblast-like MC3T3-E1 cells but general phenomena in osteoblasts.
However, there is no report showing the roles of MAP kinases in the PGF2α-stimulated IL-6 synthesis in osteoblasts as far as we know. In addition, we have reported that Rho kinase inhibitors significantly suppress the synthesis of IL-6 and the phosphorylation of p38 MAP kinase induced by PGF2α without affecting the levels of total p38 MAP kinase in these cells, suggesting that Rho-kinase plays a role in PGF2α-stimulated IL-6 synthesis as an upstream regulator of p38 MAP kinase [25]. These findings lead us to speculate that HSP90 could regulate the IL-6 synthesis stimulated by PGF2α in osteoblasts.[](https://www.ncbi.nlm.nih.gov/mesh/D011453)
In the present study, we investigated whether HSP90 is involved in the PGF2α-induced IL-6 synthesis in osteoblast-like MC3T3-E1 cells. We herein show that HSP90 inhibitors enhance the PGF2α-stimulated IL-6 synthesis in these cells, and that the amplifying effect is exerted through up-regulating p38 MAP kinase activation.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
## Materials and methods
*In the **Materials and methods** section:*
## Materials
*In the **Materials** section:*
Geldanamycin and PGF2α were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA). 17-AAG, 17-DMAG and SB203580 were obtained from Calbiochem-Novabiochem Co. (La Jolla, CA). Onalespib was purchased from Selleckchem (Houston, TX). Mouse interleukin-6 enzyme-linked immunosorbent assay (ELISA) kit was purchased from R&D System, Inc. (Minneapolis, MN). Phospho-specific p44/p42 MAP kinase antibodies, p44/p42 MAP kinase antibodies, phospho-specific p38 MAP kinase antibodies, p38 MAP kinase antibodies, phospho-specific myosin phosphatase targeting subunit (MYPT-1) antibodies, MYPT-1 antibodies, RhoA antibodies and Rho-kinase (ROCK1) antibodies were obtained from Cell Signaling Technology, Inc.
(Beverly, MA). HSP90 antibodies and Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). An ECL Western blotting detection system was obtained from GE Healthcare Life Sciences (Chalfont, UK). Control short interfering RNA (siRNA; Silencer Negative Control no.1 siRNA) and HSP90-siRNA (Silencer select Pre-designed siRNA, s67897 and s67898, presented as #1 and #2) were purchased from Ambion (Austin, TX). Other materials and chemicals were obtained from commercial sources.
PGF2α was dissolved in ethanol. Geldanamycin, 17-AAG, 17-DMAG, onalespib and SB203580 were dissolved in dimethyl sulfoxide. The maximum concentration of ethanol or dimethyl sulfoxide was 0.1%, which did not affect the assay for IL-6, real-time RT-PCR or Western blot analysis.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
## Cell culture
*In the **Cell culture** section:*
Cloned osteoblast-like MC3T3-E1 cells, an immortalized cell line which had been derived from newborn mouse calvaria [26] were maintained as previously described [27]. Briefly, the cells were cultured in α-minimum essential medium (α-MEM) containing 10% fetal bovine serum (FBS) at 37°C in a humidified atmosphere of 5% CO2/95% air. The cells in early passage until 21 passages were seeded into 35-mm diameter dishes (5 x 104 cells/dish) or 90-mm diameter dishes (2 x 105 cells/dish) in α-MEM containing 10% FBS. After 5 days, the medium was exchanged for α-MEM containing 0.3% FBS.
The cells were used for experiments after 48 h.[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
## siRNA transfection
*In the **siRNA transfection** section:*
To knock down HSP90 in osteoblast-like MC3T3-E1 cells, the cells were transfected with HSP90-siRNA or negative control siRNA utilizing siLentFect according to the manufacturer’s protocol. In brief, the cells (2 x 105 cells) were seeded into 90-mm diameter dishes in α-MEM containing 10% FBS, and sub-cultured for 48 h. The cells were then incubated at 37°C with 10 nM siRNA-siLentFect complexes (#1) or 30 nM siRNA-siLentFect complexes (#2). After 24 h, the medium was exchanged for α-MEM containing 0.3% FBS. The transfected cells were then stimulated by 10 μM of PGF2α or vehicle in α-MEM containing 0.3% FBS for the indicated periods.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
## Assay for IL-6
*In the **Assay for IL-6** section:*
The cultured cells were stimulated by 10 μM of PGF2α or vehicle in 1 ml of α-MEM containing 0.3% FBS for the indicated periods. When indicated, cells were pretreated with various doses of geldanamycin, 17-AAG, 17-DMAG or onalespib for 60 min. The preincubation with 30 μM of SB203580 or vehicle was performed for 60 min prior to the pretreatment. The cells were stimulated by 10 μM of PGF2α or vehicle in 1 ml of α-MEM containing 0.3% FBS for the indicated periods. The conditioned medium was collected at the end of incubation, and the IL-6 concentration in the medium was then measured using the mouse IL-6 ELISA kit according to the manufacturer’s protocol.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
## Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)
*In the **Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)** section:*
The cultured cells were pretreated with 0.3 μM of geldanamycin, onalespib or vehicle for 60 min, and then stimulated by 10 μM of PGF2α or vehicle in α-MEM containing 0.3% FBS for 3 h. Total RNA was isolated and reverse transcribed into complementary DNA using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Heysham, Lancashire, UK) and Omniscript Reverse Transcriptase kit (Qiagen Inc., Valencia, CA), respectively. RT-qPCR was performed in capillaries using a LightCycler system with the LightCycler FastStart DNA Master SYBR Green I (Roche Diagnostics, Basel, Switzerland). Sense and antisense primers for mouse IL-6 mRNA and GAPDH mRNA were purchased from Takara Bio Inc. (Tokyo, Japan) (primer set ID: MA039013 or RA015380, respectively).
The amplified products were determined by melting curve analysis and agarose electrophoresis. The IL-6 mRNA levels were normalized to those of GAPDH mRNA.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
## Western blot analysis
*In the **Western blot analysis** section:*
The cultured cells were pretreated with various doses of geldanamycin, 17-AAG or onalespib for 60 min, and then stimulated by 10 μM of PGF2α or vehicle in α-MEM containing 0.3% FBS for the indicated periods. As for the HSP90 knockdown cells, the cells transfected with siRNA were stimulated by 10 μM of PGF2α or vehicle in 1 ml of α-MEM containing 0.3% FBS for the indicated periods. The cells were then washed twice with phosphate-buffered saline, and then lysed, homogenized and sonicated in a lysis buffer containing 62.5 mM Tris/HCl, pH 6.8, 2% sodium dodecyl sulfate (SDS), 50 mM dithiothreitol and 10% glycerol. SDS-polyacrylamide gel electrophoresis (PAGE) was performed by the method of Laemmli [28] in 10% polyacrylamide gels. The protein was fractionated and transferred onto an Immun-Blot PVDF membrane (Bio-Rad, Hercules, CA).
The membranes were blocked with 5% fat-free dry milk in Tris-buffered saline-Tween (TBS-T; 20 mM Tris-HCl, pH 7.6, 137 mM NaCl, 0.1% Tween 20) for 1 h before incubation with primary antibodies. Western blot analysis was performed as described previously [29] using phospho-specific p44/p42 MAP kinase antibodies, p44/p42 MAP kinase antibodies, phospho-specific p38 MAP kinase antibodies, p38 MAP kinase antibodies, HSP90 antibodies, phospho-specific MYPT-1 antibodies, MYPT-1 antibodies, RhoA antibodies, ROCK1 antibodies and GAPDH antibodies as primary antibodies with peroxidase-labeled antibodies raised in goat against rabbit IgG (KPL, Inc., Gaithersburg, MD) used as secondary antibodies. The primary and secondary antibodies were diluted at 1:1000 with 5% fat-free dry milk in TBS-T. The peroxidase activity on the PVDF sheet was visualized on X-ray film by means of the ECL Western blotting detection system.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
## Densitometric analysis
*In the **Densitometric analysis** section:*
A densitometric analysis of the Western blots was performed using a scanner and image analysis software program (image J version 1.48, National Institutes of Health, Bethesda, MD). The phosphorylated protein levels were calculated as follows: the background-subtracted signal intensity of each phosphorylation signal was respectively normalized to the total protein signal and plotted as the fold increase in comparison to that of the control cells without stimulation.
## Statistical analysis
*In the **Statistical analysis** section:*
The data were analyzed by ANOVA followed by Bonferroni method for multiple comparisons between pairs, and p<0.05 was considered to be statistically significant. All data are presented as the mean ± S.E.M. of triplicate determinations from three independent cell preparations.
## Results
*In the **Results** section:*
## Effects of HSP90 inhibitors on the PGF2α-stimulated IL-6 release in MC3T3-E1 cells
*In the **Effects of HSP90 inhibitors on the PGF2α-stimulated IL-6 release in MC3T3-E1 cells** section:*
It is recognized that geldanamycin, a benzoquinone ansamycin antibiotic, binds to the N-terminal domain ATP binding site of HSP90, inhibiting ATP-dependent HSP90 chaperone activity [7,10], and its less toxic derivative 17-AAG and 17-DMAG also bind specifically to HSP90 in a manner similar to geldanamycin [8,9]. In order to investigate the involvement of HSP90 in the PGF2α-induced synthesis of IL-6 in osteoblast-like MC3T3-E1 cells, we first examined the effects of these HSP90 inhibitors on the PGF2α-stimulated IL-6 release. Geldanamycin significantly amplified the PGF2α-stimulated IL-6 release in a time-dependent manner up to 48 h (Fig 1A). In addition, the enhancing effect of geldanamycin was dose-dependent in the range between 0.1 and 1 μM (Fig 1B). The maximum effect of geldanamycin was observed at 1 μM, which caused an approximately 330% increase in the PGF2α-effect.
Additionally, 17-AAG significantly enhanced the PGF2α-stimulated IL-6 release (Fig 2A). The effect of 17-AAG (0.1 μM) on the IL-6 release caused an approximately 150% increase in the PGF2α-effect. Furthermore, 17-DMAG significantly amplified the IL-6 release (Fig 2B). The amplifying effect of 17-DMAG (0.1 μM) caused an approximately 240% increase in the PGF2α-effect.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
Effect of geldanamycin on the PGF2α-stimulated IL-6 release in MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
(A) The cultured cells were pretreated with 1 μM of geldanamycin (●,○) or vehicle (■,□) for 60 min, and then stimulated by 10 μM of PGF2α (●,■) or vehicle (○,□) for the indicated periods. (B) The cultured cells were pretreated with various doses of geldanamycin for 60 min, and then stimulated by 10 μM of PGF2α (●) or vehicle (○) for 48 h. IL-6 concentrations in the conditioned medium were determined by ELISA. Each value represents the mean ± S.E.M. of triplicate determinations from three independent cell preparations. (A) *p<0.05 compared to the value of control. **p<0.05 compared to the value of PGF2α alone. (B) *p<0.05 compared to the value of PGF2α alone.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
Effects of 17-AAG, 17-DMAG or onalespib on the PGF2α-stimulated IL-6 release in MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C112765)
The cultured cells were pretreated with 1 μM of 17-AAG (A), 1 μM of 17-DMAG (B) or vehicle for 60 min, and then stimulated by 10 μM of PGF2α or vehicle for 48 h. (C) The cultured cells were pretreated with various doses of onalespib for 60 min, and then stimulated by 10 μM of PGF2α (●) or vehicle (○) for 48 h. IL-6 concentrations in the conditioned medium were determined by ELISA. Each value represents the mean ± S.E.M. of triplicate determinations from three independent cell preparations. (A,B) *p<0.05 compared to the value of control. (A,B) **p<0.05 compared to the value of PGF2α alone. (C) *p<0.05 compared to the value of PGF2α alone.[](https://www.ncbi.nlm.nih.gov/mesh/C112765)
It is known that 17-AAG and 17-DMAG belong to geldanamycin analogues [30], and onalespib is another type HSP90 inhibitor different from geldanamycin analogues, which binds to the N-terminal domain ATP binding site of HSP90 [31]. We further examined the effect of onalespib on the PGF2α-induced IL-6 release in osteoblast-like MC3T3-E1 cells. Onalespib, which by itself had little effect on the IL-6 release, significantly amplified the PGF2α-stimulated IL-6 release as well as geldanamycin analogues (Fig 2C). The enhancing effect of onalespib was dose-dependent in the range between 0.01 and 0.1 μM. Onalespib at 0.1 μM caused an approximately 300% increase in the PGF2α-effect.[](https://www.ncbi.nlm.nih.gov/mesh/C112765)
## Effects of geldanamycin or onalespib on the PGF2α-induced expression of IL-6 mRNA in MC3T3-E1 cells
*In the **Effects of geldanamycin or onalespib on the PGF2α-induced expression of IL-6 mRNA in MC3T3-E1 cells** section:*
In order to clarify whether the amplifying effects of geldanamycin or onalespib on the PGF2α-stimulated IL-6 release is mediated through transcriptional events, we next examined the effects of geldanamycin or onalespib on the PGF2α-induced IL-6 mRNA expression. We found that geldanamycin and onalespib, which by themselves had little effect on the mRNA levels of IL-6, significantly up-regulated the expression levels of mRNA induced by PGF2α (Fig 3A and Fig 3B).[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
Effects of geldanamycin or onalespib on the PGF2α-induced expression levels of IL-6 mRNA in MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
The cultured cells were pretreated with 0.3 μM of geldanamycin (A), onalespib (B) or vehicle for 60 min, and then stimulated by 10 μM of PGF2α or vehicle for 3 h. The respective total RNA was then isolated and transcribed into cDNA. The expressions of IL-6 mRNA and GAPDH mRNA were quantified by RT-qPCR. The IL-6 mRNA levels were normalized to those of GAPDH mRNA. Each value represents the mean ± S.E.M. of triplicate determinations from three independent cell preparations. *p<0.05 compared to the value of control. **p<0.05 compared to the value of PGF2α alone.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
## Effects of HSP90 inhibitors on the PGF2α-induced phosphorylation of p44/p42 MAP kinase in MC3T3-E1 cells
*In the **Effects of HSP90 inhibitors on the PGF2α-induced phosphorylation of p44/p42 MAP kinase in MC3T3-E1 cells** section:*
Regarding the intracellular signaling pathway of PGF2α in osteoblasts, we have shown that p44/p42 MAP kinase acts as a positive regulator in the PGF2α-stimulated IL-6 synthesis in osteoblast-like MC3T3-E1 cells [23,24]. Therefore, we examined the effects of HSP90 inhibitors on the PGF2α-induced phosphorylation of p44/p42 MAP kinase in these cells. However, geldanamycin failed to affect the PGF2α-induced phosphorylation of p44/p42 MAP kinase (Fig 4A). In addition, 17-AAG had no effect on the PGF2α-induced phosphorylation of p44/p42 MAP kinase (Fig 4B).[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
Effects of geldanamycin or 17-AAG on the PGF2α-induced phosphorylation of p44/p42 MAP kinase in MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
The cultured cells were pretreated with various doses of geldanamycin (A) or 17-AAG (B) for 60 min, and then stimulated by 10 μM of PGF2α or vehicle for 20 min. The cell extracts were then subjected to SDS-PAGE with subsequent Western blot analysis with antibodies against phospho-specific p44/p42 MAP kinase or p44/p42 MAP kinase. The histogram shows the quantitative representations of the levels of phosphorylated p44/p42 MAP kinase after normalization with respect to p44/p42 MAP kinase obtained from laser densitometric analysis. The levels were expressed as the fold increase to the basal levels presented as lane 1. Each value represents the mean ± S.E.M. of triplicate determinations from three independent cell preparations. *p<0.05 compared to the value of control. N.S. designates no significant difference between the indicated pairs.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
## Effects of HSP90 inhibitors on the phosphorylation of p38 MAP kinase induced by PGF2α in MC3T3-E1 cells
*In the **Effects of HSP90 inhibitors on the phosphorylation of p38 MAP kinase induced by PGF2α in MC3T3-E1 cells** section:*
In our previous study [24], we demonstrated that PGF2α stimulates IL-6 synthesis at least in part through p38 MAP kinase in addition to p44/p42 MAP kinase in osteoblast-like MC3T3-E1 cells. In order to clarify whether the activation of p38 MAP kinase is implicated in the enhancement by HSP90 inhibitors of the PGF2α-induced IL-6 synthesis in MC3T3-E1 cells, we examined the effects of HSP90 inhibitors on the PGF2α-induced phosphorylation of p38 MAP kinase. Geldanamycin, which alone had little effect on the p38 MAP kinase phosphorylation, significantly strengthened the PGF2α-induced phosphorylation of p38 MAP kinase in a dose-dependent manner in the range between 0.3 and 1.0 μM (Fig 5A). Additionally, 17-AAG as well as geldanamycin, which by itself did not affect the p38 MAP kinase phosphorylation, markedly enhanced the PGF2α-induced phosphorylation of p38 MAP kinase (Fig 5B). The amplifying effect of 17-AAG on the p38 MAP kinase phosphorylation was dose-dependent in the range between 0.3 and 1.0 μM. Furthermore, we found that onalespib (1 μM), as well as geldanamycin and 17-AAG, which by itself had little effect of the p38 MAP kinase phosphorylation, markedly enhanced the PGF2α-induced phosphorylation of p38 MAP kinase (Fig 5C).[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
Effects of geldanamycin, 17-AAG or onalespib on the phosphorylation of p38 MAP kinase induced by PGF2α in MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
The cultured cells were pretreated with the indicated doses of geldanamycin (A) or 17-AAG (B) or onalespib (C) for 60 min, and then stimulated by 10 μM of PGF2α or vehicle for 10 min. The cell extracts were then subjected to SDS-PAGE with subsequent Western blot analysis with antibodies against phospho-specific p38 MAP kinase or p38 MAP kinase. The histogram shows the quantitative representations of the levels of phosphorylated p38 MAP kinase after normalization with respect to p38 MAP kinase obtained from laser densitometric analysis. The levels were expressed as the fold increase to the basal levels presented as lane 1. Each value represents the mean ± S.E.M. of triplicate determinations from three independent cell preparations. *p<0.05 compared to the value of control. **p<0.05 compared to the value of PGF2α alone.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
## Effects of HSP90 inhibitors on the PGF2α-induced phosphorylation of MYPT-1, and the protein levels of MYPT-1, RhoA and Rho-kinase in MC3T3-E1 cells
*In the **Effects of HSP90 inhibitors on the PGF2α-induced phosphorylation of MYPT-1, and the protein levels of MYPT-1, RhoA and Rho-kinase in MC3T3-E1 cells** section:*
In addition to our previous findings in the PGF2α-stimulated IL-6 synthesis in osteoblast-like MC3T3-E1 cells, we have demonstrated that Rho-kinase positively regulates at a point upstream of p38 MAP kinase [25]. It is generally recognized that MYPT-1, which is a component of myosin phosphatase, is a substrate of Rho-kinase [32,33]. We next examined the effects of HSP90 inhibitors on the PGF2α-induced phosphorylation of MYPT-1, and the protein levels of MYPT-1, RhoA and Rho-kinase in MC3T3-E1 cells.
Geldanamycin failed to affect the PGF2α-induced phosphorylation of MYPT-1 (Fig 6A). We found that the protein levels of MYPT-1, RhoA or Rho-kinase were not affected by geldanamycin with or without PGF2α stimulation (Fig 6A). 17-AAG, as well as geldanamycin, had no effect on the PGF2α-induced phosphorylation of MYPT-1 (Fig 6B). In addition, we confirmed that 17-AAG did not affect the amounts of MYPT-1, RhoA or Rho-kinase (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
Effects of geldanamycin or 17-AAG on the phosphorylation of MYPT-1, the amounts of MYPT-1, RhoA and Rho-kinase induced by PGF2α in MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
The cultured cells were pretreated with various doses of geldanamycin (A) or 17-AAG (B) for 60 min, and then stimulated by 10 μM of PGF2α or vehicle for 2 min. The cell extracts were then subjected to SDS-PAGE with subsequent Western blot analysis with antibodies against phospho-specific MYPT-1, MYPT-1, RhoA or Rho-kinase. (a) The histogram shows the quantitative representations of the levels of phosphorylated MYPT-1 after normalization with respect to MYPT-1 obtained from laser densitometric analysis. The levels were expressed as the fold increase to the basal levels presented as lane 1. (b),(c),(d) The histogram shows the quantitative representations of the levels of MYPT-1 (b), RhoA (c) and Rho-kinase (d) after normalization with respect to GAPDH obtained from laser densitometric analysis, respectively. The levels were expressed as the ratio to the levels presented as lane 1. Each value represents the mean ± S.E.M. of triplicate determinations from three independent cell preparations. *p<0.05 compared to the value of control. N.S. designates no significant difference between the indicated pairs.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
## Effect of PGF2α on the phosphorylation of p38 MAP kinase in HSP90 knockdown MC3T3-E1 cells
*In the **Effect of PGF2α on the phosphorylation of p38 MAP kinase in HSP90 knockdown MC3T3-E1 cells** section:*
To further investigate whether HSP90 affects the PGF2α-induced phosphorylation of p38 MAP kinase in osteoblast-like MC3T3-E1 cells, we examined the effect of HSP90-siRNA on the phosphorylation of p38 MAP kinase induced by PGF2α. We found that the levels of HSP90 were not significantly but slightly reduced in the HSP90-siRNA (#1 and #2)-transfected cells (Fig 7A and Fig 7B). The levels of phosphorylated p38 MAP kinase induced by PGF2α in the HSP90-siRNA transfected cells (#1 and #2) were significantly amplified compared to those in the control cells (Fig 7A and Fig 7B). Thus, it seems likely that HSP90-siRNA has little effect on HSP90 protein levels but reduces HSP90 activity in MC3T3-E1 cells. The protein levels of p38 MAP kinase were not affected by HSP90-siRNA (#1 and #2).[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
Effect of PGF2α on the phosphorylation of p38 MAP kinase in HSP90 knockdown MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
(A) The cultured cells were transfected with 10 nM negative control siRNA (Neg) or 10 nM HSP90-siRNA (#1). (B) The cultured cells were transfected with 30 nM negative control siRNA (Neg) or 30 nM HSP90-siRNA (#2). Twenty-four hours after transfection, the cells were stimulated by 10 μM PGF2α or vehicle for 10 min. The cell extracts were then subjected to SDS-PAGE with subsequent Western blot analysis with antibodies against phospho-specific p38 MAP kinase, p38 MAP kinase, HSP90 or GAPDH. (a) The histogram shows the quantitative representations of the levels of phosphorylated p38 MAP kinase after normalization with respect to GAPDH obtained from laser densitometric analysis. The levels were expressed as the fold increase to the basal levels presented as lane 1. (b),(c) The histogram shows the quantitative representations of the levels of (b) total p38 MAP kinase and (c) HSP90αβ after normalization with respect to GAPDH obtained from laser densitometric analysis, respectively. The levels were expressed as the ratio to the levels presented as lane 1. Each value represents the mean ± S.E.M. of triplicate determinations from three independent cell preparations. *p<0.05 compared to the value of control. **p<0.05 compared to the value of PGF2α alone.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
## Effect of SB203580 on the enhancement by HSP90 inhibitors of the PGF2α-stimulated IL-6 release in MC3T3-E1 cells
*In the **Effect of SB203580 on the enhancement by HSP90 inhibitors of the PGF2α-stimulated IL-6 release in MC3T3-E1 cells** section:*
Furthermore, we examined the effect of SB203580, a p38 MAP kinase inhibitor [34], on the enhancement by HSP90 inhibitors of the PGF2α-stimulated IL-6 release in osteoblast-like MC3T3-E1 cells. We found that SB203580, which by itself had little effect on IL-6 levels, truly suppressed the PGF2α-induced IL-6 release as our previous report [25] (Fig 8). SB203580 markedly reduced the amplification by geldanamycin, 17-AAG or 17-DMAG of the PGF2α-stimulated IL-6 release (Fig 8). SB203580 (30 μM) caused an approximately 90%, 90% and 95% decrease in the effect of PGF2α with geldanamycin, 17-AAG and 17-DMAG, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/C093642)
Effect of SB203580 on the enhancement by geldanamycin, 17-AAG or 17-DMAG of the PGF2α-induced IL-6 release in MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C093642)
The cultured cells were preincubated with 30 μM of SB203580 or vehicle for 60 min, subsequently pretreated with 1 μM of geldanamycin, 1 μM of 17-AAG, 1 μM of 17-DMAG or vehicle for 60 min, and then stimulated by 10 μM of PGF2α or vehicle for 48 h. IL-6 concentrations of the conditioned mediums were determined by ELISA. Each value represents the mean ± S.E.M. of triplicate determinations from three independent cell preparations. *p<0.05 compared to the value of control. **p<0.05 compared to the value of PGF2α alone. ***p<0.05 compared to the value of PGF2α with the pretreatment of each HSP90 inhibitor.[](https://www.ncbi.nlm.nih.gov/mesh/C093642)
## Discussion
*In the **Discussion** section:*
In the present study, we demonstrated that HSP90 inhibitors including geldanamycin, 17-AAG and 17-DMAG significantly enhanced the PGF2α-stimulated release of IL-6 in osteoblast-like MC3T3-E1 cells. We also found that onalespib, another type HSP90 inhibitor [31], amplified the IL-6 release induced by PGF2α in these cells. In addition, we showed that the expression levels of IL-6 mRNA induced by PGF2α were markedly amplified by geldanamycin and onalespib. Thus, it seems likely that the amplifying effect of HSP90 inhibitors on the PGF2α-stimulated IL-6 release is mediated through the gene transcription in MC3T3-E1 cells. Based on our findings, it is probable that HSP90 plays an inhibitory role in the PGF2α-stimulated IL-6 synthesis in osteoblast-like MC3T3-E1 cells.
To the best of our knowledge, this is probable the first report showing the suppression by HSP90 of IL-6 synthesis in osteoblasts. Therefore, we next investigated the exact mechanism behind the suppression by HSP90 of the PGF2α-stimulated IL-6 synthesis in osteoblast-like MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C001277)
Regarding the intracellular signaling system of PGF2α in osteoblasts, we have previously shown that PGF2α stimulates the synthesis of IL-6 through p44/p42 MAP kinase and p38 MAP kinase in osteoblast-like MC3T3-E1 cells [23,24]. Additionally, in our previous study [25], we have reported that Rho-kinase positively regulates PGF2α-stimulated IL-6 synthesis at a point upstream of not p44/p42 MAP kinase but p38 MAP kinase in MC3T3-E1 cells. Thus, we investigated whether HSP90 inhibitors affect the PGF2α-stimulated activation of p44/p42 MAP kinase or p38 MAP kinase in these cells. However, neither geldanamycin nor 17-AAG affected the phosphorylation of p44/p42 MAP kinase. On the contrary, we demonstrated that geldanamycin, 17-AAG and onalespib significantly enhanced the PGF2α-stimulated phosphorylation of p38 MAP kinase.
Therefore, it is probable that the enhancement by HSP90 inhibitors of PGF2α-stimulated IL-6 synthesis is due to up-regulating the activation of p38 MAP kinase but not p44/p42 MAP kinase in MC3T3-E1 cells. In addition, we demonstrated that either geldanamycin or 17-AAG failed to affect the PGF2α-induced phosphorylation of MYPT-1, a substrate of Rho-kinase [33]. We also found that HSP90 inhibitors had no effect of the amounts of MYPT-1, RhoA or Rho-kinase. It is generally recognized that Rho and the down-stream effector, Rho-kinase play crucial roles in a variety of cellular functions such as cell motility and smooth muscle contraction [32,33].
As for osteoblasts, it has been shown that the inhibition of RhoA-Rho-kinase signaling influences osteoblast adhesion, differentiation and mineralization [35]. In the present study, we showed that HSP90 inhibitors failed to affect the phosphorylation of MYPT-1, a target of Rho-kinase. Based on our findings, it seems unlikely that the amplification by HSP90 inhibitors of PGF2α-stimulated IL-6 synthesis is mediated through the enhancement of RhoA-Rho-kinase activity. Our findings suggest that HSP90 might function as a negative regulator in the PGF2α-stimulated IL-6 synthesis in osteoblast-like MC3T3-E1 cells, and that the effect of HSP90 on the IL-6 synthesis is exerted at the point between Rho-kinase and p38 MAP kinase. In order to further elucidate whether HSP90 regulates the PGF2α-induced activation of p38 MAP kinase in MC3T3-E1 cells, we examined the effect of HSP90-siRNA on the PGF2α-induced phosphorylation of p38 MAP kinase.
The PGF2α-induced levels of phosphorylated p38 MAP kinase were significantly enhanced by HSP90-siRNA. Additionally, we clearly demonstrated that SB203580 reduced the amplification by geldanamycin, 17-AAG or 17-DMAG of the PGF2α-stimulated IL-6 release in MC3T3-E1 cells. Taking these findings into account, it is most likely that HSP90 limits the PGF2α-stimulated IL-6 synthesis in osteoblast-like MC3T3-E1 cells, and the suppressive effect of HSP90 is exerted at the point between Rho-kinase and p38 MAP kinase. The potential mechanism underlying amplification by HSP90 inhibitors of the PGF2α-stimulated IL-6 synthesis in osteoblasts is summarized as Fig 9.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
Schematic illustration of the regulatory mechanism underlying the amplifying effect of HSP90 inhibitors on the PGF2α-induced IL-6 synthesis in osteoblast-like MC3T3-E1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
PGF2α, prostaglandin F2α; MAPK, mitogen-activated protein kinase; HSP90, heat shock protein 90; IL-6, interleukin-6.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
In bone metabolism, it has been generally recognized that IL-6 acts as a potent bone resorptive agent and to promote osteoclast formation [17]. In addition to the IL-6-effect on bone resorption, IL-6 is currently considered as an osteotropic factor under the condition of increased bone turnover, and is inducible even bone formation [19]. Bone resorption is the primary step of bone remodeling, and bone formation is subsequently initiated [36]. To obtain the quantity of bone and the quality, the adequate handling of bone remodeling process performed by osteoclasts and osteoblasts is essential. Although there is no doubt that IL-6 stimulates osteoclastic bone resorption, IL-6 is recognized to act as a bone remodeling mediator from the viewpoint of whole bone metabolism.
As for HSP90 in osteoblasts, we have shown that the expression levels of HSP90 protein are quite high in osteoblast-like MC3T3-E1 cells [37]. Therefore, our present findings, showing that HSP90 inhibitors could function as an up-regulator with regard to the PGF2α-stimulated IL-6 synthesis in osteoblasts, probably provides a novel insight with regard to HSP90 as an essential regulator of bone remodeling. Several HSP90 inhibitors including geldanamycin have been recognized as an anticancer drug, and the effects of HSP90 inhibitors have been tested in the clinical trial [38]. Based on our present findings, it is probable that HSP90 inhibitors, in addition to the anticancer chemotherapeutics, could affect bone metabolism as a bone remodeling modulator through the enhancement of IL-6 synthesis in osteoblasts. Therefore, HSP90 inhibitors might lead a new therapeutic strategy for acceleration of fracture healing and bone metabolic diseases such as osteoporosis.
Further investigations would be required to clarify the details underlying the effects of HSP90 on bone metabolism.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
In conclusion, our results strongly suggest that HSP90 negatively regulates the PGF2α-stimulated IL-6 synthesis in osteoblasts, and that the effect of HSP90 on the IL-6 synthesis is exerted through regulating p38 MAP kinase activation.[](https://www.ncbi.nlm.nih.gov/mesh/D015237)
# References
*In the **References** section:*
|
# Introduction
Data driven polypharmacological drug design for lung cancer: analyses for targeting ALK, MET, and EGFR
# Abstract
*In the **Abstract** section:*
Drug design of protein kinase inhibitors is now greatly enabled by thousands of publicly available X-ray structures, extensive ligand binding data, and optimized scaffolds coming off patent. The extensive data begin to enable design against a spectrum of targets (polypharmacology); however, the data also reveal heterogeneities of structure, subtleties of chemical interactions, and apparent inconsistencies between diverse data types. As a result, incorporation of all relevant data requires expert choices to combine computational and informatics methods, along with human insight. Here we consider polypharmacological targeting of protein kinases ALK, MET, and EGFR (and its drug resistant mutant T790M) in non small cell lung cancer as an example. Both EGFR and ALK represent sources of primary oncogenic lesions, while drug resistance arises from MET amplification and EGFR mutation. A drug which inhibits these targets will expand relevant patient populations and forestall drug resistance. Crizotinib co-targets ALK and MET. Analysis of the crystal structures reveals few shared interaction typ[es, highli](https://www.ncbi.nlm.nih.gov/mesh/C551994)ghting proton-arene and key CH–O hydrogen bonding interactions. These are not typically encoded into molecular mechani[cs fo](https://www.ncbi.nlm.nih.gov/mesh/D006841)rce fields. Ch[eminform](https://www.ncbi.nlm.nih.gov/mesh/D006859)atics analyses of binding data show EGFR to be dissimilar to ALK and MET, but its structure shows how it may be co-targeted with the addition of a covalent trap. This suggests a strategy for the design of a focussed chemical library based on a pan-kinome scaffold. Tests of model compounds show these to be compatible with the goal of ALK, MET, and EGFR polypharmacology.
Electronic supplementary material
The online version of this article (doi:10.1186/s13321-017-0229-8) contains supplementary material, which is available to authorized users.
## Background
*In the **Background** section:*
The importance and proven druggability of protein kinases as targets in cancer [, ], inflammation [], and other disease areas have transformed antikinase drug discovery into an information driven research area of unprecedented scale []. Public and proprietary databases contain binding data for hundreds of thousands of active compounds []. Crystal structures are publicly available for some 3000 protein kinase inhibitor complexes in the Protein Database (www.rcsb.org) [].
This data begins to enable “polypharmacological” targeting of multiple kinases [–], which may more effectively modify network behavior [], or forestall drug resistance [, ], or provide broader applicability against heterogeneous cancers or larger patient groups. Such approaches [] may involve “retargeting” via modification of known compounds, or simply “repurposing” known compounds to new applications when target profiles are suitable. Practical approaches to polypharmacological design include both experimental and computational methods [–]. There is however no single strategic approach to modify such starting compounds to achieve the final selectivity profile; this depends on the availability, identification and understanding of the essential selectivity determinants of the relevant targets, as we examine with example of this paper.
Overview of current molecular targets in advanced/metastatic non-small-cell lung cancer (NSCLC). Protein kinase inhibitors for NSCLC therapy, either approved or in advanced clinical trials, are shown in green boxes; antibodies against cytokine targets HGF asnd VEGF are gray. EGFR, MET and ALK are labeled with blue circles
Non-small-cell lung cancer (NSCLC) represents a collection of diverse molecular pathologies. Most types are relatively insensitive to chemotherapy, but the identification of genomic abnormalities in subpopulations of NSCLC patients [–] have led to the development of protein kinase inhibitors against EGFR [, ] (gefitinib, 2003; erlotinib, 2004; afatinib, 2013) and ALK (crizotinib, 2011; ceritinib, 2014; alectinib, 2015), see Fig. 1. These inhibitors specifically target either EGFR or ALK, but not both; cross-reactive inhibitors are under investigation however [].
Analogous to the results of imatinib and ABL inhibition as therapy for CML, treatment with gefitinib and erlotinib is associated with acquired resistance []. Unlike ABL inhibition, resistance development to EGFR inhibitors seems universal. The most common resistance mechanism is a secondary mutation of the gatekeeper residue (for EGFR, predominantly T790M); afatinib appears less [] but not completely [] insensitive to this mutation. Additional mechanisms of acquired resistance include the amplification of MET [], HGF
[], or HER2 []. The universal appearance of drug resistance via diverse mechanisms following EGFR inhibition therapy has generated widespread interest in polypharmacological or combinatorial treatment strategies [–]. In this paper we examine the potential of polypharmacological targeting EGFR, ALK, and MET.[](https://www.ncbi.nlm.nih.gov/mesh/C419708)
As is typical for protein kinase inhibitors, compounds known to inhibit ALK, MET, and EGFR most potently bind at the ATP site, where they are anchored to the interdomain hinge via one or more hydrogen bonds []. Together, they represent a small subset of the known scaffolds for hinge/ATP site binders [], which are already a restricted set [] compared to proposed extents of possible scaffold diversity []. The EGFR inhibitor lapatinib was one of the first nearly monospecific kinase inhibitors
[]; others may have much broader selectivity profiles []. Knowledge of the determinants of selectivity for specific protein kinases or subfamilies [, ] assists target specific or polypharmacological drug design. These include the “gatekeeper” residue at the hinge, and infrequent occurrences of unique residues, such as glycine [] or cysteine [, ]. The ongoing expansion of public [, –] and proprietary [] target-ligand binding data begins to enable “machine learning” prediction of target inhibition profiles [–]. Such methods, similar to the more common structure based methods [, ], generally do not allow precise (e.g. binding constant error <10-fold) prediction of binding properties of individual compound-target interactions, but they do provide guidance to focus efforts on compound classes or libraries with the best chances for success [–].[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
In addition to the currently approved drugs and known inhibitors, many crystal structures are available to support drug design against NSCLC targets []. For EGFR, ALK, and MET, truncated kinase domain structures have been determined many times, uncomplexed, in complexes with ATP, ATP analogs, and inhibitors, including mutants and truncation variants. These structures show considerable diversity of active and inactive conformations [, ], and increasingly enable sophisticated drug design approaches. The target EGFR has become the primary example for targeting cysteine residues for irreversible inhibition [, , ] and, along with ABL, for the design of broadened selectivity profiles to overcome or forestall drug resistance [].
A growing catalog of resistance mutants appearing in ALK [–] and MET also belong to the set of targets to be considered in general polypharmacological targeting strategies against NSCLC [–].[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
In this paper, we examine combinations of structure, binding, and target validation data to suggest a strategy for polypharmacological targeting of ALK, MET, and EGFR. We use cheminformatics methods to analyse the similarities of the targets. Inhibitor binding data shows a high degree of correlation between ALK and MET with respect to inhibition profiles, but also an essential dissimilarity with EGFR. The drug resistant mutant EGFR-T790M is intermediate between the two groups. We compare crystal structures of ALK and MET, considering especially the cross-inhibitory compound crizotinib, in an attempt to identify the structural origins of the similarities.
Despite the dissimilarity of EGFR, the availability of a cysteine residue in the ATP pocket provides an orthogonal approach for polypharmacological optimization: the ALK-MET similarity may be exploited for optimization, while the addition of a covalent trap to inhibitors may add effective EGFR inhibition to the profile. Finally, we test compounds synthesized with these properties to verify the approach.[](https://www.ncbi.nlm.nih.gov/mesh/C551994)
## Results and discussion
*In the **Results and discussion** section:*
## Cheminformatics show similarities of ALK and MET, dissimilarity of EGFR, and intermediate similarity of EGFR mutant T790M
*In the **Cheminformatics show similarities of ALK and MET, dissimilarity of EGFR, and intermediate similarity of EGFR mutant T790M** section:*
The “activity homology” (AH) similarity measure [] as applied to ALK, MET, and EGFR. a Fractions of the sets of tight binding compounds of a reference PK target that also tighly bind to the tested PK are plotted for the ca 400 PKs of the test set. The curves are color coded according to the reference PK “A”: black for ALK, red for Met, blue for EGFR, and yellow for the drug resistant mutant EGFR-(L858R, T790M), which is abbreviated EGFR-LR/TM on the plot. The PKs of the test set are ordered according to the AH with ALK. Thus, Met-M1250T has the highest ALK AH (black curve) of the three Met variants (45%), EGFR is low (<5%), while EGFR-(L858R, T790M) is relatively high (30%). RON has the highest AH to MET, while TAO1 has the highest AH with EGFR-LR/TM. The peaks with high homology to EGFR marked with an asterisk are EGFR mutants other than EGFR-LR/TM, and have high AH similarity to EGFR (but not EGFR-LR/TM). b The same data, shown as a heirarchically clustered heat map. The mutant labelled L858RT represents the EGFR-(L858R, T790M) mutant, and is more similar to Alk and Met than to the other EGFR forms
Several measures are available to evaluate the similarities of kinase drug targets [, , –], including sequence, structure, and inhibitor properties. For drug discovery purposes, experimental binding data regarding cross-reactivity of inhibitors may be the most relevant, although this data may be generated in different ways, with diversity arising from choice of binding or activity assays, conditions, target protein form, etc. Significant discrepancies between assay formats are to be expected [, , ] which are more critically reflected in disparities between in vitro measurement conditions and their applicability in vivo. Based on single concentration measurement data from Ambit BioSciences, including estimated IC50 values for the binding of >20,000 compounds to 300–400 protein kinases, researchers at Bristol-Myers Squibb evaluated inhibitor selectivity profiles with an “activity homology” (AH) score
[] (see “Methods”). By this measure, ALK and MET are the most similar of the kinases considered in this manuscript, while EGFR is distinctly different (Fig. 2). Of the 400 protein kinases in the test set, some 80 kinases are more similar to ALK than MET, including the gatekeeper mutant M1250T of MET (approximately at position 50 of the 400 kinases). About 35% of the potent ALK inhibitors are shared with the MET inhibitor set (and about 43% with MET-M1250T). Less than 5% of the ALK inhibitors bind potently to EGFR and its mutants, with the notable exception of the double mutant EGFR-(L858R, T790M).
This combination of the cancer primary mutation L858R and drug resistance gatekeeper mutation T790M is potently inhibited by 30% of the ALK inhibitors. (Figure 2 shows ALK4 as similar to EGFR, sensitive to over 50% of the EGFR inhibitors. This kinase “Activin-Like receptor Kinase” belongs to the Tyrosine Kinase Like (TKL) subfamily of the kinome, and is not related to ALK “Anaplastic Lymphoma Kinase” studied in this work).
Correlations of inhibition profiles of the Ambit 2011 kinase profiling dataset []. Disk sizes and colors (red 100%, magenta 80%, blue 50%, green 20%) show the correlations of inhibition profiles of individual PKs with that of the PK of interest. a Correlations with ALK. b Correlations with EGFR
A related measure of similarity that also uses inhibitor binding data is the correlations of inhibitor binding profiles between pairs of kinases. Highly correlated targets share similar sensitivities to changes in the inhibitors. Unlike the “activity homology” described above, correlation compares the pattern of variation of inhibition strengths, and not the absolute values. Thus, two kinases may have highly correlated inhibition profiles even if the inhibition pattern is significantly weaker for one kinase. This may occur, for example, if the overall shapes of the inhibitor binding sites are similar, but one of the kinases may lack one important binding feature.
For drug polypharmacology design purposes, it may be advantageous to enhance recognition of correlated sensitivities to ligand variation. Using the binding data of the Ambit study of 72 inhibitors and their interactions with 442 kinases [], correlation analysis highlights the similarity of ALK and MET, and the dissimilarity of EGFR (Fig. 3). The inhibitor set of the study shows a large number of protein kinases, widely distributed across the kinome, with moderate similarity to ALK. The kinases with the most correlated inhibition profiles are, like ALK, tyrosine kinases, and include the closely related LTK, INSR, IRR and IGF1R kinases, but also FAK, PYK2, FER, FES, MER, and AXL. MET is only moderately correlated, and EGFR has low correlation.
Indeed, very few kinases are correlated with EGFR; of the test panel, only HER4 and HER2 are strongly correlated, while HER3, a few TKs, and the less related RIPK2 and GAK kinases show moderate correlation similarity.
PCA transformation of the Ambit 2011 dataset, highlighting Alk, Met, and EGFR kinases. Of the first three dimensions of the principal component transformation of the dataset, prinicpal component #3 clearly distinguishes EGFR and variants from the other kinases, while ALK, MET, and EGFR all are similar with respect to PC #1. PC #2 distinguishes the T790M mutants from the other EGFR forms
A third measure of similarity is principle component analysis [, ] of multiple target-multiple inhibitor binding matrices (see “Methods”). Applied to the 2011 Ambit study [], the protein kinase targets form a broad cluster, extended along the dimension of the first principal component (Fig. 4). Considering the PCA axes to represent “pseudo-inhibitors” as described above, there is a roughly Gaussian distribution for a majority of kinases around a value representing weak to moderate binding to “pseudo-inhibitor 1”, with some kinases in a skewed distribution toward tight binding. In order to maximize the variance in the first coordinate, the PCA method has constructed a “pseudo-inhibitor” that combines tight binding for the largest number of kinases possible (the skewed distribution). This favors the selection of targets which may be inhibited tightly by many inhibitors in the dataset, i.e. “generic” targets with high propensity for inhibition.
ALK, MET, and EGFR are near the middle of the distribution in PC coordinate #1. The second PCA dimension similarly spreads the kinases into a moderately inhibited cluster, skewed toward a smaller set of kinases that are tightly inhibited; here,this includes EGFR and several of its variants, but not the two T790M mutants. The third PC dimension clearly separates EGFR and all of its mutants from the rest of the kinases, including ALK and MET. Thus, the activity homology (AH) data, the distributions of inhibitor correlation data on kinome plots, and PCA analysis of the inhibition data all show the statistical similarity of ALK and MET, the dissimilarity of EGFR, and an intermediate similarity for EGFR-T790M. The PCA analysis also identifies the inhibitors responsible for the clustering of EGFR and mutants away from the other kinases (see discussion below).
## Heterogeneity of crystal structures obscures polypharmacology prediction
*In the **Heterogeneity of crystal structures obscures polypharmacology prediction** section:*
Because the structures of the target proteins determine the binding strengths of the ligands, it is natural to expect crystal structures to reveal target similarities and to aid the formulation of polypharmacological targeting strategies. Accordingly, considering the previous section, the ALK and MET ATP binding sites would be expected to appear similar to each other than to EGFR, with the EGFR-T790M mutants somewhere in between. However, examination of the crystal structures available for these targets reveals more the difficulties of structure based prediction of comparative binding strengths than mechanisms for doing so. This is due in large part to the structural flexibility and plasticity of proteins, leading to signficant ligand induced structural changes, but also derives from structural distributions that are affected by the crystallization constructs, conditions and crystal packing arrangements. Considering all structures available for the targets illustrates this.[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
The ALK structures are most homogeneous set; they superimpose with an average Cα RMS of 0.11Å, have the same essential crystal packing arrangement and, with one exception (4FNY), share an active “DFG-in” conformation of the activation loop (with the DFG phenylalanine in its hydrophobic spine position []). The activation loop (A-loop) adopts a unique conformation, however: The A-loop phosphorylation site Tyr1278 is anchored below the “C-helix” on the side opposite to the ATP pocket, analogous to active TK structures first observed for insulin receptor kinase [, ], but in ALK with a unique alpha-helical secondary structure. The exceptional ALK DFG-out conformation structure shares the crystal packing arrangement of the other structures, but at a lower symmetry, with the asymmetric unit comprising what was a crystallographic symmetry related pair in the other structures.[](https://www.ncbi.nlm.nih.gov/mesh/C553185)
The greatest number of structures is available for EGFR (more than 100 PDB entries when including disease-related and other mutations). The largest group of these structures share an active conformation, whereby pairs of monomers related by a crystallographic three-fold symmetry operation (space group I23) form an “asymmetric dimer” that represents a structural model of the active form []. Other EGFR structures show variations in C-helix positions (in = active, out = inactive); one set of structures (e.g. 2JIU) has an asymmetric unit consisting of a dimer with an apparently active geometry. There are no observations of a “DFG-out” geometry among the EGFR structures. However, there are two clusters of conformations: one with the usual active DFG-in conformation that places the DFG Phe into the hydrophobic spine, and another conformation with altered main chain angles and position of the activation loop.[](https://www.ncbi.nlm.nih.gov/mesh/C553185)
The MET structures show the greatest conformational diversity. Of the three kinases studied here, MET has crystallized in the largest number of space groups, and the N- and C-lobe “open-closed” variations are largest. The DFG states observed include DFG-in, DFG-out, and intermediate states. The C-helix is seen in both “in” and “out” geometries. The activation loop conformations are highly varied, including many that could not be resolved in the crystal structures.
Many of the variations also involve inhibitor interactions, and one common inhibitor binding surface is quite conserved as aromatic, but is seen formed variously by three different residues with four or more distinct geometries.[](https://www.ncbi.nlm.nih.gov/mesh/C553185)
## Do crizotinib co-crystal structures explain cross-reactivity and reveal ALK and MET similarities?
*In the **Do crizotinib co-crystal structures explain cross-reactivity and reveal ALK and MET similarities?** section:*
Superposition of structures of crizotinib in complexes with ALK (PDB: 2XP2; orange/brick) and MET (PDB: 2WGJ, violet/indigo). Side chains within a contact distance of 4 Å are shown as sticks, while main chain hydrogen bonding contact atoms are shown as small spheres. A dashed line indicates the approximate position of the disordered glycine-rich loop of Alk. The side chains of the activation loop phosphorylation sites are widely separated, with Tyr-1230 of MET in a pi–pi interaction with crizotinib, and Tyr-1278 of ALK anchored away from the ATP pocket by a helical conformation of the activation loop. The structure of crizotinib in complex with ALK mutant L1196M is similar to 2XP2, excepting the gatekeeper mutation[](https://www.ncbi.nlm.nih.gov/mesh/C551994)
Even if flexibility prevents reliable prediction, it seems reasonable to expect that the structures of cross-reactive inhibitors in their different targets would identify the basis for the cross-reactivity. This would enable structure based design of e.g. an inhibitor library focussed on the likelihood of cross-reactivity. For ALK and MET cross reactivity, the low nanomolar inhibition of both ALK and MET by crizotinib is likely the clearest and best known measure of similarity between the two targets [] Besides ALK (3 nM) and MET (2 nM), crizotinib also shows single digit nM binding (KD) to protein kinases ROS1 (4 nM), MER (4 nM), EPHB6 (6 nM), and AXL (8 nM) [], depending of course on assay conditions.
Crystal structures of crizotinib in protein kinases in the PDB comprise 2WGJ (c-MET KD), 2XP2 (ALK-KD), 2YFX (L1196M), 3ZBF (ROS1), 4ANQ (G1269A), 4ANS (L1196M, G1269A), and 4C9W (NUDT1). (The S-stereoisomer of crizotinib is co-crystallized with NUDT1 also). Superposition of co-crystal structures of crizotinib with ALK (2XP2 []) and MET (2WGJ []) reveals more how the binding energies that correspond to the highly selective and nanomolar ALK and MET co-inhibition depend on interactions that are not readily identified with standard structural biology or bioinformatic methods (Fig. 5). Interacting side chains differ at many key sites, including: the residues that sandwich the adenine binding site (MET vs. Leu from the C-lobe, and Leu vs. Ile from the glycine-rich loop), the gatekeeper+2 (the site two residues C-terminal to the gatekeeper residue) aromatic/hydrophobic side chain at the hinge (Tyr vs Leu), and a key pi–pi stacking interaction with the activation loop phosphorylation site tyrosine that is observed only in the MET structure. Amino acid type specific crizotinib interactions that are shared between ALK and MET comprise the gatekeeper (Leu), the C-terminal ATP site anchor of the glycine-rich loop (Val), an alanine residue two positions N-terminal to the active site lysine, and a pyrazole–proton interaction at a gatekeeper+6 glycine residue.
The importance of the gatekeeper and pyrazole-glycine interactions are highlighted by the occurrence of resistance via mutation at these sites []. While the shared interactions are consistent with the observed co-inhibition, these residues are highly conserved in the kinome, so that these interactions are not predictive of the selectivity of the co-inhibition. Other shared but non-residue specific interactions include the anchoring hydrogen bonds at the hinge, and a key interaction with a main chain carbonyl group. The latter forms a CH–O hydrogen bond from an aromatic ring hydrogen that has been polarized by the fluorine substituent at the adjacent site on the ring []. The pi–pi stacking interaction with Tyr1230 appears important for MET inhibition, but no comparable interaction is seen in the ALK structure.
This difference was proposed to account in part for the tighter binding of crizotinib to MET vs ALK, along with differences of the backbone peptide orientation at G1269 (ALK) and A1221 (MET) []. Crizotinib binding to the MET mutant Y1230C is weakened 15-fold in a cellular assay [], supporting the view that the pi stacking interaction in important in vivo. Taken together, these details indicate that the crystal structures would not have allowed the prediction of high affinity cross-reactivity, but that prior knowledge of the cross-reactivity enables the study of the structures to identify the key candidate interactions responsible for it.[](https://www.ncbi.nlm.nih.gov/mesh/C551994)
## Do the crystal structures in general reveal polypharmacological potential?
*In the **Do the crystal structures in general reveal polypharmacological potential?** section:*
The importance of the CH–O, CH-arene, and arene–arene interactions for the total binding energetics (and, ultimately, therapeutic properties) of crizotinib is not entirely clear, but they highlight the scoring and searching dilemma that prevents in silico methods to predict ligand binding properties from target structures: The level of theory and concomitant CPU power required for computation of such binding features to provide effective scoring prevents their prediction a priori if binding geometries are unknown. (For example, the strength of the pyrazole–proton interaction described above depends on the electron-richness of the pyrazole and on energetic penalties of rotation away from the energy minimum, phenomena requiring quantum mechanics level calculations for evaluation []). Comparisons across all PDB structures provide some measure of the range of structural variability, but do not enable the calculation of binding energies, while inhibitor binding studies provide averaged binding energies averaged over structural distributions in the assay environmental conditions. In the PDB, there are currently 51 MET and 36 ALK structures; of these, only 3 MET structures are of the phosphorylated protein. For crizotinib, both MET and ALK were nonphosphorylated forms of truncated kinase domains, and the co-crystal structures showed inactive geometries.
(MET has significant residual activity when unphosphorylated, and is activated 160-fold upon autophosphorylation []). For receptor TKs, phosphorylation often occurs autocatalytically in trans as a consquence of ligand binding to extracellular domains and oligomerization []. The mechanism of activation of tyrosine kinases by phosphorylation of the activation loop is usually considered to involve destabilization of possibly rigid inactive states of the kinase.[](https://www.ncbi.nlm.nih.gov/mesh/D006841)
Structural variabilities and activation states of the targets. a A cartoon style plot showing the activation loops of representative crystal structures for MET (green PDBIDs 1R0P, 3DKC, 3F66, 3L8V, 3Q6W and 3RHK) and ALK (red both monomers from PDBID 4CNH) shows the diversity of positions for the tyrosine side chain suitable for crizotinib pi-stacking interactions (MET: Tyr1230, aligned position in ALK: Tyr1278). The ALK structures are two monomers from a single PDB structure that show different activation loop geometries. b Different shapes of the ATP binding sites arise from activation loop structural differences, among other differences. Here, structures for crizotinib-ALK (red PDBID: 2YFX), a pyrido-pyrimidine inhibitor and MET (violet PDBID: 4GG7), and ARQ197-MET (turquoise PDBID: 3RHK) are superimposed. c Distributions of activation relevant geometries for PDB structures of ALK, MET, and EGFR. The axes represent the first two dimensions of a PLS analysis to identify the geometric parameters most relevant to identify active and inactive states (see text)[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
Oncogenic activation via mutation or fusion often disrupts inactivating geometries. Partially as a result of this, and partially due to their inherent plasticity, protein kinase cancer drug target structures are highly flexible; prioritization for in silico drug discovery purposes may be difficult []. Superposition of ALK and MET structures (Fig. 6) from the PDB illustrate this. The MET structures show great diversity in conformations of the activation loop, and the tyrosine that is involved in pi–pi stacking interactions with crizotinib (Tyr1035) is found distributed across the entire space accessible to the activation loop (Fig. 6a). The structures for ALK are more homogeneous, and cluster into one major group and one minor group.
Most ALK structures show a helical conformation for the activation loop following the DFG sequence, anchored to the C-helix by two arginine residues flanking the conserved glutamic acid of the C-helix, and by hydrophobic and aromatic interactions involving Tyr1278 (Fig. 6a). For the exceptional geometry, the activation loop retains a helical conformation and salt bridge interactions with helix C, but the Tyr1278 anchoring is lost, and the helix is rotated away from the C-helix. Prediction of ATP site binding will usually depend critically on the choice of the “correct” target structure (Fig. 6b), but if the “correct” structure is induced by inhibitor, it will obviously not be available for a priori searches. Statistical analysis of the protein kinases currently in the PDB provides an excellent illustration of the structural diversity currently observed to date []. PLS analysis [] to identify the geometric measures that most strongly differentiate active and inactive geometries [], and cluster crystal structures accordingly, shows how the structures for ALK, MET, and EGFR are distributed between apparent activity states.
The MET structures are mostly inactive and broadly distributed, the ALK structures are mostly active (or close to it), and the EGFR structures are both active and inactive (Fig. 6c). For ALK, the position on the horizontal axis (with the coordinate definition dominated by DFG related geometries) does not clearly mark them as active. However, the vertical axis and its inclusion of helix C position parameters, coupled with “nearly active” DFG geometries, clusters ALK together with the active group.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
Variability of aromatic side chain positions. a A view (from the N-lobe side of the ATP pocket) showing clustering of aromatic amino acids. In MET structures, Phe1223 (green) of the “DFG” motif clusters into the DFG-in or DFG-out positions, with an intermediate position also represented. Phe1089 (cyan) from the glycine-rich loop of MET usually clusters near the tip of an extended glycine-rich loop, but is also seen in several structures to occupy nearly exactly the position adopted by the DFG Phe1223 in protein DFG-out configurations. Tyr1230, when interacting with inhibitors, forms a tight cluster at a single position, but is absent when Phe1089 or Phe1223 occupies an adjacent site. The PDBID codes for these structures are 2RFN, 2RFS, 2WD1, 2WGJ, 2WKM, 3A4P, 3C1X, 3CCN, 3CD8, 3CE3, 3CTH, 3CTJ, 3DKF, 3DKG, 3EFK, 3F66, 3F82, 3I5N, 3L8V, 3LQ8, 3Q6W, 3QTI, 3R7O, 3RHK, 3U6H, 3U6I, 3ZXZ, 3ZZE, 4AOI, 4AP7, 4DEG, 4DEH, 4DEI, 4EEV, 4GG5, 4GG7, and 2YFX. b Inhibitor types cluster according to the clustering of the target structures. Generally, type II inhibitors (cyan), with some diversity of chemical scaffolds, often bind in part via packing against Phe1223 in its typical DFG-out position, while type I inhibitors (violet) often bind via pi–pi stacking interactions against Tyr1230. The exceptional geometry of both protein and inhibitor for Arq197 (yellow) is apparent. The PDBID codes for these structures are: 2RFN, 2RFS, 2WD1, 2WGJ, 2WKM, 3A4P, 3C1X, 3CCN, 3CD8, 3CE3, 3CTH, 3CTJ, 3DKF, 3EFJ, 3EFK, 3F66, 3F82, 3I5N, 3L8V, 3LQ8, 3QTI, 3RHK, 3VW8, 3ZXZ, 3ZZE, 4AOI, 4AP7, 4DEG, 4DEH, 4DEI, 4EEV, 4GG5, and 4GG7. c The pyrimidone inhibitor of MET structure 3EFJ binds via pi–pi stacking interactions, but the dimer resolved in the crystal shows that the protein can provide the partner for this interactions with two different side chains, with identical inhibitor binding geometries[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
A closer look at the structural diversity reveals several interesting phenomena. One is the clustering of aromatic side chains at the ATP pocket (Fig. 7). Many of the MET structures show nearly identical positions for Tyr1230, similar to the geometry seen in the crizotinib complex. These are compatible with DFG-in geometries. Standard DFG-out geometries do not allow Tyr1230 to occupy that space, but replace it nearly exactly with Phe1223 of DFG, in place for inhibitor packing interactions.
As a third alternative, this space may be occupied by the aromatic Phe1089 from the glycine-rich loop, represented by a small cluster of three structures in this superposition. There is also a unique positions for the DFG and glycine-rich loop positions. Inhibitor types are associated with the clusters of arenes at this site (Fig. 7b). Interactions with the DFG Phe in the DFG-out configurations are dominated by single aromatic rings from Type II inhibitors that occupy the deep pocket (vacated by the DFG-out Phe). Interactions with Tyr1230 of the activation loop involve a small variety of arenes: several halogenated or nitrated phenyl groups and a larger number of fused 5- and 6-membered heteroatomic aromatic ring systems, mostly in the same space.
(One structure is displaced (3ZZE), but pi interactions may be maintained via resonance across nitrogen and amide bond linkages). One structure is unique: the complex with ARQ197 (PDBID: 3RHK []) shows a fused three-ring system sandwiched between Phe1089 and Phe1223 in unique positions. Finally, significant structural variation may be observed within a single crystal structure. The two monomers of the MET kinase domain in a crystal structure of a complex with a pyrimidone containing type II inhibitor (PDBID: 3EFJ []) shows that group to interact with the protein via pi–pi stacking interactions, whereby the interacting partner is the DFG Phe1223 for one monomer, but is the glycine-rich loop Phe1089 for the other (Fig.
7c).[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
Study of the variabilities shown by the crystal structures reveal physiologically relevant properties of the individual proteins and ligands studied, but their interpretation with respect to therapeutic properties requires much more experimental information and careful analyses of the differing environments in crystallo and in vivo. For protein kinases, crystal structures are commonly truncated single domain proteins with specific phosphorylation states, and the structures of flexible elements such as the activation loop may be determined by crystal packing interactions. In contrast, the disease targets are usually larger, multidomain proteins, often in larger assemblies, and with heterogeneities of chemical modifications and cellular environments. Highly potent inhibitors can compete with many of these effects, such that co-crystal structures generally reveal the key target-inhibitor interactions faithfully. But the crystal structures may also capture both the protein and ligand in states that are unimportant for therapeutic properties.
One uncertainty for protein kinase co-crystallography is the activation state of the enzyme. Crizotinib binding in MET described above involves pi–pi stacking with Tyr1230. However, the three activated MET structures in the PDB, phosphorylated on Tyr1235, show an activation loop structure with Tyr1230 far removed from the ATP binding pocket. The apparent structural homogeneity of ALK is also deceptive. The oncologically relevant forms of ALK are most commonly fusion proteins
[, ] that remove membrane attachment and render ALK constitutively active. Dimer- or oligomerization is thereby essential for cell transformation, but the details of the structures and mechanism of activation are not known []. Mutations that confer resistance to crizotinib [] include several that may destabilize the intramolecular A-loop helix packing that is apparently part of the inactivation mechanism for ALK []. The helix packs most prominently against the C-helix, with an Arg + Glu + Arg triplet, likely to stabilize a helical conformation, slotting into a space between two Glu residues extending from adjacent turns of the C-helix (of these, one is the usual partner to the active site lysine of active protein kinase structures. In addition, the phosphorylation site Tyr1278, which is adjacent to the anchoring Arg at the terminus of the activation loop helix, contributes to anchoring the helix via an edge-face pi–pi stacking interaction with Tyr1096.T his residue is found in a sequence N-terminal to the ALK kinase domain.
The proximity of this anchor to the fusion position for the activation fusions with e.g. EML4- or NPM-suggest that the fusion may activate the protein by weakening the inactivating AL-helix interactions. For MET as a target in NSCLC, it is the wt protein which is of greatest relevance, although NSCLC related MET mutations have also been observed [].[](https://www.ncbi.nlm.nih.gov/mesh/C551994)
Summarizing the structures analysed here, we have seen most significantly: (1) that the diversity of activation forms of ALK, MET, and EGFR show how crystal structures cluster according to successful crystallization conditions, which is difficult to relate to the distribution of structures that are most relevant physiologically or biochemically (for in vitro binding measurements), (2) that knowledge of the cross-reactivity of crizotinib is a prerequisite for identifying key binding interactions, due to the divergence of sequences at the binding site, and (3) that these are most likely special interactions of arene groups and polarized bonds which would be overlooked by simplified molecular mechanics methods that are developed for rapid in silico approaches.[](https://www.ncbi.nlm.nih.gov/mesh/C551994)
## How can cheminformatics inform crystallography?
*In the **How can cheminformatics inform crystallography?** section:*
It is clear that structures need to be interpreted with respect to binding data. Inhibition and/or binding data [, ] (including variation of ATP concentration and by single site mutation) are available from a variety of protein forms and assay formats [, , , , –]. Crizotinib binds to an inactive conformation of MET []. These data show insensitivity to phosphorylation in Abl, expected for type I inhibitors []).
It is however affected by resistance mutations, whereby the T315I mutant of Abl is most tightly bound (ca 10 nM), and the gatekeeper+2 mutations F317I or F317L are most weakly inhibited (3000 and 600 nM, respectively). This sensitivity is interesting for considerations of ALK and MET binding, as described above. It should be readily appreciated that prediction of the ALK-MET cross-reactivity by automated bioinformatics or structural analysis methods seems highly unlikely. Whether more general machine-learning approaches could do so, presumably by identifying underlying and subtly interlinked selectivity determinants without recourse to model assumptions, remains to be seen. In any case, it will not be a competition between cheminformatics and crystallography, but will involve the integration of structural data into informatics techniques.[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
Does methionine as gatekeeper correlate generally with the selectivity properties of EGFR inhibitors? a Stereo plot of the first three PCA dimensions of the protein kinase inhibition data of the Ambit panel of 2011 (Fig. 4), with the protein kinases colored according to gatekeeper (red methionine, gray not methionine). EGFR and mutants are distinguished from the other kinases principally by PC axis 3, while the two gatekeeper mutant (T790M) forms of EGFR are distinguished from the other EGFR forms by PC axis 2. b The inhibitors contributing to the composition of PC axes 2 and 3 (loading plot) highlight the inhibitors that are potent for EGFR and most potent for the T790M mutations (upper left), those that are potent for EGFR but less potent for the T790M mutations (upper right), and inhibitors that are less potent or antiselective for EGFR (lower two quadrants)[](https://www.ncbi.nlm.nih.gov/mesh/D008715)
Many statistical questions simply identify correlations, which may be recognized even with relatively sparse datasets. These may then generate hypotheses for more detailed study. The principal component analysis of the Ambit kinase and inhibitor panel of 2011 [] generated transformed coordinates that indicated unique clustering for both EGFR and its drug resistant mutant T790M (Figs. 4, 8). The coordinate transformation highlights the inhibitors especially responsible for the unique position of EGFR. Early profiling data already indicated that EGFR could be targeted with unique selectivity [].
The PCA transform of the 2011 data, and in particular the 2nd and 3rd dimensions, identifies inhibitors with particularly notable EGFR inhibition properties. PC dimension #2 (corresponding to a pseudo-inhibitor, see “Methods”), which separates EGFR T790M variants from the other EGFR mutants, also “bundles” selectivity determinants into the corresponding pseudo-inhibitor that have broad applicability to the rest of the kinome. One of these is most obviously the occurrence of methionine as gatekeeper, and the PCA plot shows the enrichment of protein kinase targets with this gatekeeper in the direction of the displacement of EGFR T790M variants relative to the remaining EGFR cluster. PC dimension #3, with less total variance than PC #2, has its clearest source of variance with the separation of all the EGFR targets. The “loading plot” for PC #2 and #3 (Fig. 8b) shows the inhibitors mostly responsible for the identification of these properties, and include highly selective EGFR inhibitors such as HKI-272, BIBW-2992, etc. (upper quadrants of Fig. 8b), and also antiselective inhibitors such as sorafenib (lower quadrants).
Similarly, inhibitors that bind the native EGFR sequence preferentially (right-most quadrants, e.g. dasatinib) or the T790M mutant (left quadrants, e.g. staurosporine) are identified by PC dimension #2.[](https://www.ncbi.nlm.nih.gov/mesh/D008715)
## Focussing libraries toward ALK + MET + EGFR polypharmacological inhibition
*In the **Focussing libraries toward ALK + MET + EGFR polypharmacological inhibition** section:*
The analyses of this work reveal statistical target similarities between Alk and Met, along with a fundamental dissimilarity of EGFR, and an intermediate position for the drug resistant EGFR-T790M mutant. They also show however how distributions of structural variations, the importance of subtle chemical interactions, and mismatches between the systems used to generate experimental binding data prevent direct design of an inhibitor with the desired polypharmacological selectivity profile. As a consequence, the design goal is to create a library of test compounds with the greatest likelihood of having the target properties. For ALK, MET, and EGFR, the challenge is to achieve cross-reactivity despite the dissimilarity of EGFR. (The dissimilarity is shown statistically, as in Fig.
2, which also shows the existence of some overlap between potent inhibitor sets for ALK or MET and EGFR. One example inhibitor is brigatinib []). The solution is simple: use the known cross-reactivity of ALK and MET, which we may consider to be based on the “shape” of the respective ATP pockets, and add covalent trapping groups at sites with a good potential for reaction with the cysteine at the gatekeeper+7 site in EGFR.[](https://www.ncbi.nlm.nih.gov/mesh/C000598580)
a Overlay of ALK (red PDBID: 3LCS), MET (blue PDBID: 2WKM), and EGFR (magenta PDBID: 3IKA) showing the relative positions of the tricycle inhibitor (here, from WZ4002 as stick model) and gatekeeper+7 targetable cysteine residue of EGFR. b Surface plot to show the relative positions of the gatekeeper+7 cysteine of EGFR (magenta surface) and crizotinib (modelled by superposition of the crizotinib-MET complex (PDBID: 2WGJ) with EGFR[](https://www.ncbi.nlm.nih.gov/mesh/C571455)
Although the unusual nature of EGFR-inhibitor interactions revealed by the cheminformatics above does not depend on its cysteine at the gatekeeper+7 position, this cysteine does provide an ideal mechanism [] for a high affinity interaction type that is essentially decoupled or “orthogonal” to shape-based similarity that characterizes ALK and MET. Thus, the detailed strategy to focus a library for ALK + MET + EGFR polypharmacological inhibition is first to identify compound classes that provide ALK + MET co-inhibition, and then to select scaffolds from these that in addition allow modification to target covalent inhibition of EGFR. The strength of the EGFR interaction need not be high, but would have to satisfy geometric and dynamic requirements for covalent binding. Many examples of suitable scaffolds have been published, and superpositions of the targets and relevant inhibitors (Fig. 9) show the viability of the approach.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Candidate chemotypes for orthogonal EGFR covalent inhibition, prioritized based on the binding data of Abbott []. Values for individual inhibitors are plotted according to ALK and MET binding strengths, with chemotypes indicated by symbol (and defined for the tightest binders) and EGFR binding strengths indicated by color (red = 1 nM, violet = 10 nM, blue = 100 nM). The complete structure of the inhibitor for which the data point is plotted, is disclosed in the analysis [], and the corresponding substituents are depicted for this chemotype in the figure at lower saturation
Crizotinib itself may be considered for this purpose, but it does not inhibit EGFR [] (although is does bind the EGFR G719C mutant at micromolar levels). Other candidate scaffolds include for example (Fig. 10) staurosporine, bisindolylmaleimides, 4‐{2‐phenylimidazo [1,2‐a] pyrazin‐3‐yl} pyrimidine, and 3‐phenyl‐1‐(4‐{thieno[3,2‐c]pyridin‐3‐yl} phenyl) urea. Examination of the literature on candidate scaffolds and their suitability for ALK + MET + EGFR polypharmacology highlights especially the tricyclic scaffold found in the covalent EGFR inhibitor WZ4002
[] (PDBID: 3IKA) and other derivatives known for covalent EGFR inhibition []. WZ4002 is known to bind both ALK [] and MET [], and the core dianilino-pyrimidine kinase binding scaffold also shows ALK and MET inhibition (Fig. 10) in other compounds []. This scaffold is well known in industry, including use as in connection with acrylamide groups for covalent binding, and a substructure search returns well over 104 compounds from patent literature.
This need not hinder further research, however, because the earliest patents have expired or are due to expire soon (for example, methoxylated forms of (2,4) dianilino 5-chloropyrimidine were patented with a priority date of 1995 []).[](https://www.ncbi.nlm.nih.gov/mesh/C551994)
Tested compounds
To test the suitability of dianilino-pyrimidine kinase binding scaffolds for ALK + MET + EGFR polypharmacology, we profiled three compounds representing the basic scaffold (including chlorine as the gatekeeper interacting atom: 5-chloro-N2,N4-diphenylpyrimidine-2,4-diamine), with additional single acrylamide functional groups as substituents on each of the candidate aromatic rings. Acrylamide substitution on the N4 phenyl group (at the meta position) places the covalent trap analogous to its position in WZ4002. Acrylamide substitution on the N2 phenyl moiety (also at the meta position) places the covalent trap at a position potentially suitable for a covalent trap to the gatekeeper+7 site in EGFR; varying the linker to the acrylamide function allows for uncertainties regarding optimal geometries and protein plasticity (Fig. 11; Additional file 1).[](https://www.ncbi.nlm.nih.gov/mesh/D002713)
Alk, Met, and EGFR binding properties of the test tricyclic compounds
Tests of the compounds confirm suitability for ALK + MET + EGFR polypharmacology optimization (Table 1). For the N2 phenyl ring substituted compounds 2a 1 and 2b, 2 Kd values as measured by KdELECT assays (DiscoverX) show submicromolar binding for ALK, MET and both tested forms of EGFR. For compound 1, 3 with the acrylamide function at the site corresponding to that of WZ4002, both ALK and MET binding were inhibited more weakly compared to EGFR binding. Retesting the compounds under scanKINETIC assay conditions for the two EGFR targets showed Kd values that were considerably tighter than in the KdELECT assays, approximately 2-fold for compound 2a, and 4-7-fold for compounds 1 and 2b. The sensitivity to the assay conditions seen for both forms of EGFR and uniformly tighter binding under the new assay conditions may indicate tighter binding for ALK and MET under these altered assay conditions as well.[](https://www.ncbi.nlm.nih.gov/mesh/D000178)
The kinetics testing results (Table 1) support several further important conclusions. First, comparison of the the apparent binding constants comparing 1-h and 6-h incubation times show that compounds 1 and 2a are relatively slow-binding, while binding of 2b is complete within 1 h. Second, comparison of the apparent binding constants after 30-fold dilution following one-hour of incubation times showed slow dissociation behavior for compounds 1 and 2a, and complete dissociation of 2b within 5 h. The simultaneous appearance of slow association and slow dissociation complicates the interpretation of the results, but the percentage values shown in Table 1 show the apparent residual amounts of compounds relative to bound values at 1 h. These values indicate retention of roughly half of compound 1, and nearly all of compound 2a, after dilution. Covalent binding is the most obvious interpretation of these data.
In contrast, compound 2b follows fast association and dissociation kinetics, with no evidence of covalent binding. The similarity of compound 2a to 2b and 1 with respect to structure and binding strength, differing only in the linker to the acrylamide binding group, is further evidence that that it is the covalent binding property that determines the difference in binding kinetics, rather than e.g. a slow conformational change of the target enzymes.[](https://www.ncbi.nlm.nih.gov/mesh/D000178)
The variations in properties of compounds 1, 2a and 2b indicate diversity of their potential application as basic scaffolds for generating libraries suitable for polypharmacological targeting of Alk, Met, and EGFR. Compound 1, with analog WZ4002 known to bind covalently to EGFR, has an intrinsic selectivity for EGFR, but still may be suitable for Alk-Met-EGFR polypharmacology, with suitable decoration. Compounds 2a and 2b, on the other hand, show greater affinity for Alk and Met, and good affinities for both EGFR variants. Both 2a and 2b are thus good candidates for optimization. However, because 2a and apparently not 2b is able to bind covalently to EGFR (both forms), the 2a scaffold seems most likely to provide good chances for polypharmacological optimization, as substituents may be chosen to optimize Alk and Met binding, using their intrinsically higher similarities, while maintaining EGFR binding via the covalent trap (providing high potency without the need for precise shape matching).[](https://www.ncbi.nlm.nih.gov/mesh/C571455)
## Conclusions
*In the **Conclusions** section:*
The chain of reasoning of this work began with the identification of a set of relevant key targets for a disease, here non small cell lung cancer (NSCLC), and narrowing the selection based on the extent of available data on compounds and targets. Targets validated with approved inhibitors include EGFR, its drug resistant gatekeeper mutant of T190M, as well as ALK and MET. Mutations or fusions of both EGFR and ALK may be primary causes of cancer, whereas both the EGFR gatekeeper mutant T190M and MET may generate drug resistance. Analysis of the relevant inhibitor binding data identified similarities between ALK and MET, and highlighted the essential dissimilarity of EGFR, with an intermediate position for the T790M mutant. A result of this analysis is the choice of a strategy for polypharmacology to optimize ALK and MET binding via shape and surface complementarity, while maintaining covalent binding to EGFR.
A suitable inhibitor scaffold was chosen, based solely on binding data. An accompanying analysis of the published crystal structures for the targets, aiming to guide optimization, highlighted mostly the intractability of mapping the inhibitor binding similarities to structural properites. This is due both to great structural variability of the targets and to the subtle nature of the essential interactions, which generally are not recognizable by molecular mechanics methods. However, the structures did highlight key details, such as the unusual binding interactions of crizotinib that are conserved between ALK and MET, despite differences in sequence, and also the binding mode of the dianilino-pyrimidine inhibitor scaffold. Thus, crystallography informs cheminformatics.
This polypharmacology targeting example, while highlighting specific characteristics of ALK, MET, and EGFR, may be generalizable to other kinase target profiles in several ways. The flexibility and large numbes of kinases means that similar intractabilities of structural analysis will occur, but also that other combinations of diverse selectivity mechanisms may be found for appropriate targeting. Automation of such approaches seem highly unlikely until these mechanisms may be catalogued in some machine-analysable form. Other non-kinase polypharmacology approaches will likely be quite different, possibly involving target profiles more diverse, more rigid, or otherwise distinct from protein kinases.[](https://www.ncbi.nlm.nih.gov/mesh/C551994)
## Methods
*In the **Methods** section:*
Activity homology data were taken from Posy et al. [], with the data for Alk, Met, EGFR and EGFR-L858R, T790M plotted with Excel (Fig. 2a), and as a heat map (Fig. 2b) with tree clustering performed by the Clustermap function of the Seaborn python package (DOI: 10.5281/zenodo.45133). This measure is defined as the percentage of potent inhibitors (those with an estimated IC50 < 150 nM) of a reference kinase “A” that are also potent inhibitors of the comparison kinase “B”. As defined, this is “the prior probability that a compound will be active for kinase B given that it is active for kinase A” [].
(Note that this defines a matrix which is not symmetric with respect to swapping the reference and compared kinases, because each will have a unique set of potent inhibitors.)
Kinome profiles of target similarity (Fig. 3) were evaluated using the data of Davis et al. [], with the nM inhibition values xi converted into logarithms to be proportional to binding energies for inhibition values tighter than the upper measurement limit of 10 micromolar, and arbitrarily set uniformly to 5 (corresponding to 100 micromolar) for values above the upper measurement limit. These values were used to calculate target similarity as Pearson’s correlation coefficient for the pairs of vectors of binding energy equivalents and plotted as disks at the kinase positions of the kinome plot of Manning et al. []. Hues and radii reflect the correlation values.
Principal components (Figs. 4, 8) were determined using Karhunen–Loeve Decomposition as implemented in Mathematica (version 10.3) with the binding energy equivalent values. This method of data analysis can be applied to an t · n matrix T ij that represents how the t targets are inhibited by n inhibitors (the targets {i} are thus t positions in an n-dimensional space of inhibitors {j}). The analysis determines the orthonormal linear transformation of the coordinate system that diagonalizes the t · t covariance or correlation matrices of the targets with respect to their inhibition profiles. This transformation into a new n dimensional space of virtual or pseudo-inhibitors {k} is done (usually with the {Ti,1, Ti,2,…} vectors normalized to unit variance and zero average) such that the variance is maximal for k = 1 (PCA dimension 1), second highest for k = 2 (PCA dimension 2), and so on. This reduces the redundancy of similarities between inhibitors, and clusters the targets in spaces whose dimensionality may be reduced according to the degree of variance desired “to explain the data”.
As the PCA transformation maximizes the spread in inhibition values by a “pseudo-inhibitor” created by the recombination of inhibitors of the dataset, inhibitors that are selective for subgroups of targets are most strongly represented in the relevant transformed coordinates, with the (mutually orthogonal) coordinates ranked (PC #1, PC #2, etc.) according to the total variance in that coordinate. The “loading plots” show the coefficients of the individual inhibitors in the “pseudo-inhibitor” principal components created by the transformation.
Superpositions of the ALK, MET, and EGFR protein kinase structures from the Protein Data Bank (PDB) (Additional file 2) were performed using PYMOL (version 1.7) and scripts that extracted protein kinase monomers from the entries. The scripts aligned the CA atoms from the gatekeeper+3 residue as hinge anchor position (1196–1199—ALK, 1158–1161—MET, 766–769 or 790–793—EGFR), along with aF helix atoms as the core of the C-lobe (1308–1324—ALK, 1262–1278—MET, 869-885 or 893–909—EGFR) residues.
SIMCA (v.13, Umetrics AB, Umeå, Sweden) was used to build a PLS regression plot [] (Fib 6c) by taking 233 parameters (distances, angles, dihedrals etc.) from the annotated dataset of Möbitz [] as independent variables and the designated active/inactive state of the kinase domains as the dependent variable. Thus, the PLS method transforms the dimensions of the individual structural parameters [] into orthogonal dimensions that maximize covariance of the transformed dimensions with the activity state of the kinase.
In the SIMCA implementation, a 7-fold cross validation technique prevents overfitting in the estimate of the number of significant components. The variables were scaled by unit variance. The PLS model built from this dataset has 4-components with both R2 (goodness of fit, maximum 1) and Q2 (predictive ability, maximum 1) values more than 0.9, where 1st and 2nd components account for about 80 and 6% of the variations respectively.
The three tricyclic compounds (1, 2a and 2b) were purchased from the company ARTTSynthesis (www.arttsynthesis.com). Tests for binding properties were performed by the company DiscoverX (www.discoverx.com), using a concentration of 1 µM against ALK, MET and EGFR variants L858R and L858R_T790M) using KdELECT and scanKINETIC from DiscoverX [, ].
## Additional file
*In the **Additional file** section:*
# Abbreviations
*In the **Abbreviations** section:*
PK
protein kinases; this work cites ABL, ALK, AXL, EGFR, EPHB2, FAK, FER, FES, GAK, HER2, IGF1R, INSR, IRR, LTK, MER, MET, PYK2, ROS1, and RIPK2
EML4
echinoderm microtubule-associated protein-like4
HGF
hepatocyte growth factor
NPM
nucleophosmin
NUDT1
hydix hydrolase 1
TK
tyrosine kinase (protein kinase group)
TKL
tyrosine kinase like (protein kinase group)
AH
activity homology
ATP[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
adenosine triphosphate[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
CML
chronic myelogenous leukemia
DFG[](https://www.ncbi.nlm.nih.gov/mesh/C553185)
Asp-Phe-Gly initiating sequence of the activation loop[](https://www.ncbi.nlm.nih.gov/mesh/C553185)
NSCLC
nonsmall cell lung carcinoma
PC
principal component
PCA
principal component analysis
AL or A-loop
activation loop
PLS
projection on to latent spaces, or equivalently partial least squares
PDBID
Protein Data Bank ID code
N‐(3‐{[5‐chloro‐4‐(phenylamino)pyrimidin‐2‐yl]amino}phenyl)prop‐2‐enamide.[](https://www.ncbi.nlm.nih.gov/mesh/D006841)
N‐[(3‐{[5‐chloro‐4‐(phenylamino)pyrimidin‐2‐yl]amino}phenyl)methyl]prop‐2‐enamide.[](https://www.ncbi.nlm.nih.gov/mesh/D006841)
N‐(3‐{[5‐chloro‐2‐(phenylamino)pyrimidin‐4‐yl]amino}phenyl)prop‐2‐enamide.[](https://www.ncbi.nlm.nih.gov/mesh/D006841)
Electronic supplementary material
The online version of this article (doi:10.1186/s13321-017-0229-8) contains supplementary material, which is available to authorized users.
Dilip Narayanan and Osman A. B. S. M. Gani contributed equally to this work
# References
*In the **References** section:*
|
# Introduction
Deconvoluting AMP-activated protein kinase (AMPK) [adenine nucleotide](https://www.ncbi.nlm.nih.gov/mesh/D000227) binding and sensing
# Abstract
*In the **Abstract** section:*
AMP-activated protein kinase (AMPK) is a central cellular energy sensor that adapts metabolism and growth to the energy state of the cell. AMPK senses the ratio of adenine nucleotides (adenylate energy charge) by competitive binding of AMP, ADP, and ATP[ to three sites (CB](https://www.ncbi.nlm.nih.gov/mesh/D000227)S1, CBS3, and CBS4) in its γ-subunit. Because these t[hre](https://www.ncbi.nlm.nih.gov/mesh/D000249)e [bin](https://www.ncbi.nlm.nih.gov/mesh/D000244)ding s[ite](https://www.ncbi.nlm.nih.gov/mesh/D000255)s are functionally interconnected, it remains unclear how nucleotides bind to individual sites, which nucleotides occupy each site under physiologi[cal conditi](https://www.ncbi.nlm.nih.gov/mesh/D009711)ons, and how binding to one site [affects bin](https://www.ncbi.nlm.nih.gov/mesh/D009711)ding to the other sites. Here, we comprehensively analyze nucleotide binding to wild-type and mutant AMPK protein complexes by quantitative competi[tion assay](https://www.ncbi.nlm.nih.gov/mesh/D009711)s and by hydrogen-deuterium exchange MS. We also demonstrate that NADPH, in addition to the known [AMPK lig](https://www.ncbi.nlm.nih.gov/mesh/D006859)a[nd NADH, ](https://www.ncbi.nlm.nih.gov/mesh/D003903)directly and competitively binds AMPK a[t the](https://www.ncbi.nlm.nih.gov/mesh/D009249) AMP-sensing CBS3 site. Our findings re[veal](https://www.ncbi.nlm.nih.gov/mesh/D009243) how AMP binding to one site affects the confor[mat](https://www.ncbi.nlm.nih.gov/mesh/D000249)ion and adenine nucleotide binding at the ot[her](https://www.ncbi.nlm.nih.gov/mesh/D000249) two sites and establish CBS3, and not CBS1, as th[e high affinity ex](https://www.ncbi.nlm.nih.gov/mesh/D000227)changeable AMP/ADP/ATP-binding site. We further show that AMP binding at CBS4 increases AMP binding [at ](https://www.ncbi.nlm.nih.gov/mesh/D000249)C[BS3](https://www.ncbi.nlm.nih.gov/mesh/D000244) [by ](https://www.ncbi.nlm.nih.gov/mesh/D000255)2 orders of magnitude and reverses t[he ](https://www.ncbi.nlm.nih.gov/mesh/D000249)AMP/ATP preference of CBS3.[ To](https://www.ncbi.nlm.nih.gov/mesh/D000249)gether, these results illustrate how the three CBS sites co[lla](https://www.ncbi.nlm.nih.gov/mesh/D000249)b[ora](https://www.ncbi.nlm.nih.gov/mesh/D000255)te to enable highly sensitive detection of cellular energy states to maintain the tight ATP homeostastis required for cellular metabolism.[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
ATP homeostasis is critically important for cellular functions and is maintained at a tight concentration range by the regulatory protein kinase AMP-activated protein kinase (AMPK), 2 a therapeutic target for metabolic diseases and cancer (). Because of this tight range, AMPK senses the energy state of the cell largely as ratio of AMP to ATP, which changes much more dramatically than ATP levels under energy stress, whereas ADP levels change relatively more modestly. High levels of AMP activate AMPK by three different mechanisms: direct allosteric activation of its kinase activity (, ), increasing activation loop phosphorylation by upstream kinase (, ), and protection of the phosphorylated activation loop against dephosphorylation (, ). High levels of ADP also protect against activation loop dephosphorylation (although 5–10 times less potently than AMP for the γ1 and γ3 isoforms) and do not stimulate direct kinase activation as AMP (, ).[](https://www.ncbi.nlm.nih.gov /mesh/D000255)
AMPK consists of three subunits, a kinase domain-containing α-subunit, a carbohydrate-binding module (CBM)-containing β-subunit, and an adenine nucleotide-binding γ-subunit. The γ-subunit contains four cystathionine β-synthetase (CBS) motifs, which function in pairs known as Bateman domains as universal adenine nucleotide-binding sites (). CBS1 and CBS2 form Bateman domain 1, whereas CBS3 and CBS4 form Bateman domain 2 (note that CBS1, -2, -3, and -4 were previously () known as sites 2, 4, 1, and 3). Crystal structures and binding assays have established that CBS2, in which the conserved, ribose-binding aspartate is mutated to arginine, fails to bind adenine nucleotides (, ). On the other hand, AMP is tightly bound to CBS4 (, ) as wild-type AMPK and cbs1 and cbs3 mutant AMPK co-purify with an equimolar amount of AMP, whereas a cbs4 mutant only purifies with trace amounts of nucleotides ().
Therefore CBS4 likely does not exchange AMP against ATP in the presence of physiological AMP levels (). However, under artificial conditions, i.e. the complete absence of AMP, prebound AMP at CBS4 can very slowly exchange against ATP in solution (∼40% exchange after removal of free nucleotides, followed by 6 h incubation with 5 mm ATP) (). The remaining two CBS sites (CBS1 and CBS3) are thought to be exchangeable as biochemical binding and competition experiments confirmed the existence of two readily exchangeable adenine nucleotide (AXP)-binding sites in purified recombinant AMPK (, , , ). Mutations in key AXP-binding residues in the γ2- and γ3-AMPK isoforms increase basal AMPK activity and are causally associated with glycogen storage diseases and Wolf-Parkinson-White cardiomyopathy ().[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
AMP-occupied CBS3 is directly bound by a loop (αRIM) adjacent to the autoinhibitory domain of the α-subunit (, ), suggesting that the αRIM senses AMP at CBS3. Consistently, mutating key αRIM residues compromises AMPK activation (). Moreover, AMP increases and ATP decreases the interaction between the isolated αRIM loop and core-AMPK (γ-subunit plus scaffolding C termini of the α- and β-subunit), and AMP stimulation of the interaction requires the intact αRIM (). In addition to mutations in CBS3, also mutations in CBS4, which does not interact with the αRIM or other parts of the α- or β-subunits, abolish allosteric kinase activation by AMP (), yet the molecular basis for the CBS4 requirement has remained elusive. When ATP was soaked into crystals of AMP-bound core-AMPK, ATP occupied CBS1 and CBS3, whereas CBS4 remained AMP bound (, ).
In contrast, when AMP-depleted core-AMPK was co-crystallized with ATP, ATP bound to CBS1 and CBS4, CBS3 was unoccupied, and the γ-subunit underwent a noticeable conformational change ().[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
One of the two exchangeable AXP-binding sites binds both AMP and Mg2+-free ATP with micromolar affinity (“high affinity site”), whereas the other site binds AXP much more weakly (hundreds of micromolar (, , )), and likely with positive cooperativity (). Recent surface plasmon resonance (Biacore) data have revealed that ATP binds AMPK in the presence of Mg2+, which shields charge interactions of ATP phosphates with CBS residues, with single site kinetics and a binding constant in the 100 μm range (). This low affinity of Mg2+-ATP would be consistent with preferential AMP-binding at the high affinity binding site to allow AMP exchange at nucleotide concentrations that reflect cellular conditions of energy excess and energy stress (). Currently, it is poorly understood which affinity state maps to which CBS site, how nucleotide binding at one site affects binding at the other site(s), and what adenine nucleotide binds to each of the three functional CBS sites at physiological concentrations of ATP, ADP, and AMP. Focusing on AMP and ATP, we combined fluorescence binding and competition experiments of wild-type and mutant AMPK with hydrogen deuterium exchange mass spectrometry (HDX-MS).
These experiments allowed us to (i) identify CBS3 as a high affinity exchangeable AXP-binding site as well as binding site for NADPH, (ii) map the CBS conformational connectivity and demonstrate that preferential AMP-binding at CBS3 requires AMP-bound CBS4, and (iii) suggest that under physiological conditions only CBS3 significantly exchanges between AMP and ATP.[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
## Results
*In the **Results** section:*
## Isolated CBS motifs bind AXP weakly and with altered AMP/ATP specificity
*In the **Isolated CBS motifs bind AXP weakly and with altered AMP/ATP specificity** section:*
The phosphate groups of the three AMP molecules in AMP-bound AMPK face toward each other and are coordinately bound by a set of charged amino acids, whereas the adenine rings face away from each other and are bound by CBS-specific residues (Fig. 1). Specifically, His151 binds the phosphate groups of both AMP at CBS1 and CBS4, whereas His298 forms hydrogen bonds with the phosphates at both CBS3 and CBS4, suggesting that binding to one CBS motif could affect binding to the other sites. To characterize nucleotide binding at individual CBS motifs in the context of holo-AMPK, but absence of other nucleotide-binding sites, we generated a set of three AMPK variants in which we mutated the ATP-binding lysine in the kinase domain (α1-K47N) plus adenine-binding residues previously shown to be required for nucleotide binding () at two of three CBS sites, thus only one functional CBS remains (Fig. 1). All mutant proteins were stable and could be purified as stoichiometric complexes from size exclusion chromatography columns (Fig. 2). Similar to a previous approach (), we determined relative nucleotide affinities by incubating wild-type and mutant proteins with a fluorescently labeled adenine nucleotide analog and competed the interaction with unlabeled AMP or ATP.
Incubation of 500 nm 3′-(7-diethylaminocoumarin-3-carbonylamin)-3′-deoxyadenosine-5′-diphosphate (deac-ADP; see “Experimental procedures”) () with 4 μm of the recombinant human α1β2γ1-isoform of AMPK resulted in a strong increase in deac-ADP fluorescence and a shift of the emission maximum to a shorter wavelength (Fig. 3A), indicating that deac-ADP binds wild-type and the three mutant AMPK proteins.[](https://www.ncbi.nlm.nih.gov/mesh/D010755)
Binding of AMP to the γ1 subunit. Stick presentation of the three AMP molecules (green carbon atoms) and key AMP-binding residues of AMPK (cyan carbon atoms). Oxygen is colored in red, nitrogen in blue, and phosphorus in orange. The phosphate groups of the 3 AMP molecules are coordinately bound by a set of charged amino acids, whereas the ribose and adenine rings face away from each other and are bound by CBS-specific residues. Mutations of the indicated adenine-binding residues therefore selectively block AMP binding to individual CBS sites. Note that the adenine-binding mutations at CBS3 and CBS4 are each sufficient, even at the high concentration of adenine nucleotides in cells, to completely abolish AMPK activation by AMP without affecting AMPK catalytic activity in the absence of AMP ().[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
Size exclusion chromatograms and SDS-PAGE of AMPK mutant proteins. A, C, and E–G, 120-ml column; B and D, 300-ml column; *, monomeric α1β2γ1 fractions; A, aggregate/oligomers.[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
Fluorescence spectra of wild-type and mutant AMPK/deac-ADP and competition by AMP and ATP. A, emission spectra of 0.5 μm deac-ADP in the absence and presence of wild-type and mutant AMPK proteins in which only one of the CBS sites remained functional (CBS1 only: α1-K47N, γ1-V276G/L277G/I312D; CBS3 only: α1-K47N, γ1-L129D/V130D/I312D; CBS4 only: α1-K47N, γ1-L129D/V130D/V276G/L277G). Excitation: 430 nm. B and C, emission spectra of 0.5 μm deac-ADP in the absence and presence of 4 μm α1β2γ1 AMPK (α1-K47N, γ1-L129D/V130D/V276G/L277G/I312D) (B) or α1β2γ1 AMPK (α1-K47N, γ1-L129D/V130D/V276G/L277G/I312D/S314A) (C) and 5 mm AMP or ATP.[](https://www.ncbi.nlm.nih.gov/mesh/D000244)
As controls, we first incubated deac-ADP with AMPK in which the adenine-binding residues at all three CBS sites were mutated (plus K47N in the kinase domain), which resulted in a small increase in fluorescence (Fig. 3B). This indicated that deac-ADP, in which the ADP-ribose is modified by a bulky coumarin derivative, has residual AMPK-binding activity. This binding signal could not be competed with a 10,000-fold molar excess of AMP, consistent with the reported inability of AMP to bind to this triple cbs mutant (cbs1, L129D/V130D; cbs2, V276G/L277G; cbs3, I312D) (). Neither could ATP compete with the increase in fluorescence, but surprisingly caused a minor increase in fluorescence that was diminished by introduction of an additional cbs4 mutation, S314A (Fig. 3C). Second, in excellent agreement with previous studies (, , ), Mg2+-AMP bound wild-type AMPK, and competed the binding of deac-ADP to wild-type AMPK, with biphasic kinetics, consistent with two exchangeable binding sites, with Kd values of 2.1 μm (high affinity site) and ∼290 μm (low affinity site) (Fig. 4, A–C), whereas Mg2+-ATP competed with monophasic kinetics, consistent with similar affinities for both exchangeable sites (Fig. 4B).[](https://www.ncbi.nlm.nih.gov/mesh/D000244)
Nucleotide binding to wild-type AMPK. A and B, competition of 0.5 μm deac-ADP bound to α1β2γ1 AMPK by unlabeled Mg2+-AMP (A) and Mg2+-ATP (B). The dashed line indicates the fluorescence signal of unbound deac-ADP. Because fluorescence competition failed to unambiguously determine the IC50 value for the high affinity AMP-binding site, we independently determined its affinity by ITC (C).[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
We then tested the ability of AMP and ATP to compete deac-ADP from the AMPK mutant proteins. In contrast to wild-type AMPK, the “non-exchangeable” AMP-binding site CBS4 became exchangeable in the cbs1/cbs3 mutant, as it was efficiently bound by deac-ADP, and binding could be competed with IC50 values of 3.8 μm (AMP) and 14.8 μm (ATP) in the absence of Mg2+, and 1.9 μm for AMP in the presence of Mg2+ (Fig. 5A; note that Mg2+-ATP inefficiently competed deac-ADP from the CBS4 only mutant, resulting in an unreliable IC50 value for this measurement). Both AMPK cbs3/cbs4 (CBS1 only) and AMPK cbs1/cbs4 (CBS3 only) bound ATP with low to moderate affinity and AMP with very low affinity. Especially CBS3 in the absence of CBS4 and CBS1 bound AMP with an IC50 value that is well below physiological AMP concentrations and with a more than 8-fold preference of ATP over AMP in the presence of Mg2+ (and more than 90-fold in its absence).
Given that the physiological concentration of ATP is close to 2 orders of magnitude higher than that of AMP, CBS3 would be constitutively bound by ATP in the absence of functional CBS4 and CBS1.[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
Nucleotide binding to individual CBSs and pairs of CBSs. Competition of 0.5 μm deac-ADP bound to 4 μm α1β2γ1 AMPK by unlabeled AMP and ATP. A, AMPK mutants with only a single functional adenine nucleotide-binding site (see Fig. 1). Note that in these experiments singly and doubly mutated CBS4 (I312D and I312D/S314A) behaved identically. The ability of Mg2+-AMP to compete deac-ADP from 4 μm AMPK with an IC50 of 1.9 μm indicates that AMP still binds CBS4 with submicromolar affinity, but no longer non-exchangeably. B, AMPK mutants with two functional adenine nucleotide-binding sites. Samples were incubated for 30 min, excited at 430 nm, and emission recorded at 470 nm. The dashed line indicates the fluorescence signal of unbound (i.e. completely competed) deac-ADP.[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
Next we determined binding of AMP and ATP to AMPK in which only the ATP-binding site in the kinase domain and a single CBS site were mutated, leaving intact two binding sites in all three combinations. Importantly, Chen et al. () had shown that in these mutants CBS4 remains tightly AMP-bound during protein purification. In agreement with this report, when we left CBS3 and CBS4 intact (α1-K47N, γ1-L129D/V130D), AMP and ATP competed a single site (i.e. data cannot be modeled to two sites), confirming that the presence of wild-type CBS3 is sufficient to allow CBS4 to bind AMP tightly enough to not noticeably exchange during competition. In the presence of wild-type CBS4, CBS3 bound AMP with relatively high affinity (16.9 μm) and ATP with much lower affinity (153 μm) (Fig. 5B), resembling the high affinity exchangeable binding site in the context of wild-type AMPK. In contrast, when we mutated the catalytic site plus CBS3 (CBS1 + CBS4 intact), the competition again fit only a single site model, yet the affinity for AMP remained weak (IC50 = 160 μm) and below that of ATP (IC50 = 78 μm) (Fig. 5B), resembling more closely the low affinity exchangeable AXP-binding site.
Finally, when only CBS1 and CBS3 remained functional, we could fit the competition data only as a single site with low affinity for ATP and very low affinity for AMP (Fig. 5B), similar to what we have found for the exchangeable low affinity binding site in wild-type AMPK. Collectively, these data indicate that (i) either a functional CBS1 or a functional CBS3 is sufficient to allow CBS4 to bind AMP non-reversibly during competition, that (ii) in the presence of functional CBS4, CBS3 resembles more the high affinity and CBS1 the low affinity exchangeable site, and (iii) that in the presence of only CBS1 and CBS3 neither formed a high affinity binding site, but both most likely formed low affinity binding sites that could not be computationally deconvoluted.[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
## AXP binding to a single CBS motif modulates the conformation of all three CBS sites as well as the CBM
*In the **AXP binding to a single CBS motif modulates the conformation of all three CBS sites as well as the CBM** section:*
To directly test conformational stabilization, we analyzed wild-type and mutant AMPK by HDX-MS. HDX-MS measures the backbone amide exchange of hydrogen against deuterium in deuterated buffers. Ligand binding typically protects binding sites against deuterium exchange and can also induce conformational changes that alter protection patterns. Incubation of wild-type AMPK with 2.5 mm AMP dramatically reduced hydrogen/deuterium exchange at CBS3 and CBS4, strongly at CBS1, and moderately at the catalytic center of the kinase domain (catalytic loop, activation loop, and parts of helices αB, αD, and αF; Fig. 6A and supplemental Tables S1–S3).
Because AMP at CBS4 exchanges very slowly, the strong decrease in HDX at CBS4 is likely predominantly or completely due to its conformational stabilization upon AMP binding to CBS1 and CBS3.[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
HDX-MS changes of wild-type and mutant AMPK upon incubation with AMP or ATP. Changes were overlaid as heat map onto the structure of α1β2γ1 AMPK (4RER). The heat map legend (% change) is shown at the bottom. A, AMPK wild-type in the presence versus absence of AMP (left) or ATP (right). B, AMPK wild-type in the presence versus absence of 991. C, D, and E, AMPK triple mutants in which only CBS4 (C), CBS3 (B), or CBS1 (E) are functional (see Fig. 1).[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
Unexpectedly, the CBM of the β-subunit was also strongly protected (Fig. 6A, left panel, and supplemental Table S3), indicating that the adenine nucleotide-binding sites in the γ-subunit and the CBM are conformationally connected, even though these domains are located on opposite sides of the AMPK heterotrimer and do not contact each other. Conversely, when we incubated AMPK with the Merck compound 991, which binds a pocket formed by the CBM and kinase domain (allosteric drug and metabolite-binding site) (), the 991-binding site and the regulatory αC-helix became strongly stabilized, whereas the distal CBS3 in the γ-subunit became mildly destabilized (Fig. 6B). AMP-dependent protection against dephosphorylation of p-Thr174 in the activation loop likely involves occluding p-Thr174 by the CBM linker ().
The observed CBM stabilization by AMP binding may therefore play a direct role in this critical AMP response. When we incubated AMPK with ATP, the overall protection was similar to that with AMP. However, protection of CBS3, CBS4, and the CBM was weaker, whereas protection of CBS1 and the ATP-binding catalytic site in the kinase domain increased (Fig. 6A, right panel, and supplemental Tables S1–S3).[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
When we incubated the AMPK mutant protein in which CBS4 is the only functional AXP-binding site (K47N/cbs1/cbs3) with AMP, CBS4 became strongly protected, confirming that CBS4 is indeed exchangeable in the absence of CBS1 and CBS3. In addition, CBS3 became protected, indicating that AMP-binding at CBS4 strongly conformationally stabilizes CBS3 (Fig. 6C), in excellent agreement with the fluorescence competition data. In contrast to CBS3, AMP-binding at CBS4 modulated accessibility at CBS1 more mildly (Fig. 6C; note that deuterium exchange both increases and weakly decreases in different CBS1 peptides), consistent with the relatively small effect of CBS4 on CBS1 AMP-binding. A similar, but milder protection of CBS3 and CBS4 occurred for the CBS3-only (K47N/cbs1/cbs4)
AMPK mutant in the presence of AMP (Fig. 6D), whereas AMP-binding to the CBS1-only (K47N/cbs3/cbs4) AMPK mutant stabilized all three binding sites (Fig. 6E).[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
## CBS occupancy in mixtures of adenine nucleotides
*In the **CBS occupancy in mixtures of adenine nucleotides** section:*
All current binding and structural studies have been performed either in the presence of high concentrations of AMP or high concentrations of ATP, whereas in cells AMPK is exposed to mixtures of AMP, ADP, and ATP. An important, but unresolved, question is therefore which CBS motifs are occupied by AMP, ADP, or ATP under physiological conditions, with AMP always present, but at levels that are 10–100-fold lower than ATP levels. To probe for occupancy under physiological conditions, we incubated AMPK with two different AXP mixtures: 4.5 mm ATP, 0.4 mm ADP, 0.04 mm AMP and 3.8 mm ATP, 1.0 mm ADP, 0.3 mm AMP, which have been reported to mimic normal cellular (non-stressed) and energy stress conditions in cells, respectively (). Consistently, the AXP ratio mimicking energy stress allosterically activated recombinantely purified AMPK 2-fold relative to the mixture mimicking energy excess (Fig. 7A), the same fold direct activation as for the standard switch from AMPK, 100 μm ATP to AMPK, 100 μm ATP + 200 μm AMP. In striking contrast to the different HDX profiles of AMPK, 2.5 mm AMP versus AMPK, 2.5 mm ATP (Fig. 6, A and B, and supplemental Tables S1–S3), the profiles in the presence of the stress and non-stress AXP mixtures were quite similar to each other and showed significant differences only in two small areas (Fig. 7B), clearly indicating that these conditions induce much more limited nucleotide exchange.
Whereas HDX-MS has the potential to distinguish AMP versus ATP occupancy due to the unique β- and γ-phosphate groups of ATP, none of the resolved β- and γ-phosphate contacting peptides makes exclusive interactions with these groups (see Fig. 8 and supplemental Table S1) and decreases in HDX can be due to both direct nucleotide binding and to conformational stabilizations. Because our fluorescence competition data are most consistent with CBS1 (binds the more abundant ATP with equal or higher affinity than AMP) being bound by ATP under both normal and energy stress conditions, and CBS4 being essentially non-exchangeably bound to AMP, only CBS3 may undergo noticeable ATP/AMP exchange at physiological levels.[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
Changes in HDX-MS protection in the presence of AXP mixtures mimicking non-stress and energy stress conditions. A, AMPK kinase assay in the presence of AXP mixtures mimicking nonstress (4.5 mm ATP, 0.4 mm ADP, 0.04 mm AMP) and energy stress (3.8 mm ATP, 1.0 mm ADP, 0.3 mm AMP) conditions. Left panel, AlphaScreen-based kinase assay; right panel, radioactive kinase assay. B, HDX-MS perturbation map of AMPK in the presence of the two different AXP mixtures. HDX-MS changes were overlaid as heat map onto the structure of α1β2γ1 AMPK (PDB code 4RER). Main areas of differential protection are highlighted by dashed outlines. Note that there is no change at the catalytic site, indicating that the catalytic site is constitutively bound by ATP as expected. The color code (% change) is shown below the structures. The relatively mild changes are consistent with noticeable ATP exchange only at CBS3, whereas the non-physiological shift from high concentration of pure AMP to high concentration of pure ATP resulted in strong HDX changes (Fig. 6A), consistent with near full nucleotide exchange at both CBS3, CBS1, and ATP binding/release at the catalytic site.[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
AXP/CBS interaction network in human AMPK-γ1. Shown are the three AMP-bound CBS sites in holo-AMPK co-crystallized with AMP. β- and γ-phosphate interacting residues (gray) are collectively based on the structures of core AMPK co-crystallized with ATP (PDB code 4EAG) () and core- AMPK soaked with ATP (PDB codes 2V92 and 4EAJ) (, ).[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
## NADPH binds AMPK at the high affinity exchangeable AXP-binding site
*In the **NADPH binds AMPK at the high affinity exchangeable AXP-binding site** section:*
We found that in addition to the known AMPK ligand NADH (), NADPH can also bind AMPK. When we incubated 5 μm NADPH with 10 μm maltose-binding protein (MBP)-tagged AMPK, NADPH fluorescence increased and shifted its emission maximum, indicative of direct binding (Fig. 9A). In contrast, incubation of NADPH with 10 μm of the MBP control did not alter the NADPH emission spectrum (Fig. 9A). To estimate the affinity for NADPH and NADH, we incubated 5 μm NADPH or 5 μm NADH with increasing concentrations of MBP-AMPK. Although the half-maximal increase of the NADPH fluorescence signal indicated a Kd of ≤20 μm, an increase of NADH fluorescence did not approach saturation at the highest AMPK concentration (100 μm), indicating that NADH binds AMPK only with low affinity (≫20 μm; Fig. 9B).
Because NADPH is an adenine dinucleotide, we speculated that it would bind to one or more of the AXP-binding sites in the γ-subunit. Indeed 10 μm AMP almost completely competed 5 μm NADPH bound to 10 μm AMPK, indicating that NADPH binds to the high affinity exchangeable AMP-binding site (Fig. 9C). Conversely, ATP was also able to compete NADPH, but only at much higher concentrations (Fig. 9D). We therefore conclude that NADPH binds to the high affinity exchangeable AMP-binding site and that this site has a preference for AMP over ATP, suggesting that NADPH binds CBS3.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
NADPH binds adenine nucleotide-binding site(s) in the γ-subunit. A, NADPH emission spectra in the absence and presence of MBP and MBP-tagged phosphorylated α1β2γ1 AMPK. The spectrum of AMPK in the absence is shown as negative control. B, estimation of NADPH and NADH binding constants. 5 μm NADPH or NADH were incubated with increasing concentrations of MBP-AMPK and MBP-AMPK concentrations were plotted against fluorescence intensity at the emission maxima. C and D, AMP and ATP compete the NADPH binding signal. Emission spectra of 5 μm NADPH, 10 μm AMPK were recorded in the absence and presence of increasing concentrations of AMP (C) and ATP (D).[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
To physically localize the NADPH-binding site, we collected HDX-MS perturbation data of AMPK in the absence versus presence of 1 mm NADPH. As shown in Fig. 10, NADPH binding selectively protected CBS3 (plus CBS4), but not CBS1, against deuterium exchange. Because CBS4 is non-exchangeable, the HDX-MS data confirm that NADPH specifically binds CBS3. Given the high affinity (low μm IC50) with which AMP competes NADPH from CBS3 and the AMP-over-ATP binding preference of CBS3, we conclude that CBS3, and not as previously proposed CBS1 (), is the high affinity exchangeable AMP-binding site.[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
Binding of NADPH protects CBS3 and CBS4. HDX-MS changes wild-type AMPK upon incubation with 1 mm NADPH. Changes were overlaid as heat map onto the structure of α1β2γ1 AMPK (PDB code 4RER). The heat map legend (% change) is shown at the bottom.[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
## Discussion
*In the **Discussion** section:*
AMPK plays critical roles in numerous aspects of physiology (, , ) and is an important target for the treatment of metabolic diseases, including diabetes and obesity (, ), as well as cancer (, , ) and heart disease (). In addition, AMPK activation mimics many of the beneficial effects of exercise and caloric restriction (). Crystal structures and biochemical binding assays have demonstrated that adenine nucleotides can competitively bind to three functional CBS motifs in the γ-subunit, of which CBS3 functions as an adenine sensor site. However, due to the technical difficulties of deconvoluting complex binding events, many aspects of AXP occupancy at individual CBS sites, and how binding to one site is communicated to the other sites, have remained unclear.[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
To deconvolute the binding events, we generated a set of mutant AMPK complexes in which only individual sites or pairs of sites remain functional, and quantitatively determined their nucleotide-binding and deuterium-exchange patterns. These experiments provided insight into functional and conformational coupling of CBS sites, occupancy in the presence of AXP mixtures, assignment of the high affinity exchangeable AXP site, and a mechanistic model for the requirement of CBS4 for AMP sensing at CBS3.[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
## Functional and conformational coupling of CBS sites
*In the **Functional and conformational coupling of CBS sites** section:*
Early binding studies using GST-CBS fusion proteins provided the first support for cooperative AXP binding (). Here we specifically demonstrate that CBS4 becomes readily exchangeable in the absence of functional CBS1 and CBS3, and AXP-binding to either CBS1 or CBS3 strongly decreased deuterium exchange at CBS4. Therefore, AXP-bound CBS1 and CBS3 can each separately increase AMP affinity at CBS4 and conformationally stabilize CBS4. Consistently, AMPK with mutation of only CBS1 or only CBS3 stably co-purifies with AMP () and AXP-binding follows a single site binding model with kinetics different from that of CBS4-only. Stabilization is likely due to His151, which forms hydrogen bonds with the AMP phosphate groups at both CBS4 and CBS1, and His298, which forms hydrogen bonds with the AMP phosphate groups at both CBS4 and CBS3 (Fig. 1).
Conversely, CBS3 becomes conformationally stabilized by CBS4 and CBS1, and AMP-binding to CBS4 (functional CBS3 + CBS4 versus functional CBS3) increased the CBS3 affinity for AMP by 2 orders of magnitude and increased the AMP/ATP binding preference by 2–3 orders of magnitude.[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
## AMP-sensing requires CBS4 to conformationally stabilize AMP-bound CBS3
*In the **AMP-sensing requires CBS4 to conformationally stabilize AMP-bound CBS3** section:*
Mutating a key adenine-binding residue of CBS4 (I312D) abolishes direct allosteric AMPK activation by AMP (). This indicates that CBS4 critically contributes to AMP-sensing at CBS3/αRIM, even though CBS4 is constitutively bound by AMP and does not directly interact with either the αRIM loop or AMP at CBS3. Our fluorescence competition and HDX-MS data suggest that CBS4 is required for AMP sensing by stabilizing AMP binding at CBS3, which increases both the affinity for AMP and the preferential binding of AMP over ATP (Fig. 5). AMP-bound CBS4 therefore allows AMP exchange at CBS3 at nucleotide concentrations that reflect cellular conditions of energy excess ([ATP]/[AMP] >100) and energy stress ([ATP]/[AMP] ∼10–20). Our finding that AMP-bound CBS4 selectively stabilizes AMP binding at CBS3 provides a rationale for the CBS4 requirement for AMP sensing.[](https://www.ncbi.nlm.nih.gov/mesh/D000225)
How does AMP binding to CBS4 mechanistically increase AMP affinity and AMP/ATP-binding preference at CBS3? Inspection of AMPK crystal structures indicates that stabilization is predominantly due to CBS4 positioning the side chain of His298. The His298 backbone amide directly forms a hydrogen bond with the phosphate group of AMP at CBS3 and positions the adjacent key residue Arg299, which forms two hydrogen bonds with the AMP phosphate group and two with the AMP adenine ring at CBS3 (Fig. 11). Similarly, αRIM Glu364 positions CBS3 Arg70 and Lys170, which form salt bridges with the phosphate group of AMP at CBS3. The phosphate group of AMP carries two negative charges to interact with the three positively charged residues stabilized by αRIM (Arg70 and Lys170) and CBS4 (Arg299) plus the partially positive backbone amide of His298.
In contrast, the corresponding α-phosphate of ATP carries only a single charge and therefore interacts less tightly with the positive charge cluster (Fig. 11A). In addition, as seen when structurally aligned with ATP-soaked core-AMPK (PDB code 2V92), the side chains of Arg70 and Lys170 need to move away in opposite directions from Glu364 to avoid clashes with the β- and γ-phosphate groups of ATP (Fig. 11B), thereby weakening the Glu364 interaction. Together, these constraints provide a structural explanation for both CBS4 and αRIM Glu364 collectively stabilizing AMP preferentially over ATP at CBS3.
Consistent with a requirement of both CBS4 and the αRIM for high affinity binding of AMP at CBS3, AMP was preferentially bound to CBS1 and CBS4 in the crystal structure of the αRIM-lacking core-AMPK (), but preferentially bound to CBS3 and CBS4 in the context of the αRIM-containing holo-AMPK (, ).[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
CBS4 and αRIM stabilize AMP-binding at CBS3 and increase the AMP/ATP-binding preference. A, CBS4-CBS3-αRIM network in AMP-bound AMPK (PDB code 4CFE). B, transparent structure from panel A overlaid with ATP, Arg70, and Lys170 from the structure of ATP-soaked core-AMPK (PDB code 2V92). Note that the α-phosphate of ATP has only a single negative charge and that the β- and γ-phosphate groups of ATP push Lys170 and Arg70 away from αRIM Glu364.[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
## CBS3 is the high affinity exchangeable AMP-binding site
*In the **CBS3 is the high affinity exchangeable AMP-binding site** section:*
Our work presents several lines of evidence in support of CBS3 being the high affinity exchangeable AMP-binding site. First, AMP competes NADPH binding with high affinity, whereas ATP competes with moderate affinity, directly indicating that the NADPH-binding site is the high affinity site. This site is CBS3, as NADPH selectively protects CBS3, but not CBS1, against hydrogen deuterium exchange. Second, CBS3 in the presence of functional CBS4 and αRIM binds AMP with high, and ATP with moderate affinity, whereas CBS1 in the presence of functional CBS4 binds AMP weakly and with preference of ATP over AMP binding. Third, HDX-MS in the presence of wild-type holo-AMPK demonstrates that AMP stabilizes CBS3 more strongly than ATP, whereas it stabilizes CBS1 less strongly than ATP, consistent with higher affinity AMP binding at CBS3 and relatively stronger ATP binding at CBS1.[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
## Under physiological conditions nucleotide exchange is likely limited to CBS3
*In the **Under physiological conditions nucleotide exchange is likely limited to CBS3** section:*
Several lines of evidence suggest that during the shift from non-stress (4.5 mm ATP, 0.4 mm ADP, 0.04 mm AMP) to energy stress (3.8 mm ATP, 1.0 mm ADP, 0.3 mm AMP) only CBS3 noticeably exchanges ATP against AMP. By mutating CBS1, we could show that in the presence of the tightly AMP-bound CBS4, CBS3 binds AMP with 10- to 100-fold higher affinity than ATP, closely matching the 10- to 100-fold higher ATP to AMP levels under different stress conditions. In contrast, under the converse conditions, CBS1 binds ATP with similar or higher affinity than AMP, suggesting that at millimolar ATP and micromolar AMP physiological levels CBS1 remains largely ATP-bound and CBS4 AMP-bound. These conclusions are supported by HDX-MS. First, the AMP/ATP preferences of CBS3 and CBS1 are reflected by CBS3 being more strongly protected by AMP than ATP and, conversely, CBS1 is more strongly protected by ATP than AMP.
Second, the AMPK HDX profiles in the sole presence of AMP versus ATP differ strongly, consistent with AMP/ATP exchange at both CBS1 and CBS3 as seen in crystal structures and competition assays, whereas the profiles in the presence of stress versus non-stress AXP levels are much smaller, indicating reduced AMP/ATP exchange.[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
ATP is the essential cofactor of numerous cellular reactions, which require ATP concentrations within a very tight range. Yet ATP is constantly and rapidly metabolized (humans hydrolyze roughly their own body weight of ATP per day ()), and therefore requires highly efficient cellular and organismal regulation of ATP homeostasis. This task is centrally performed by AMPK and depends on the ability of AMPK to precisely and sensitively detect AXP ratios under different physiological conditions to modulate its kinase activity. Our study, together with previous binding assays and crystal structures of holo-AMPK bound to AMP and core-AMPK bound to ATP, provides a first view of the intricate arrangement and interactions between nucleotide-binding sites to allow precise and highly sensitive sensing of AXP ratios under physiological stress and non-stress conditions (Fig. 12). Understanding how the sensed nucleotides modulate kinase activity and activation loop accessibility will require the missing key structure of holo-AMPK in its inactive ATP-bound state.[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
Schematic model of AXP occupancy under energy-stress (A) and non-stress (B) conditions. Positive and negative charges are indicated by blue “+” and red “−” signs. A-circled P, AMP; A-3 circled P, ATP.[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
## Experimental procedures
*In the **Experimental procedures** section:*
## DNA constructs and reagents
*In the **DNA constructs and reagents** section:*
The human wild-type His6-α1(11–550)β2γ1 AMPK expression construct was described previously (). Mutations were introduced by site-directed mutagenesis using the QuikChange method (Stratagene) or standard PCR-based methods. All expression constructs and mutations were confirmed by DNA sequencing.
## Protein expression and purification
*In the **Protein expression and purification** section:*
Expression plasmids were transformed into Escherichia coli BL21(DE3). Cells were grown to an A600 of ∼1 at 28 °C and induced with 100 μm isopropyl β-d-thio-galactopyranoside at 16 °C overnight. Cell pellets were resuspended in 25 mm Tris, pH 8.0, 300 mm NaCl, 25 mm imidazole, 10% glycerol, 5 mm β-mercaptoethanol, and lysed by French Press with pressure set to 900 pascal. Lysates were cleared by centrifugation for 30 min at 20,000 × g, passed over a 50-ml nickel-chelating HP-Sepharose column (GE Healthcare), and eluted with 25 mm Tris, pH 8.0, 300 mm NaCl, 500 mm imidazole, 10% glycerol, 5 mm β-mercaptoethanol. The eluted His-tagged AMPK was further purified by size-exclusion chromatography through a 300-ml HiLoad 26/60 Superdex 200 column (GE Healthcare) in 10 mm Tris, pH 8.0, 150 mm NaCl, 5 mm MgCl2, 1 mm EDTA, 10% glycerol, 2 mm DTT.
The proteins eluted from the gel filtration column at a volume corresponding to the size of a monomeric complex at a purity ≥95% as judged by SDS-PAGE (Fig. 2).[](https://www.ncbi.nlm.nih.gov/mesh/D007544)
## Fluorescence binding and competition assays
*In the **Fluorescence binding and competition assays** section:*
Deac-ADP was synthesized by the method of Webb et al. (). 4 μm AMPK wild-type and mutant proteins were incubated with 0.5 μm deac-ADP in 10 mm Tris, pH 8.0, 150 mm NaCl, 1 mm EDTA, 10% glycerol, 2 mm DTT in the presence and absence of 5 mm MgCl2 for 30 min at room temperature. The emission spectrum of deac-ADP was recorded using a SpectraMax M2e reader (Molecular Devices Corporation). For deac-ADP competition assays, AMP or ATP were added at increasing concentrations, whereas deac-ADP was kept at a constant concentration (0.5 μm). The IC50 values were obtained from curve fitting to the competitive inhibitor model using GraphPad Prism.[](https://www.ncbi.nlm.nih.gov/mesh/D000244)
For NADPH emission spectra, 5 μm NADPH were incubated with either 10 μm MBP or 10 μm MBP-AMPK and excited at 340 nm using the SpectraMax M2e cuvette mode. For binding isotherms, NADP and NADPH (5 μm each) were incubated with increasing concentrations of AMPK, excited at 340 nm, and full emission spectra were recorded. Fluorescence emission maxima were plotted against AMPK concentration and EC50 values were obtained from curve fitting to the one-site binding model using GraphPad Prism.[](https://www.ncbi.nlm.nih.gov/mesh/D009249)
## HDX coupled with mass spectrometry
*In the **HDX coupled with mass spectrometry** section:*
Solution-phase amide HDX experiments were carried out using a fully automated system as described previously (, ). Five microliters of a 10 μm protein solution were mixed with 20 μl of D2O containing HDX buffer (20 mm KPO4, pH 7.4, 50 mm KCl) and incubated at 4 °C for 10, 30, 60, 300, 900, and 3,600 s. Following on exchange, unwanted forward or back exchange was minimized and the protein was denatured by dilution with 25 μl of quench solution (0.1%, v/v, trifluoroacetic acid (TFA) in 3 m urea). Samples were then passed through an immobilized pepsin column (prepared in house) () at 200 μl/min (0.1%, v/v, TFA, 15 °C) and the resulting peptides were trapped on a C8 trap column (Hypersil Gold, Thermo Fisher). The bound peptides were then gradient eluted (4–40%, w/v, CH3CN and 0.3%, w/v, formic acid) across a 2 × 50-mm C18 high performance liquid chromatography column (Hypersil Gold, Thermo Fisher) for 5 min at 4 °C. The eluted peptides were then subjected to electrospray ionization directly coupled with a high resolution Orbitrap mass spectrometer (Exactive, Thermo Fisher).
Each HDX experiment was carried out in triplicate. Peptide ion signals were confirmed if they had a MASCOT score of 20 or greater and had no ambiguous hits using a decoy (reverse) sequence in a separate experiment using a 60-min gradient. The intensity-weighted average m/z value (centroid) of each isotopic envelope of the peptide was calculated with in house HDX Workbench software () and corrected for back-exchange on an estimated 70% recovery and accounting for the known deuterium content of the on-exchange buffer.[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
## Isothermal titration calorimetry (ITC)
*In the **Isothermal titration calorimetry (ITC)** section:*
ITC was performed in a MicroCal PEAQ iTC200 (Malvern) with 18 injections at room temperature. 30 μm AMPK was loaded into the cell and 300 μm AMP in the syringe. Data analysis was performed with the MicroCal PEAQ-ITC instrument control software and yielded: n = 0.856 ± 0.02 sites, Kd = 2.06 ± 0.538 μm, ΔH = 12.6 ± 0.757 kcal/mol, ΔG = −7.76, −TΔS = 4.88 kcal/mol.[](https://www.ncbi.nlm.nih.gov/mesh/D000249)
## Kinase assays
*In the **Kinase assays** section:*
For the radioactive kinase assay, 10 nm phosphorylated AMPK were incubated with 15 μm biotin-SAMS peptide, 2 mm DTT, and 0.25 μl [γ-32P]ATP per 15-μl reaction in kinase buffer (25 mm Tris, pH 7.4, 12 mm MgCl2, 1 mm Na3VO4, 5 mm NaF) in the presence of AXP mixtures mimicking non-stress (4.5 mm ATP, 0.4 mm ADP, 0.04 mm AMP) and energy-stress (3.8 mm ATP, 1.0 mm ADP, 0.3 mm AMP) conditions for 30 min at room temperature. Reactions were terminated by addition of 0.5 volumes of 7.5 m guanidine hydrochloride solution in water and reactions were spotted on a SAM 2® Biotin Capture Membrane (Promega). The membrane was washed once for 30 s with 2 m NaCl, 3 times for 2 min with 2 m NaCl, 4 times for 2 min with 2 m NaCl in 1% H3PO4, and 2 times for 30 s with deionized water to remove unbound reaction components. After drying the membrane at room temperature for 30–60 min, signals were quantitated by phosphorimager analysis.[](https://www.ncbi.nlm.nih.gov/mesh/D001710)
The luminescence-proximity AlphaScreen-based kinase assay will be described in detail in a future publication and is based on a previously described AMPK activity reporter assay (). Briefly, 10 nm phosphorylated AMPK were incubated with 50 μm of a biotinylated peptide substrate (biotin-GSTKMRRVATLVDLGYKK) on ice for 10 min. The reaction was terminated by 1000-fold dilution in AlphaScreen buffer (50 mm MOPS, pH 7.4, 50 mm NaF, 50 mm CHAPS, and 0.1 mg/ml of bovine serum albumin) and further incubated in the presence of 50 nm His6-Rad53 FHA domain protein, 5 μg/ml each of AlphaScreen streptavidin donor beads and nickel-acceptor beads at room temperature for 90 min in the presence of AXP mixtures mimicking non-stress (4.5 mm ATP, 0.4 mm ADP, 0.04 mm AMP) or energy-stress (3.8 mm ATP, 1.0 mm ADP, 0.3 mm AMP) conditions. Donor beads contain a photosensitizer, which can convert ambient oxygen into short-lived singlet oxygen upon light activation at 680 nm. When the acceptor beads are brought close enough to the donor beads by interaction between His6-tagged FHA domain protein and AMPK-phosphorylated biotinylated peptide substrate, singlet oxygen can diffuse from the donor to the acceptor beads and transfer energy to the thioxene derivatives of the acceptor beads, resulting in light emission at 520 to 620 nm, which was measured in an Envision (PerkinElmer Life Sciences) plate reader.[](https://www.ncbi.nlm.nih.gov/mesh/D001710)
## Author contributions
*In the **Author contributions** section:*
X. G., Y. Y., S. J. N., A.
K., D. G., J. K., M. H. E.
T., P. W. deW., L. W., and X. L.
conducted the experiments, M. R. W., P. R. G., H.
E. X., and K. M. analyzed the results, and K. M.
wrote the paper with comments from all authors.
## Supplementary Material
*In the **Supplementary Material** section:*
This work was supported by the Francis Crick Institute with funding from Cancer Research UK Grant FC001211, United Kingdom Medical Research Council Grant FC001211, Wellcome Trust Grant FC001211 (to M. R. W.), grants from the Van Andel Research Institute (to H. E. X. and K. M.), Ministry of Science and Technology (China) Grants 2012ZX09301001, 2012CB910403, 2013CB910600, XDB08020303, and 2013ZX09507001, National Science Foundation Grant 91217311 (to H. E. X.), and National Institute of Health Grants DK071662 (to H. E. X.) and GM102545 and GM104212 (to K. M.). The authors declare that they have no conflicts of interest with the contents of this article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
This article contains supplemental Tables S1–S3.
AMPK
AMP-activated protein kinase
αRIM
α-regulatory subunit interaction motif
AXP[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
AMP/ADP/ATP adenine nucleotides[](https://www.ncbi.nlm.nih.gov/mesh/D000244)
CBM
carbohydrate-binding module
CBS
cystathionine β-synthetase
deac-ADP[](https://www.ncbi.nlm.nih.gov/mesh/D000244)
3′-(7-diethylaminocoumarin-3-carbonylamin)-3′-deoxyadenosine-5′-diphosphate[](https://www.ncbi.nlm.nih.gov/mesh/D000244)
HDX-MS
hydrogen deuterium exchange mass spectrometry[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
ITC
isothermal titration calorimetry
MBP
maltose-binding protein
PDB
Protein Data Bank.
The abbreviations used are:
# References
*In the **References** section:*
|
# Introduction
[Glycine](https://www.ncbi.nlm.nih.gov/mesh/D005998)-Binding Site Stimulants of NMDA Receptors Alleviate Extrapyramidal Motor Disorders by Activating the Nigrostriatal Dopaminergic Pathway
# Abstract
*In the **Abstract** section:*
Dysfunction of the N-methyl-d-aspartate (NMDA) receptor has been implicated in the pathogenesis of schizophrenia. Although agonists for the glycine-binding sites of NMDA receptors have potential as new medication for schizophrenia, their modulation of antipsychotic-induced extrapyram[idal si](https://www.ncbi.nlm.nih.gov/mesh/D005998)de effects (EPS) has not yet been clarified. We herein evaluated the effects of glycine-binding site stimulants of NMDA receptors on antipsychotic-induced EPS in mice and rats. d-cycloserine (DCS) and d-serine significantly [improve](https://www.ncbi.nlm.nih.gov/mesh/D005998)d haloperidol (HAL)-induced bradykinesia in mice, whereas glycine showed no effects. Sodiu[m benzoate, a](https://www.ncbi.nlm.nih.gov/mesh/D003523) d[-am](https://www.ncbi.nlm.nih.gov/mesh/D003523)ino ac[id oxida](https://www.ncbi.nlm.nih.gov/mesh/D012694)se inhibitor, also atten[uated HAL-i](https://www.ncbi.nlm.nih.gov/mesh/D006220)nd[uce](https://www.ncbi.nlm.nih.gov/mesh/D006220)d bradykinesia. Improvements in HAL-indu[ced bra](https://www.ncbi.nlm.nih.gov/mesh/D005998)dykinesia by DCS wer[e antagonized b](https://www.ncbi.nlm.nih.gov/mesh/D020160)y the NMDA antagonist dizocilpine or nitric oxide sy[nth](https://www.ncbi.nlm.nih.gov/mesh/D006220)ase inhibitor L-NG-Nitro-l-arginine met[hyl](https://www.ncbi.nlm.nih.gov/mesh/D006220) ester. In addition, DCS [sig](https://www.ncbi.nlm.nih.gov/mesh/D003523)nificantly reduced HAL-in[duce](https://www.ncbi.nlm.nih.gov/mesh/D016202)d Fos expres[sion in the](https://www.ncbi.nlm.nih.gov/mesh/D016291) dorsolateral striatum without affec[ting that in the nucleus accumbens](https://www.ncbi.nlm.nih.gov/mesh/D019331). Furthermore, [a m](https://www.ncbi.nlm.nih.gov/mesh/D003523)icroinjection of DCS in[to ](https://www.ncbi.nlm.nih.gov/mesh/D006220)the substantia nigra pars compacta significantly inhibited HAL-induced EPS concomitant with elevations in dopamine release in the stria[tum](https://www.ncbi.nlm.nih.gov/mesh/D003523). The present results demonstrated for the first time that stimul[ati](https://www.ncbi.nlm.nih.gov/mesh/D006220)ng the glycine-binding sites of NMDA recepto[rs allev](https://www.ncbi.nlm.nih.gov/mesh/D004298)iates antipsychotic-induced EPS by activating the nigrostriatal dopaminergic pathway, suggesting th[at glyc](https://www.ncbi.nlm.nih.gov/mesh/D005998)ine-binding site stimulants are beneficial not only for efficacy, but also for side-effect management.[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
## 1. Introduction
*In the **1. Introduction** section:*
Schizophrenia is a heterogenous disease with diverse psychotic symptoms including positive and negative symptoms, neurocognitive impairments, and mood disturbances [,,,]. It is well-known that hyperactivity of the meso-limbic dopaminergic system is involved in the pathogenesis of schizophrenia (dopamine hypothesis), and numerous first-generation antipsychotics, which commonly antagonize dopamine D2 receptors, have been developed [,,]. These agents effectively improve positive symptoms (e.g., hallucinations, delusion, and excitement) in patients with schizophrenia through D2 receptor blockade in the limbic regions (e.g., the nucleus accumbens) [].
However, they frequently induce extrapyramidal side effects (EPS) by blocking D2 receptors in the basal ganglia (e.g., the striatum). In addition, first generation antipsychotics were only effective for positive symptoms, not negative symptoms or cognitive impairments, suggesting that multiple mechanisms (e.g., serotonergic and glutamatergic systems) other than the dopaminergic system are also involved in the generation of schizophrenia symptoms [,,,].[](https://www.ncbi.nlm.nih.gov/mesh/D004298)
N-methyl-d-aspartate (NMDA) receptors are heteromeric tetramer proteins composed of GluN1 and GluN2 subunits containing d-serine/glycine- and glutamate-binding sites, respectively [,]. Besides d-serine/glycine- and glutamate-binding sites, they also possess several regulatory sites sensitive to polyamines, Zn2+, protons, and glutathione [,]. NMDA receptors are involved in the etiology and treatment of various neuropsychiatric disorders (e.g., schizophrenia, depression, Alzheimer’s disease, and ischemic stroke)
[,,,,,]. Previous studies proposed that the dysfunction of NMDA receptors is involved in the pathogenesis of schizophrenia (the glutamate hypothesis in schizophrenia) [,,]. This hypothesis was derived from agents including phencyclidine (PCP) and dizocilpine (MK-801), which block the function of NMDA receptors and cause psychosis in humans, inducing schizophrenia-like symptoms [,,,]. It is also supported by previous findings showing decreased glutamate levels in the cerebrospinal fluid [] and the down-regulation of brain NMDA receptor expression in patients with schizophrenia [].
Furthermore, based on the glutamate hypothesis in schizophrenia, several agents (e.g., d-cycloserine (DCS), d-serine, and sodium benzoate), which stimulate NMDA receptor functions, are expected to have potential as new medication for schizophrenia [,,,]. However, their actions regarding the induction and/or modulation of EPS are unknown and remain to be clarified.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
In the present study, we performed behavioral and immunohistochemical studies in mice and rats to evaluate the effects of the glycine-binding site stimulants of NMDA receptors on antipsychotic-induced EPS (i.e., bradykinesia and catalepsy) and elucidate their action mechanisms.[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
## 2. Results
*In the **2. Results** section:*
## 2.1. Effects of Glycine-Site Stimulants of N-Methyl-d-aspartate (NMDA) Receptors on Haloperidol-Induced Bradykinesia
*In the **2.1. Effects of Glycine-Site Stimulants of N-Methyl-d-aspartate (NMDA) Receptors on Haloperidol-Induced Bradykinesia** section:*
We examined the effects of the glycine-site stimulants of NMDA receptors on haloperidol (HAL)-induced bradykinesia using the mouse pole test. The glycine-site agonist of NMDA receptors, DCS (3–30 mg/kg, i.p.), significantly improved HAL (1 mg/kg, i.p.)-induced bradykinesia in a dose-dependent manner (Tturn: F(3,44) = 7.8073, p = 0.0003, Ttotal: F(3,44) = 6.7772, p = 0.0007) (Figure 1A). d-serine (100–1000 mg/kg, i.p.) significantly attenuated HAL-induced bradykinesia (Tturn: X2 = 8.4239, df = 3, p = 0.0380), whereas glycine (30–300 mg/kg, i.p.) showed no effects (Figure 1B,C). In addition, the d-amino acid oxidase inhibitor, sodium benzoate (600 mg/kg, i.p.), also significantly reduced HAL-induced bradykinesia (Ttotal: X2 = 8.7330, df = 5, p = 0.0481) (Figure 1D).[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
We subsequently examined the effects of the NMDA receptor antagonist dizocilpine and nitric oxide synthase (NOS) inhibitor L-NG-Nitro-l-arginine methyl ester (L-NAME) on the antibradykinetic action of DCS. Dizocilpine (0.01 mg/kg, i.p.), L-NAME (10 mg/kg, i.p.), or vehicle was simultaneously injected with DCS (30 mg/kg, i.p.) 15 min before the HAL injection (1 mg/kg, i.p.). Under these conditions, the improvement of HAL-induced bradykinesia by DCS was significantly antagonized by dizocilpine (0.01 mg/kg, i.p.) (Tturn: U = 47, p = 0.0066) (Figure 2).
Similarly, L-NAME (10 mg/kg, i.p.) also significantly antagonized improvements by DCS (Tturn: U = 144.5, p = 0.0137) (Figure 2).[](https://www.ncbi.nlm.nih.gov/mesh/D016291)
## 2.2. Effects of d-Cycloserine on Haloperidol-Induced Fos Expression
*In the **2.2. Effects of d-Cycloserine on Haloperidol-Induced Fos Expression** section:*
It is well-known that HAL evokes Fos protein expression, a biological marker of neural excitation [], in the forebrain (e.g., striatum and nucleus accumbens) via dopamine D2 receptor blockade [,]. Therefore, we examined the effects of an anti-bradykinetic dose of DCS on HAL-induced Fos expression in the dorsolateral striatum (dlST) and shell region of the nucleus accumbens (AcS).[](https://www.ncbi.nlm.nih.gov/mesh/D006220)
As shown in Figure 3A, we confirmed that HAL (1 mg/kg, i.p.) markedly increased Tturn and Ttotal values, which were significantly reversed by DCS (30 mg/kg, i.p.). Brain samples were then obtained from these animals 2 h after the HAL injection and subjected to Fos immunohistochemistry. Under these conditions, control (vehicle + vehicle) and DCS (vehicle + DCS) animals showed negligible Fos expression in the dlST and AcS. The number of Fos immunoreactivity (IR)-positive cells was markedly increased by HAL (vehicle + HAL, dlST: F(3,21) = 14.9794, p = 0.0001, AcS: F(3,21) = 8.3832, p = 0.0030) (Figure 3B–D).
However, HAL-induced Fos expression was significantly inhibited by DCS in the dlST (p = 0.0324). The number of Fos-IR-positive cells in the dlST was approximately 31 cells/grid with vehicle + HAL; however, this value was reduced to approximately 16 cells/grid by the combined treatment with DCS (Figure 3B,C). Interestingly, DCS did not significantly affect HAL-induced Fos expression in the AcS (Figure 3B,D).[](https://www.ncbi.nlm.nih.gov/mesh/D006220)
## 2.3. Microinjection and In Vivo Microdialysis Studies with d-Cycloserine
*In the **2.3. Microinjection and In Vivo Microdialysis Studies with d-Cycloserine** section:*
In order to investigate the action sites of DCS, we conducted microinjection studies using rats. DCS (10 μg/site) or vehicle was microinjected into the bilateral substantia nigra pars compacta (SNc) or dlST 15 min after the HAL (1 mg/kg, i.p.) treatment. HAL-induced EPS was evaluated by the catalepsy test 30 min after the HAL injection. Under these conditions, microinjections of DCS into the SNc or dlST significantly attenuated HAL-induced catalepsy (Figure 4). The catalepsy time with HAL was significantly reduced from 243.6 ± 28.0 to 82.7 ± 39.9 (SNc: U = 17, p = 0.0123) and 283.8 ± 16.2 to 140.5 ± 56.0 (dlST: U = 7, p = 0.0495) by the microinjection of DCS into the SNc and dlST, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
We further conducted in vivo microdialysis studies to evaluate extracellular dopamine release in the dlST following the microinjection of DCS (10 μg/site) into the ipsilateral SNc (Figure 5). The microinjection of DCS into the SNc significantly enhanced dopamine release in the ipsilateral dlST. Extracellular dopamine levels were significantly elevated by approximately 25% by the microinjection of 10 μg DCS into the SNc (100 min: T = 2.2701, df = 10, p = 0.0466).[](https://www.ncbi.nlm.nih.gov/mesh/D004298)
## 3. Discussion
*In the **3. Discussion** section:*
The present study demonstrated that the glycine-binding site agonists of NMDA receptors, DCS and d-serine, significantly alleviated HAL-induced bradykinesia with relative potencies in the order of DCS > d-serine. Sodium benzoate, an inhibitor of the d-amino acid oxidase catalyzing the oxidative metabolism of d-serine [,], also alleviated HAL-induced bradykinesia. Thus, sodium benzoate appears to alleviate HAL-induced bradykinesia by increasing extracellular d-serine levels[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
Although d-serine-independent mechanisms (e.g., NMDA receptor redox-site intervention) are also proposed [,]. Among the glycine-binding site stimulants tested, glycine was inactive. The lack of efficacy with glycine may have been due to its poor penetration into the brain or an insufficient dosage. However, it was not possible to test higher doses of glycine because of its acute toxicity (occasional death at 1000 mg/kg).[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
Antibradykinetic doses of the glycine-binding site stimulants of NMDA receptors in the pole test were 3–30 mg/kg (i.p.) for DCS, 300 mg/kg (i.p.) for d-serine, and 600 mg/kg (i.p.) for sodium benzoate. These doses were similar to those producing efficacy in animal models of schizophrenia with NMDA antagonists (e.g., phencyclidine and dizocilpine), namely, 10–30 mg/kg (s.c.) for d-cycloserine, 600 mg/kg (i.p.) for d-serine, and 300–1000 mg/kg (p.o.) for sodium benzoate [,,]. Therefore, the glycine-binding site stimulants of NMDA receptors are expected to reduce EPS associated with antipsychotic treatments in clinical settings.[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
It has been well-documented that antipsychotics elevate the regional expression of the Fos protein, a biological marker of neural activation, both in the nucleus accumbens and striatum by blocking D2 receptors []. Furthermore, D2 receptor-mediated Fos expression in the nucleus accumbens and striatum are considered to reflect the antipsychotic action and EPS liability of antipsychotics, respectively [,,,,]. Second-generation antipsychotics (atypical antipsychotics) with fewer EPS commonly lead to reduced Fos expression in the striatum [,,,]. In the present study, we showed that DCS significantly reduced Fos expression in the dlST. This evidence further supports DCS counteracting striatal D2 receptor blockade by HAL to attenuate the induction of EPS.
The effects of DCS on Fos expression were region-specific and did not significantly alter HAL-induced Fos expression in the AcS. These results suggest that a combination of the glycine-binding site stimulants of NMDA receptors with antipsychotics preferentially attenuates EPS (D2 blocking action in the striatum) without interfering with the therapeutic action of antipsychotics.[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
In order to elucidate the action mechanisms of DCS, we first examined the effects of dizocilpine (NMDA antagonist) and confirmed that it antagonized improvements in HAL-induced bradykinesia by DCS, indicating that DCS alleviated EPS by activating NMDA receptors. Since NMDA receptor-mediated neurotransmission is known to be mediated by NO synthesis, we also examined the effects of the NOS inhibitor L-NAME and showed that it antagonized DCS-induced improvements in HAL-induced bradykinesia. Therefore, NMDA receptor-mediated NO production appears to be involved in the antiparkinsonian action of DCS. In addition, we performed microinjection studies in combination with in vivo microdialysis measurements of dopamine release in the striatum. Our results demonstrated that microinjections of DCS (10 µg/site/4 min) into the SNc or dlST significantly improved HAL-induced catalepsy in rats, indicating that SNc and dlST are both, at least partly, involved in the anti-cataleptic action of DCS.
In vivo microdialysis results also revealed that DCS locally injected into the SNc significantly enhanced dopamine release in the dlST. These results suggest that the stimulation of glycine-binding sites by DCS in the SNc activates the nigrostriatal dopamine pathway, which reduces EPS by elevating striatal dopamine increases. On the other hand, the mechanisms of action of DSC in the striatum currently remain unknown. Since the suppression of striatal medium spiny neurons leads to the amelioration of extrapyramidal disorders [], the activation of NMDA receptors by DCS may inhibit medium spiny neurons via inhibitory GABAergic interneurons.
Further studies are needed in order to elucidate the action mechanisms of DSC in the striatum in more detail.[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
In conclusion, we herein investigated the actions of glycine-binding site stimulants of NMDA receptors in the modulation of antipsychotic-induced EPS and showed that DCS, d-serine, and sodium benzoate significantly attenuated HAL-induced EPS. The anti-EPS action of DCS was antagonized by dizocilpine and L-NAME, suggesting that the activation of NMDA receptors and subsequent induction of NO were involved in the amelioration of EPS by DCS. In addition, DCS counteracted HAL-induced Fos expression in the striatum, but not in the nucleus accumbens, suggesting that the glycine-binding site stimulants of NMDA receptors preferentially attenuate the D2 blocking action of antipsychotics in the striatum compared with that in the striatum. Furthermore, the microinjection of DCS into the SNc effectively improved HAL-induced EPS concomitant with elevations in dopamine release in the striatum. The present results suggest that the stimulation of glycine-binding sites of NMDA receptors alleviates antipsychotic-induced EPS by activating nigrostriatal dopamine neurons.
Based on the glutamate hypothesis, the glycine-binding site stimulants of NMDA receptors are expected to become a new medication for schizophrenia []. Clinical studies showed that several agents including DCS and sodium benzoate improved negative symptoms and/or cognitive impairment in patients with schizophrenia [,,,]. The present results suggest that glycine-binding site agonists are beneficial not only for efficacy, but also side-effect management in the treatment of schizophrenia. However, due to the limitations associated with animal experiments, further clinical studies are needed in order to validate the EPS liability and/or antiparkinsonian effects of glycine-binding site stimulants of NMDA receptors.[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
## 4. Materials and Methods
*In the **4. Materials and Methods** section:*
## 4.1. Animals
*In the **4.1. Animals** section:*
Male ddY mice and SD rats (Japan SLC, Shizuoka, Japan) at 8–10 weeks of age were used. Animals were kept in air-conditioned rooms (24 ± 2 °C and 50 ± 10% relative humidity) under a 12-h light/dark cycle (light on: 8:00–20:00) and allowed free access to food and water. Animal care methods complied with the Guide for the Care and Use of Laboratory Animals of the Ministry of Education, Science, Sports and Culture of Japan, and experimental protocols were approved by the Experimental Animal Research Committee at Osaka University of Pharmaceutical Sciences (#17, 30 March 2015).[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## 4.2. Evaluation of Bradykinesia
*In the **4.2. Evaluation of Bradykinesia** section:*
The pole test was performed as described previously []. Mice were placed at the top (head-upward) of a pole (diameter: 8 mm and height: 45 cm). The time for the animal to rotate downward (Tturn) and descend to the floor (Ttotal) was then measured with a cut-off time of 90 s. Only mice that showed Tturn < 8 s and Ttotal < 18 s in the pre-test trial (typically performed 2 h before the test trial) were used.
The glycine-site stimulants of NMDA receptors, DCS (3–30 mg/kg, i.p.), d-serine (100–1000 mg/kg, i.p.), and glycine (30–300 mg/kg, i.p.), and the d-amino acid oxidase inhibitor, sodium benzoate (10–600 mg/kg, i.p.) were administered to animals 15 min before the HAL injection, and the pole test was performed 30 min later. In experiments using dizocilpine or l-NAME, mice first received dizocilpine (0.01 mg/kg, i.p.), l-NAME (10 mg/kg, i.p.), or vehicle simultaneously injected with DCS or vehicle 15 min before the HAL injection (1 mg/kg, i.p.). The pole test was performed 30 min after the HAL injection.[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
## 4.3. Analysis of Fos Protein Expression
*In the **4.3. Analysis of Fos Protein Expression** section:*
Regarding Fos immunohistochemical staining, brain samples were obtained from mice 120 min after the HAL injection. Under pentobarbital (80 mg/kg, i.p.) anesthesia, all mice were transcardially perfused with ice-cold phosphate-buffered saline (PBS), which was followed by 4% formaldehyde perfusion. Brains were removed from the skull and stored in fresh fixative for at least 24 h.[](https://www.ncbi.nlm.nih.gov/mesh/D006220)
Fos immunohistochemical staining was performed using previously reported methods [,]. Coronal sections (thickness: 30 μm) were cut from the brain using a Microslicer (DSK-3000, Dosaka, Kyoto, Japan). Slices were incubated for 2 h with 2% normal rabbit serum, and with goat c-Fos antiserum (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) for an additional 18–36 h. Sections were then incubated with a biotinylated rabbit anti-goat IgG secondary antibody (Vector Laboratories, Burlingame, CA, USA) for 2 h. After a 30 min incubation with 0.3% hydrogen peroxide for 30 min to inactivate endogenous peroxidase, sections were incubated for 2 h with an avidin–biotinylated horseradish peroxidase complex (Vectastain ABC Kit, Vector Laboratories, Burlingame, CA, USA). Fos-IR was visualized using the diaminobenzidine–nickel staining method and quantified by counting the number of Fos-IR positive nuclei in dlST and AcS.[](https://www.ncbi.nlm.nih.gov/mesh/D006861)
## 4.4. Microinjection Study
*In the **4.4. Microinjection Study** section:*
DCS was microinjected into the SNc (−6.0 mm anterior to the bregma, ±2.2 mm lateral to the midline, 6.2 mm inferior to the brain surface) or dlST (+1.0 mm anterior to the bregma, ±1.0 mm lateral to the midline, 3.5 mm inferior to the brain surface) [] in rats, according to a previously reported method []. Under pentobarbital (80 mg/kg, i.p.) anesthesia, male SD rats were fixed in a stereotaxic frame (Narishige, SR-6, Tokyo, Japan). Stainless steel-guide cannulae were then inserted into a position 1 mm above the bilateral SNc or dlST and fixed to the skull with dental cement. After a recovery period (ca. 1 week), animals were subjected to microinjection experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
On the day of the experiment, injection cannulae were inserted into the SNc or dlST through the guide cannulae. Under freely-moving conditions, DCS (10 μg/1 μL/site) was injected into the SNc or dlST at a flow rate of 0.25 μL/min for 4 min using a microinjection pump (KDS220; Kd Scientific Inc., Holliston, MA, USA). Control animals were given the same volume of vehicle alone. Fifteen min after the DCS microinjection, animals were injected with HAL (1 mg/kg, i.p.) and, 30 min later, subjected to the catalepsy test in order to evaluate the induction of EPS [].
When the same animals were treated with a different drug (or vehicle) solution, the microinjection study was performed after a drug withdrawal period of at least 4 days. After experiments, the brain was removed from animals under pentobarbital (80 mg/kg, i.p.) anesthesia, and the position of each injection site was checked.[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
## 4.5. In Vivo Microdialysis Study
*In the **4.5. In Vivo Microdialysis Study** section:*
Under pentobarbital (40 mg/kg, i.p.) anesthesia, male SD rats were fixed in a stereotaxic instrument (Narishige, SR-6, Tokyo, Japan). A guide cannula (diameter: 1 mm) was inserted into a position 1 or 2 mm above the unilateral SNc or dlST and fixed to the skull using dental cement. After a recovery period (ca. 1 week), animals were subjected to in vivo microdialysis experiments. Briefly, a dialysis probe (Eicom, A-I-6-02, Kyoto, Japan) was inserted into the dlST through a guide cannula and artificial cerebrospinal fluid (aCSF), containing NaCl 140 mM, KCl 2.4 mM, MgCl2 1.0 mM, CaCl2 1.2 mM, and NaHCO3 5.0 mM, was perfused at a flow rate of 1.5 μL/min using a microperfusion pump (Eicom, ESP-32, Kyoto, Japan). Under freely-moving conditions, animals were given DCS: 10 μg/1 μL/site was slowly injected into the SNc at flow rate of 0.25 μL/min for 4 min using a microinfusion pump (KDS220; Kd Scientific Inc., Holliston, MA, USA).
Dialysate samples were collected into a microtube every 10 min (15 μL/sample), and analyzed for dopamine levels using an HPLC-ECD system. The mobile phase consisted of 0.1 M acetic acid-citric acid buffer, 190 mg/L 1-octanesulfonic acid sodium, 5 mg/L EDTA 2 Na, pH 3.5, with 16% methanol pumped at a flow rate of 230 μL/min. All data were analyzed using eDAQ Power Chrom (eDAQ Pty Ltd., Denistone East, NSW, Australia). Extracellular dopamine levels were expressed as a percentage of the basal control level (steady state), which was the mean of six points before the application of DCS, in each animal.[](https://www.ncbi.nlm.nih.gov/mesh/D010424)
After experiments, animals were deeply anesthetized with pentobarbital (80 mg/kg, i.p.) and the brain was removed from the skull. Coronal sections (thickness of 100 µm) were prepared from each brain using a microslicer (DSK, Kyoto, Japan) and the position of each injection site was checked.[](https://www.ncbi.nlm.nih.gov/mesh/D010424)
## 4.6. Drugs
*In the **4.6. Drugs** section:*
HAL, DCS, d-serine, glycine, sodium benzoate, dizocilpine, L-NAME, and CNQX were purchased from Sigma-Aldrich (St. Louis, MO, USA). The Vectastain ABC kit and DAB substrate were purchased from Vector Laboratories (Burlingame, CA, USA). All other reagents were obtained from commercial sources. HAL was dissolved in 1% lactate solution and then diluted with physiological saline. Other agents were dissolved in physiological saline.
All drugs were injected intraperitoneally in a volume of 5 mL/kg into mice or 1 mL/kg into rats.[](https://www.ncbi.nlm.nih.gov/mesh/D006220)
## 4.7. Statistical Analysis
*In the **4.7. Statistical Analysis** section:*
Data are expressed as the mean ± S.E.M. The significance of differences among multiple groups was assessed by a one-way ANOVA followed by Tukey’s test or Kruskal−Wallis test (nonparametric one-way ANOVA) followed by the Steel−Dwass post-hoc test. Comparisons between only two groups were performed by the non-parametric Mann-Whitney U-test or parametric Student’s t-test. A p-value of less than 0.05 was considered significant.
# Author Contributions
*In the **Author Contributions** section:*
Yukihiro Ohno and Saki Shimizu designed the experiments. Saki Shimizu, Shunsaku Sogabe, Ryoto Yanagisako, Akiyoshi Inada, Megumi Yamanaka and Higor A. Iha performed the experiments and analyzed the data. Saki Shimizu and Yukihiro Ohno wrote the manuscript.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflicts of interest.
# Abbreviations
*In the **Abbreviations** section:*
AcS Shell region of the nucleus accumbens DCS d-Cycloserine dlST Dorsolateral striatum EPS Extrapyramidal side effects HAL Haloperidol IR Immunoreactivity L-NAME L-NG-Nitro-l-arginine methyl ester NMDA N-Methyl-d-aspartate NOS Nitric oxide synthase SNc Substantia nigra pars compacta[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
# References
*In the **References** section:*
Effects of glycine-site agonists of N-methyl-d-aspartate (NMDA) receptors and the d-amino acid oxidase inhibitor on haloperidol (HAL)-induced bradykinesia. (A–D) Glycine-site agonists of NMDA receptors, d-cycloserine (3–30 mg/kg, i.p.), d-serine (100–1000 mg/kg, i.p.), and glycine (30–300 mg/kg, i.p.), and the d-amino acid oxidase inhibitor, sodium benzoate (10–600 mg/kg, i.p.), were administered to animals 15 min before the HAL injection. The pole test was performed 30 min after the HAL injection. Each column represents the mean ± S.E.M. of 5–13 mice. These data were analyzed using the Kruskal−Wallis and Steel−Dwass tests. * p < 0.05, ** p < 0.01, significantly different from the value with HAL alone.[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
Effects of dizocilpine or L-NG-Nitro-l-arginine methyl ester (L-NAME) on the ameliorative action of d-cycloserine (DCS) against haloperidol (HAL)-induced bradykinesia. (A,B) Mice received dizocilpine (0.01 mg/kg, i.p.), L-NAME (10 mg/kg, i.p.), or vehicle simultaneously with DCS (30 mg/kg, i.p.) or vehicle 15 min before the HAL injection (1 mg/kg, i.p.). The pole test was performed 30 min after the HAL injection. Each column represents the mean ± S.E.M. of 12–23 mice. These data were analyzed using the Mann-Whitney U-test. * p < 0.05: Significantly different from the value with HAL alone. # p < 0.05, significantly different from the value with HAL + DCS.[](https://www.ncbi.nlm.nih.gov/mesh/D016291)
(A) Effects of d-cycloserine (DCS) on haloperidol (HAL)-induced Fos expression in the dorsolateral striatum (dlST) and accumbens shell (AcS). (A) DCS (30 mg/kg, i.p.) or vehicle was administered to animals 15 min before the HAL injection, which was followed by the pole test 30 min later; (B) Photographs illustrating Fos-IR-positive cells in the dlST and AcS (left panel: vehicle + HAL (1 mg/kg, i.p.)-treated mice, right panel: DCS (30 mg/kg, i.p.) + HAL (1 mg/kg, i.p.)-treated mice). Scale bar: 100 μm; (C,D) Effects of DCS (30 mg/kg, i.p.) on HAL (1 mg/kg, i.p.)-induced Fos expression in the dlST (C) and AcS (D). The brain was removed from animals 2 h after the HAL injection. Each column represents the mean ± S.E.M. of 6–7 mice. These data were analyzed using the Kruskal−Wallis and Steel−Dwass tests (behavioral test) or one-way ANOVA and Tukey’s test (Fos analysis). * p < 0.05, ** p < 0.01, significantly different from the value for vehicle + vehicle. # p < 0.05, ## p < 0.01, significantly different from the value for vehicle + HAL.[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
Effects of intranigral and intrastriatal microinjections of d-cycloserine (DCS) on haloperidol (HAL)-induced catalepsy in rats. (A,B) The effects of DCS (10 µg/1 µL/side) microinjected into the substantia nigra compacta (SNc) or dorsolateral striatum (dlST) against HAL-induced catalepsy were examined. Each dose of HAL was administered 15 min after each DCS microinjection and, 30 min later (45 min after the DCS microinjection), the catalepsy time was measured. Schematic drawings of a rat brain section illustrating DCS or vehicle injection sites (filled circles) in the SNc (A) or dlST (B) are shown at the top. Each column represents the mean ± S.E.M. of 6–11 rats. These data were analyzed by the Mann-Whitney U-test. * p < 0.05, significantly different from the control value with vehicle + HAL.[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
Effects of an intranigral microinjection of d-cycloserine (DCS) on dopamine release in the rat striatum. (A) Schematic drawing illustrating the microinjection site of DCS in the substantia nigra pars compacta (SNc) and microdialysis site in the dorsolateral striatum (dlST); (B) Extracellular levels of dopamine were analyzed in 10-min dialysate samples. Data were normalized to the mean value of the first six 10-min samples (basal value). Each column shows the mean ± S.E.M. of 6 rats. These data were analyzed by the Student’s t-test. * p < 0.05, ** p < 0.01, significantly different from the basal value.[](https://www.ncbi.nlm.nih.gov/mesh/D003523)
|
# Introduction
Enhanced Susceptibility of Ogg1 Mutant Mice to Multiorgan Carcinogenesis
# Abstract
*In the **Abstract** section:*
The role of deficiency of oxoguanine glycosylase 1 (Ogg1) Mmh homolog, a repair enzyme of the 8-hydroxy-2’-deoxyguanosine (8-OHdG) residue in DNA, was investigated usi[ng the multiorgan carcinoge](https://www.ncbi.nlm.nih.gov/mesh/C067134)ne[sis bi](https://www.ncbi.nlm.nih.gov/mesh/C067134)oassay in mice. A total of 80 male and female six-week-old mice of C57BL/6J background carrying a mutant Mmh allele of the Mmh/Ogg1 gene (Ogg1−/−) and wild type (Ogg1+/+) mice were administered N-diethylnitrosamine (DEN), N-methyl-N-nitrosourea (MNU), N-butyl-N-(4-hy[droxybutyl) nitrosam](https://www.ncbi.nlm.nih.gov/mesh/D004052)in[e (](https://www.ncbi.nlm.nih.gov/mesh/D004052)BBN[), N-bis (2-hydroxypro](https://www.ncbi.nlm.nih.gov/mesh/D008770)py[l) ](https://www.ncbi.nlm.nih.gov/mesh/D008770)nit[rosamine (DHPN) and 1,2-dimethylhydraz](https://www.ncbi.nlm.nih.gov/mesh/D002085)in[e d](https://www.ncbi.nlm.nih.gov/mesh/D002085)ihy[drochloride (DMH) (DMBDD) to induce](https://www.ncbi.nlm.nih.gov/mesh/C012457) c[arci](https://www.ncbi.nlm.nih.gov/mesh/C012457)nogene[sis in multiple organs, and observed ](https://www.ncbi.nlm.nih.gov/mesh/D019813)up[ to](https://www.ncbi.nlm.nih.gov/mesh/D019813) 34[ week](https://www.ncbi.nlm.nih.gov/mesh/C012457)s. Significant increase of lung adenocarcinomas incidence was observed in DMBDD-treated Ogg1−/− male mice, but not in DMBDD-administered Ogg1+/+ an[imals](https://www.ncbi.nlm.nih.gov/mesh/C012457). Furthermore, incidences of lung adeno[mas w](https://www.ncbi.nlm.nih.gov/mesh/C012457)ere significantly elevated in both Ogg1−/− males and females as compared with respective Ogg1−/− control and DMBDD-treated Ogg1+/+ groups. Incidence of total liver tumors (hepatocell[ular ](https://www.ncbi.nlm.nih.gov/mesh/C012457)adenomas, hemangiomas and hemangiosarcomas) was significantly higher in the DMBDD-administered Ogg1−/− males and females. In addition, in DMBDD-treat[ed ma](https://www.ncbi.nlm.nih.gov/mesh/C012457)le Ogg1−/− mice, incidences of colon adenomas and total c[olon ](https://www.ncbi.nlm.nih.gov/mesh/C012457)tumors showed a trend and a significant increase, respectively, along with significant rise in incidence of simple hyperplasia of the urinary bladder, and a trend to increase for renal tubules hyperplasia in the kidney. Furthermore, incidence of squamous cell hyperplasia in the forestomach of DMBDD-treated Ogg1−/− male mice was significantly higher than that of Ogg[1+/+ ](https://www.ncbi.nlm.nih.gov/mesh/C012457)males. Incidence of small intestine adenomas in DMBDD Ogg1−/− groups showed a trend for increase, as compared to the wild[ type](https://www.ncbi.nlm.nih.gov/mesh/C012457) mice. The current results demonstrated increased susceptibility of Ogg1 mutant mice to the multiorgan carcinogenesis induced by DMBDD. The present bioassay could become a useful tool to examine the inf[luenc](https://www.ncbi.nlm.nih.gov/mesh/C012457)e of various targets on mouse carcinogenesis.
## 1. Introduction
*In the **1. Introduction** section:*
DNA damage and disruption of DNA repair are considered key factors in the susceptibility of mammals to endogenous and exogenous carcinogens, as well as processes of aging and cancer development []. The oxidative DNA damage includes a variety of oxidative lesions in DNA and the main attack site of reactive oxygen species (ROS) is at the 8 position of guanine, producing strongly mutagenic base 8-hydroxy-2′-deoxyguanosine (8-OHdG) []. 8-OHdG is used as an oxidative DNA damage marker which mispairs with adenine (A) residues, thus resulting in increase of spontaneous G:C to T:A transversion mutations [].[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
Three DNA repair enzymes from various bacteria and Saccharomyces cerevisiae, namely, the MutM (Fpg), MutY and MutT DNA glycosylase homologs are known to prevent spontaneous mutagenesis induced by 8-OHdG []. In mammalian cells, the MutM homolog (MMH; the glycosylase/apurinic, apyrimidinic (AP) lyase), MutY and MutT homolog enzymes have also been identified [,,]. In both mammalian and yeast cells, cloned human and mouse cDNAs encode distinct nuclear and mitochondrial forms of the DNA glycosylase, the product of the Ogg1 gene, which is generated by alternative RNA splicing [,,].
MutY and MutM homologs prevent G:C to T:A transversions in DNA, while MutT protein hydrolyzes 8-oxo-dGTP to 8-oxo-dGMT and pyrophosphate, thus avoiding the occurrence of A:T to C:G transversion mutations during DNA replication [,]. Analysis of the mutation spectrum revealed that the frequency of G:C to T:A transversions increased five-fold in Ogg1 mutant mice compared with wild-type animals [].[](https://www.ncbi.nlm.nih.gov/mesh/C067134)
Mmh/Ogg1 homozygous mutant (Ogg1−/−) mice used in our studies have physically normal appearance but exhibit three- and seven-fold increased accumulation of 8-OHdG adduct at 9 and 14 weeks of age, respectively, in comparison with or heterozygous or wild-type animals []. We have previously demonstrated that treatment of Ogg1−/− mice with dimethylarsinic acid (DMA) and phenobarbital (PB) for 78 weeks resulted in enhancement of lung and liver carcinogenesis, respectively [,]. The tremendous increase of 8-OHdG levels with consequent G:C to T:A transversions and deletions in the kidney DNA of Ogg1−/− mice were reported following administration of potassium bromate (KBrO3) []. Furthermore, a significant increase of mutation frequency in Ogg1−/− mice livers was observed during liver regeneration after partial hepatechtomy following KBrO3 treatment [].
In addition, Sakumi et al. and Xie et al. demonstrated spontaneous development of lung, ovary tumors and lymphomas in Myh and Ogg1 knockout mice [,]. However, it is still unknown how the ablation of these enzymes affects the tumorigenicity of various chemical carcinogens.[](https://www.ncbi.nlm.nih.gov/mesh/C067134)
Previously, several in vivo bioassay systems for carcinogenicity detection of test compounds have been developed. However, these bioassays usually predict carcinogenicity of test chemicals only in single organs with known strategies of carcinogenesis initiation. To develop the experimental approach for the determination of carcinogenicity in numerous target organs, multiorgan wide-spectrum initiation bioassay (namely, the multiorgan carcinogenicity bioassay: DMBDD model) has been established [,,,]. This bioassay was applied in rats and included treatment with five genotoxic carcinogens, N-diethylnitrosamine (DEN), N-methyl-N-nitrosourea (MNU), N-butyl-N-(4-hydroxybutyl) nitrosamine (BBN), N-bis (2-hydroxypropyl) nitrosamine (DHPN) and 1,2-dimethylhydrazine dihydrochloride (DMH) (DMBDD), as initiators of liver, lungs, kidneys, urinary bladder, stomach, small intestine, colon and thyroid gland carcinogenesis [,,].
It has been demonstrated that DMBDD-induced organ-specific DNA damage could be attributed to free radicals, methylation, and accumulation of non-repaired DNA damage []. In rats, DEN is usually used as initiator of liver carcinogenesis, BBN as initiator of bladder carcinogenesis, DMH as initiator of intestine carcinogenesis, and MNU as initiator of stomach, bladder and liver carcinogenesis []. DHPN is a wide-spectrum carcinogen in rats, which induces lung, thyroid, kidney, bladder and liver cancers [,]. In previous studies, DMBDD treatment has been proposed to inactivate the tumor suppressor p53 in the bladder tumors of Zucker diabetic rats []. However, to our knowledge, the DMBDD model was never applied in mice and there is no information how the DMBDD treatment influences oncogenes and tumor suppressor genes.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
The aim of the present study was to investigate the differences in susceptibility of Ogg1 mutant and wild type mice of C57BL/6J background to the treatment with five types of genotoxic carcinogens (DEN, MNU, BBN, DHPN and DMH: DMBDD) by applying the multiorgan carcinogenesis bioassay in mice.[](https://www.ncbi.nlm.nih.gov/mesh/D004052)
## 2. Results
*In the **2. Results** section:*
## 2.1. General Observations
*In the **2.1. General Observations** section:*
All control Ogg1−/− male or female mice were alive at the end of the study. They were healthy and long-lived as compared to the control Ogg1+/+ mice. Three DMBDD-treated Ogg1 knockout male and three female mice were found moribund at Weeks 11, 12, and 17, and 10, 20 and 25, respectively. The causes of death of Ogg1−/− male mice were malignant lymphomas/leukemia, lung adenocarcinoma and fibrosarcoma, while Ogg1−/− female mice died due to the development of lymphoma/leukemia. Four DMBDD-administered Ogg1+/+ male and one female mice died at Weeks 16, 29, 33, 35 and 32, respectively.
The main causes of death in male and female wild type mice were malignant lymphoma/leukemia, T cell lymphoma and bladder transitional cell carcinoma (TCC). One non-treated control Ogg1+/+ male mouse was found moribund at Week 27 due to a urinary tract infection.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
As lung, liver and colon tumors were observed in the DMBDD-treated Ogg1−/− mice that were found moribund during the study, effective number of animals used for the histopathological analysis included all mice. Body weight and survival curves, final body weight and absolute and relative organ weights of mice are shown in Table 1 and Figure 1. Body weights of control Ogg1 mutant male and female mice were significantly lower than those of wild type mice all through the experiment. DMBDD treatment induced significant decrease of body weight of both Ogg1 mutant and wild type mice, However, at Experimental Week 14, mean body weight of Ogg1−/− mice became equal to that of the corresponding control animals of the same genotype, while the body weight of the DMBDD-administered Ogg1+/+ mice continued to be significantly lower compared to the control Ogg1+/+ until the end of the study (Figure 1A). Therefore, at Week 34, final body weight of the DMBDD-treated Ogg1+/+ but not Ogg1−/− mice were significantly decreased as compared with the control mice of the same genotype.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
DMBDD administration inhibited food intake of the Ogg1+/+, but not the Ogg1−/− mice compared with the control mice of the same genotype (Figure S1A). Water intakes were similarly decreased in all DMBDD-treated Ogg1−/− and Ogg1+/+ animals (Figure S1B). Thus, the body weight of the DMBDD-administered Ogg1+/+ mice appeared to be significantly lower than the control Ogg1+/+ due to the inhibited food intake. The absolute and relative liver, kidneys and spleen weights of the control Ogg1−/− male and female mice were significantly lower than those of the control Ogg1+/+ mice (Table 1). DMBDD treatment induced significant increases of relative liver, kidneys and spleen weights of Ogg1 mutant but not wild type both male and female mice in comparison with corresponding Ogg1−/− controls.
The absolute and relative weights of the lungs were significantly increased in both Ogg1−/− and Ogg1+/+ male and female mice (Table 1).[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
## 2.2. Survival Curves
*In the **2.2. Survival Curves** section:*
Survival curves for the DMBDD-administered and control Ogg1−/− and Ogg1+/+ mice are presented in Figure 1B. In the present model, there were no significant differences in survival between the control Ogg1 homozygous mutant and wild type mice. Trends for decrease in survival were observed in both Ogg1−/− and Ogg1+/+ DMBDD-treated animals. However, in DMBDD-treated Ogg1−/− male and female mice earlier decrease in survival (males: Week 11; females: Week 10), respectively, as compared to the wild type mice (males: Week 16; females: Week 32) was found (Figure 1B). Importantly, the earlier development of tumors in DMBDD-administered Ogg1−/− males and females, mostly malignant lymphomas/leukemias, lung adenocarcinoma and subcutaneous tumors (fibrosarcomas) was the reason for their earlier mortality.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
## 2.3. Results of Histopathological Examination
*In the **2.3. Results of Histopathological Examination** section:*
Table 2 summarizes the data on the incidence of preneoplastic, neoplastic and some non-neoplastic lesions and general distribution of tumors induced by DMBDD administration in Ogg1 knockout and wild type mice. Representative pictures of neoplastic lesions observed in the lungs, livers and colons of mice are presented in Figure 2. Neoplastic nodules induced in the DMBDD-treated group of Ogg1−/− and Ogg1+/+ mice were mainly lung, liver, colon, small intestine, urinary bladder tumors, malignant lymphomas/leukemias and subctaneous tumors (fibrosarcomas). In male Ogg1+/+ mice higher number of tumors were induced by the DMBDD treatment as compared to the Ogg1+/+ females likely due to the lower susceptibility to genotoxic carcinogens in females.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Macroscopically, no tumors were found in the non-treated control Ogg1+/+ mice, however, spontaneous development of lung nodules (hyperplasia and adenoma) was detected in the control Ogg1−/− animals. Furthermore, DMBDD-treated Ogg1−/− mice were more susceptible to the induction of different tumors as compared to Ogg1−/− control and wild type mice. DMBDD treatment induced elevation of total tumor incidence and number of tumor bearing mice in mutant, predominantly Ogg1−/− female animals (males, 100%, 5.9 ± 3.5/mouse; females, 100%, 4.6 ± 2.1/mouse, p < 0.0001), as compared to the wild type mice (males, 85%, 3.5 ± 2.8/mouse; females, 40%, 0.7 ± 0.9/mouse). All DMBDD-treated male and female Ogg1−/− mice developed many nodules in the lungs, while incidences and multiplicities of lung nodules in DMBDD-treated Ogg1+/+ was lower as compared to the Ogg1−/− DMBDD-administered animals. Furthermore, the incidence of liver lesions in Ogg1−/− mice, as well as their multiplicity was also increased after carcinogens treatment as compared to the corresponding controls.
Moreover, in male, but not female DMBDD-treated Ogg1−/− mice, incidences of colon tumors and fibrosarcomas, were elevated.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Histopathological examination demonstrated significant elevation of incidences and multiplicities of lung adenoma and total lung tumors in DMBDD-treated Ogg1−/− male and female mice lungs (total tumors: males, 100%, 4.1 ± 2.7/mouse; females, 100%, 4.1 ± 2.2/mouse; adenoma: males, 100%, 4.0 ± 2.3/mouse; females, 95%, 2.7 ± 2.8/mouse) as compared to both respective Ogg1−/− controls (total tumors: males, 5%, 0.1 ± 0.2/mouse; females, 0%, 0/mouse; adenoma: males, 5%, 0.1 ± 0.2/mouse; females, 0%, 0/mouse ) and DMBDD-treated Ogg1+/+ mice (total tumors: males, 75%, 3.0 ± 2.7/mouse; females, 60%, 1.4 ± 2.2/mouse; adenoma: males, 70%, 2.7 ± 2.8/mouse; females, 60%, 1.4 ± 2.0/mouse) (Table 2). Interestingly, significant increase of lung adenocarcinoma incidence and multiplicity was found in DMBDD-administered Ogg1−/− male mice (35%, 0.5 ± 0.7/mouse, p < 0.01), but not in the wild type males (10%, 0.1 ± 0.3/mouse) as compared to corresponding controls of the same genotype. Furthermore, incidences and multiplicities of lung adenocarcinoma were higher in female Ogg1−/− mice of DMBDD group as compared to the DMBDD-treated wild type counterparts. In addition, increases of lung hyperplasia incidences due to the DMBDD application were observed in both Ogg1 homozygous mutant and wild type mice.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
In the liver of Ogg1 knockout and wild type mice, DMBDD treatment caused development of putative preneoplastic foci of mostly basophilic phenotype, hepatocellular adenomas (HCAs), hemangiomas and hemangiosarcomas. Importantly, hemangiosarcomas were detected only in the Ogg1−/− mice (males, 15%, 0.6 ± 1.4/mouse; females, 5%, 0.1 ± 0.2/mouse). No hepatocellular carcinomas were apparent in DMBDD-treated Ogg1 knockout and wild type animals. Increases of HCA incidences (males, 20%; females, 20%; p = 0.05) and multiplicities (males: 0.3 ± 0.6/mouse, p = 0.05; females, 0.2 ± 0.4) were detected in livers of the DMBDD-initiated Ogg1−/− mice as compared to the Ogg1−/− controls (males, 0%, 0.1 ± 0.2/mouse; females, 0%, 0/mouse) and DMBDD-treated Ogg1+/+ groups (males, 5%, 0.1 ± 0.2/mouse; females, 0%, 0/mouse). Significant elevations of total liver tumor incidences were observed in Ogg1−/− males (25%, p < 0.05) and females (25%, p < 0.05), but not Ogg1+/+ mice as compared to the corresponding controls of the same genotype (0%).
Furthermore, a trend for increase and a significant elevation of total liver tumors multiplicity were observed in DMBDD-treated Ogg1−/− males (0.9 ± 2.0/mouse, p = 0.05) and females (0.3 ± 0.6/mouse, p < 0.05), in respect of control Ogg1−/− group. In addition, DMBDD administration caused elevation of bile duct proliferation in the liver of Ogg1−/− mice as compared to the Ogg1−/− and Ogg1+/+ counterparts. Significant increases of biliary cysts formation in the DMBDD groups were observed in the liver of both Ogg1−/− and Ogg1+/+ animals in respect of corresponding controls (Table 2).[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
In kidneys, significant increase of renal tubular degeneration (80%, p < 0.05) and a trend for increase of tubular renal cell HPL was found in the DMBDD-treated Ogg1−/− male mice as compared to corresponding controls of the knockout and wild type genotypes (Table 2). Only one Ogg1+/+ DMBDD-treated mouse developed renal adenoma.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
In the urinary bladder, significant increase of simple hyperplasia (25%, p < 0.05) and a trend for increase of papillary and nodular (PN) hyperplasia (20%) incidences as compared to the corresponding controls of the same genotype was detected (Table 2).
The incidence of total colon tumors (25%, p < 0.05) was significantly increased in the DMBDD-treated Ogg1−/− male mice but not in the DMBDD-administered wild type males (10%). Development of adenocarcinoma (5%) was found in one male Ogg1−/− mouse of the DMBDD group. Furthermore, incidences of small intestine total tumors showed a trend for increase in the DMBDD-administered Ogg1−/− male mice (20%) as compared to the Ogg1−/− control group and DMBDD-treated Ogg1+/+ animals (5%). In females, incidences and multiplicities of colon tumors induced by DMBDD were comparable with that of observed in wild type mice, pointing out the sex differences in susceptibility to colonic tumorigenesis.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
In the forestomach, the DMBDD treatment resulted in significant elevation of the squamous cell HPL incidence in Ogg1−/− (male, 70%, p < 0.0001; female, 55%, p < 0.001) and Ogg1+/+ (male, 35%, p < 0.01; female, 40%, p < 0.01) mice as compared to the corresponding controls of the same genotype. Furthermore, in male mice, it was significantly increased in comparison to wild type DMBDD-treated males (p < 0.05).[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Trends for increase of malignant lymphomas/leukemias were observed in Ogg1 homozygous mutant males (15%) and females (20%) treated with DMBDD, as compared to wild type mice (Table 2). One Ogg1+/+ female mouse in DMBDD group developed T cell lymphoma (5%).[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
## 2.4. Blood Biochemistry
*In the **2.4. Blood Biochemistry** section:*
The results of the blood biochemistry analysis are shown in Table 3. Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels in the blood of both Ogg1 null and wild type mice showed strong trends for increase, or were significantly elevated by the DMBDD treatment. Furthermore, this induction was higher in Ogg1−/− animals. Serum sodium (Na) levels were elevated by the DMBDD administration in both Ogg1 mutant and wild type mice. Moreover, creatinine level was higher in the blood of DMBDD-treated Ogg1−/− mice as compared to the respective Ogg1−/− control groups.
Alkaline phosphatase (ALP), T-cholesterol and chloride (Cl) levels were lowered in the wild type DMBDD-treated animals, but not altered in DMBDD Ogg1−/− mice. Serum calcium (Ca) level was significantly decreased in the DMBDD-treated Ogg1 knockout male mice as compared to the wild type males administered DMBDD. In addition, inorganic phosphorus (IP) levels showed a trend for increase or the significant elevation in the blood of DMBDD-treated and control Ogg1−/− male and female mice, respectively, as compared to the wild type groups receiving the same treatment.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
In the blood serum of Ogg1 mutant and wild type mice, levels of total protein showed a trend (Ogg1−/− males, females and Ogg1+/+ males) and a significant decrease (Ogg1+/+ females) as compared to the non-treated respective control groups. Furthermore, albumin levels were lower in DMBDD-treated Ogg1−/− and Ogg1+/+ groups, with significant differences observed for Ogg1−/− and Ogg1+/+ DMBDD-administered males. Albumin/globulin (A/G) ratio was significantly lower in the Ogg1+/+ male DMBDD group.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
## 3. Discussion
*In the **3. Discussion** section:*
The present study revealed that Ogg1 mutant mice are more susceptible to the induction of tumors due to the treatment with DMBDD, than wild type C57Bl/6J mice. In the DMBDD-treated Ogg1−/− mice, main causes of death besides malignant lymphoma/leukemia were lung adenocarcinoma and skin/subcutis fibrosarcoma, while Ogg1+/+ animals died from malignant lymphoma/leukemia and urinary bladder carcinoma. Furthermore, the earlier mortality of DMBDD-administered Ogg1−/− mice appeared to be due to the earlier tumor development. Importantly, DMBDD caused significant increases of incidences and multiplicities of lung adenocarcinoma in Ogg1−/− males, liver tumors in Ogg1−/− males and females and colon tumors in Ogg1−/− male mice as compared to the Ogg1−/− controls. In the kidneys, urinary bladder, stomach, small intestine and subcutis of Ogg1 mutant mice, increases of carcinogenicity as compared to the DMBDD-treated wild type animals were obvious.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Lungs of Ogg1 null mice were strongly affected by DMBDD initiation, which could be concluded from significant increases of lung adenocarcinoma incidence in DMBDD-treated Ogg1−/− male mice and incidences and multiplicities of adenomas and total lung tumors in Ogg1−/− males and females. As lungs are strongly exposed to molecular oxygen, it is likely the most carcinogenicity sensitive organ in Ogg1 knockouts. Furthermore, in the lung of non-treated Ogg1−/− animals, spontaneously developed tumors were observed, possibly due to the accumulation of non-repaired oxidative DNA base modifications even in the absence of initiation. Several authors have reported an increase of spontaneous lung tumors in MutM, MutY and MutT-deficient mice [,,,]. Previously, lung tumors were also shown to be significantly induced in Ogg1−/− mice by the DMA treatment
[]. Furthermore, significant enhancement of spontaneous lung tumorigenesis was observed when the Ogg1 mutation was combined with a MutY homolog (MUTYH) or MSH2-deficient condition, and the G:C to T:A transversions in the K-ras gene were detected in the lung tumors []. In our previous study, genes related to cancer, cellular growth, proliferation and cell cycle (e.g., polymerase (DNA-directed), delta 4 (Pold4), cyclin C and mitogen activated protein kinase 8) and angiogenesis (e.g., matrix metalloproteinases 13, 14, and 17) were found to be up-regulated in non-treated Ogg1−/− mice lungs, but those involved in free radical scavenging, lipid metabolism, drug and endocrine system development and function were suppressed comparing to the Ogg1+/+ case []. From the present and previous results, MutM, MutY and MutT homologs responsible for the repair of oxidative DNA modifications are extremely important for suppression of lung tumorigenesis in mammal.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
DMBDD treatment induced significant increases of relative liver weights in Ogg1 homozygous mutant but not wild type mice. Furthermore, higher elevation of AST and ALT serum levels reflecting the pathological processes in the liver supported our histopathological findings in the DMBDD-treated Ogg1 knockout mice. The mechanism of DMBDD carcinogenicity in the liver of Ogg1−/− mice might be accumulation of non-repaired oxidative base modifications in DNA leading to increase of cell proliferation, occurrence of mutations and further elevation of cell proliferation, resulting in promotion and progression of liver carcinogenesis []. Interestingly, hemangiosarcomas were detected only in the DMBDD-treated Ogg1−/− mice. Hemangiomas and hemangiosarcomas are known to arise as primary vascular neoplasms in the liver and could be initiated in mice by DHPN [].
They are usually not sharply demarcated from the surrounding parenchyma and the neoplastic cells are generally elongated or spindle-shaped and may form solid areas occupying dilated hepatic sinusoids and are typically locally invasive (Figure 3). Tsutsumi et al. previously demonstrated that incidences and multiplicities of hemangiomas and hemangiosarcomas in the liver were markedly higher in the poly(ADP-ribose) polymerase-1 (Parp-1)-null mice, while Parp-1 is one of the poly(ADP-ribose) polymerase family proteins taking part in genomic stability, DNA repair and cell death triggered by DNA damage []. Thus, the relationship between defective DNA repair and development of hemangiosarcomas may exist.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Increase of DNA 8-OHdG levels has been previously reported by the DEN treatment in the livers of rats and mice [,]. Furthermore, mutT-deficient mice were also reported to be susceptible to liver carcinogenesis []. Moreover, the effect of potassium bromate, which has been reported to induce oxidative stress, was investigated in Ogg1−/− mouse liver after partial hepatectomy [], and the results indicated a significant increase of mutation frequency and liver tumorigenicity being consistent with our present and previous data showing the promotion and progression of hepatocarcinogenesis in DMBDD and PB-treated Ogg1−/− mice []. Arai et al. suggested that high levels of cell proliferation are very important for the fixation of mutations induced by oxidative stress conditions in the liver
[]. Furthermore, in our previous study, it has been detected that cell proliferation and DNA 8-OHdG levels in the liver of Ogg1−/− mice treated with PB are much higher than that of wild type mice. Therefore, they are highly susceptible to the carcinogens treatment []. Thus, it could be suggested that accumulation of unrepaired 8-OHdG in the livers of DMBDD-treated Ogg1−/− animals might cause a significant increase of cellular proliferation, resulting in acceleration of hepatocarcinogenesis. With regard to specific elevation of cell proliferation in DMBDD target organs, elevation of cell proliferation has been previously shown in the lung, liver, colon, urinary bladder, thyroid and kidney by initiation with BBN, DEN, DMH, DHPN and MNU
[].[](https://www.ncbi.nlm.nih.gov/mesh/C067134)
From our previous results, in contrast to the wild type mice, in the livers of Ogg1-deficient animals, Nrf2 phosphorylation, and likely, its transformation to the nuclear did not occur, resulting in increase of oxidative stress and DNA damage of liver cells []. The accumulation of reactive oxygen species and non-repaired DNA oxidative base modifications in the Ogg1−/− livers, thus, could become the reason of higher susceptibility to liver tumorigenesis. At present, Nrf2 is recognized as important protein involved in regulation of broad transcriptional response preventing DNA, proteins and lipids damage, recognition, repair and removal of macromolecular damage, and tissue renewal after application of toxic substance. Mice that lack the Nrf2 transcription factor were more sensitive to the genotoxic and cytotoxic and effects of foreign chemicals and oxidants than wild-type animals []. Multiple studies demonstrated enhanced tumorigenicity in Nrf2-disrupted mice compared to wild-type in models of lung disease and cancer, hepatocarcinogenesis, colon cancer, stomach cancer, bladder cancer, mammary cancer, skin cancer, and inflammation [].
Furthermore, Nrf2 has been shown to upregulate the activity of multiple DNA repair, including the process of removal of oxidative stress-induced endogenous DNA interstrand cross-links [,].[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
The histological examination revealed a trend and a significant increase of renal tubular hyperplasia and degeneration, respectively, in DMBDD-treated Ogg1−/−, predominantly male mice. The observation of an increased kidney weights and blood biochemistry data supported the finding concerning serious kidneys dysfunction in Ogg1 mutant mice administered DMBDD. It has been previously suggested that Ogg1 plays a major role in renal tumorigenesis [], thus the observed increase of renal tubular hyperplasia could be related to the insufficient repair of 8-OHdG in kidneys. Furthermore, significantly elevated sodium (Na), creatinine and IP level and lowered calcium (Ca) in the blood serum of DMBDD-initiated Ogg1−/− mice signified about the impaired kidney function or kidney disease.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
It has been reported that the incidence of bladder cancer induced by BBN is significantly higher in C57BL/6 mouse strains []. In this study, we observed development of simple, preneoplastic nodular (PN) hyperplasia and TCC in both DMBDD-administered Ogg1 homozygous mutant and wild type mice. However, a trend either significant increase for PN and simple urinary bladder hyperplasia incidences was observed, indicating increased susceptibility to bladder carcinogenesis in DMBDD-treated Ogg1−/− male and female mice.[](https://www.ncbi.nlm.nih.gov/mesh/D002085)
In DMBDD-treated Ogg1−/− male mice, significantly enhanced incidence and multiplicity of colon tumors as compared to the Ogg1−/− control has been found. However, in females, inductions of colon tumors induced by DMBDD in Ogg1−/− and Ogg1+/+ animals were comparable, pointing out the sex differences in susceptibility to colonic tumorigenesis. Furthermore, in our study, an increase of carcinogenicity in the small intestine of Ogg1−/− male mice was also observed. Previously, the MutY homolog (MUTYH)-null mice have been reported to have a higher susceptibility to intestinal adenoma and adenocarcinoma [].
Thus, both the MutM and MutY deficiency leading to high levels of 8-OHdG in the colonic mucosa could be responsible for the tumorigenesis in the colon and small intestine.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
In the study with mutT homolog-1 (MTH1)-deficient mice, 18 months after birth, increases of tumorigenicity were also detected in stomachs, as compared with wild type mice []. These data support our results on enhancement of forestomach squamous cell HPL in DMBDD- treated Ogg1−/− male mice, suggesting that mutT and Ogg1 deficiency may promote carcinogenesis in the forestomach.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Ogg1-null mice have been reported to show an increased susceptibility to UVB-induced skin tumorigenesis []. They developed more malignant tumors (squamous cell carcinomas and sarcomas) than did wild-type mice. In line with these results, in the present study, we observed increase of fibrosarcoma incidence in DMBDD-treated Ogg1−/− male mice. Furthermore, trends for increase of incidences of malignant lymphomas/leukemias induced by the DMBDD treatment in Ogg1−/− mice as compared to the Ogg1−/− controls and DMBDD-treated Ogg1+/+ animals were found. One DMBDD-treated female Ogg1+/+ mouse developed T cell lymphoma, likely due to the MNU treatment, as previously reported in C57Bl/6J mice
[], but in our study no such thymic lymphomas were observed in Ogg1−/− mice. It is necessary to mention, that no increased risk of thyroid cancer in DMBDD-treated Ogg1 knockout mice was found in this study.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
It is important to note that mutations in the tumor genome induced by the Ogg1 deficiency could also cause tumors to express large number of mutant tumor specific proteins (neoantigens) which have been recently demonstrated to become one of key elements for efficacy of immuno-checkpoint inhibitors as anticancer therapeutics [].
In conclusion, this study provides the experimental evidence for a strong relationship between repair of the oxidative base modifications and multiorgan carcinogenesis. The mechanism of DMBDD carcinogenicity in the tissues of Ogg1−/− mice could be related to the accumulation of non-repaired oxidative DNA modifications leading to mutations and elevation of cell proliferation what likely resulted in promotion and progression of carcinogenesis. The multiorgan carcinogenesis bioassay is concluded to become an important tool to examine the effects of different factors on carcinogenicity in mice.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
## 4. Materials and Methods
*In the **4. Materials and Methods** section:*
## 4.1. Chemicals
*In the **4.1. Chemicals** section:*
DEN, BBN and DMH (purity ≥ 98%) were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). DHPN and MNU were purchased from Nacalai Tesque Inc. (Kyoto, Japan) and Wako Pure Chemicals Industries (Osaka, Japan), respectively. Other chemicals were from Sigma or Wako Pure Chemical Industries (Osaka, Japan).[](https://www.ncbi.nlm.nih.gov/mesh/D004052)
## 4.2. Animals
*In the **4.2. Animals** section:*
Mmh/Ogg1 homozygous mutant (Ogg1−/−) generated previously [] and wild type mice (Ogg1+/+) of C57Bl/6J background were bred and placed in an environmentally controlled room maintained at a constant temperature of 22 ± 1 °C, relative humidity of 44 ± 5% and 12 h (7:00–19:00) light/dark cycle. During all the experimental period they were given free access to drinking water and food (Oriental CE-2 pellet diet, Oriental Yeast Co., Tokyo, Japan). Mice body weights, food and water consumptions were measured weekly for the first 12 weeks of the study and subsequently once every 4 weeks. The time when the animal should be euthanized was decided due to the specific signs, such as no response to stimuli or the comatose condition, loss of body weight loss and related changes in food and water consumption, hypothermia, heart rate and external physical appearance changes, dyspnea and prostration. The experiments were performed according to the Guidelines of the Public Health Service Policy on the Humane Use and Care of Laboratory Animals and approved by the Institutional Animal Care and Use Committee of Osaka City University Graduate School of Medicine (Approval No.597; 26 November 2015).[](https://www.ncbi.nlm.nih.gov/mesh/D060766)
## 4.3. Experimental Design
*In the **4.3. Experimental Design** section:*
We developed the new protocol for multiorgan carcinogenicity bioassay which could be applied in mice (Figure 3). In the present study, Ogg1−/− (80) and Ogg1+/+ (80) six-week-old male and female mice were randomly divided into 4 groups each comprising of 20 mice. The treatment with five genotoxic carcinogens, including DEN, MNU, BBN, DMH, and DHPN, was performed as followers: DEN at a dose of 400 ppm was administered in a drinking water for 3 days from the very beginning of the experiment. We decided to perform DEN treatment in drinking water, as in the preliminary experiment too strong toxic effect was observed with Ogg1−/− mice after the DEN intraperitoneal (i.p.) injection. After finishing DEN administration, four i.p. injections of MNU (20 mg/kg b.w.) were done (2 times/week), following by six subcutaneous (s.c.) injections of DMH (10 mg/kg b.w.) during Weeks 3 and 4.
BBN at a dose of 0.05% was administered in drinking water for 4 weeks starting immediately after finishing the DEN treatment, and 0.1% DHPN was applied for 2 weeks in drinking water during Weeks 4 and 5. Mice in the control groups were administered saline as injections (i.p. or s.c.) or the tap water for drinking. Animals were observed every day and euthanized in case of becoming moribund during the study, or at the end of the experiment at Week 34. All surviving mice were killed under the isofluorene treatment and the DMBDD target organs, including liver, lung, kidneys, urinary bladder, small intestine and colon and thyroid gland, were immediately excised and fixed in 10% phosphate-buffered formalin, Thereafter, tissues were embedded in paraffin, sections of 4 μm in thickness were prepared and stained with hematoxylin and eosin (H & E) for the routine histology. We assessed the incidences of hyperplasia (HPL), adenoma and adenocarcinomas in the lungs, putative preneoplastic foci (PPFs), tumors, bile duct proliferation and biliary cysts in the liver, incidences of adenoma and adenocarcinoma in the small intestine and colon.
Intestines were excised and intraluminally injected and fixed with 10% phosphate-buffered formalin.[](https://www.ncbi.nlm.nih.gov/mesh/D004052)
## 4.4. Blood Biochemical Analysis
*In the **4.4. Blood Biochemical Analysis** section:*
Blood was collected via the abdominal aorta from 7–9 mice per group per sex at the end of the study period after overnight fasting. Automatic analyzer (Olympus AJ-5200, Tokyo, Japan) was employed for the blood biochemical analysis to detect total protein (T-protein, g/dL), albumin/globulin ratio (A/G ratio), albumin (g/dL), total bilirubin (T-bil, mg/dL), aspartate aminotransferase (AST, IU/L), alanine aminotransferase (ALT, IU/L), γ-glutamyl transpeptidase (γ-GTP, IU/L), alkaline phosphatase (ALP, IU/L), triglycerides (TG, mg/dL), total cholesterol (T-chol, mg/dL), blood urea nitrogen (BUN, mg/dL), creatinine (mg/dL), chloride (Cl), sodium (Na), potassium (K), calcium (Ca) and inorganic phosphorus (IP) (mEq/L).[](https://www.ncbi.nlm.nih.gov/mesh/D001663)
## 4.5. Statistical Analysis
*In the **4.5. Statistical Analysis** section:*
The statistical analysis of the significance of differences between mean values was performed with the StatLight-2000(C) program (Yukms corp, Tokyo, Japan). The inter group differences detected for the incidences of histopathological findings were analyzed with the by χ2 test Fisher’s exact probability test (two-sided). Kaplan–Meier analysis was used to examine the changes in survival rates of Ogg1 knockout and wild type mice. Homogeneity of variance between of Ogg1−/− and Ogg1+/+ groups was detected by the F test. Student’s t-test (two-sided) was applied in the case the data were homogeneous; otherwise, Welch test was used.
p Values less than 0.05 were considered significant.
# Supplementary Materials
*In the **Supplementary Materials** section:*
Supplementary materials can be found at .
# Author Contributions
*In the **Author Contributions** section:*
Conception and design of the experiments: Anna Kakehashi, and Hideki Wanibuchi. Performance of the experiments: Anna Kakehashi. Analysis of the data: Anna Kakehashi and Naomi Ishii. Writing of the paper: Anna Kakehashi, Naomi Ishii, Takahiro Okuno, Masaki Fujioka, Min Gi.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
Body weight (A); and survival (B) curves for DMBDD-treated and control Ogg1−/− and Ogg1+/+ male and female mice. *** p < 0.001 significantly different vs. respective control group of the same genotype; a p < 0.05 and c p < 0.001 significantly different vs. the respective Ogg1+/+ control groups.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Representative histopathological pictures (H&E staining) of: lung hyperplasia (a); adenoma (b); adenocarcinoma (c); liver PPF (basophilic foci) (d); HCA (e); hemangioma (f); hemangiosarcoma (g); colon adenoma (h); and adenocarcinoma (i) developed in DMBDD-treated Ogg1−/− mice. HPL, hyperplasia; HCA, hepatocellular adenoma; AdCa, adenocarcinoma; PPFs, putative preneoplastic foci.[](https://www.ncbi.nlm.nih.gov/mesh/D006416)
Experimental protocol of medium-term multiorgan carcinogenesis bioassay applied in Ogg1−/− and Ogg1+/+ mice. wks: weeks.
Final survival ratios, final body and relative organ weights of Ogg1−/− and Ogg1+/+ mice.
Data are Mean ± SD for the surviving animals at the end of the study. Relative organ weights were calculated with the following equation: Absolute organ weight/final body weight × 100. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001: significantly different vs. the respective control groups of the same genotype. a p < 0.05; b p < 0.01; c p < 0.001; d p < 0.0001 significantly different vs. the respective Ogg1+/+ DMBDD-treated or control groups. e Final body weights of all survived mice at the termination of the experiment.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Neoplastic and preneoplastic proliferative lesions in male and female Ogg1−/− and Ogg1+/+ mice.
* p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001 and (i) p = 0.05 vs. respective control mice of the same genotype. a p < 0.05; b p < 0.01; d p < 0.0001 vs. wild type control or DMBDD-treated mice. HPL, hyperplasia; HCA, hepatocellular adenoma; AdCa, adenocarcinoma; PN, papillary or nodular; TCC, transitional cell carcinoma, PPFs, putative preneoplastic foci.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Blood biochemistry data of DMBDD-treated and control Ogg1 knockout and wild type mice.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
Values are means ± SD; * p < 0.05; ** p < 0.01; (i) p = 0.05 vs. respective control group of mice of same genotype. a p < 0.05; b p < 0.01 vs. respective wild type control or DMBDD group. TG, triglycerides; T-Bil, T-bilirubin; IP, inorganic phosphorus.[](https://www.ncbi.nlm.nih.gov/mesh/C012457)
|
# Introduction
[Phosphine](https://www.ncbi.nlm.nih.gov/mesh/D010720)-catalyzed enantioselective [3 + 2] cycloadditions of [γ-substituted allenoates](https://www.ncbi.nlm.nih.gov/mesh/C110055) with [β-perfluoroalkyl enones](https://www.ncbi.nlm.nih.gov/mesh/D007659)†
# Abstract
*In the **Abstract** section:*
Herein we present a phoshine-catalyzed highly regio-, diastereo- and enantioselective [3 + 2] cycloaddition of γ-substituted allenoates with [β-perflu](https://www.ncbi.nlm.nih.gov/mesh/D010720)oroalkyl enones, delivering a wide range of densely functionalized perfluoroalkylat[ed cyclopentenes with th](https://www.ncbi.nlm.nih.gov/mesh/C110055)ree co[ntiguous chiral stereoc](https://www.ncbi.nlm.nih.gov/mesh/D007659)enters.[](https://www.ncbi.nlm.nih.gov/mesh/D003517)
The enantioselective construction of densely functionalized cyclopentene bearing contiguous three stereocenters has been a challenging task in organic synthesis. Herein, we present a phoshine-catalyzed highly regio-, diastereo- and enantioselective [3 + 2] cycloaddition of γ-substituted allenoates with β-perfluoroalkyl enones, delivering a wide range of densely functionalized perfluoroalkylated cyclopentenes with three contiguous chiral stereocenters.[](https://www.ncbi.nlm.nih.gov/mesh/D003517)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Cyclopentenes (or cyclopentanes) are valuable skeletons found in several natural products and pharmaceuticals (Fig. 1). Among existing methodologies for their preparation, phosphine-catalyzed [3 + 2] cycloaddition of allenoates with electron-deficient olefins was first reported by Lu in 1995 as a powerful and straightforward strategy for the construction of functionalized cyclopentene rings., As a result of tremendous effort from numerous research groups, Lu's enantioselective [3 + 2] cycloaddition reaction of terminal allenoates with electron-deficient olefins has been well established over the past years. However, asymmetric [3 + 2] cycloaddition reaction of γ-substituted allenoates with electron-deficient olefins has been less explored despite the increase in stereochemical diversity of the cycloaddition products. In 2007, Miller's group first realized a unique “derac[emization” re](https://www.ncbi.nlm.nih.gov/mesh/D003517)actio[n upon cycloa](https://www.ncbi.nlm.nih.gov/mesh/D003517)ddition of chalcone with racemic γ-methyl allenoates but requisite the use of a stoichiometric amount of chiral phosphine catalyst A (Scheme [1a).
Subs](https://www.ncbi.nlm.nih.gov/mesh/D010720)equently, Fu and co-workers have acc[omplished ](https://www.ncbi.nlm.nih.gov/mesh/C110055)the cycloaddition reactio[n of ra](https://www.ncbi.nlm.nih.gov/mesh/D000475)cemic γ-substituted allenoates with heteroatom-bearing olefins with the use of a catalytic amount of chiral phosphin[e B, furnish](https://www.ncbi.nlm.nih.gov/mesh/D003517)ing a facile access to functionalized cyclopentenes with two adjacent stereo centers (Scheme 1b). Recently, Marinetti and coworkers have r[eported a ](https://www.ncbi.nlm.nih.gov/mesh/C110055)highly enantioselective [[3 + 2] ](https://www.ncbi.nlm.nih.gov/mesh/D000475)cycloaddition of γ-substituted allenoates with benzylidenemalononitrile by utilizing chiral phosphahel[icenes catalyst C (Schem](https://www.ncbi.nlm.nih.gov/mesh/C110055)e 1b).[](https://www.ncbi.nlm.nih.gov/mesh/D000475)
Selected natural products and pharmaceuticals contain cyclopentene or cyclopentane rings.[](https://www.ncbi.nlm.nih.gov/mesh/D003517)
[3 + 2] cycloaddition reaction of γ-substituted allenoates and olefins.[](https://www.ncbi.nlm.nih.gov/mesh/C110055)
Despite this progress, the scope of γ-substituted allenoates and electron-deficient olefin partner for enantioselective Lu's annulation is very limited, and the construction of cyclopentene derivatives with three contiguous chiral stereocenters has been a major challenge but a highly desirable task. Moreover, introduction of perfluoroalkylated, particularly trifluoromethylated, stereocenters into chiral compounds have garnered special attention in pharmaceutical and pesticide industry since the polarity, bioavailability, metabolic stability and other properties of the parent molecules could be influenced greatly by these perfluoroalkyl groups. During the course of our continuous interest in design, synthesis and application of novel chiral β-aminephosphines, in asymmetric catalysis and the synthesis of enantio-enriched trifluoromethylated building blocks,, we envisaged that the challenging enantioselective [3 + 2] cycloadditions of γ-substituted allenoates with β-perfluoralkyl α,β-enones might be addressed by systematic screening of known phosphines or rational design of new catalysts (Scheme 1b). In the present study, we report our efforts in addressing this challenging reaction by identifying two phosphine catalysts, commercially available bisphosphine (R,R)-DIPAMP and novel multifunctional (S)-P3 which have been developed in our group. Further control experiments have shown that the reaction under the catalysis of (R,R)-DIPAMP was a deracemization process, while the kinetic resolution reaction was observed under the multifunctional phosphine catalyst (Scheme 1).[](https://www.ncbi.nlm.nih.gov/mesh/C110055)
## Results and discussion
*In the **Results and discussion** section:*
In order to validate the feasibility of the asymmetric [3 + 2] cycloaddition of γ-substituted allenoates with β-perfluoralkyl α,β-enones, allenoate 2a and enone 1a were exposed to a range of commercially available chiral bisphosphine catalysts (Table 1). A small amount of the desired 3aa was observed when (S,S)-DIOP or (R,R)-Et-DUPHOS was utilized as the catalyst (Table 1, entries 1 and 2). Interestingly, (R,R)-Et-BPE exhibited a promising level of reactivity with 64% yield and stereoinduction with 39% ee (Table 1, entry 3). Fortunately, 86% yield of 3aa with 89% ee was obtained using (R,R)-DIPAMP as a catalyst (Table 1, entry 4). It can be noted that multifunctional chiral phosphines (S)-P1–P6 bearing hydrogen bond donors, such as amide and (thio) urea groups, could deliver higher chemical yield but with unacceptable enantioselectivity (Table 1, entries 5–10).
Gratifyingly, the enantioselectivity was improved to 92%, albeit with a slightly lower yield when decreasing the reaction temperature from 25 °C to –20 °C (Table 1, entries 11–13). However, much lower reaction temperature was not beneficial for enantioselectivity and reactivity (Table 1, entry 14). In addition, much lower yield and enantioselectivity was observed when (Z)-1a was utilized in the reaction, indicating that the configuration of enone also affected the reaction significantly (Table 1, entry 15). Further screening of solvents demonstrated that toluene was the best reaction medium for this transformation (see ESI† for details). Then, the optimized reaction conditions were identified: 10 mol% (R,R)-DIPAMP as the catalyst and toluene as the reaction medium at –20 °C.[](https://www.ncbi.nlm.nih.gov/mesh/C110055)
Optimization of reaction conditions a
aUnless otherwise specified, all reactions were carried out with (E)-1a (0.1 mmol), racemic 2a (0.15 mmol) in toluene (1 mL).[](https://www.ncbi.nlm.nih.gov/mesh/D014050)
bYield of isolated products; d.r. and r.r. > 20 : 1.
cDetermined by HPLC analysis.
d(Z)-1a was used.
Under optimal reaction conditions, we investigated the scope of the enantioselective [3 + 2] cycloaddition reaction (Scheme 2). Remarkably, a wide range of β-trifluoromethyl substituted enones containing different electron nature functional groups worked well with allenoate 2a, thereby resulting in a highly regioselective α-addition products 3ba–3ha in good yields with 88–94% ee. However, the introduction of an ortho substituent, such as Cl and Br, to the phenyl ring of enone led to dramatic decrease in the enantioselectivity (3ia and 3ja). To our delight, naphthyl- and heteroaryl-containing substrates 1k–1o were also compatible, efficiently furnishing a set of trifluoromethylated cyclopentenes containing naphthyl- and heteroaryl frameworks 3ka–3oa. In addition, the present protocol could be readily extended to the challenging synthesis of cyclohexenyl and cyclohexyl based trifluoromethyl enone 1p and 1q.
It was noteworthy that both β-pentafluoroethyl and β-heptafluoropropyl enone were particularly effective in the present transformation, forming valuable perfluoroalkyl substituted cyclopentene 3ra and 3sa in good yields with 94% ee. Furthermore, γ-aryl allenoates 2b–2d with substituted aryl and hetereoaryl groups were well applicable and formed corresponding products 3ab–3ad with high regioselectivity and enantioselectivity. The absolute configuration of product 3aa was confirmed by single-crystal X-ray diffraction analysis.[](https://www.ncbi.nlm.nih.gov/mesh/D007659)
Enantioselective [3 + 2] cycloadditions of γ-aryl substituted allenoates with β-perfluoro substituted enonea.[](https://www.ncbi.nlm.nih.gov/mesh/C110055)
After intensive screening of various chiral phosphine catalysts, it was found that multifunctional phosphine (S)-P3 displayed good performance in the substrates with ortho-substituent, and the desired products 3ia and 3ja could be isolated in 85–88% yields with 96% and 99% ee, respectively (Scheme 3).[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
(S)-P3 catalysed enantioselective [3 + 2] cycloadditions of 1i and 1j with 2a.[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
Unfortunately, the performance of (R,R)-DIPAMP in the cycloaddition of γ-alkyl substituted allenoates was not as good as that in the cases of γ-aryl substituted allenoates. For example, the reaction of 2e with 1c resulted in the formation of desired 3ce in 67% yield but with only 86% ee. After further screening of a series of chiral phosphine catalysts, solvents and reaction temperature, it was found that (S)-P3 was a privileged catalyst for cycloaddition of γ-alkyl allenoates. In general, allenoates 2e–2g with different alkyl substituents at γ position participated in the annulation process with good regio- and enantioselectivity. In addition, diverse alkyl substituents such as benzyl, halogen and ester group were well tolerant, furnishing the corresponding cycloadducts 3ch–3cj in moderate to good yields with high enantioselectivity.
Furthermore, allenoates with bulky substituents such as isopropyl, cyclopentyl and cyclohexyl at γ position worked well, thereby forming the desired 3ck–3cm in good yields with 92–94% ee. Good to excellent regioselectivity and enantioselectivity were also obtained in the cycloaddition reactions of allenoate 2g with a wide range of β-trifluoromethyl substituted enones (Scheme 4).[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
Enantioselective [3 + 2] cycloadditions of γ-alkyl substituted allenoates with β-perfluoro substituted enonea.[](https://www.ncbi.nlm.nih.gov/mesh/C110055)
Next, we turned our attention to gain insight into catalytic process for the proposed [3 + 2] cycloaddition reaction. In case of (R,R)-DIPAMP catalysed cycloaddition of 1d and racemic 2a, the starting material 2a was recovered in 38% yield (based on 2a) with 0% ee (eqn (1)). Furthermore, when optically active allenoate (+)-2a (76% ee) served as the substrate, ee of 3da did not improve but the recovered (+)-2a had a higher ee (eqn (2)). These results have supported that a deracemization process was followed in the (R,R)-DIPAMP catalysed cycloaddition of 1d and 2a.[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
To examine both the phosphines in (R,R)-DIPAMP induce enantioselectivity independently or cooperatively, (R,R)-SDIPAMP that contained only one nucleophilic phosphine was synthesized and subjected to the reaction of 1d and racemic 2a (Scheme 5). Although the reaction became slower, enantio-selectivity of 3da remained unchanged, demonstrating that both the phosphines in (R,R)-DIPAMP might induce enantioselectivity independently (Scheme 5b).[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
Synthesis of (R,R)-SDIPAMP and its application in the asymmetric [3 + 2] cycloaddition of 2a and 1d.[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
Based on the abovementioned results and earlier reports, a plausible catalytic cycle for (R,R)-DIPAMP catalysed asymmetric [3 + 2] cycloaddition reaction of γ-aryl allenoates with trifluoromethyl enones has been illustrated in Scheme 6. The zwitterionic intermediate I was formed through nucleophilic addition of (R,R)-DIPAMP to racemic 2a. The deracemization process resulted from the same nucleophilic attack rate (K 1 = K 2) of (R,R)-DIPAMP to both the enantiomers of allenoates 2a. The subsequent [3 + 2] cycloaddition favoured α-addition to provide intermediate II, which then underwent proton transfer to provide intermediate III. Finally, (R,R)-DIPAMP and cyclopetene 3da were released from intermediate III.[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
Possible catalytic cycle for (R,R)-DIPAMP catalysed asymmetric [3 + 2] cycloaddition.[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
In contrast to (R,R)-DIPAMP, a kinetic resolution reaction takes place with multifunctional chiral phosphine (S)-P3 as the catalyst and (+)-2a and (+)-2g is recovered in 76% ee (in toluene, 77% ee in CHCl3) and 81% ee respectively (eqn (3) and (4)). In order to confirm the possible hydrogen-bonding interaction during the catalytic process, (S)-P7 without hydrogen-bond donor was synthesized and subjected to the cycloaddition reaction (Scheme 7). The conversion decreased dramatically under higher catalyst loading and higher reaction temperature. The ee value of the recovered 2g also vanished (Scheme 7b). These results demonstrated that the hydrogen-bond donor in (S)-P3 was crucial for enantioselective formation of cycloaddition product via kinetic resolution process.[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
Synthesis of (S)-P7 and its application in the asymmetric [3 + 2] cycloaddition of 2g and 1d.[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
On the basis of above control experiments and recent excellent mechanistic studies on the [3 + 2] cycloaddition of allenoates with electron-deficient olefins, a tentatively proposed catalytic cycle for (S)-P3 catalysed asymmetric [3 + 2] cycloaddition reaction of racemic allenoate with trifluoromethyl enone is shown in Scheme 8. (–)-2 might prefer a configuration that would facilitate hydrogen-bonding interactions of N–H and carbonyl group (Scheme 8, TS-1). On the other hand, the nucleophilic attack of (S)-P3 with (+)-2 might be suppressed by the steric interaction of the bulky R 2 group with the phenyl moiety (Scheme 8, TS-2). Accordingly, different nucleophilic attack rates (K 1 > K 2) of (S)-P3 to both the enantiomers of allenoates 2 contribute to the kinetic resolution process. It should be note that further experimental and theoretical studies are required to gain insights into kinetic resolution process.[](https://www.ncbi.nlm.nih.gov/mesh/C110055)
Possible catalytic cycle for (S)-P3 catalysed asymmetric [3 + 2] cycloaddition reaction of racemic allenoate with trifluoro-methyl enone.[](https://www.ncbi.nlm.nih.gov/mesh/D010720)
## Conclusions
*In the **Conclusions** section:*
In conclusion, we have developed a highly regio-, diastereo- and enantioselective [3 + 2] cycloaddition of γ-substituted allenoates with β-perfluoroalkyl enones with (R,R)-DIPAMP or (S)-P3 as a catalyst; it provides a facile access to a wide range of trifluoromethylated cyclopentenes with three contiguous chiral centers (up to 88% yield with 99% ee). In case of γ-aryl allenoate, commercially available chiral phosphine (R,R)-DIPAMP was identified as an efficient catalyst. In contrast, presently developed multifunctional phosphine (S)-P3 has displayed high performance in the asymmetric cycloaddition of γ-alkyl allenoates with trifluoromethyl enones. In addition, control experiments have demonstrated that under the catalysis of (R,R)-DIPAMP, racemic allenoate reacted with trifluoromethyl enone through a “deracemization” process, whereas a clearly kinetic resolution reaction takes place with multifunctional chiral phosphine (S)-P3 as a catalyst due to the hydrogen-bonding interaction between catalyst and the allenoate. Efforts toward other transformations of allenoate under the catalysis of our developed catalysts P1–P6 are currently underway and will be reported in due course.[](https://www.ncbi.nlm.nih.gov/mesh/C110055)
|
# Introduction
Use of [N‐methyliminodiacetic acid boronate esters](https://www.ncbi.nlm.nih.gov/mesh/C533766) in suzuki‐miyaura cross‐coupling polymerizations of [triarylamine](https://www.ncbi.nlm.nih.gov/mesh/D000588) and [fluorene](https://www.ncbi.nlm.nih.gov/mesh/C041509) monomers
# Abstract
*In the **Abstract** section:*
ABSTRACT
Polytriarylamine copolymers can be prepared by Suzuki‐Miyaura cross‐coupling reactions of bis N‐methyliminodiacetic acid (MIDA) boronate ester sub[stituted arylamines with di](https://www.ncbi.nlm.nih.gov/mesh/D000588)bromo arenes. The roles of solvent composition, temperature, re[action time, and co‐monomer structure were examined ](https://www.ncbi.nlm.nih.gov/mesh/C533766)and (co)polym[ers prepar](https://www.ncbi.nlm.nih.gov/mesh/D000588)ed con[taining 9, 9‐d](https://www.ncbi.nlm.nih.gov/mesh/D006841)ioctylfluorene (F8), 4‐sec‐butyl or 4‐octylphenyl diphenyl amine (TFB), and N, N′‐bis(4‐octylphenyl)‐N, N′‐diphenyl phenylenediamine (PTB) u[nits, using a Pd(OAc](https://www.ncbi.nlm.nih.gov/mesh/D005449))2[/2](https://www.ncbi.nlm.nih.gov/mesh/D005449)‐di[cyclohexylphosphino‐2′,6′‐dimethoxybiphenyl](https://www.ncbi.nlm.nih.gov/mesh/D004159) ([SPh](https://www.ncbi.nlm.nih.gov/mesh/C533766)os) cat[alyst system. The performance of a di‐functionalized MID](https://www.ncbi.nlm.nih.gov/mesh/D010655)A [bor](https://www.ncbi.nlm.nih.gov/mesh/D010655)onate ester monom[er was c](https://www.ncbi.nlm.nih.gov/mesh/C516071)o[mpared with that of an equivalent pinacol boron](https://www.ncbi.nlm.nih.gov/mesh/D001713)at[e est](https://www.ncbi.nlm.nih.gov/mesh/D001713)er. Higher molar mass polymers were produced from reaction[s starting with a d](https://www.ncbi.nlm.nih.gov/mesh/C533766)ifunctionalized pinacol boronate ester monomer th[an the equivalent difu](https://www.ncbi.nlm.nih.gov/mesh/C000621940)nctionalized MIDA boronate ester monomer in biphase solvent mixtures (toluene/dioxane/water[). Matrix‐assisted las](https://www.ncbi.nlm.nih.gov/mesh/C000621940)er desorption/ionization mass spectroscopic an[alysis revealed tha](https://www.ncbi.nlm.nih.gov/mesh/C533766)t polymeric structures rich in residue[s assoc](https://www.ncbi.nlm.nih.gov/mesh/D014050)i[ated wi](https://www.ncbi.nlm.nih.gov/mesh/C025223)t[h the](https://www.ncbi.nlm.nih.gov/mesh/D014867) starting MIDA monomer were present, suggesting that homo‐coupling of the boronate ester must be occurring to the detriment of cross‐coupling in the step‐gr[owth](https://www.ncbi.nlm.nih.gov/mesh/C533766) polymerization. However, when comparable reactions of the t[wo boronate mo](https://www.ncbi.nlm.nih.gov/mesh/D001897)nomers with a dibromo fluorene monomer were completed in a single phase solvent mixture (dioxane + water), high molar mass polymers wit[h relati](https://www.ncbi.nlm.nih.gov/mesh/D001897)vely narrow distr[ibution ranges w](https://www.ncbi.nlm.nih.gov/mesh/D005449)ere obtained after only 4 h of reaction. © 2017 The Authors[. Journ](https://www.ncbi.nlm.nih.gov/mesh/C025223)al [of Po](https://www.ncbi.nlm.nih.gov/mesh/D014867)lymer Science Part A: Polymer Chemistry Published by Wiley Periodicals, Inc. J. Polym. Sci., Part A: Polym. Chem. 2017, 55, 2798–2806
## INTRODUCTION (cont.)
*In the **INTRODUCTION (cont.)** section:*
Boronic acids and esters are widely used as intermediates in the syntheses of pharmaceuticals, natural products, and organic materials via Suzuki‐Miyaura reactions.1 In the materials field, these types of reactions often provide the best approach to obtain a range of conjugated polymers for electronic applications.2 However, the fidelity of these cross‐coupling polymerizations is often compromised by the protodeboronation of monomers or growing chains under the reaction conditions. A variety of boron protecting groups,3 such as trifluoroborate salts,4 trialkyoxyborate salts,5 and N‐methyliminodiacetic acid (MIDA) boronate esters6, 7, 8, 9, 10, 11, 12, 13, 14 have been used to extend the range of molecules amenable to high fidelity Suzuki‐Miyaura cross coupling. The potential for the synthesis of conjugated polymers is supported by recent work on the use of trifluoroborates for coupling reactions of electron deficient monomers.15[](https://www.ncbi.nlm.nih.gov/mesh/D001897)
MIDA boronate esters are of particular interest as they are cheap to synthesize, air stable, and the hydrolysis can be controlled to slowly release boronic acids for effective cross‐coupling reactions.6 We have reported the first use of a MIDA boronate ester protecting group on a bifunctional thienyl (AB type) monomer in Suzuki‐Miyaura polymerizations to synthesize highly regioregular poly(3‐hexylthiophene‐2,5‐diyl) (rr‐P3HT).16 This work was facilitated by facile production of the monomer in high yield by amine‐mediated electrophilic borylation which provides a direct route from aryl‐H to aryl‐B(OR)2 compounds.17, 18, 19, 20 This process produced MIDA boronate esters without requiring the synthesis and isolation of arylboronic acid intermediates, some of which can be particularly susceptible to protodeboronation.6, 7, 21, 22 The slow hydrolysis of thienyl bis MIDA boronate ester (AA type) monomers has most recently been used in Suzuki‐Miyaura copolymerizations to produce a range of thiophene containing polymers (Scheme 1).23[](https://www.ncbi.nlm.nih.gov/mesh/C533766)
Example of Suzuki‐Miyaura copolymerization of thienyl bis MIDA boronate ester with dibromoarene.[](https://www.ncbi.nlm.nih.gov/mesh/C533766)
Polytriarylamines (PTAAs) are amorphous semiconducting polymers of interest in organic electronics,24 as they can be readily processed from solution and show stable performance in air, with moderate charge‐carrier mobilities in organic field effect transistors (up to 0.05 cm2 V−1 s−1). They have been used very successfully in blends with small molecule organic semiconductors to deliver very high performance, robust, reproducible organic field‐effect transistor (OFET) devices.25 In general, the highest performing PTAAs have bridged phenyl units, such as fluorenes or indenofluorenes in the polymer backbone and these polymers are routinely synthesized by Suzuki‐Miyaura reactions.[](https://www.ncbi.nlm.nih.gov/mesh/D000588)
This contribution discusses the utility of using bis MIDA boronate ester substituted arylamines to produce high molar mass arylamine (co)polymers in Suzuki‐Miyaura cross‐coupling reactions. The two step, one pot electrophilic borylation process can be used to produce both bis pinacol and bis MIDA boronate ester protected arylamine monomers at room temperature, as outlined in Scheme 2.[](https://www.ncbi.nlm.nih.gov/mesh/C533766)
Electrophilic diborylation of arylamine monomers with protection steps. [Color figure can be viewed at wileyonlinelibrary.com][](https://www.ncbi.nlm.nih.gov/mesh/D000588)
The structures of the monomers used in this study are presented in Scheme 3, with the structures of the polymers produced presented in Scheme 4. The performance of the bis MIDA boronate ester monomers (e.g., 1b) were compared against the equivalent triarylamine bis pinacol boronate esters in cross‐coupling reactions with the dibromo co‐monomer(s) (3 and 4).[](https://www.ncbi.nlm.nih.gov/mesh/C533766)
Structures of boronate and dibromo arene monomers used in the Suzuki‐Miyaura cross‐coupling reactions.[](https://www.ncbi.nlm.nih.gov/mesh/D001897)
Structures of polymers synthesized in the Suzuki‐Miyaura cross‐coupling polymerizations (R = octyl).
## EXPERIMENTAL
*In the **EXPERIMENTAL** section:*
## Materials
*In the **Materials** section:*
Monomers, 1a, 2a, 2c, 3, and 4 were kindly supplied by Cambridge Display Technology Ltd. Bifunctionalized BMIDA 4‐octyl phenyl diphenyl amine (TFB) and N, N′‐bis(4‐octylphenyl)‐N, N′‐diphenyl phenylenediamine (PFB) monomers, 1b and 2b, were synthesized using a published route.22 The procedures are outlined in the Supporting Information. 1H NMR spectra of these monomers are presented in Supporting Information Figures S1 and S2. Pd(OAc)2 was supplied by Acros Organics.
The ligand, 2‐dicyclohexylphosphino‐2′,6′‐dimethoxybiphenyl (SPhos), potassium phosphate tribasic (K3PO4), 1,4‐dioxane, and toluene were all purchased from Aldrich Ltd. and used as supplied.[](https://www.ncbi.nlm.nih.gov/mesh/C533766)
## NMR Studies of the Hydrolysis of Bis MIDA Monomer, 1b
*In the **NMR Studies of the Hydrolysis of Bis MIDA Monomer, 1b** section:*
A J. Young's NMR tube was charged under inert atmosphere with 1b, (1.0 equiv., 14.0 mg, 0.021 mmol), mesitylene as internal reference (1.0 μL), and suspended in anhydrous d8‐tetrahydrofuran (d8‐THF) (0.6 mL). Subsequently, D2O (30.0 equiv. per BMIDA moiety), was added [1b] = 3.5 × 10−2 M, and the reaction mixture was vigorously shaken to homogenize before recording its NMR spectrum (t 0). Then the tube was rotated at ambient temperature (10 rpm) or heated in an oil bath at 60 °C, and followed by NMR (1H and 11B) spectroscopy at different reaction times.[](https://www.ncbi.nlm.nih.gov/mesh/C010219)
## Representative Polymerization Procedure (Entry 19)
*In the **Representative Polymerization Procedure (Entry 19)** section:*
Equimolar amounts of monomers, 1b (66.74 mg, 0.1 mmol) and 3 (45.92 mg, 0.1 mmol), were placed into a Radley's carousel tube (Table 1). Small amounts of toluene (0.50 mL, 0.43 g) and dioxane (1.00 mL, 1.03 g) were next washed into the tube. Stock solutions of the monomers in solvents were not prepared owing to the poor solubility of the BMIDA monomers at room temperature in the desired solvents. A stock solution of K3PO4 was prepared (consisting of 0.1592 g, 0.8 mmol K3PO4 per 1 mL water). An amount of this base solution (1.1592 g) was transferred into the reaction tube.
The tube contents were then stirred with a magnetic flea while being thoroughly degassed by bubbling nitrogen gas through the solution for 20 min. A stock solution of the palladium (II) acetate Pd(OAc)2/SPhos catalyst system in toluene was prepared, composed of: Pd(OAc)2 (3.36 mg, 0.015 mmol), SPhos (12.33 mg, 0.030 mmol), and toluene (1.50 mL, 1.30 g). The catalyst solution was stirred with a magnetic flea for about 20 min while being thoroughly degassed by three repeated cycles of evacuation followed by replenishment with nitrogen gas. The previously degassed reaction tube and contents were then placed into a carousel reactor maintained at 90 °C. After 10 min, a glass syringe was used to inject 0.5 mL of the catalyst solution into the now heated reaction tube contents while maintaining a nitrogen environment.
The reaction tube was stirred under a nitrogen environment at 90 °C for a period of 24 h.[](https://www.ncbi.nlm.nih.gov/mesh/D014050)
Suzuki Cross‐coupling Polymerizations of Triarylamine Boronate Monomers (1a, 1b, 2a, 2b) with Dibromo Monomers (2c, 3, 4), Using Pd(OAc)2/SPhos Catalyst and K3PO4 as Base, in Mixed Solvent Systemsa[](https://www.ncbi.nlm.nih.gov/mesh/D001896)
Reaction conditions: equimolar quantities of monomers, M1 + M2, total monomer = 0.2 mmol.
Dissolved in mixed solvent system, where T is toluene, D is dioxane, and W is H2O (mL); 4 equivalent mmol of K3PO4 (8 equiv. entries 14 and 15).[](https://www.ncbi.nlm.nih.gov/mesh/D014050)
Molar mass determined by GPC in THF versus polystyrene standards.[](https://www.ncbi.nlm.nih.gov/mesh/C018674)
Reactions in which the polymer precipitated out of the solvent mixture during the reaction time. Large amounts of polymeric material precipitated out of most of the dioxane rich PTFB and PPFB reactions (entries 1–4, 9) within 10 min of start of reaction. Only Entry 10 reaction remained homogeneous for a few hours. GPC analysis of these entries refers to the molar mass of the precipitated polymer recovered after the reaction time. The remaining GPC analysis of entries 5–8, 11–23 refers to molar masses of all polymeric material (both precipitated and soluble) recovered after the respective reaction times.[](https://www.ncbi.nlm.nih.gov/mesh/C025223)
## Characterization of the Reaction Products (Entry 19)
*In the **Characterization of the Reaction Products (Entry 19)** section:*
A sample of the reaction mixture (0.5 mL) was removed after 5 h under a nitrogen atmosphere using a glass syringe and placed into a small amount of toluene (1.0 mL) (Table 1). The sample solution was allowed to cool, before being added drop wise to a stirred excess amount of chilled methanol (5 mL) to precipitate the polymer. The polymer formed a fine dispersion in the stirred methanol. A pipette was used to transfer samples of this dispersal into a pair of vials suitable for use in a centrifuge. The vial samples were placed in a centrifuge at 14,000 rpm for 10 min.
The supernatant layer was removed from above the separated polymer. The polymer samples were then dried and redissolved in THF for GPC analysis. After 24 h, upon completion of the reaction, the remaining tube contents were added to a small amount of toluene (2.0 mL). The diluted reaction solution was then allowed to cool, before being added drop wise to a stirred excess amount of chilled methanol (20 mL) to precipitate the polymer.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
Samples of the dispersion were then treated similarly to prepare samples for GPC analysis. Polymer molar mass and molar mass distribution was determined by GPC in THF solution using a Viscotek GPCmax VE2001 and a Viscotek VE3580 RI detector (referenced to polystyrene standards). The sample solutions were made up in THF (1 mg polymer per mL solvent) and filtered before injection. Matrix‐assisted laser desorption/ionization time‐of‐flight (MALDI‐TOF) mass spectrometry was carried out using a Shimadzu Biotech AXIMA Confidence MALDI mass spectrometer in reflectron mode or linear (positive) mode, referencing against either PEG 1K or PPG 4K. Polymer samples where dissolved in THF to a concentration of 10 mg mL−1. A similar solution was prepared of the matrix (dithranol).
Polymer sample solution of 1 µL was mixed with 10 µL of matrix solution. The solution was mixed and approximately 0.5 µL was spotted on to the sample plate. 1H NMR spectra of the copolymers recovered in each case were recorded with a Bruker 400 MHz NMR instrument in deuterated chloroform (CDCl3) using tetramethylsilane as an internal standard (examples are presented in Supporting Information Figs. S3–S6).[](https://www.ncbi.nlm.nih.gov/mesh/C018674)
## RESULTS AND DISCUSSION
*In the **RESULTS AND DISCUSSION** section:*
## Hydrolysis of Bis MIDA Monomer, 1b
*In the **Hydrolysis of Bis MIDA Monomer, 1b** section:*
Hydrolysis of the BMIDA moiety by water under neutral conditions has been observed for a range of aryl‐ and heteroaryl‐BMIDA boronate esters.23, 26 To understand the species involved in the Suzuki‐Miyaura reaction, the hydrolysis of monomer 1b was studied in THF/water mixtures. The respective boronic acid (Scheme 5) 1k is formed with the MIDA diacid precipitating in each case under these conditions. The bis MIDA monomer 1b is not completely soluble in THF at RT or 60 °C at the initial concentration used in the polymerization. However, the formed boronic acid, 1k, is soluble in THF.
As the hydrolysis to the boronic acid proceeded at 60 °C, eventually all of 1b present dissolved in the solution. This occurred after 2 h in the in situ 1H NMR experiment [Fig. 1(a)], as from this point onwards the combined integral area of resonances (between δ = 6.9–7.1 ppm) attributed to the 8 aryl protons (2H1 + 2H2 + 4H3) that are unchanged in the hydrolysis remains constant relative to integral area of the internal standard (mesitylene) peak at 6.7 ppm. The extent of hydrolysis was determined from comparing the integral area of the aryl hydrogens unaffected by hydrolysis (δ = 6.9–7.1 ppm) against those signals associated with the four aryl protons (4Hk) directly next to the B(OD)2 groupings in 1k (δ = 7.7 ppm). Compound 1b is hydrolyzed by 48% and 85% after 2 and 8 h, respectively. There is no evidence of protodeboronation of 1b or 1k (by 1H NMR spectroscopy).
Boric acid (δ 11B = 20.1) is observed as a very minor by‐product [Fig. 1(b)]. No significant changes observed from 24 to 48 h of reaction, where 1b is almost fully consumed.[](https://www.ncbi.nlm.nih.gov/mesh/C533766)
Conditions used in hydrolysis studies of bis MIDA monomer, 1b.[](https://www.ncbi.nlm.nih.gov/mesh/C533766)
(a) Collated aryl proton regions of 1H NMR spectra in d 8‐THF of the hydrolysis of 1b to 1k. Reaction conditions as described in Scheme 5. Mesitylene (Mes) was added as internal standard. (b) Collated 11B NMR spectra in d 8‐THF of the hydrolysis of 1b to 1k. Reaction conditions as described in Scheme 5. [Color figure can be viewed at wileyonlinelibrary.com][](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Polymerizations in Dioxane/Water Mixtures
*In the **Polymerizations in Dioxane/Water Mixtures** section:*
Initial Suzuki‐Miyaura cross‐coupling polymerizations of the pinacol boronate ester monomers (1a, 2a) and the MIDA boronate ester monomers (1b, 2b) with their respective dibromo co‐monomers (2c, 3) to produce PTFB and PPFB (Table 1, entries 1–4) were performed in dioxane/water mixtures similar to those outlined by Burke and coworkers.6 The BMIDA monomers proved to be only partially soluble in dioxane at room temperature, with the reactions initially taking on the appearance of an emulsion. However, all the boronate monomers proved to react very rapidly with their respective dibromo comonomers. For example, the reaction between PFB MIDA boronate ester monomer, 2b, and dibromo PFB monomer, 2c, (Table 1, entry 4), in dioxane/water solvent mixture (5: 1 mL) at 90 °C, proceeded with PFB polymer precipitating out of the solvent mixture within 10 min (similar timescales observed for reactions 1–3). The precipitated PPFB material was recovered from the solution, redissolved in hot toluene and reprecipitated in excess methanol.
GPC analysis of this sample indicated that polymer of number average molar mass (M n), M n = 14,300, had been formed before precipitation. The molar mass distributions of polymer in the remaining reaction solution were sampled after 24 h and 48 h. A narrower molar mass distribution of polymer in solution was evident after 48 h (Ð = 1.6) owing to the longer growing polymer chains having by this stage precipitated out of solution, leaving predominantly inactive oligomers (M n ∼ 5,000) still in solution. The MALDI‐TOF mass spectrum of the precipitated PPFB polymer recovered from reaction 4 is presented in Figure 2, with assignment details outlined in Supporting Information Table S1. The most significant distribution of mass peaks can be attributed to PPFB terminated on both ends by hydrogen atoms, that is, [H–(2)n–H]+, with structures equating to at least n = 15 evident.
The next most significant series can be attributed to PPFB end terminated with one bromine and one hydrogen atom, that is, [H–(2)n–Br]+, with structures equating at least n = 14 evident. The insolubility of the higher molecular weight PTFB and PPFB polymers in these dioxane/water mixtures (Table 1, entries 1–4) results in premature precipitation of the polymers before the reaction is complete.[](https://www.ncbi.nlm.nih.gov/mesh/C000621940)
MALDI‐TOF mass spectrum of precipitated PPFB polymer recovered from cross‐coupling reaction of bis BMIDA PFB monomer (2b) with dibromo PFB monomer (2c) (Table 1, entry 4). [Color figure can be viewed at wileyonlinelibrary.com][](https://www.ncbi.nlm.nih.gov/mesh/D010655)
Suzuki‐Miyaura cross‐coupling reactions of the TFB or PFB monomers (1, 2) with the dibromo fluorene monomer, 4, gave poly (9,9‐dioctylfluorene [F8] ‐alt‐ 4‐octylphenyl diphenyl amine [TFB]) (PF8TFB) and poly(9,9‐dioctylfluorene [F8] ‐alt‐ N,N′‐bis(4‐octylphenyl)‐N,N′‐diphenyl phenylenediamine) [PFB] (PF8PFB). The solvent composition initially used was dioxane/water, 4:1 (mL) and the reaction is conducted at 90 °C (Table 1, entries 5–8). Copolymerizations with the dibromo fluorene monomer, 4, improved the solubility of the polymers in the dioxane/water mixtures, with the respective polymers only precipitating out of solution after 4 h. The overall M n of all PF8TFB material collected from reactions with pinacol ester and BMIDA functionalized monomers (Table 1, entries 5 and 6) were 14,800 and 17,200, respectively, after this short reaction time.
This single phase system yielded polymers exhibiting very similar molar mass distributions regardless of whether the starting boronate monomer was either pinacol or BMIDA based (Ð = 2.2). A closer comparison of the molar mass distributions of the polymers recovered from the two reactions clearly indicates that the BMIDA reaction (Table 1, entry 6) is marginally shifted toward higher molar mass material, with a new peak emerging equating to a molar mass of 60,000, as outlined in Figure 3. The doubling in molar mass suggests late stage chain coupling as previously observed in the synthesis of P3HT from BMIDA monomers.16[](https://www.ncbi.nlm.nih.gov/mesh/C533766)
GPC molar mass distributions of PF8TFB polymers obtained from cross‐coupling reactions of pinacol ester or MIDA boronate ester TFB monomers (1a or 1b) with dibromo F8 monomer (4) in dioxane/water, 4: 1 (mL) solvent mixtures at 90 °C after 4 h (Table 1, entries 5 and 6).[](https://www.ncbi.nlm.nih.gov/mesh/D004159)
## Polymerizations in Biphasic Conditions with Toluene
*In the **Polymerizations in Biphasic Conditions with Toluene** section:*
Toluene was added to the solvent mixtures to maintain the solubility of the polymers during the course of the reaction. Reactions were conducted between PPFB monomers, 2a or 2b with 2c, in solvent mixtures that included 1 mL of toluene at the expense of the dioxane content. The reaction with the PFB pinacol ester again resulted in rapid precipitation of polymer (Table 1, entry 9), whereas the reaction with the MIDA boronate ester remained homogeneous for a few hours, before material precipitated out of solution (Table 1, entry 10). Polymers of intermediate molecular weights (M n = 14,000–15,000) were obtained after 5 h in each case, with broader mass distributions than those prepared in toluene free reactions (Table 1, entries 3 and 4). The proportion of toluene present in the solvent mixtures was further increased to toluene (T)/dioxane (D)/water (W), 3:2:1 (mL) and PTFB polymers (Table 1, entries 11 and 12) prepared by reactions at 90 °C for 24 h.
The polymers remained in solution throughout the reaction but were recovered with low M n = 6,000–7,000. The GPC trace of the polymer from reaction 12 also indicated conversion to polymer of 96%. A similar reaction with the bis MIDA boronate ester completed at 60 °C (Table 1, entry 13) produced polymer of similar molar mass (M n = 6,000) with a lower Ð of 1.7, while showing only low overall conversion to polymer of 66% (Supporting Information Fig. S7). This suggests that the two phase solvent reactions are not consistent with a conventional step‐growth polymerization.
The low conversion to polymer evident after 24 h at 60 °C indicated that, even in presence of base, hydrolysis and cross‐coupling of the bis MIDA monomer under these biphasic conditions was relatively slow; hence, all future reactions were completed at higher temperatures. Lower molar mass polymer, M n = 5,000 was obtained from reactions completed in a T/D/W, 3:1:1 (mL) solvent system at 90 °C (Table 1, entry 15), even using double the amount of catalyst (5 mol %), ligand (10 mol %), and base (8 equivalents). The cross‐coupling reaction of the bis pinacol boronate ester TFB monomer (1a) with the dibromo monomer (3) in a similar solvent mixture at 90 °C (Table 1, entry 14) yielded polymer of higher molar mass, M n = 11,500. The MALDI‐TOF mass spectrum of PFTB polymer produced from the cross‐coupling reaction of the bis pinacol boronate ester FTB monomer (1a) with dibromo FTB monomer (3) (reaction 14) is presented in Figure 4(a), with assignment details outlined in Supporting Information Table S2. The FTB residue masses for 1 and 3 repeat units are 356 and 299 mass units, respectively.
The main series () are associated with even numbered residue PFTB chain structures terminated at both ends by hydrogen atoms, that is, [H–(1 – 3)n–H]+ (Table 2). The two next most significant series are attributed to odd numbered residue polymers in which both ends are terminated with the same monomer residues. The most prominent of these two series () can be attributed to structures containing residues of monomer 1 next to ends terminated with hydrogen atoms, that is, [H–(1 – 3)n–1 – H]+. The other series () can be attributed to the equivalent for 3, that is, [H – 3–(1 – 3)n–H]+. The MALDI‐TOF mass spectrum of PFTB produced from the reaction of bis BMIDA TFB monomer (1b) with 3 (reaction 15) is presented in Figure 4(b), with assignment details outlined in Supporting Information Table S3.
The predominant series of peaks () would appear to equate to PTFB structures rich in residues of 1, for example, [H–(1)3–(1 – 3)n–H]+. The remaining distribution series are present in similar smaller amounts, with four more also associated with polymer structures containing excess residue 1 than would be expected in a classical cross coupling polymerization, either [H–(1)x–(1 – 3)n–H]+, where x = 1, 2, 4 (, , ), or a doubly bromine terminated version of the main peak distribution, defined as [Br – 3–(1)4–(1 – 3)n–Br]+ (). The prominence of these polymeric structures in the material obtained from cross‐coupling reaction with the BMIDA FTB monomer (1b) indicates that homocoupling of boron end groups resulting from the BMIDA monomer is competing with the cross‐coupling reaction resulting in a reduction in the molar mass of PFTB polymer obtained (M n = 5,100, Table 1, entry 15), relative to a similar reaction with the pinacol boronate ester FTB monomer (1a) (M n = 11,500, Table 1, entry 14). A significant amount of homocoupling, originating from boronate ester monomers used in Suzuki polycondensations of P(Cbz‐alt‐TBT) and PCDTBT, has recently been observed.27, 28 A difference in either reactivity or solubility of the two boronate ester monomers (1a and 1b) in these multiphase solvent mixtures may also contribute to changes in the relative rates of homo‐ versus heterocoupling in these particular reactions. As we discussed earlier, in some single phase (dioxane + water) reactions, there was little difference in the polymer produced from either boronate monomer in comparable reactions (Table 1, entries 5 and 6).
These results indicated that while adding toluene to the solvent mixture aided polymer solubility during the course of the reaction, creating a new phase had a detrimental effect on the molar mass of polymers obtained from MIDA boronate ester monomers, which could not be arrested by increasing the amount of the catalyst system or partially reducing the overall reaction volume.[](https://www.ncbi.nlm.nih.gov/mesh/D014050)
MALDI‐TOF mass spectra of PTFB polymers recovered from cross‐coupling reactions of (a) bis pinacol ester TFB monomer (1a) with dibromo TFB monomer (3) (Table 1, entry 14) and (b) bis BMIDA TFB monomer (1b) with dibromo TFB monomer (3) (Table 1, entry 15). [Color figure can be viewed at wileyonlinelibrary.com][](https://www.ncbi.nlm.nih.gov/mesh/D000611)
PTFB Polymer Structures Assigned from MALDI Mass Spectra of Polymers Recovered from Cross‐coupling Reactions of Pinacol Ester TFB (1a) or BMIDA (1b) Monomer with Dibromo TFB Monomer (3) (Table 1, Entries 14 and 15) [Color table can be viewed at wileyonlinelibrary.com][](https://www.ncbi.nlm.nih.gov/mesh/D000611)
The molar mass distributions of polymers obtained from cross‐coupling reactions of the pinacol boronate ester monomer (1a) or BMIDA monomer (1b) with dibromo FTB monomer (3) in directly comparable reactions (Table 1, entries 14 and 15) at 90 °C are outlined in Figure 5. The reaction progress with time was also monitored, including at 80 °C (Table 1, entries 16 and 19; Supporting Information Fig. S8). In each instance, the polymer chains created in reactions starting with a BMIDA‐based monomer showed no signs of cross‐coupling further to create longer chains of higher molar mass beyond 5 h.[](https://www.ncbi.nlm.nih.gov/mesh/C000621940)
GPC molar mass distributions of PTFB polymers obtained from cross‐coupling reactions of bis pinacol ester or bis BMIDA functionalized TFB monomers (1a or 1b, respectively) with the dibromo TFB monomer (3) in T:D:W, 3:1:1 (mL), solvent mixtures at 90 °C (Table 1, entries 14 and 15) after different time intervals. [Color figure can be viewed at wileyonlinelibrary.com][](https://www.ncbi.nlm.nih.gov/mesh/D000611)
The step growth polymerizations involving the pinacol boronate ester monomer still contained active chain ends after 5 h that reacted further to create longer chains by GPC after 24 to 48 h. In one particular reaction containing a BMIDA monomer (Table 1, entry 19), a second batch of catalyst/ligand dissolved in toluene (similar to initial charge) was added to the reaction mixture after 5 h to see whether this would induce further cross‐coupling between the existing polymer chains. No change in molar mass distribution of polymer was observed even after a total reaction time of 24 h.[](https://www.ncbi.nlm.nih.gov/mesh/C000621940)
Cross‐coupling of the TFB boronate ester monomers (1a or 1b) with dibromo monomer 3 in a T/D/W, 1:1:1 mL at 90 °C (Table 1, entries 18, 19) gave a marked increase in the molar mass of polymer, with polymers of M n = 13,700 (Ð = 2.2) produced from the cross‐coupling reaction of monomer, 1a, with dibromo monomer, 3 (Table 1, entry 18). As before, higher molar mass PFTB polymers were produced in the cross‐coupling reactions of the TFB pinacol boronate ester monomer, 1a, with the dibromo monomer, 3, than were achieved with the BMIDA TFB monomer, 1b, in comparable multiphase reactions at 80 and 90 °C (Fig. 6). Reactions with the dibromo fluorene monomer (4) were also completed in T:D:W, 1:1:1 (mL) solvent mixtures (Table 1, entries 20–23 and Fig. 7). The inclusion of the dibromo F8 monomer (4) as a co‐monomer, instead of dibromo FTB monomer (3), increased the molar masses of polymers obtained in the reactions with both the pinacol boronate ester and BMIDA monomers: peak molar masses (M p) of the polymers obtained in reactions with 1a increased from 30,000 to 90,000, and in reactions with 1b from 25,000 to 84,000. Cross‐coupling reactions of the MIDA boronate ester TFB or PFB monomer (1b or 2b) with dibromo F8 monomer (4) (Table 1, entries 21 and 23) both yielded polymers with M n of approximately 20,000.
A scale‐up reaction of entry 23 produced PF8PFB in yield of 0.98 g (80%) exhibiting similar molecular weight (experimental details in Supporting Information Table S4 and Fig. S9).[](https://www.ncbi.nlm.nih.gov/mesh/D001897)
GPC molar mass distributions of PTFB polymers obtained from cross‐coupling reactions of pinacol ester or BMIDA functionalized TFB monomers (1a or 1b, respectively) with the dibromo TFB monomer (3) in T:D:W, 1:1:1 (mL), solvent mixtures at 80 °C and 90 °C (Table 1, entries 16–19). [Color figure can be viewed at wileyonlinelibrary.com][](https://www.ncbi.nlm.nih.gov/mesh/D000611)
GPC molar mass distributions of polymers obtained from cross‐coupling reactions of pinacol ester or BMIDA monomers (1a or 1b) with dibromo TFB or F8 monomers (3 or 4) in T:D:W, 1:1:1 (mL), solvent mixture at 90 °C after 24 h (Table 1, entries 18–21). [Color figure can be viewed at wileyonlinelibrary.com].[](https://www.ncbi.nlm.nih.gov/mesh/C000621940)
## CONCLUSIONS
*In the **CONCLUSIONS** section:*
Suzuki‐Miyaura cross‐coupling reactions of bis pinacol boronate ester or bis MIDA boronate ester monomers, with dibromo comonomers, produced high molar mass conjugated polymers after optimization of the reaction conditions. It appears that polymer chains, generated from cross‐coupling reactions of bis BMIDA monomer (1b) with dibromo comonomer (3) at 90 °C, stop growing within a reaction time of 5 h using this catalyst system in biphasic solvent mixtures. MALDI‐TOF mass spectral evidence suggest homocoupling of residues associated with the MIDA boronate ester monomer is occurring which could contribute to limiting the achievable molar mass of polymer.[](https://www.ncbi.nlm.nih.gov/mesh/C000621940)
Cross‐coupling reactions undertaken in dioxane + water mixtures resulted in rapid precipitation of the (co)polymer often before the reaction had reached maximum molar mass or high conversion. However, the BMIDA monomer (1b) proved as successful as the pinacol ester (1a) in these single phase copolymerizations with dibromo fluorene monomer (4) to produce PF8TFB. A polymer of high molar mass, M n = 17,000 (Ð = 2.2) was precipitated from the solution within only 4 h. Optimum reaction conditions for maintaining the polymer in solution, to achieve higher molar mass, were achieved in T:D: W (1:1:1) solvent mixtures in biphasic reactions. PF8TFB and PF8PFB polymers (approximately M n = 20,000) with broad molar mass distributions, Ð = 3.0, were obtained from reactions starting with the respective BMIDA monomers in this solvent mixture after 24 h (Table 1, entries 21 and 23).[](https://www.ncbi.nlm.nih.gov/mesh/C025223)
## Supporting information
*In the **Supporting information** section:*
# REFERENCES AND NOTES
*In the **REFERENCES AND NOTES** section:*
|
# Introduction
NAAG Peptidase Inhibitors Act via mGluR3: Animal Models of Memory, Alzheimer’s, and [Ethanol](https://www.ncbi.nlm.nih.gov/mesh/D000431) Intoxication
# Abstract
*In the **Abstract** section:*
Glutamate carboxypeptidase II (GCPII) inactivates the peptide neurotransmitter N-acetylaspartylglutamate (NAAG) following synaptic release. Inhibitors of GCPII increase extracellular N[AAG levels and are effica](https://www.ncbi.nlm.nih.gov/mesh/C027172)ci[ous ](https://www.ncbi.nlm.nih.gov/mesh/C027172)in animal models of clinical disorders via NAAG activation of a group II [meta](https://www.ncbi.nlm.nih.gov/mesh/C027172)botropic glutamate receptor. mGluR2 and mGluR3 knock-out (ko) mice were[ use](https://www.ncbi.nlm.nih.gov/mesh/C027172)d to test the hypothesis that mGluR3 mediates the activity of GCPII inhibitors ZJ43 and 2-PMPA in animal models of memory and memory loss. Short- (1.5 h) and long- (24 h) term novel ob[ject](https://www.ncbi.nlm.nih.gov/mesh/C486211) reco[gnitio](https://www.ncbi.nlm.nih.gov/mesh/C402107)n tests were used to assess memory. Treatment with ZJ43 or 2-PMPA prior to acquisition trials increased long-term memory in mGluR2, but not mGluR3, ko mice.[ Nin](https://www.ncbi.nlm.nih.gov/mesh/C486211)e mo[nth-ol](https://www.ncbi.nlm.nih.gov/mesh/C402107)d triple transgenic Alzheimer’s disease model mice exhibited impaired short-term novel object recognition memory that was rescued by treatment with a NAAG peptidase inhibitor. NAAG peptidase inhibitors and the group II mGluR agonist, LY354740, reversed the short-term memory deficit induced by acute ethanol administration in wild type mic[e. 2-PMP](https://www.ncbi.nlm.nih.gov/mesh/C104753)A also moderated the effect of ethanol on short-term memor[y in mG](https://www.ncbi.nlm.nih.gov/mesh/D000431)luR2 ko mice but failed to do so in[ mGluR](https://www.ncbi.nlm.nih.gov/mesh/C402107)3 ko mice. LY354740 and ZJ43 b[locked ](https://www.ncbi.nlm.nih.gov/mesh/D000431)ethanol-induced motor activation. Both GCPII inhibitors and LY354740 also signi[ficantly](https://www.ncbi.nlm.nih.gov/mesh/C104753) mode[rate](https://www.ncbi.nlm.nih.gov/mesh/C486211)d the los[s of mo](https://www.ncbi.nlm.nih.gov/mesh/D000431)tor coordination induced by 2.1 g/kg ethanol treatmen[t. These](https://www.ncbi.nlm.nih.gov/mesh/C104753) data support the conclusion that inhibitors of glutamate carboxypeptidase II are[ effica](https://www.ncbi.nlm.nih.gov/mesh/D000431)cious in object recognition models of normal memory and memory deficits via an mGluR3 mediated process, actions that could have widespread clinical applications.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
N-Acetylaspartylglutamate (NAAG), a prevalent and widely distributed peptide co-transmitter, is inactivated by glutamate carboxypeptidase II (GCPII) following synaptic release []. Inhibitors of GCPII [, ] are effective in animal models of several clinical conditions [reviewed in –]. These inhibitors enhance long-term memory in the 24 h delay novel object recognition test [], improve memory in an animal model of multiple sclerosis [], rescue behaviors and short-term memory impairment in animal models of schizophrenia [–]. Consistent with these results, mice that are null mutant for GCPII demonstrate full memory in the 24 h delay novel object recognition test, while their heterozygous littermates and wild type C57Bl mice exhibit no significant recall in this test of long-term memory []. GCPII inhibitors also are analgesic in models of inflammatory and neuropathic pain [–] and reduce the effects of traumatic brain injury
[] while GCPII knockout (ko) mice are protected from peripheral neuropathy and ischemic and traumatic brain injury [–].[](https://www.ncbi.nlm.nih.gov/mesh/C027172)
NAAG reduces transmitter release from neurons and synaptosomes via a group II mGluR receptor [, ]. Inhibitors of GCPII elevate extracellular levels of NAAG and also reduce the release of glutamate and other transmitters [, , ]. These neurochemical actions of the peptidase inhibitors and their positive effects in animal models are blocked by the group II mGluR antagonist LY341495. While a substantial body of data supports the conclusion that the peptide activates a group II metabotropic receptor [, ], highly purified NAAG fails to activate mGluR2 or mGluR3 receptors expressed in transfected cells [, ], results suggesting that some reports of NAAG activation of a group II mGluR were due to the presence of low levels of residual glutamate (≤0.5% []) in commercially available NAAG.
In contrast, data from other studies preclude the conclusion that the NAAG activity is due to this level of contaminating glutamate []. Consistent with an action of NAAG at mGluR3, NAAG peptidase inhibition blocks the motor activation effects of phencyclidine in mGluR2, but not mGluR3, ko mice []. In those studies where the effects of NAAG or NAAG peptidase inhibition have not been shown to be blocked by a group II mGluR antagonist, it is possible that NAAG is acting as an antagonist at a subclass of NMDA receptors [].[](https://www.ncbi.nlm.nih.gov/mesh/C027172)
In order to further test the hypothesis that a group II mGluR, specifically mGluR3, mediates the procognitive efficacy of GCPII inhibitors, these compounds were tested across a series of animal models that included short- and long-term novel object memory, Alzheimer’s disease, and acute alcohol intoxication, using a group II antagonist in wild type mice and testing mice that are null mutant for mGluR2 and mGluR3.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## Methods
*In the **Methods** section:*
## Animals
*In the **Animals** section:*
The experimental protocols used in this research were approved by the Georgetown University Animal Care and Use Committee and are consistent with guidelines of the US National Institutes of Health. Efforts were made to reduce animal suffering and to minimize the number of animals used. Adult male C57BL/6NCr mice from the National Cancer Institute, Frederick Research Center were tested once at 2–4 months of age and used for all studies except those involving the knock out or transgenic mice. The mGluR2 and mGluR3 ko mice (knock out R1 cell lines into C58Bl/6 mice and backcrossed into ICR[CD1] mice) [] were provided by Eli Lilly Pharmaceuticals and tested twice at 3–5 months of age in a novel object recognition study.
Wildtype littermates for these ko mouse colonies were not available. There are no published studies that compare the performances of these ko mice with their wild type littermates in the novel object recognition test. The triple-transgenic mouse model (3xTg line) that expresses three genes associated with familial Alzheimer’s disease, namely APPSwe, PS1M146V, and tauP301L [] were from Jackson Labs (Strain: B6;129-Psen1 Tg(APPSwe,tauP301L)1Lfa/Mmjax; genetic background: (129 × 1/SvJ x 129S1/Sv)F1-Kitl<+>; JAX MMRRC Stock# 034830) and tested at 2–9 months of age in the short-term novel object recognition test. Mice were housed 5 to a cage and maintained on a 12:12 h light–dark cycle with food and water available ad libitum. Behavioral testing was performed during the light cycle between 10 am and 4 pm.
## Drugs
*In the **Drugs** section:*
The GCPII/NAAG peptidase inhibitor ZJ43 (ZJ43 (N-[[[(1S)-1-carboxy-3-methylbutyl]amino]carbonyl]-l-glutamic acid) was synthesized as previously described [] and provided by Alan Kozikowski. The GCPII inhibitor 2- ((2-(phosphonomethyl)pentane-1, 5-dioic acid) [, ]) was from Reagents4Research, LLC (Hangzhou, CN). LY341495, a selective group II mGluR antagonist [], and LY354740, a heterotropic group II mGluR agonist [], were from Tocris Cookson Ltd. (Bristol, UK).
All compounds were dissolved in saline and pH was adjusted to 7.4 prior to i.p. injection. Ethanol (2.1 g/l, ip) was given as a concentration of 20% v/v in saline. Doses of ZJ43, 2-PMPA, LY341495, LY354740 and ethanol were based on data from published and preliminary studies.[](https://www.ncbi.nlm.nih.gov/mesh/C486211)
## Novel Object Recognition Test
*In the **Novel Object Recognition Test** section:*
Novel object recognition is a validated and widely used test for assessing recognition memory [–], including in studies of aging [, ] and Alzheimer’s disease mouse models [, ]. Individual mice (3–4 month old) were placed in a 22 × 32 × 30 cm testing chamber with beige walls for a 5 min habituation interval followed by i.p. injection with saline, 2-PMPA (50 mg/kg) or ZJ43 (150 mg/kg), with or without LY341495, and returned to home cage. Thirty minutes later, mice were placed in the testing chamber for 10 min with two identical objects (acquisition session). Mice were returned to home cages and 1.5 h (short-term memory) or 24 h (long-term memory) later were returned to testing chamber in the presence of one of the original objects and one novel object (recognition session) for 10 min. Wild type mice exhibit short-term but not long-term memory in this test [, ].
The original objects consisted of two smooth surfaced weighted red cylinders 7 cm high × 4 cm diameter at base. The novel object consisted of a blue, 7 cm high × 5 cm diameter (base) round pyramid. The acquisition and recognition sessions were video recorded and the time mice spent exploring each object was assessed by an observer who was blinded to drug treatment and genotype. The chambers and objects were cleaned with ethanol between trials. Exploratory behavior was defined as sniffing, touching and directing attention to the object.
In preliminary studies, naïve mice exhibited no significant preference for the red cylinder or the blue pyramid. Exploration time (Table 1) is expressed as the mean ± the standard error of the mean (SEM). For the acquisition session, the recognition index was calculated as (time exploring one of the objects/the time exploring both objects) × 100. For the recognition session, the RI was calculated as (time exploring the novel object/the time exploring both the familiar and novel object)
× 100.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
Exploration time data from short-term (Figs. 2, 3, and 4) and long-term novel object recognition (Fig. 1)
Long-term novel object recognition memory test in mGluR2 and mGluR3 KO mice. In this and the following novel object recognition figures: for the acquisition session, the recognition index (RI) was calculated as (time exploring one of the objects/the time exploring both objects) × 100. For the recognition session, the RI was calculated as (time exploring the novel object/the time exploring both the familiar and novel object) × 100. During the acquisition phase, each group of mice explored each of the two identical objects about the same amount of time (recognition index ~50). During the recognition phase 24 h later, the mGluR2 KO mice (m2ko) treated with saline explored the novel and familiar object similar amounts of time while those treated with 2-PMPA (100 mg/kg) or ZJ43 (150 mg/kg) explored the novel object twice as often as the original object (recognition index ~70), while the NAAG peptidase inhibitors had no procognitive effect in the mGluR3 ko mice (m3ko). m2ko/saline, n = 11; m2ko/PMPA, n = 12; m2ko/ZJ43, n = 11; m3ko/saline, n = 11; m3ko/PMPA, n = 11; m3ko/ZJ43, n = 12. *p < 0.05, **p < 0.01, ***p < 0.001 for comparison between acquisition session and recognition session within treatment group in Figs. 1, 2 and 3[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
Short-term novel object recognition memory test in triple transgenic Amice. Eight week old AD mice explored the novel object significantly more than the familiar object while the 9 month old AD mice failed to discriminate between the novel and familiar object. 2-PMPA (100 mg/kg) restored the ability of the older mice to discriminate between the novel and familiar object. 8 week, n = 15; 9 mos/saline, n = 10; 9 mos/PMPA, n = 10[](https://www.ncbi.nlm.nih.gov/mesh/C402107)
Ethanol impairment of short-term novel object recognition is reversed by NAAG peptidase inhibitors and the group II mGluR agonist LY354740. Mice treated with two injections of saline before the acquisition phase of the test, explored the novel object significantly more than the familiar object 1.5 h later (recognition index ~70). Ethanol (2.1 g/kg) blocked discrimination of the novel object in the retention session. Pretreatment with ZJ43 (150 mg/kg) and 2-PMPA (100 mg/kg) reversed the cognitive deficits induced by ethanol. Pretreatment with LY354740(LY40) dose dependently reversed the effects of ethanol. S-S, n = 10; S-EtOH, n = 11; ZJ43-EtOH, n = 6; 2-PMPA-EtOH, n = 6; LY40(2)-EtOH, n = 12; LY40(5)-EtOH, n = 12; LY40(10)-EtOH, n = 7[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
mGluR3 mediates the procognitive effects of NAAG peptidase inhibition in the ethanol treatment model. mGluR2 (m2KO) and mGluR3 KO (m3KO) mice exhibited short-term (1.5 h) memory in the novel object memory test. Mice of both strains failed to discriminate between the novel and familiar object when treated with 2.1 g/kg ethanol prior to the acquisition trial. 2-PMPA (100 mg/kg) partially reversed the effect of ethanol in mGluR2 but not mGluR3 mice. mGluR2ko-S, n = 11; mGluR2ko-EtOH, n = 12; mGluR2ko-PMPA + EtOH, n = 10; mGluR3ko-S, n = 9; mGluR3ko-EtOH, n = 9; mGluR3ko-PMPA + EtOH, n = 10[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
To study the effects of ethanol on short-term memory, mice were placed the testing chamber for a 5 min habituation interval followed by injection with saline, ZJ43 (150 mg/kg), 2-PMPA (100 mg/kg) or LY354740 (2, 5, 10 mg/kg) and returned to home cage. Thirty minutes later mice were injected with ethanol (2.1 g/kg, i.p.) and returned to their home cage for 10 min after which they were placed in a testing chamber for 10 min with two identical objects (acquisition session). Mice were returned to home cages and 1.5 h later were placed back into the testing chamber for 10 min in the presence of one of the original objects and one novel object (recognition session).[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## Open Field Motor Activation Test
*In the **Open Field Motor Activation Test** section:*
High doses of ethanol induces increases in motor activity (40) and loss of coordination (41) in mice. In the present study, mice were habituated to an open field chamber (Med Associates, St., Albans, Vermont, ENV-515 43 × 43 cm, with infrared beams and detectors) for 30 min prior to i.p. injection with saline, ZJ43 (150 mg/kg)with or without LY341495 (3 mg/kg), or with LY354740, returned to the chamber for 15 min, and injected (i.p.) with 2.1 g/kg of ethanol. Locomotor activity was then recorded as distance travelled during 10 min in the open field chamber.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## Rotorod Test
*In the **Rotorod Test** section:*
The rotorod was used to assess motor coordination and balance. Mice were injected (i.p.) with saline, ZJ43 (150 g/kg), 2-PMPA (10, 50, 100 mg/kg) with or without LY341495, or with LY354740 (10 mg/kg), returned to their home cage in the testing room and 15 min later were injected (i.p.) with 2.1 g/kg ethanol. Forty-five minutes later mice were placed on the drum (70 mm dia) facing away from the direction of the rotation so they can walk forward at constant speed (4 rpm) for 10 s of habituation. The drum was then accelerated over 3 min, from 0 to 40 rpm (with cut off time = 3 min) and the latency to fall from the drum was recorded. Each animal was tested three times with 15 min between trials.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## Statistical Analysis
*In the **Statistical Analysis** section:*
For the novel object recognition test, the time spent exploring each object was analyzed by two-way repeated measures ANOVA, with session as within-subject factor and treatment as a between-subject factor. Discrimination ratio data were analyzed by one-way ANOVA followed by Student–Newman–Keuls post-hoc test. Motor activation data and rotorod data were analyzed with GLM ANOVA followed by post-hoc Tukey test.
## Results
*In the **Results** section:*
## Total Exploration Times
*In the **Total Exploration Times** section:*
The time exploring individual objects during acquisition trials and recognition trials for each treatment group are presented in Table 1. Within treatment groups, there was a wide range of total exploration times in the acquisition and retention sessions. Drug treatments combined with ethanol tended to result in less attention to the objects during the acquisition trials. This was particularly evident in the LY354740 with ethanol (2.1 g/kg) treatment groups but also observed in the ko mice treated with ethanol. An additional anomaly is the substantial difference in the total exploration times of mGluR2 ko/saline treated mice versus the mGluR2ko/ethanol treated mice.
Despite these differences between groups, there are clear and significant drug effects in the recognition sessions. The reliability of the novel object recognition data is supported by the fact that, despite differences in the total exploration times among the treatment groups, both objects are nearly equally attended during the acquisition session across all groups and that the recognition data fall clearly into two categories: nearly equal attention to both objects (failed memory) or significantly greater attention to the novel object than familiar object (memory). There were no apparent effects of 2-PMPA or ZJ43 on mean exploration times during the acquisition trials relative to saline treated mice.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## NAAG Peptidase Inhibitors ZJ43 and 2-PMPA are Procognitive for Long-term Memory in mGluR2 and but not mGluR3 Knockout Mice
*In the **NAAG Peptidase Inhibitors ZJ43 and 2-PMPA are Procognitive for Long-term Memory in mGluR2 and but not mGluR3 Knockout Mice** section:*
Mice lacking functional mGluR2, like wild type C57BL mice [], explored the two similar objects to the same extent during the acquisition trial and failed to discriminate the between the novel and familiar objects when tested one day later (Fig. 1). Also as in wild type mice [], both NAAG peptidase inhibitors, ZJ43 (150 mg/kg, i.p.) and 2-PMPA (50 mg/kg, i.p.) increased (p < 0.001) exploration of the novel object when the drugs were given prior to the acquisition trial on day 1.[](https://www.ncbi.nlm.nih.gov/mesh/C486211)
mGluR3 knockout mice also spent similar amounts of time exploring the two identical objects during the acquisition trial and the novel and familiar object during the retention trial. However, neither ZJ43 nor 2-PMPA was procognitive in these mice based on their failure to elicit greater exploration of the novel object in the retention trial.[](https://www.ncbi.nlm.nih.gov/mesh/C486211)
## NAAG Peptidase Inhibitor 2-PMPA is Procognitive for Short-Term Memory in Triple Transgenic Alzheimer’s Mice
*In the **NAAG Peptidase Inhibitor 2-PMPA is Procognitive for Short-Term Memory in Triple Transgenic Alzheimer’s Mice** section:*
The triple-transgenic (3xTg line) mice express three genes associated with familial Alzheimer’s disease, APPSwe, PS1M146V, and tauP301L. The same mice were tested at 2, 5 and 9 months of age in the novel object test of short-term memory. At 2 months of age, these mice demonstrated short-term memory while they failed in this test at 5 and 9 months of age (Fig. 2). The 9-month old mice had a high level of exploratory behavior in both the acquisition and retention trials (Table 1). Treatment with 100 mg/kg of 2-PMPA prior to the acquisition phase reversed this memory deficit (p = 0.05) in the 9-month old mice as demonstrated by their level of exploration of the novel object relative to the familiar object.[](https://www.ncbi.nlm.nih.gov/mesh/C402107)
## Acute Ethanol Intoxication and Short-Term Memory
*In the **Acute Ethanol Intoxication and Short-Term Memory** section:*
C576BL/6 mice (3–4 month old) treated with saline prior to the acquisition trial explored the two identical objects about the same amount of time (Fig. 3) and when presented with one novel object and one familiar object 1.5 h later, they explore the novel object about twice as frequently as the familiar object []. In this study, there was a main effect of treatment and session (F(6,59) = 4.51, p < 0.01, F(1,59) = 199.32, p < 0.001) and a treatment x session interaction (F(6,59) = 3.73, p < 0.01).
Mice treated with ethanol (2.1 g/kg, i.p.) prior to the acquisition trial also explored the two objects about the same amount of time. However, in contrast to the saline treated control mice, the ethanol treated mice explored the novel and familial objects equally in the retention trial suggesting a failure to recall or recognize the familiar object relative to the novel one. The NAAG peptidase inhibitors ZJ43 (150 mg/kg), 2-PMPA (100 mg/kg) and the type 2/3 metabotropic glutamate receptor agonist LY354740 (10 mg/kg) reverse the effects of ethanol on novel object recognition (p < 0.01, p < 0.001 and p < 0.01 respectively). A low dose of the type 2/3 metabotropic glutamate receptor antagonist LY341495 (2 mg/kg) failed to block the effects of 2-PMPA on ethanol treatment in this assay.
When given alone to control mice prior to acquisition, 2 and 3 mg/kg of LY341495 reduced memory on the retention trial and could not be used to confirm the role of NAAG at the group II metabotropic glutamate receptors in this study. Mice given 1.7 g/kg (i.p.) ethanol did not exhibit a significant loss of short-term memory in this assay (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
In the absence of confirmation that the mGluR mediated the efficacy of NAAG peptidase inhibition in the ethanol intoxication model, the study was repeated using mGluR2 and mGluR3 ko mice treated with saline or 2.1 mg/kg (i.p.) ethanol (Fig. 4). Saline treated mice of both strains showed a significant level of recognition of the novel object and this short-term memory was blocked by pretreatment with ethanol. mGluR2 ko mice pretreated with 2-PMPA (100 mg/kg) prior to ethanol administration demonstrated significant memory (p < 0.05) in the recognition trials while 2-PMPA was without a significant effect in the ethanol treated mGluR3 ko mice.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## NAAG Peptidase Inhibitors Block Ethanol-Induced Motor Activation
*In the **NAAG Peptidase Inhibitors Block Ethanol-Induced Motor Activation** section:*
Ethanol (2.1 g/kg, i.p.) induced a prompt increase (p < 0.001) in motor activation in mice placed in an open field chamber to which they previously had been habituated (Fig. 5). There was a main effect of drug (saline, ZJ43 or LY doses, F(6,80) = 4.679, p < 0.001) and treatment (saline or ethanol, F(1,80) = 15.503, p < 0.001). Pretreatment with ZJ43 (50, 100 and 150 mg/kg, i.p.) dose dependently reduced motor activation during the 10 min interval immediately following ethanol administration (50 mg/kg, p < 0.01 and 150 mg/kg, p < 0.001).
The group II metabotropic glutamate receptor agonist LY354740 (10 mg/kg) reversed the effects of ethanol (p < 0.001). To confirm the role of NAAG and a type 2 or 3 metabotropic glutamate receptor in mediating the effects of NAAG peptidase inhibition, ZJ43 (150 mg/kg) was co-administered with the group II antagonist LY341495 (3 mg/kg). The antagonist reversed the effect of 150 mg/kg ZJ43 (p < 0.05). The group II mGluR agonist LY354740 (10 mg/kg, i.p.) also blocked the ethanol effect.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
Ethanol-induced motor activation reversed by ZJ43 and group II agonist LY354740. Ethanol (2.1 g/kg, i.p.) increases motor activation in open field test. Pretreatment with ZJ43 (50, 100 and 150 mg/kg, i.p.) dose dependently reduced motor activation during the 10 min interval immediately following ethanol administration. The group II metabotropic glutamate receptor agonist LY354740 (10 mg/kg) was similarly effective in moderating the effects of ethanol. The group II metabotropic glutamate receptor antagonist LY341495 reversed the effect of ZJ43 (150 mg/kg) on ethanol-induced motor activation. S-saline. ZJ = ZJ43, LY95 = LY341495, LY40 = LY354740. N: s/s (11), ZJ150/s, ZJ100/Et-OH, LY95 + ZJ150/Et-OH (9), s/Et-OH (20), ZJ50/Et-OH, ZJ150/Et-OH, LY40/Et-OH (10). *p < 0.05, **p < 0.01, ***p < 0.001[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## Rotorod Test of Ethanol-Induced Loss of Motor Coordination or Balance
*In the **Rotorod Test of Ethanol-Induced Loss of Motor Coordination or Balance** section:*
In the rotorod test, mice given with ethanol (2.5 g/kg, i.p.). There was a significant effect of drug (F(8,100) = 3.458, p < 0.01) and treatment (F(1,100) = 22.409, p < 0.001). Ethanol treatment produced a 55% reduction in latency to fall from the rotorod relative to saline treated mice (p < 0.001) (Fig. 6). ZJ43 (150 mg/kg, i.p.) blocked this effect of ethanol.
2-PMPA (10, 50 and 100 mg/kg) blocked the effect of 2.1 g/kg ethanol in a dose-dependent manner (p < 0.01 for 100 mg/kg and p < 0.05 for 50 mg/kg comparing with ethanol group). The effects of ZJ43 and 2-PMPA were reversed by the group II mGluR antagonist LY341495 (3 mg/kg, p < 0.05) while this antagonist alone had no significant effect on the ethanol-induced loss of motor coordination. The group II mGluR agonist LY354740 (LY40, 10 mg/kg) also reduced the effects of 2.1 g/kg ethanol on latency to fall (both p < 0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
ZJ43 and 2-PMPA moderate ethanol-induced loss of balance on Rotarod Test. a Ethanol (2.1 g/kg, i.p.) significantly reduced latency to fall relative to saline treated mice (p < 0.001). 2-PMPA reversed the effect in dose dependent manner (p < 0.01 for 100 mg/kg and p < 0.05 for 50 mg/kg comparing with ethanol group). ZJ43 (150 mg/kg) and LY354740 (10 mg/kg) also reduced the effects of 2.1 g/kg ethanol on latency to fall (both p < 0.05). The group II antagonist LY341495 (3 mg/kg) blocked the effect of 2-PMPA (100 mg/kg) and ZJ43 (150 mg/kg), both at p < 0.05 versus 2-PMPA and ZJ43. S-saline, LY95 = LY341495, LY40 = LY354740. All groups are compared individually for statistical significance versus the saline-ethanol group. N: s/s(10), s/EtOH(12), ZJ43-EtOH(12), ZJ43 + LY95/EtOH(17), 2-PMPA100-EtOH(8), 2-PMPA50-EtOH(8), 2-PMPA10-EtOH(8), 2-PMPA100 + LY95(3)EtOH(8), LY95/s(10), LY40/EtOH(8), LY95/EtOH(10)[](https://www.ncbi.nlm.nih.gov/mesh/C486211)
## Discussion
*In the **Discussion** section:*
## mGluR3 Mediates Procognitive Efficacy of NAAG Peptidase Inhibition
*In the **mGluR3 Mediates Procognitive Efficacy of NAAG Peptidase Inhibition** section:*
The purpose of this study was to further test the procognitive effects of NAAG peptidase inhibition across models of normal memory and in clinically relevant models of cognitive deficits. In parallel, the goal was to test the hypothesis that these procognitive effects are mediated by NAAG activation of the group II metabotropic glutamate receptor, mGluR3. While many strains of mice are reported to demonstrate short-term (1.5 h) memory in the novel object recognition test, there are no reports of mice exhibiting long-term (24 h) memory in this test. Similarly, the mGluR2 and mGluR3 ko mice in this study showed significant memory in the short-term test (Fig. 4) and the absence of memory of the familiar object in the long-term memory test (Fig. 1). In previous studies, NAAG peptidase inhibition reversed short-term memory deficits elicited by a low dose of the NMDA antagonist MK801
[] and had procognitive activity in the 24 h delay test []. Both of these actions were blocked by the group II mGluR antagonist LY341495, consistent with a series of studies that support the conclusion that the peptide mediates group II mGluR activation [, , ]. The finding that the procognitive effects of 2-PMPA and ZJ43 in the 24 h delay novel object recognition test were observed in mGluR2 but not mGluR3 ko mice (Fig. 1), supports the conclusion that mGluR3 is the group II receptor mediating these procognitive actions. This conclusion is further strengthened by the efficacy of 2-PMPA in partially reversing ethanol-induced cognitive impairment of short-term memory in mGluR2 but not mGluR3 ko mice (Fig. 4). A similar result in support of a role for mGluR3 in the efficacy of NAAG peptidase inhibition was observed in a mouse model of schizophrenia [].
Central to the conclusion that these effects of the peptidase inhibitors are mediated by the activation of mGluR3 by elevated levels of synaptically released NAAG, rather than by the drugs directly, are the reports that high levels (100 UM) of ZJ43 and 2-PMPA do not activate group II receptors in vitro [, ].[](https://www.ncbi.nlm.nih.gov/mesh/C027172)
These data on the mGluR3 mediated procognitive actions of NAAG peptidase inhibition also are consistent with the broader hypothesis that this receptor plays a more general role in memory formation or retrieval. Supporting this view, mGluR3 ko mice showed deficits in working memory when tested in T- and Y mazes and polymorphisms in mGlur3 are associated with cognitive deficits in schizophrenia patients [, ].
## NAAG as mGluR3 Agonist
*In the **NAAG as mGluR3 Agonist** section:*
A rigorous study of the effect of purified NAAG on hippocampal slices and cells transfected with mGluR2 or mGluR3 [] found no evidence of the peptide activating a group II receptor and suggested that some prior reports of NAAG activation of these receptors could have been due to a previous report that commercial NAAG was contaminated with 0.3–0.4% glutamate []. While this glutamate effect cannot be discounted for some studies, it does not seem consistent with other results []. For example, NAAG and glutamate dose responses for group II mGluR activation differed by no more than threefold when tested against cerebellar astrocytes, cells that expressed mGluR3 message but little if any mGluR2 []. The failure of NAAG to activate group II receptors in transfected cells [, ] further contrasts the report that NAAG is several orders of magnitude more potent than glutamate in reducing transmitter release from spinal cord synaptosomes
[], an action that was blocked by the group II antagonist LY341495 and the mGluR3 selective antagonist beta-NAAG []. Very high levels of NAAG and NAAG peptidase activity are expressed in spinal cord and spinal sensory neurons [, ] and peptidase inhibition moderated the effect of spinal cord trauma [].[](https://www.ncbi.nlm.nih.gov/mesh/C027172)
At the moment, there are not sufficient data to resolve the apparent conflict among the data on the failure of NAAG to activate mGluR3 in transfected cells and the studies presented here and elsewhere [, , , , , , ] in which the effects of NAAG and NAAG peptidase inhibition are blocked by group II mGluR antagonists and are absent in mGluR3 ko mice. One possibility is that the mechanism of expression or dimerization of mGluR3 following its transfection into non-neuronal or -glial cell lines differs from that in vivo, in spinal cord synaptosomes or in cultured astrocytes.[](https://www.ncbi.nlm.nih.gov/mesh/C027172)
## NAAG Peptidase Inhibition in Alzheimer’s Disease Model Mice
*In the **NAAG Peptidase Inhibition in Alzheimer’s Disease Model Mice** section:*
The APPSwe, PS1M146V, and tauP301L transgenic mouse line [] captures both the beta-amyloid and Tau neuropathology found in Alzheimer’s disease [] and thus represents one of the most widely studied animal models of this disorder. Age-dependent behavioral changes have been characterized in this mouse line including deficits in novel object recognition and attention [–]. The novel object recognition test also has been used to characterize other transgenic animal models of Alzheimer’s disease
[, ] and recognition memory for novel objects serves as a marker for clinical diagnosis of this disorder []. This test is particularly useful in distinguishing cognitive loss in normal aging versus loss in Alzheimer’s disease model mice inasmuch as different strains of mice retain short-term memory in the novel object recognition test well beyond 9 months of age even while other cognitive functions are declining [] In the present study, short-term novel object recognition was observed in the transgenic mice at 2 months but not at 5 and 9 months of age. Consistent with prior reports of procognitive effects of NAAG peptidase inhibition [, , ], 2-PMPA significantly improved performance on this task in the 9-month old triple mutant Alzheimer’s disease mice. While normal 9- month old wild type mice were not tested in this study, a substantial literature demonstrates mice of different strains, including non-transgenic colony mates of triple transgenic Alzheimer’s mice do not exhibit a decline in short-term novel object recognition as late as 22 months of age [, , , –].
Thus, while the behavioral test presented here does not speak to cognitive deficits associated with normal aging in mice, it will be interesting to determine if NAAG peptidase inhibitors are procognitive in other behavioral tests in which normal mice demonstrate aging-related cognitive deficits. In a similar study, we found that ZJ43 also reversed the cognitive deficit of aged triple transgenic mice in the novel object recognition test and the efficacy of ZJ43 was reversed by the group II mGluR antagonist (Olszewski and Neale, work in progress).[](https://www.ncbi.nlm.nih.gov/mesh/C402107)
## Broader Impact of NAAG Peptidase Inhibition in Cognition
*In the **Broader Impact of NAAG Peptidase Inhibition in Cognition** section:*
The observations that NAAG peptidase inhibition and deletion of GCPII are procognitive in control conditions, where the mice have not been cognitively challenged via a drug treatment ([] and Fig. 1), suggest that the procognitive actions of these peptidase inhibitors in animal models of clinical conditions, such as Alzheimer’s disease (Fig. 2), schizophrenia [, ] and ethanol intoxication (Fig. 3), might not be specifically reversing the neurochemical processes that underlie these clinical models but rather be generally procognitive. However, while the conditions that induce these models are clearly different, they can be linked by the common element of increases in glutamate release and NAAG peptidase inhibitors have been consistently shown to reduce synaptic release of glutamate [, , , , –]. In any case, the procognitive efficacy of NAAG peptidase inhibition in long-term novel object recognition, demonstrate that NAAG’s role on cognition is not limited to deficits induce by excess glutamate release.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
In human studies, treatment with growth hormone-releasing hormone increased NAAG levels in the prefrontal cortex and improved performance of subjects exhibiting mild cognitive impairment []. Similarly, NAAG levels in the hippocampus positively correlated with cognitive functioning in multiple sclerosis patients [].[](https://www.ncbi.nlm.nih.gov/mesh/C027172)
## Other Group II mGluR Ligands in Cognition Studies
*In the **Other Group II mGluR Ligands in Cognition Studies** section:*
The mGluR2/3 agonist LY379268 had a procognitive effect in the novel object recognition test [] while heterotropic group II mGluR agonists and an mGluR2 positive allosteric modulator have procognitive effects in other behavioral tests [, ]. Yet in other studies, agonists including the LY354740 impair rather than enhance attention and working memory [–]. Interpretation of these reports is complicated by the use of agonists and antagonists that interact with both mGluR2 and mGluR3 in vitro and in vivo [, , , ]. In studies using ko mice, some heterotrophic group II agonists have been shown to work via mGluR2 rather than mGluR3
[, , ]. This leads to the possibility that the contrasting effects of heterotrophic group II agonists on cognition may reflect differences in the behavioral tests that were used or their differential actions on mGluR2 and mGluR3 receptors. Evidence that the procognitive effects of NAAG peptidase inhibition could be specific to the type of memory being tested comes from the report that 100 mg/kg of 2-PMPA in mice does not affect long-term memory in the step through passive avoidance test or spatial working memory in the Y maze [].[](https://www.ncbi.nlm.nih.gov/mesh/C118218)
## NAAG Peptidase Inhibition and Ethanol Induced Cognitive Deficits in Short-Term Novel Object Recognition
*In the **NAAG Peptidase Inhibition and Ethanol Induced Cognitive Deficits in Short-Term Novel Object Recognition** section:*
NAAG peptidase inhibition consistently reversed the cognitive impairment induced by ethanol in short-term novel object recognition (Fig. 3 and4). We previously reported that NAAG peptidase inhibition alone had no effects on cognition in this test []. The data in Fig. 4 clearly demonstrate that the efficacy of these inhibitors in this short-term memory test require mGluR3. While the effect of 2-PMPA in the mGluR2 ko mice does not reflect a complete reversal of the action of ethanol, the recognition index for this group if mice is significantly different from the acquisition session.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## Ethanol, Glutamate and Group II mGluR
*In the **Ethanol, Glutamate and Group II mGluR** section:*
Glutamate appears to have a central role in drug addiction and alcoholism []. The efficacy of NAAG peptidase inhibition in reversing the effects of ethanol treatment (Figs. 3, 4, 5, 6) is consistent with the view that group II receptors, particularly mGluR3, are among the more promising targets for the development of drugs to treat alcohol addiction [–] Heterotropic group II glutamate receptor agonists reduce drug seeking, conditioned place preference and stress-induced reinstatement in animal models of alcohol addiction [, , ]. The agonist LY379268 also blocks the effects of alcohol on discriminative stimulus testing [], alcohol self-administration and reinstatement []. Additionally, epistatic effects of genetic variants of the mGluR3 and COMT genes have been associated with hippocampal volume in alcohol-dependent patients but not in controls [].
The activity of NAAG peptidase inhibition in the present study might again be related to the efficacy of the peptide in reducing release of glutamate inasmuch as ethanol induces glutamate release in the nucleus accumbens, hippocampus and posterior ventral tegmental area, brain regions known to mediate some of the clinical effects of alcohol consumption [–].[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
The stimulant effects of alcohol in adolescents [] are modeled in mice by open field motor activation induced immediately after administration []. In the present study using mice that previously had been habituated to the open field chamber, ethanol (2.1 g/kg) induced an increase in open field activity relative to control mice over the first 10 min minutes after injection (Fig. 5). Pretreatment of mice with NAAG peptidase inhibitors reduced this initial motor activation. Such a result might be taken to indicate that NAAG peptidase inhibitors have a sedative effect as they similarly block motor activation induced by PCP and d-amphetamine [–, ].
However, these inhibitors, given alone, do not affect motor activity in control mice habituated or unhabituated to open field conditions [].[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
Data from animal models suggest that mGluR3 agonists might be useful in treatment of ethanol addiction [–]. The data presented here demonstrate that drugs that elevate NAAG levels also moderate motor activation, cognitive and balance effects of ethanol intoxication and support the conclusion that mGluR3 mediates these effects. The breadth of actions of NAAG peptidase inhibitors in these studies suggests that they may be affecting a central mechanism underlying ethanol-induced intoxication.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
## Conclusion
*In the **Conclusion** section:*
NAAG peptidase inhibition improves cognition in the novel object recognition tests and reduces the cognitive and motor deficits induced by ethanol via a cellular mechanism that involves activation of mGluR3. The GCPII inhibitor 2-PMPA also reverses the cognitive deficit observed in an animal model of Alzheimer’s disease. These and other data support the conclusion that GCPII is a significant target for the development of procognitive drugs.[](https://www.ncbi.nlm.nih.gov/mesh/D000431)
# Abbreviations
*In the **Abbreviations** section:*
NAAG[](https://www.ncbi.nlm.nih.gov/mesh/C027172)
N-Acetylaspartylglutamate[](https://www.ncbi.nlm.nih.gov/mesh/C027172)
mGluR
Metabotropic glutamate receptor
mGluR2
Metabotropic glutamate receptor type 2
mGluR3
Metabotropic glutamate receptor type 3
ko
Knockout
GCPII
Glutamate carboxypeptidase II
# References
*In the **References** section:*
|
# Introduction
Metalloradical activation of [α-formyldiazoacetates](https://www.ncbi.nlm.nih.gov/mesh/D003979) for the catalytic asymmetric radical cyclopropanation of [alkenes](https://www.ncbi.nlm.nih.gov/mesh/D000475)†
# Abstract
*In the **Abstract** section:*
For the first time, α-formyldiazoacetates (FDA), have been successfully applied for asymmetric olefin cyclopropanation via Co(ii)-based meta[lloradical catalysis.](https://www.ncbi.nlm.nih.gov/mesh/D003979)[](https://www.ncbi.nlm.nih.gov/mesh/D003979)
For the first time, α-formyldiazoacetates have been successfully applied for the asymmetric cyclopropanation of alkenes via Co(ii)-based metalloradical catalysis. The cobalt(ii) complex of the D 2-symmetric chiral amidoporphyrin [Co(3,5-DitBu-ChenPhyrin)] is an effective metalloradical catalyst that c[an activate α-formyld](https://www.ncbi.nlm.nih.gov/mesh/D003979)iazoacetates to cyclopropanate both aromatic and aliphatic olefins with[ varied](https://www.ncbi.nlm.nih.gov/mesh/D000475) elec[tronic](https://www.ncbi.nlm.nih.gov/mesh/D003035) properties, affording the synthetica[lly useful](https://www.ncbi.nlm.nih.gov/mesh/D003035) 1,1-cyclopropaneformylesters in high[ yields with b](https://www.ncbi.nlm.nih.gov/mesh/D011166)ot[h high diastereo- and en](https://www.ncbi.nlm.nih.gov/mesh/D011166)antioselectivity.[](https://www.ncbi.nlm.nih.gov/mesh/D003979)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Radical reactions have been increasingly exploited as attractive tools in modern organic synthesis as they exhibit a number of unique features, including tolerance of functional groups. To address the existing challenges such as the control of enantioselectivity, metalloradical catalysis (MRC) has rendered a fundamentally new approach that enables the catalytic generation of metal-stabilized organic radicals as well as the selective control of their subsequent radical reactions., As stable metalloradicals, cobalt(ii) complexes of the D 2-symmetric chiral amidoporphyrins [Co(D 2-Por*)] have emerged as a class of effective catalysts for asymmetric olefin cyclopropanation through a distinct radical process involving the catalytic generation of α-metalloalkyl radicals as the key intermediates (Scheme 1: A). It has been suggested that the unusual capability of [Co(D 2-Por*)] in activating acceptor/acceptor-substituted diazo reagents as well as regulating the reactivity and selectivity of the radical processes is further enhanced by the postulated double hydrogen-bonding interactions between the amide N–H donors on the amidoporphyrin ligand and the two acceptors on the C-centered radical moiety (Scheme 1).,, Consi[derin](https://www.ncbi.nlm.nih.gov/mesh/D008670)g the demonstrated functional group tolerance of Co(ii)-based metalloradical catalysis (Co(ii)-MRC),,,– we were attracted to the [possibilit](https://www.ncbi.nlm.nih.gov/mesh/D003035)y of accessing a new type of α-metalloa[lkyl radical be](https://www.ncbi.nlm.nih.gov/mesh/D011166)ar[ing both α-f](https://www.ncbi.nlm.nih.gov/mesh/D011166)ormyl and α-alkoxycarbonyl functionalities from the metalloradic[al act](https://www.ncbi.nlm.nih.gov/mesh/D000475)ivation of α-formyldiazoacetates (FDA). Despite the fact that free α-formylalkyl radicals a[re scarce and prone to ](https://www.ncbi.nlm.nih.gov/mesh/D009942)H-atom abstraction because of the weak aldehydic C–H bonds, we reasoned that this type of α-met[alloalkyl ra](https://www.ncbi.nlm.nih.gov/mesh/D011166)dical might be accessible on the basis of the [combined effec](https://www.ncbi.nlm.nih.gov/mesh/D001391)ts of metal stabilization, double H-bonding interaction, and protection by the well-defined cavity of the ligand system (Sch[eme 1: A](https://www.ncbi.nlm.nih.gov/mesh/D006859)).
Assuming that the α-formyl-α-al[koxyc](https://www.ncbi.nlm.nih.gov/mesh/D000577)arbonyl-α-Co(iii)-a[lkyl radicals ](https://www.ncbi.nlm.nih.gov/mesh/D011166)(A) are capable of undergoing stereoselective radical addition with olefins, followed by the effective 3-exo-tet radical cyclization of [the co](https://www.ncbi.nlm.nih.gov/mesh/D003035)rresponding γ-Co(iii)-alkyl radic[als (B](https://www.ncbi.nlm.nih.gov/mesh/D003035)), we anticipated the potential development of a new catalytic process for[ the asymmetric synthe](https://www.ncbi.nlm.nih.gov/mesh/D009942)sis of optically active cyclopropanes bearing both aldehyde and ester functionalities, which would[ be valuable for ster](https://www.ncbi.nlm.nih.gov/mesh/D003979)eo[sel](https://www.ncbi.nlm.nih.gov/mesh/D003979)ective organic synthesis (Sche[me 1).](https://www.ncbi.nlm.nih.gov/mesh/D009930)cbi.nlm.nih.gov/mesh/D006859)
Working proposal for the radical cyclopropanation of alkenes with FDA via Co(ii)-MRC.[](https://www.ncbi.nlm.nih.gov/mesh/D000475)
The catalytic asymmetric cyclopropanation of alkenes with diazo reagents represents the most general approach for the stereoselective synthesis of optically active cyclopropanes. While a number of diazo reagents have been successfully employed, there is no previous report of a catalytic system that is effective for asymmetric olefin cyclopropanation with α-formyldiazoacetates (FDAs). The development of this catalytic process apparently confronts the formidable challenges associated with the inherent low reactivity of the acceptor/acceptor-substituted diazo reagents and the incompatibility of the aldehyde functionality with existing catalytic systems. Recently, Fokin and coworkers developed a Rh2-catalyzed system for a highly asymmetric cyclopropanation with N-sulfonyl-1,2,3-triazoles for the production of cyclopropyl imines, which could be subsequently transformed into the corresponding formyl cyclopropane derivatives. While this offers a valuable alternative for the preparation of optically active formyl cyclopropanes, the direct synthesis via asymmetric cyclopropanation with α-formyl diazo reagents is an appealing process that remains to be developed.
As a new application of Co(ii)-MRC, we herein wish to report the first catalytic system based on [Co(D 2-Por*)] that is highly effective in activating FDA for asymmetric cyclopropanation. This asymmetric radical process is generally applicable for a broad scope of alkenes, offering a direct method for the high-yielding synthesis of 1,1-cyclopropaneformylesters with excellent control of the diastereo- and enantioselectivity. The products can be readily transformed into other chiral 1,1-bifunctionalized cyclopropanes and chiral dihydrofurans.[](https://www.ncbi.nlm.nih.gov/mesh/D000475)
## Results and discussion
*In the **Results and discussion** section:*
Initial experiments were carried out with styrene as the model substrate to examine the suitability of FDA for the catalytic radical cyclopropanation by Co(ii)-MRC (Table 1). While [Rh2(OAc)4] was indeed incompatible (entry 1), [Co(TPP)] only produced trace amounts of the corresponding cyclopropane from ethyl α-formyldiazoacetate (EFDA) (entry 2). Remarkably, when the Co(ii) complex of the D 2h-symmetric achiral amidoporphyrin [Co(P1)] was used as the catalyst, the reaction proceeded successfully to form the desired (E)-1,1-cyclopropaneformylester in a 46% yield (entry 3). The dramatic difference in the catalytic activity between [Co(TPP)] and [Co(P1)] is in alignment with the hypothesized role of the double H-bonding interaction in activating EFDA and stabilizing the resulting intermediate A (Scheme 1).
By switching to [Co(P2)], the reactivity was further enhanced with the observation of a significant level of enantioselectivity (entry 4). Of the solvents examined, toluene was proven to be the medium of choice (entries 4–8). Lowering the reaction temperature further increased the enantioselectivity, but decreased the yield (entries 8–10). The diastereoselectivity was greatly improved when the bulkier tert-butyl α-formyldiazoacetate (t-BFDA) was used, affording cyclopropane 1a in a 78% yield with 95 : 5 dr and 96% ee (entry 11). The product yield could be further improved to 84% by increasing the catalyst loading to 5 mol% while maintaining the high level of diastereo- and enantioselectivity (entry 12).[](https://www.ncbi.nlm.nih.gov/mesh/D020058)
The catalytic asymmetric cyclopropanation of styrene with FDA a[](https://www.ncbi.nlm.nih.gov/mesh/D020058)
aCarried out in one-portion under N2 with [olefin] = 0.20 M.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
bIsolated yields.
cee of major (E)-diastereomer determined by chiral HPLC.
dDetermined by 1H-NMR.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
eWith 5 mol% of catalyst for 20 h.
Under the optimized conditions, the scope of this Co(ii)-based asymmetric radical cyclopropanation was investigated (Table 2). Like styrene, its derivatives bearing substituents with varied electronic and steric properties could be cyclopropanated by [Co(P2)] with t-BFDA. For example, p- and m-alkyl styrenes were cyclopropanated to formylcyclopropanes 1b–1d in high yields with excellent diastereo- and enantioselectivity (entries 1–3). Halogenated (entries 4–6) and electron-deficient (entries 7 and 8) styrene derivatives could also undergo high-yielding cyclopropanation, producing 1e–1i with high stereoselectivities. The configurations of the two contiguous chiral centers in 1h were established as [1R,2S] by X-ray crystal structural analysis (see ESI†).
The cyclopropanation was also suitable for other aromatic olefins as exemplified with 2-naphthalene for near quantitative formation of cyclopropane 1j (entry 9). In addition, 1,1-disubstituted olefins such as α-methylstyrene could also be effectively employed, affording (E)-formylcyclopropane 1k in a 93% yield with remarkable control of both the diastereo- and enantioselectivity of the two newly-generated contiguous all-carbon quaternary stereogenic centers (entry 10). To demonstrate the functional group tolerance of the Co(ii)-based radical cyclopropanation, m-formylstyrene could be effectively cyclopropanated to cyclopropane 1l in a high yield with high diastereo- and enantioselectivity (entry 11). Notably, the two unprotected formyl groups were well tolerated by the metalloradical system. It is also worth mentioning that the Co(ii)-catalyzed cyclopropanation process could be scaled up ten-fold as demonstrated with the high-yielding synthesis of the cyclopropane 1d on a 1.0 mmol scale without affecting the excellent stereoselectivity (entry 3).[](https://www.ncbi.nlm.nih.gov/mesh/D003035)
The asymmetric cyclopropanation of alkenes with t-BFDA by [Co(P2)] a , b , c[](https://www.ncbi.nlm.nih.gov/mesh/D000475)
aCarried out in one-portion under N2 with [olefin] = 0.20 M.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
bIsolated yields.
cee of major (E)-diastereomer determined by chiral HPLC.
dee determined upon derivatization.
eResults in the parentheses were obtained for the reaction performed on a 1.0 mmol scale.
fAbsolute configuration determined by X-ray diffraction as [1R,2S].
g5 equiv. olefin.[](https://www.ncbi.nlm.nih.gov/mesh/D000475)
hee determined by chiral GC.
iNeat condition.
The Co(ii)-based radical cyclopropanation was further highlighted for its exceptional reactivity toward electron-deficient olefins, which are typically problematic substrates for catalytic systems involving electrophilic metallocarbene intermediates. For example, [Co(P2)] could catalyze the CC cyclopropanation of acrylonitrile with t-BFDA to form 1,1,2-cyclopropaneformylesternitrile 1m in a high yield with good enantioselectivity (entry 12), leaving the cyano group untouched. In marked contrast, when treated with Rh2-based catalyst, acrylonitrile was previously shown to react with the CN bond of FDA to form oxazoles. Other electron-deficient olefins such as ethyl and methyl acrylates could also be cyclopropanated to form the 1,1,2-cyclopropaneformyldiesters 1n and 1o in good yields with 96% ee and 97% ee, respectively, although with diminished control of diastereoselectivity (entries 13 and 14). The presence of three electron-withdrawing groups in the cyclopropanes 1m–1o renders them highly electrophilic, making them valuable intermediates for synthetic applications.
Furthermore, aliphatic olefins, another class of challenging substrate for asymmetric cyclopropanation, could also be cyclopropanated by [Co(P2)] as exemplified by the high-yielding reaction of 1-octene under neat condition, forming 1p with high stereoselectivity (entry 15).[](https://www.ncbi.nlm.nih.gov/mesh/D003035)
As an initial exploration of applications, the formyl unit of the resulting chiral 1,1-cyclopropaneformylesters could be readily converted into other functional groups, forming various cyclopropane derivatives while retaining high enantiopurity. For example, the formyl group in (E)-1a could be transformed into a trans-vinyl unit via the Horner–Wadsworth–Emmons reaction, affording (E)-1,1-cyclopropanevinylester 2 in a 78% yield with full retention of both the relative and absolute configurations (eqn (1)). When treated with Bestmann–Ohira reagent, the formyl group in (E)-1a could be smoothly converted to a terminal alkyne functionality, resulting in chiral (E)-1,1-cyclopropaneethynylester 3 in a 70% yield without any diminishment of the original stereochemistry (eqn (2)). This transformation provides an alternative way to direct asymmetric cyclopropanation with α-ethynyldiazoacetates for chiral 1,1-cyclopropaneethynylesters. It is noted that α-ethynyldiazoacetates containing terminal alkyne units seem synthetically inaccessible.
While the [Co(P2)]-catalyzed cyclopropanation with FDA generally forms (E)-cyclopropanes, the (Z)-diastereoisomers could be conveniently accessed through the stereospecific epimerization previously reported. As demonstrated with (E)-1a, treatment with 5 equivalents of NaI at room temperature resulted in the formation of (Z)-1a as the major diastereomer with only partial loss of the original optical purity (eqn (3)). Interestingly, when (E)-1g was treated with 10 equivalents of NaI at an elevated temperature, a ring-expansion involving the formyl group occurred instead, affording 2,3-dihydrofuran 4 in a 74% yield (eqn (4)). In the absence of any external chiral induction, the enantiopurity appeared to be largely retained during the rearrangement.[](https://www.ncbi.nlm.nih.gov/mesh/D003521)
## Conclusions
*In the **Conclusions** section:*
In summary, we have demonstrated that the metalloradical catalyst [Co(P2)] can effectively activate α-formyldiazoacetates (FDAs) for a highly asymmetric olefin cyclopropanation, without affecting the otherwise reactive aldehyde functionality. This represents the first application of α-formyldiazo reagents for metal-catalyzed asymmetric cyclopropanation. The Co(ii)-based radical cyclopropanation with FDA can be successfully applied to a broad scope of olefin substrates, permitting the direct synthesis of chiral 1,1-cyclopropaneformylesters in high yields with high diastereo- and enantioselectivity. Given that the resulting enantioenriched cyclopropanes contain two contiguous chiral centers in the ring structure, including one all-carbon quaternary stereogenic center bearing both aldehyde and ester functionalities, this new Co(ii)-based asymmetric radical cyclopropanation process should find wide applications in stereoselective organic synthesis.[](https://www.ncbi.nlm.nih.gov/mesh/D011166)
|
# Introduction
CFTR-NHERF2-LPA2 Complex in the Airway and Gut Epithelia
# Abstract
*In the **Abstract** section:*
The cystic fibrosis transmembrane conductance regulator (CFTR) is a cAMP- and cGMP-regulated chloride (Cl−) and bicarbonate ([HCO3](https://www.ncbi.nlm.nih.gov/mesh/D000242)−) cha[nnel](https://www.ncbi.nlm.nih.gov/mesh/D006152) localized [primaril](https://www.ncbi.nlm.nih.gov/mesh/D002712)y [at ](https://www.ncbi.nlm.nih.gov/mesh/D002712)the ap[ical plasma](https://www.ncbi.nlm.nih.gov/mesh/D001639) m[embra](https://www.ncbi.nlm.nih.gov/mesh/D001639)ne of epithelial cells lining the airway, gut and exocrine glands, where it is responsible for transepithelial salt and water transport. Several human diseases are asso[ciat](https://www.ncbi.nlm.nih.gov/mesh/D012492)ed wi[th al](https://www.ncbi.nlm.nih.gov/mesh/D014867)tered CFTR channel function. Cystic fibrosis (CF) is caused by the absence or dysfunction of CFTR channel activity, resulting from mutations in the gene. Secretory diarrhea is caused by the hyperactivation of CFTR channel activity in the gastrointestinal tract. CFTR is a validated target for drug development to treat CF, and extensive research has been conducted to develop CFTR inhibitors for therapeutic interventions of secretory diarrhea. The intracellular processing, trafficking, apical membrane localization, and channel function of CFTR are regulated by dynamic protein–protein interactions in a complex network. In this paper, we review the current knowledge of a macromolecular complex of CFTR, Na+/H+ exchanger regulatory factor 2 (NHERF2), and lysophosphatidic acids (LPA) receptor 2 (LPA2) at the apical plasma membrane of airway and gut epithelial cells, and discuss its relevance in human physiology and diseases. We also explore the possibilities of targeting this complex to fine tune CFTR channel activity, with a hope to open up new avenues to develop novel therapies for CF and secretory diarrhea.
## 1. Introduction
*In the **1. Introduction** section:*
Protein–protein interactions regulate virtually all cellular processes by promoting the proper cellular localization of regulatory partners and by facilitating the signaling through pathways to achieve exquisite spatiotemporal control. The formation of multiple protein complexes at discrete subcellular microdomains increases the specificity and efficiency of cell signaling [,,].
CFTR is a cAMP- and cGMP-regulated chloride (Cl−) and bicarbonate (HCO3−) channel localized primarily at the apical surfaces of epithelial cells lining the airways, gut, and exocrine glands, where it is responsible for the transepithelial salt and water transport [,,]. CFTR is a member of the ATP-binding cassette (ABC) transporter superfamily and consists of 1480 amino acids. CFTR is composed of two repeated motifs; each consisting of a six-helix membrane-spanning domain (MSD) and a cytosolic nucleotide binding domain (NBD). These two motifs are linked by a cytoplasmic regulatory (R) domain, which contains multiple consensus phosphorylation sites (Figure 1)
[,,].[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
The R domain is a unique feature of CFTR within the ABC superfamily. Both the amino and carboxyl termini of CFTR mediate its interactions with a wide variety of binding partners []. CFTR channel is activated by phosphorylation of its R domain and binding and hydrolysis of ATP at NBDs. CFTR channel activity is determined by the quantity of channels at the plasma membrane, the open probability of these channels, and their single channel conductance.
Mutations in the CFTR gene alter one or more of these parameters, causing the impairment or loss of the channel activity. More than 2000 mutations have been identified in the CFTR gene [], which are traditionally grouped into six classes based on the nature of the defect(s) [,]. Class I mutations have defects in biosynthesis, resulting in low levels of truncated and/or dysfunctional CFTR proteins.
Class II mutations have defects in folding or maturation, causing no to very little CFTR protein to reach the cell surface. Class III mutations encode CFTR proteins that have defects in channel gating, and Class IV mutations encode proteins that have reduced capacity to transport Cl−. Class V mutations have reduced mRNA stability. Class VI mutations encode CFTR proteins with decreased stability and increased turnover at the cell surface [,]. Because some mutations have multiple defects, an expanded classification method was also proposed [].
One such mutation is Phe508del (deletion of a phenylalanine residue at position 508 on CFTR protein), which is the most prevalent CFTR mutation with approximately 90% of CF patients carrying it on at least one allele. Phe508del is a class II mutation. However, upon reaching the cell surface following rescue procedures, it displays characteristics of class III and VI mutations [].[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
The intracellular processing, trafficking, apical plasma membrane localization and channel function of CFTR are regulated by dynamic protein–protein interactions in a complex network (CFTR interactome). A wide variety of CFTR-interacting partners have been identified, including receptors, scaffolding proteins, channels, transporters, etc. [,]. Several CFTR-containing macromolecular complexes at the apical plasma membrane of epithelial cells have been characterized; examples include (i) the complex of β2-adrenergic receptor (β2-AR), Na+/H+ exchanger regulatory factor 1, and CFTR at the apical surfaces of airway epithelial cells, which couples β2-AR signaling to CFTR channel function [], (ii) the complex of multidrug resistance protein 4 (MRP4), PDZ-containing kidney protein 1, and CFTR at the apical surfaces of intestinal epithelial cells, which couples the cAMP transporter activity of MRP4 to CFTR channel function
[], and (iii) the complex of LPA2, NHERF2, and CFTR at the apical surfaces of airway and intestinal epithelial cells, which couples the LPA2-mediated signaling to CFTR channel function []. In this article, we review the current knowledge of CFTR-NHERF2-LPA2 complex at the apical plasma membrane of airway and gut epithelial cells and its relevance in human physiology and diseases. We also explore the possibilities, and provide our perspectives, on how to target this complex to fine tune CFTR channel activity, with a hope to open up new avenues to develop novel therapeutics for CFTR-associated diseases.[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
## 2. CFTR-NHERF2-LPA2 Complex in Airway and Gut Epithelial Cells
*In the **2. CFTR-NHERF2-LPA2 Complex in Airway and Gut Epithelial Cells** section:*
## 2.1. Characterization of CFTR-NHERF2-LPA2 Complex
*In the **2.1. Characterization of CFTR-NHERF2-LPA2 Complex** section:*
NHERF2 is a postsynaptic density-95, discs large, zona occludens-1 (PDZ) domain-containing protein and primarily localizes at the apical plasma membrane of epithelial cells. NHERF2 has 337 amino acids and contains two PDZ domains and an ezrin/radixin/moesin (ERM) domain at the C-terminus. The ERM domain mediates the interaction of NHERF2 with merlin/ERM proteins and links NHERF2 to the actin cytoskeleton []. NHERF2 has been shown to cluster signaling molecules into macromolecular complexes [,,,,].[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
LPA are growth-factor-like phospholipids present in every tissue and most biological fluids at nanomolar to micromolar concentrations. LPA mediate diverse cellular responses, such as proliferation, migration, survival, angiogenesis, inflammation and much more []. At least six G-protein-coupled LPA receptors have been identified, which couple to Gs, Gi, Gq and/or G12/13 to activate various signaling pathways [,]. The LPA receptor LPA2 contains 351 amino acids and belongs to the endothelial differentiation gene family.
LPA2 is structurally unique at the C-terminus, in which it contains a dileucine motif and several putative palmitoylated cysteine residues in the proximal region that are responsible for binding to several zinc-finger proteins [,]. The last four amino acids of LPA2, Asp-Ser-Thr-Leu (DSTL), form a class I PDZ domain-binding motif and mediate the interaction of LPA2 with several PDZ proteins, including NHERF2 []. Through the interaction with LPA2, NHERF2 regulates the LPA-mediated phospholipase C-β3 (PLC-β3) signaling pathway and the activation of extracellular signal-regulated kinases and Akt [,].
LPA has been reported to induce the formation of a ternary complex containing LPA2, thyroid-hormone-receptor-interacting protein 6, and NHERF2 at microdomains on the plasma membrane, which regulates the anti-apoptotic signaling of LPA2 [].[](https://www.ncbi.nlm.nih.gov/mesh/D010743)
Li et al. [] demonstrated for the first time that (i) wild-type (WT)-CFTR forms a macromolecular complex with NHERF2 and LPA2 at the apical plasma membrane of intestinal epithelial cells (HT29-CL19A) and airway epithelial cells (Calu-3); (ii) LPA inhibits the CFTR Cl− channel function (using a CFTR-mediated iodide efflux assay) through an LPA2-mediated Gi pathway; (iii) LPA inhibits the CFTR-dependent short-circuit currents (Isc) in polarized epithelial cells in a compartmentalized manner; (iv) LPA inhibits the CFTR-dependent Isc in mouse intestinal epithelia tissues; (v) Administration of LPA substantially reduced the cholera toxin (CTX)-induced and CFTR-mediated intestinal fluid secretion in mice; and (vi) disruption of this complex using an LPA2-specific peptide, which contains the last 11 amino acids at the C-terminus of LPA2 and serves as a disruptor of LPA2-NHERF2 interaction in cells, reversed the LPA2-mediated inhibition of CFTR channel function in cells []. This study not only discovered the molecular mechanism underlying the coupling of LPA2-mediated inhibitory signaling to CFTR Cl− channel function at the apical plasma membrane of airway and intestinal epithelial cells (Figure 2), but demonstrated that LPA substantially reduced the CTX-induced intestinal fluid secretion in vivo, which had clinical implications for treating human diseases associated with the hyperactivation of CFTR channel function (e.g., secretory diarrhea).[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
In a study to investigate the roles of NHERF1/2/3 in regulating the CFTR-dependent duodenal HCO3− secretion in mice, Singh et al. found that the forskolin (FSK)-stimulated HCO3− secretion was significantly increased in Nherf2−/− mice and that NHERF2 is required for the LPA2-mediated inhibition of HCO3− secretion []. The data supported the findings of Li and colleagues’ [] and together suggested that by attenuating this LPA2-mediated inhibitory signaling, CFTR channel function could be potentially augmented.[](https://www.ncbi.nlm.nih.gov/mesh/D001639)
Phe508del-CFTR protein has multiple defects, including a folding defect which results in the protein to get trapped in the endoplasmic reticulum and targeted for degradation, with little to no protein trafficked to the cell surface. Although the mutant protein can be partially rescued by exposure to low temperature and/or using CFTR correctors, it is unstable at the cell surface and exhibits impaired channel activity [,,]. Recently, we found that when rescued to the plasma membrane, Phe508del-CFTR also complexes with NHERF2 and LPA2 in CF bronchial epithelial cells (CFBEo−-Phe508del-CFTR cells) and in intestinal enterospheres developed from Phe508del−/− mice []. By formation of such a complex, the LPA2-mediated signaling could exert its inhibitory effect on the rescued Phe508del-CFTR at the cell surface (Figure 2).[](https://www.ncbi.nlm.nih.gov/mesh/D010649)
## 2.2. The Involvment of CFTR in Two Major Human Diseases: Cystic Fibrosis (CF) and Secretory Diarrhea
*In the **2.2. The Involvment of CFTR in Two Major Human Diseases: Cystic Fibrosis (CF) and Secretory Diarrhea** section:*
Several human diseases are associated with altered channel function of CFTR, including CF and secretory diarrhea [,].
CF is a life-shortening autosomal recessive inherited disease caused by the absence or dysfunction of CFTR channel activity, resulting from mutations in the CFTR gene [,]. There are approximately 70,000 CF patients worldwide []. Clinically, CF affects multiple organs, including the lungs (chronic lung disease causes most of the CF-associated morbidity and mortality), upper airway (e.g., sinusitis), pancreas (e.g., pancreatic insufficiency, CF-related diabetes mellitus), sweat glands (elevated sweat chloride level), intestines (e.g., meconium ileus, constipation, distal intestinal obstruction syndrome), liver (e.g., cholestasis, cirrhosis), and vas deferens (male infertility) [].
In the CF lungs, the loss of CFTR function causes a cascade of pathological events: the depletion of airway surface liquid (ASL), mucus plugging the airways, failure of mucociliary clearance, chronic bacterial infections, and excessive and ineffective inflammation (which fails to eradicate pulmonary pathogens). This may cause bronchiectasis and progressive airway destruction, eventually leading to the loss of pulmonary function [,]. Other genetic and environmental factors (e.g., modifier genes, socioeconomic status) also strongly influence the severity of the disease []. LPA levels have been found elevated in the bronchoalveolar lavage (BAL) fluid of subjects with CF
[], suggesting that LPA and its receptors may play a role in the pathogenesis of CF.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
Secretory diarrhea involves the hyperactivation of CFTR channel in the gastrointestinal tract []. When the gut lumen is exposed to certain types of stimuli (e.g., toxins secreted by the colonizing pathogenic microorganisms Escherichia coli, Vibrio cholera []), the intracellular second messengers cAMP and/or cGMP are excessively produced, causing the hyperactivation of CFTR channel. This hyperactivation increases the electrical and osmotic driving forces for the parallel flows of Na+ and water and inhibits the fluid absorption processes mediated by Na+/H+ exchangers (e.g., NHE3) and epithelial sodium channel. The net result is the excessive fluid secretion into the intestine lumen, which overwhelms the reabsorbing capacity of the colon and leads to fluid loss and dehydration [,].
Because CFTR is the primary chloride channel at the apical membrane of intestinal epithelial cells and plays a critical role in intestinal fluid secretion and homeostasis, extensive research has been conducted to develop CFTR inhibitors (channel blockers) as potential anti-diarrheal agents, including those derived from natural products (e.g., crofelemer, tannic acids, steviodides, etc.) and small molecules identified by using high throughput screening (e.g., CFTRinh-172, GlyH-101, PPQ-102, BPO-27, iOWH032 etc.) [,,].[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
## 3. Strategies to Target CFTR-NHERF2-LPA2 Complex for Possible Therapeutic Interventions of CF and Secretory Diarrhea
*In the **3. Strategies to Target CFTR-NHERF2-LPA2 Complex for Possible Therapeutic Interventions of CF and Secretory Diarrhea** section:*
## 3.1. Strategies to Target CFTR-NHERF2-LPA2 Complex for Possible Therapeutic Interventions of CF
*In the **3.1. Strategies to Target CFTR-NHERF2-LPA2 Complex for Possible Therapeutic Interventions of CF** section:*
Traditional CF therapies target the downstream disease consequences/symptoms, including improving the mucociliary clearance (e.g., restoration of ASL, mucus alteration), anti-inflammatory, anti-infective, improving nutrition, and lung transplantation []. Recently, the development of CFTR modulators to restore CFTR function has fundamentally changed CF disease management. Two types of these modulators are potentiators and correctors. CFTR potentiators are molecules that increase the channel gating of CFTR, while CFTR correctors are molecules that rescue the folding and/or trafficking of CFTR to increase the cell surface density.
Kalydeco® (ivacaftor, a CFTR potentiator) was approved by U.S. Food and Drug Administration to treat CF patients with G551D and other nine class III and IV mutations (G178R, S549N, S549R, G551S, G1244E, S1251N, S1255P, G1349D and R117H) []. Orkambi® (a combination of ivacaftor and lumacaftor; a CFTR corrector) was approved to treat CF patients age 6 years and older with two copies of Phe508del [].[](https://www.ncbi.nlm.nih.gov/mesh/C545203)
Based on the findings that (i) LPA levels are elevated in BAL fluid of subjects with CF [], (ii) CFTR-NHERF2-LPA2 complex at the apical plasma membrane of airway epithelial cells functionally couples LPA2-mediated signaling to CFTR channel function, and (iii) airway epithelial cells play a central role in regulating mucociliary clearance and modulating the innate and adaptive immune responses after infection [], we explored the possibility of targeting CFTR-NHERF2-LPA2 complex to tackle two major pathologies associated with CF: loss or dysfunction of CFTR channel function, and excessive inflammation.[](https://www.ncbi.nlm.nih.gov/mesh/C032881)
## 3.1.1. Disruption of NHERF2-LPA2 Interaction to Potentiate CFTR Channel Function
*In the **3.1.1. Disruption of NHERF2-LPA2 Interaction to Potentiate CFTR Channel Function** section:*
Because disruption of NHERF2-LPA2 interaction using an LPA2-specific peptide reversed the LPA2-mediated inhibition of CFTR channel function [], one straightforward approach was to develop small-molecule mimics of this LPA2 peptide, with more favorable drug-like properties such as improved potency and specificity, bioavailability, and being metabolically stable. As a proof-of-concept study, we developed an amplified luminescent proximity homogeneous assay to screen for compounds that specifically disrupt the LPA2–NHERF2 interaction []. From a library of 80 available inhibitors (disruptors) of PDZ domain-mediated protein–protein interactions, one hit compound (named CO-068) was found to have the highest potency (IC50 = 63 μM) and specificity []. Used at 50 μM concentration, CO-068 was found to (i) specifically disrupt the LPA2–NHERF2 interaction in cells, without affecting CFTR-NHERF2 or NHERF2-PLC-β3 interactions; (ii) augment the basal CFTR-mediated Isc in polarized Calu-3 cells; (iii) augment the FSK-induced and CFTR-mediated Isc in polarized Calu-3 cells; (iv) increase both the basal and FSK-stimulated CFTR-dependent submucosal glands fluid secretion in an ex vivo pig model, and (v) elevate the compartmentalized cAMP levels in cells [].
The study was the first to demonstrate that specific disruption of NHERF2-LPA2 interaction potentiates the WT-CFTR channel function under the basal and stimulated conditions (Figure 3a).[](https://www.ncbi.nlm.nih.gov/mesh/C561063)
To explore the possibility of using this approach for CF therapy, we tested the effect of CO-068 on the channel function of the rescued Phe508del-CFTR in two assays. We found that CO-068 potentiated the Isc of the rescued Phe508del-CFTR in polarized CFBEo−-Phe508del-CFTR cells and induced the swelling of intestinal enterospheres developed from WT mice and from Phe508del−/− mice []. The data suggests that, when in combination with a CFTR corrector (e.g., lumacaftor), specific disruption of NHERF2-LPA2 interaction can potentiate the channel function of Phe508del-CFTR.[](https://www.ncbi.nlm.nih.gov/mesh/C561063)
Regarding this disruptor approach (Figure 3a), it should be noted that (i) CO-068 has a IC50 value of 63 μM, which could account for the moderate potentiating effect on CFTR channel function we observed, and also highlight the need for chemical optimization to discover more potent and specific lead compounds; (ii) because this approach abolishes an LPA2-mediated inhibitory signaling on CFTR, in addition to WT- and Ph508del-CFTR, it could augment the channel activities of other CFTR mutants that complex with LPA2, therefore representing a mutation non-specific approach; and (iii) more work is needed to investigate the relevance of this approach in CF therapy, e.g., testing its efficacy in primary human CF airway epithelial cells and tissues, toxicological studies, etc.[](https://www.ncbi.nlm.nih.gov/mesh/C561063)
Another disruptor approach has also been explored to increase the channel activity of Phe508del-CFTR. This approach was designed to disrupt the interaction between CFTR and the CFTR-associated ligand (CAL). CAL is a PDZ domain-containing protein and a negative regulator of Phe508del-CFTR surface abundance []. Cushing and colleagues [] developed a decametric peptide iCAL36 that selectively binds the PDZ domain of CAL. The authors showed that iCAL36 enhanced the functional stability of Phe508del-CFTR and had complementary rescue effect with CFTR corrector corr-4a
[]. This group and their collaborators then used computational approach to study the binding of iCAL36 and derivatives to CAL to develop more specific and efficient inhibitors [,]. More recently, they developed a cell-permeable peptidyl inhibitor and demonstrated that this inhibitor disrupted the CFTR/CAL-PDZ interaction and increased the channel activity of Phe508del-CFTR in combination with CFTR correctors [].[](https://www.ncbi.nlm.nih.gov/mesh/D010649)
Taken together, these disruptor studies provide proof-of-principle data for targeting PDZ domain-mediated protein-protein interactions within the CFTR-containing macromolecular complexes for the possible therapeutic intervention of CF.
## 3.1.2. Targeting CFTR-NHERF2-LPA2 Complex to Suppress the Release of Interleukin 8 (IL-8)
*In the **3.1.2. Targeting CFTR-NHERF2-LPA2 Complex to Suppress the Release of Interleukin 8 (IL-8)** section:*
Given the critical role of IL-8 and the IL-8 receptor signaling pathway in the pathogenesis of CF lung inflammation [], we performed a proof-of-concept study to test whether CFTR-NHERF2-LPA2 complex regulates the IL-8 secretion from airway epithelial cells (CFBEo−-Phe508del-CFTR cells), thereby playing a role in the pathological cascade of excessive inflammation which fails to eradicate pulmonary pathogens in CF.[](https://www.ncbi.nlm.nih.gov/mesh/D010649)
We found that (i) LPA2 expression level was elevated in CFBEo−-Phe508del-CFTR cells compared with that in CFBEo−-WT-CFTR cells; (ii) CFBEo−-Phe508del-CFTR cells secreted IL-8 under basal (un-stimulated) conditions and responded to IL-1β stimulation to secret large amount of IL-8; (iii) Rescue of Phe508del-CFTR by using lumacaftor slightly decreased the basal level of secreted IL-8 and significantly decreased the IL-1β-stimulated IL-8 secretion (11%); (iv) Low temperature (28 °C) rescue of Phe508del-CFTR dramatically decreased IL-8 secretion from CFBEo−-Phe508del-CFTR cells; and (v) Antagonism of LPA2 using a specific inhibitor decreased the basal level (44%) and IL-1β-stimulated IL-8 (~39%) secretion [].[](https://www.ncbi.nlm.nih.gov/mesh/D010649)
The data suggests that in addition to its role in regulating airway fluid homeostasis, CFTR-NHERF2-LPA2 complex may also play a role in modulating IL-8 secretion from airway epithelial cells (Figure 3b). Rescue of Phe508del-CFTR and antagonism of LPA2 attenuate the IL-8 release from CFBEo−-Phe508del-CFTR cells, which could suppress the initiation of inflammatory response in CF and therefore inhibit the excessive infiltration of neutrophils and the subsequent inflammation. Of course, this hypothesis needs to be tested in other CF cell lines, especially in primary human CF airway epithelial cells, and thereafter in animal studies. In addition, studies are needed to test whether rescue of Phe508del-CFTR and antagonism of LPA2 affect the release of other chemokines/cytokines from airway epithelial cells.[](https://www.ncbi.nlm.nih.gov/mesh/D010649)
Theoretically, antagonism of LPA2 will suppress the LPA2-mediated Gi signaling on AC and increase the compartmentalized cAMP level in proximity to CFTR, and therefore potentiate CFTR channel function. Currently, we are testing if these antagonists potentiate CFTR channel activity in a variety of airway epithelial cell lines.[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
## 3.2. Targeting CFTR–NHERF2–LPA2 Complexes in the Gut Epithelia for Possible Therapeutic Interventions of Secretory Diarrhea
*In the **3.2. Targeting CFTR–NHERF2–LPA2 Complexes in the Gut Epithelia for Possible Therapeutic Interventions of Secretory Diarrhea** section:*
In addition to directly blocking CFTR channel, CFTR-NHERF2-LPA2 complex could also be targeted to tackle the abnormal fluid homeostasis in secretory diarrhea. One potential approach is to use LPA, because LPA not only inhibits the CFTR-dependent intestinal fluid secretion through LPA2-mediated Gi pathway [], but stimulates the intestinal Na+ and fluid absorption by activating NHE3 through an LPA5-mediated signaling cascade []. These two effects are beneficial to secretory diarrhea therapy. In this regard, LPA-rich food (e.g., hen egg yolk and white, and LPA-enriched soy lipid extract) could be an inexpensive alternative to mitigate secretory diarrhea.[](https://www.ncbi.nlm.nih.gov/mesh/C032881)
Similarly, it can be envisioned that specific and potent LPA2 agonists would inhibit the CFTR channel function and have the potential to inhibit intestinal fluid secretion (Figure 4). As a proof-of-concept study, we tested the efficacy of a specific LPA2 agonist, GRI977143 [], on the CFTR Cl− channel function. We found that (i) GRI977143 significantly inhibited the FSK-induced and CFTR-mediated Isc in polarized human gut epithelial cells (HT29-CL19A cells); (ii) GRI977143 significantly inhibited the CTX-induced and CFTR-dependent intestinal fluid secretion in a closed-loop fluid secretion mouse model; and (iii) GRI977143 inhibited both the basal and FSK-induced cAMP levels at the plasma membrane of HT29-CL19A cells []. Currently, we are testing whether GRI977143 affects the IL-8 and other chemokines/cytokines release from intestinal epithelial cells and whether GRI977143 affects the integrity of tight junction in these cell lines.[](https://www.ncbi.nlm.nih.gov/mesh/D002712)
## 4. Conclusions
*In the **4. Conclusions** section:*
CF and secretory diarrhea are two major human diseases associated with dysregulated CFTR channel activities. Since the discovery of the CFTR gene in 1989, significant progress has been made in understanding the CF pathogenesis and in development of effective CF therapies. The approval of Kalydeco® and Orkambi® has proven that CFTR is a validated drug target.
Because the channel function of CFTR is regulated by dynamic protein–protein interactions in a complex network, identification of these CFTR-containing macromolecular complexes can further our understanding of CFTR biology in health and in disease, and open up potential avenues for drug development to treat CFTR-associated diseases. Accumulating evidence suggests that CFTR complexes with NHERF2 and LPA2 at the apical plasma membrane of airway and gut epithelial cells, which couples the LPA2-mediated signaling to CFTR channel function. Several approaches could be explored to target this complex to fine tune CFTR channel activity, including using specific disruptors of LPA2–NHERF2 interaction and specific LPA2 modulators. With the technological advancements in drug discovery, we believe that more potent and specific small molecules will be identified and tested, which will provide us more tools to gain insights into the pathophysiology of CFTR and to combat CFTR-associated diseases.
# Author Contributions
*In the **Author Contributions** section:*
Weiqiang Zhang, Anjaparavanda P. Naren, Zhihong Zhang, and Yanhui H. Zhang wrote the paper.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# Abbreviations
*In the **Abbreviations** section:*
ABC ATP-binding cassette AC Adenylyl cyclase β2-AR β2-Adrenergic receptor ASL Airway surface liquid BAL Bronchoalveolar lavage CAL CFTR-associated ligand CF Cystic fibrosis CFBEo−- CF bronchial epithelial cells CFTR CF transmembrane conductance regulator CTX Cholera toxin ERM Ezrin/radixin/moesin FSK Forskolin IL-8 Interleukin 8 Isc Short-circuit currents LPA Lysophosphatidic acids LPA2 LPA receptor 2 MRP4 Multidrug resistance protein 4 MSD Membrane-spanning domain NBD Nucleotide binding domain NHE Na+/H+ exchanger NHERF2 Na+/H+ exchanger regulatory factor 2 PDZ Postsynaptic density-95, discs large, zona occludens-1 PLC-β3 Phospholipase C-β3 R domain Regulatory domain WT Wild type[](https://www.ncbi.nlm.nih.gov/mesh/D005576)
# References
*In the **References** section:*
The putative domain structure of the cystic fibrosis transmembrane conductance regulator (CFTR) protein. CFTR is composed of two membrane-spanning domains (MSD1 and MSD2), two nucleotide binding domains (NBD1 and NBD2), and a regulatory domain (R). NH2: amino terminal tail; COOH: carboxyl terminal tail.[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
Once residing at the plasma membrane, the cystic fibrosis transmembrane conductance regulator (CFTR) forms a complex with Na+/H+ exchanger regulatory factor 2 (NHERF2) and lysophosphatidic acid receptor 2 (LPA2), which couples the LPA2-mediated signaling with CFTR channel activity in a compartmentalized manner. Upon activation of LPA2, adenylyl cyclase (AC) is inhibited through Gi pathway, leading to a decreased cAMP level in proximity to CFTR and consequently inhibiting CFTR channel function. For clarity, only the major signaling molecules involved in this macromolecular complex are depicted here.[](https://www.ncbi.nlm.nih.gov/mesh/D000242)
The CFTR-NHERF2-LPA2 complex could be targeted to attenuate the LPA2-mediated inhibitory signaling on CFTR channel function and/or suppress the LPA2-mediated IL-8 release from airway epithelial cells. These two effects are beneficial to CF therapy. (a) The approach to disruption of NHERF2-LPA2 interaction; (b) The approach to antagonism of LPA2.
In the gut, the CFTR-NHERF2-LPA2 complex could be targeted by using LPA or LPA2 agonists to inhibit CFTR channel activity. This inhibition would be beneficial for the therapeutic intervention of secretory diarrhea.
|
# Introduction
Inhibition of RhoA/Rho kinase pathway and smooth muscle contraction by [hydrogen sulfide](https://www.ncbi.nlm.nih.gov/mesh/D006862) Inhibition of RhoA/Rho kinase pathway and smooth muscle contraction by hydrogen sulfide
# Abstract
*In the **Abstract** section:*
Abstract
Hydrogen sulfide (H2S) plays an important role in smooth muscle relaxation. Here, we investigated the expression of enzymes in H2S synthesis and the mechanism regulating colonic smooth [muscle function ](https://www.ncbi.nlm.nih.gov/mesh/D006862)by[ H2](https://www.ncbi.nlm.nih.gov/mesh/D006862)S. Expression of cystathionine‐γ‐lyase (CSE), but not cystathionine‐β‐synthase (CBS), was identified in th[e c](https://www.ncbi.nlm.nih.gov/mesh/D006862)olonic smooth muscle of rabbit, mouse, and human. Carbachol (CCh)‐induced [con](https://www.ncbi.nlm.nih.gov/mesh/D006862)traction in rabbit muscle strips and isolated muscle cells was inhibited by l‐cysteine (substrate of CSE) and NaHS (an exogenous H2S donor) in a concentration[‐dependen](https://www.ncbi.nlm.nih.gov/mesh/D002217)t [fas](https://www.ncbi.nlm.nih.gov/mesh/D002217)hion. H2S induced S‐sulfhydration of RhoA that was associated with inhibition of RhoA act[ivity. CCh](https://www.ncbi.nlm.nih.gov/mesh/D003545)‐induced Rho kinase acti[vity](https://www.ncbi.nlm.nih.gov/mesh/C025451) also was inhib[ite](https://www.ncbi.nlm.nih.gov/mesh/D006862)d by l‐cysteine and NaHS in a concentration‐de[pen](https://www.ncbi.nlm.nih.gov/mesh/D006862)dent fashion. Inhibition of CCh‐induced contraction by l‐cysteine was blocked by the CS[E i](https://www.ncbi.nlm.nih.gov/mesh/D002217)nhibitor, dl‐propargylglycine (DL‐PPG) in dispersed[ muscle ce](https://www.ncbi.nlm.nih.gov/mesh/D003545)lls. [Inhi](https://www.ncbi.nlm.nih.gov/mesh/C025451)bition of CCh‐induced Rho kinase activity by l‐cystei[ne ](https://www.ncbi.nlm.nih.gov/mesh/D002217)was blocked by CSE siRNA[ in cultur](https://www.ncbi.nlm.nih.gov/mesh/D003545)ed cells and DL‐PPG in dispersed mu[scle cells. Stimula](https://www.ncbi.nlm.nih.gov/mesh/C009055)ti[on of ](https://www.ncbi.nlm.nih.gov/mesh/C009055)Rho kinase activity and muscle contraction [in ](https://www.ncbi.nlm.nih.gov/mesh/D002217)response to CCh was also inhibit[ed by l‐cy](https://www.ncbi.nlm.nih.gov/mesh/D003545)steine or NaHS in colonic muscle cells from mous[e and ](https://www.ncbi.nlm.nih.gov/mesh/C009055)human. Collectively, our studies identified the expression of CSE in colonic smooth muscle and determ[ine](https://www.ncbi.nlm.nih.gov/mesh/D002217)d that sulfhydration of[ RhoA by H](https://www.ncbi.nlm.nih.gov/mesh/D003545)2S l[eads](https://www.ncbi.nlm.nih.gov/mesh/C025451) to inhibition of RhoA and Rho kinase activities and muscle contraction. The mechanism identified may provide novel therapeutic approaches to mitigate gastrointestinal motility[ di](https://www.ncbi.nlm.nih.gov/mesh/D006862)sorders.
# Abbreviations
*In the **Abbreviations** section:*
CBS
cystathionine‐β‐synthase
CCh[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
carbachol[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
CSE
cystathionine‐γ‐lyase
DL‐PPG[](https://www.ncbi.nlm.nih.gov/mesh/C009055)
dl‐propargylglycine[](https://www.ncbi.nlm.nih.gov/mesh/C009055)
MLCK
myosin light‐chain kinase
MLCP
myosin light‐chain phosphatase
NaHS[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
sodium hydrosulfide[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Hydrogen sulfide (H2S) is produced via both enzymatic and nonenzymatic pathways in mammalian cells, but most of the H2S levels in tissues are attributed to enzymatic synthesis (Abe and Kimura 1996; Kamoun 2004; Caliendo et al. 2010; Wang 2012). The main enzymes responsible for H2S generation, cystathionine‐β‐synthase (CBS) and cystathionine‐γ‐lyase (CSE), are expressed in several tissues, but the pattern of expression has been reported to be tissue‐specific. Expression of CBS is mainly found in the central and peripheral nervous system, whereas expression of CSE is mainly found in the vascular system and liver (Bao et al. 1998; Renga 2011; Wang 2012). The H2S levels in the gastrointestinal (GI) system include two sources: a luminal sulfate‐reducing bacterial source in the large intestine and an endogenous generation by different cells within the wall of GI tract (Linden et al. 2010; Farrugia and Szurszewski 2014). Regulation of GI functions including motility and secretion by H2S has been reported previously (Distrutti et al. 2006; Linden et al. 2008; Hennig and Diener 2009; Xu et al. 2009; Wallace 2010; Strege et al. 2011; Sha et al. 2014; Nalli et al. 2015).[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Relaxation of vascular muscle in response to H2S was documented in the aorta, portal vein, and mesenteric artery (Wang 2012). H2S also inhibits the motility in different regions of the GI tract in different species. In the GI tract, studies using whole segments of the intestine or muscle strips from rat, mouse, guinea pig, and human have shown that H2S inhibits contraction (Dhaese and Lefebvre 2009; Gil et al. 2011, 2013; Kasparek et al. 2012; Martinez‐Cutillasa et al. 2015). The spontaneous contractions in colonic muscle strips from mouse, rat, and human were inhibited by NaHS (Gallego et al. 2008). Inhibition of spontaneous contractions by NaHS was not affected by the neural blocker, tetrodotoxin, but significantly reduced by KATP channel blocker, glibenclamide, and apamin suggesting that the effect of NaHS was independent of neural activation and probably involved activation of KATP channels as well as apamin‐sensitive small conductance K+ channels (Gallego et al. 2008).
In circular muscle of rat ileum, but not longitudinal muscle, the effect of NaHS was partly inhibited by glibenclamide (Nagao et al. 2011, 2012). In longitudinal muscle of rat jejunum, the inhibitory effect of NaHS on stimulated contractile activity was blocked by glibenclamide without affecting the spontaneous contractile activity (Kasparek et al. 2012). In contrast, studies by Teague et al. (2002) in rabbit and guinea pig ileum demonstrated that NaHS caused inhibition of acetylcholine‐ or electrical field‐stimulated muscle contraction and the effect was independent of KATP channel activation as glibenclamide had no effect on NaHS‐induced decrease in muscle contraction.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Expression of CSE and/or CBS has been reported in different cell types in the GI tract including enteric neurons, interstitial cells of Cajal (ICC), and epithelial cells (Schicho et al. 2006; Martin et al. 2010; Kasparek et al. 2012; Sha et al. 2014; Quan et al. 2015). Hence, the effect of H2S could be due to release of neurotransmitters from the enteric neurons, inhibition of pacemaker activity of ICC, and/or its direct inhibitory effect on smooth muscle function. Although the evidence for the enzymatic production of H2S by GI tissues was produced, the evidence for the regulation of synthesis is not clear. The rate of H2S production in intact mouse colonic tissue in the presence of 10 mmol/L cysteine in Kreb's solution was shown to be around 0.6 pmol/min/mg (Linden et al. 2008, 2010). In the small and large intestine of rat, the basal level of H2S was reported to be around 60 nmol/g/h and the response to 10 mmol/L cysteine in these tissues ranged from 300 to 500 nmol/g/h, a 5‐ to 10‐fold increase (Martin et al. 2010).[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Phosphorylation of serine 19 on the 20‐kDa regulatory light‐chain of myosin II (MLC20) is an essential step in smooth muscle contraction (Hartshorne et al. 1998; Somlyo and Somlyo 2003; Murthy 2006; de Godoy and Rattan 2011). Phosphorylation of MLC20 is regulated by a Ca2+/calmodulin‐dependent MLC kinase (MLCK), which initiates phosphorylation of MLC20 and MLC phosphatase (MLCP), which dephosphorylates MLC20. MLCP is a heterotrimer with an 110‐ to 130‐kDa regulatory subunit (myosin phosphatase target subunit 1 (MYPT1), a 37‐kDa catalytic subunit of type 1 phosphatase (PP1cδ), and a 20‐kDa subunit of unknown function (Hartshorne et al. 1998).[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
The RhoA/Rho kinase signaling pathway regulates muscle contraction. RhoA is a small G protein with inherent GTPase activity and cycles between two states: an inactive GDP‐bound state and an active GTP‐bound state. Activated RhoA binds to the Rho‐binding domain of Rho kinase, causing the enzyme to unfold and freeing its catalytic activity. Phosphorylation of MYPT1 at Thr696 by Rho kinase causes its dissociation from, and inhibition of, the catalytic subunit resulting in MLC20 phosphorylation and muscle contraction (Hartshorne et al. 1998; Somlyo and Somlyo 2003; Murthy 2006; de Godoy and Rattan 2011).[](https://www.ncbi.nlm.nih.gov/mesh/D006153)
Our results demonstrate selective expression of CSE in colonic smooth muscle cells of rabbit, mouse, and human, and addition of l‐cysteine or NaHS causes S‐sulfhydration of RhoA that lead to inhibition of RhoA and Rho kinase activities and muscle contraction in response to contractile agonists.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Materials and Methods
*In the **Materials and Methods** section:*
## Reagents
*In the **Reagents** section:*
Cystathionine‐β‐synthase (CBS) and cystathionine‐γ‐lyase (CSE) antibodies were purchased from Proteintech (Chicago, IL); [32P]ATP was purchased from PerkinElmer (Cambridge, MA); HRP‐conjugated secondary antibodies were obtained from Cell Signaling Technology (Danvers, MA); PVDF membranes were obtained from Millipore (Billerica, MA); Effectene Transfection Reagent, QIAEX®II was from Qiagen (Germantown, MD); culture medium (Dulbecco's modified Eagle's medium) was from Fisher Scientific (Ashville, NC); l‐Cysteine and dl‐propargylglycine (PPG) were from Sigma (St. Louis, MO). All other reagents were from Sigma (St. Louis, MO).[](https://www.ncbi.nlm.nih.gov/mesh/C000615311)
Rabbits (New Zealand white male) weighing 4–5 lbs were purchased from RSI Biotechnology (Clemmons, NC), and mice (male C57BL/6 strain) were purchased from Jackson Laboratories (Bar Harbor, ME). Rabbits and mice were acclimatized at the facility administered by the Division of Animal Resources, Virginia Commonwealth University. The Institutional Animal Care and Use Committee of Virginia Commonwealth University approved all the procedures conducted. Colons from normal human subjects were obtained from a nonprofit organization known as National Disease Research Interchange (NDRI, Philadelphia, PA) that provides human organs and tissue. The studies involving human tissues are approved as exempt from VCU Institutional Review Board.
## Isolation of smooth muscle cells
*In the **Isolation of smooth muscle cells** section:*
The colon from rabbit, mouse, and human were dissected out and after emptying the contents was placed in oxygenated Kreb's solution composed of 118 mmol/L NaCl, 4.75 mmol/L KCl, 1.19 mmol/L KH2PO4, 1.2 mmol/L MgSO4, 2.54 mmol/L CaCl2, 25 mmol/L NaHCO3, 11 mmol/L glucose at 37°C, and pH 7.4. Small colonic segments were threaded onto a glass rod. The longitudinal muscle with adherent myenteric plexus was removed by radial abrasion with Kim wipes. Circular muscle layer of the colon was used to isolate smooth muscle cells as described previously (Murthy et al. 2003a,b; Rajagopal et al. 2013; Nalli et al. 2015). Briefly, colonic circular muscle strips were incubated for 30 min at 31°C in 15 mL of HEPES medium (120 mmol/L NaCl, 4 mmol/L KCl, 2.6 mmol/L KH2PO4, 0.6 mmol/L MgCl2, 25 mmol/L HEPES, 14 mmol/L glucose, 2.1% (v/v) Eagle's essential amino acid mixture, 0.1% collagenase
[type II], and 0.1% soybean trypsin inhibitor). After the 30‐min digestion period, tissues were washed with collagenase‐free medium (50 mL) and muscle cells were allowed to disperse spontaneously. Cells were collected by filtration through 500 μm Nitex followed by centrifugation twice at 350g for 10 min. Smooth muscle cells were cultured in DMEM containing 10% fetal bovine serum and cells passaged once after attaining confluence. For experiments, muscle cells were used in the first passage.[](https://www.ncbi.nlm.nih.gov/mesh/C051865)
## Transfection of CSE siRNA
*In the **Transfection of CSE siRNA** section:*
The eukaryotic expression vector pcDNA3 was used to subclone CSE siRNA into the multiple cloning sites (EcoRI). Smooth muscle cells in culture were transiently transfected for 48 h with recombinant plasmid cDNAs. To monitor transfection efficiency, muscle cells were cotransfected with 2 μg pcDNA3 vector and 1 μg of pGreen Lantern‐1 DNA (Rajagopal et al. 2013; Nalli et al. 2015).
## Western blot analysis
*In the **Western blot analysis** section:*
Expression of CSE and CBS was measured by western blot as described previously (Murthy et al. 2003a,b; Rajagopal et al. 2013; Nalli et al. 2015). Smooth muscle cells were homogenized in lysis buffer containing Triton X‐100 and protease and phosphatase inhibitors. The lysates were centrifuged at 20,000g for 10 min at 4°C and the supernatants were collected. Aliquots containing an equal amount of protein (50 μg) were resolved on 10% SDS‐PAGE and the proteins were transferred to PVDF membranes. The membranes were incubated overnight with CSE or CBS antibodies followed by incubation with appropriate secondary antibody conjugated with horseradish peroxidase.
Enhanced chemiluminescence was used to visualize protein bands on the membrane.[](https://www.ncbi.nlm.nih.gov/mesh/D017830)
## Assay for RhoA GTPase activation
*In the **Assay for RhoA GTPase activation** section:*
RhoA activation was evaluated in pull‐down assays using specific anti‐RhoA‐GTP antibody and protein A/G beads (Rajagopal et al. 2013). In brief, cultured smooth muscle cells were incubated with carbachol (CCh, 1 μmol/L) for 10 min. In some experiments, CCh was added after treatment with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min. In some cases, the cells were preincubated with CSE inhibitor, DL‐PPG (1 mmol/L) for 10 min before the addition of CCh in the presence or absence of l‐cysteine or NaHS. Cells were homogenized in the lysis buffer and GTP‐bound RhoA was immunoprecipitated using monoclonal antibody (NewEast Biosciences, Malvern, PA) that specifically recognizes RhoA‐GTP.
RhoA‐GTP bound antibody were pulled down by protein A/G, washed with lysis buffer, and processed for separation by SDS‐PAGE. The level of activated RhoA was evaluated by western blot analysis.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
## Biotin switch assay to detect S‐sulfhydration
*In the **Biotin switch assay to detect S‐sulfhydration** section:*
S‐sulfhydration was measured by biotin switch assay as described previously (Jaffrey and Snyder 2001; Mustafa et al. 2009; Kang et al. 2015) with modification. HEK293 cells transfected with RhoA cloned pcDNA 3 vector were homogenized in a medium A (250 mmol/L HEPES‐NaOH [pH 7.7], 1 mmol/L EDTA, 2.5% SDS, 0.1 mmol/L neocuproine) containing 100 μmol/L deferoxamine. Protein samples (250 μg) were then treated in the presence or absence of NaHS (0.1 mmol/L and 1 mmol/L) for 15 min and then incubated at 50°C for 20 min with blocking buffer (medium A adjusted to 2.5% SDS and 20 mmol/L methyl methane thiosulfonate) with frequent vortexing. Proteins were precipitated using acetone and incubated at 37°C for 3 h in medium A adjusted to 1% SDS with 4 mmol/L biotin‐HPDP (N‐[6‐(biotinamido) hexyl]‐3′‐(2′‐pyridyldithio)‐propionamide) in dimethyl formamide.
The biotinylated proteins were precipitated by streptavidin‐agarose beads, washed with medium A, and resolved by SDS‐PAGE. After the transfer of proteins onto PVDF membranes, RhoA was analyzed by western blot.[](https://www.ncbi.nlm.nih.gov/mesh/D001710)
## Assay for Rho kinase activity
*In the **Assay for Rho kinase activity** section:*
Rho kinase activity was measured by immunokinase assay as previously described (Murthy et al. 2003a,b; Rajagopal et al. 2013; Nalli et al. 2015). Smooth muscle cells were treated with different concentrations of NaHS or l‐cysteine for 10 min and then with CCh for 10 min. The cells were homogenized in medium containing 50 mmol/L Tris‐HCl (pH 7.5), 0.1% SDS, 0.5% sodium deoxycholate, 1% Nonidet P‐40, 150 mmol/L NaCl, 1 mmol/L PMSF, 10 μg/mL aprotinin, 10 μg/mL pepstatin A, and 10 μg/mL leupeptin. Aliquots containing an equal amount of protein (50 μg) were incubated overnight at 4°C with Rho kinase‐2 antibody and protein A/G agarose. The immunoprecipitates collected by centrifugation were washed twice with a medium containing 10 mmol/L MgCl2 and 40 mmol/L HEPES (pH 7.4) followed by incubation for 5 min at 4°C with myelin basic protein (MBP) (1 mg/mL).
The reaction was initiated by the addition of 10 μCi of [32P]ATP (3000 Ci/mmol) and 20 μmol/L ATP at 37°C and terminated after 10 min by spotting the reaction mixture onto phosphocellulose disks to capture phosphorylated MBP. The disks were washed with 75 mmol/L phosphoric acid to remove free radioactivity. Phosphorylation of MBP was determined from the radioactivity on disks by liquid scintillation.[](https://www.ncbi.nlm.nih.gov/mesh/C025451)
## Measurement of contraction in muscle strips
*In the **Measurement of contraction in muscle strips** section:*
Rabbit colonic muscle strips cut in the direction of circular muscle layer from rabbit colon were rinsed immediately in Kreb's medium containing 118 mmol/L NaCl, 4.8 mmol/L KCl, 1 mmol/L MgSO4, 1.15 mmol/L NaH2PO4, 15 mmol/L NaHCO3, 10.5 mmol/L glucose, and 2.5 mmol/L CaCl2. Strips with the aid of silk threads were suspended vertically in 5 mL tissue bath containing oxygenated (95% O2/5% CO2) Kreb's medium (pH of 7.4) at 37°C. The isometric force generated by circular muscle was measured by mounting the tissue between a glass rod and isometric transducer (Grass Technologies, Quincy MA) connected to a computer recording system. A resting tension of 1 g was given and the muscle strips were allowed to equilibrate. Muscle strips were contracted with 10 μmol/L CCh in the presence or absence of different concentrations of l‐cysteine or NaHS pretreatment for 10 min.
In some experiments, muscle strips were pretreated with 10 μmol/L glibenclamide (a KATP channel blocker) before the addition of NaHS or l‐cysteine. Tissue weight was recorded after the experiment and contraction was calculated as area under the curve in response to CCh alone and compared with CCh with l‐cysteine or NaHS (Nalli et al. 2015).[](https://www.ncbi.nlm.nih.gov/mesh/C051865)
## Measurement of contraction in muscle cells
*In the **Measurement of contraction in muscle cells** section:*
Contraction in dispersed muscle cells was determined by scanning micrometry as described previously (Murthy et al. 2003a,b; Rajagopal et al. 2013; Nalli et al. 2015). An aliquot of muscle cells (0.4 mL containing 104/cell mL) was preincubated with a different concentration of l‐cysteine or NaHS for 10 min and then with CCh for another for 10 min. A final concentration of 1% acrolein was used to terminate the reaction and a drop of cell suspension was placed on a slide under a cover slip. In some experiments, muscle cells were pretreated with 10 μmol/L glibenclamide (a KATP channel blocker) before the addition of NaHS or l‐cysteine. Cell length was measured by scanning micrometry.
Cell length in the absence of any treatment was taken as resting cell length. Contraction was expressed as a decrease in mean cell length from control cell length.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Statistical analysis
*In the **Statistical analysis** section:*
Data are expressed as mean ± SEM. “n” represents average values of one sample run in duplicate or triplicate from one animal. Comparisons between groups were analyzed by Student's t‐test. A P < 0.05 was considered statistically significant. Statistical analyses were performed using the Prism software program (GraphPad Software, San Diego, CA).
## Results
*In the **Results** section:*
## Expression of CSE protein in smooth muscle cells
*In the **Expression of CSE protein in smooth muscle cells** section:*
Expression studies by western blot demonstrated the presence of CSE protein (66 kDa) in cultured muscle cells of the colon of mouse, rabbit, and human (Fig. 1). Under similar conditions, there was no detectable expression of CBS protein in these samples. However, expression of CBS protein was demonstrated using CBS‐specific antibody in mouse brain (Fig. 1). The presence of smooth muscle‐specific γ‐actin and absence of markers for interstitial cells of Cajal and endothelial cells determined the purity of these cultured smooth muscle cells in previous studies (Teng et al. 1998). Selective expression of CSE in smooth muscle cells is consistent with the tissue‐specific expression of CSE and CBS (Schicho et al. 2006; Linden et al. 2010; Wang 2012).
Expression of cystathionine‐γ‐lyase (CSE) in colonic smooth muscle. Smooth muscle cells were isolated from the colon of rabbit, mouse, and human, and cultured in DMEM‐10. Lysates were prepared from cultured muscle cells and expression of CSE and CBS were analyzed by western blot. Expression of CSE (66 kDa), but not CBS, was detected in colonic smooth muscle from rabbit, mouse, and human, and also in mouse brain. Expression of CBS (61 kDa) was detected in lysates derived from the mouse brain. Representative images from four separate experiments are shown in the figure.
## Inhibition of contraction by H2S in muscle strips
*In the **Inhibition of contraction by H2S in muscle strips** section:*
Isometric contraction was measured in muscle strips isolated from the colon of rabbit. Muscle strips were equilibrated to a passive tension of 1 g for 1 h before experiments were conducted. Area under curve (AUC) measurement representing force versus time in response to CCh (10 μmol/L) was defined as muscle contraction above basal. AUC was recorded for the first 2 min in response to CCh. The strips were incubated with various concentrations of l‐cysteine (0.1 μmol/L to 10 mmol/L) or NaHS (0.01 μmol/L to 1 mmol/L) for 10 min before the addition of CCh and measurements of AUC were recorded again to estimate the amount of inhibition of contraction in response to l‐cysteine or NaHS.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
As shown previously (Nalli et al. 2015), CCh (10 μmol/L) induced a contraction of 367 ± 26 gram‐seconds above basal tension (n = 4). CCh‐induced contraction was inhibited by l‐cysteine and NaHS in a concentration‐dependent fashion (Fig. 2A). In vascular and visceral smooth muscle, the inhibitory effect of H2S was shown to be mediated by activation of plasma membrane KATP channels and hyperpolarization (Zhao et al. 2001; Zhao and Wang 2002; Tang et al. 2005; Gallego et al. 2008; Mustafa et al. 2011; Wang 2012; Gade et al. 2013). Incubation of colonic muscle strips with glibenclamide (10 μmol/L) for 10 min did not affect the inhibitory effect of l‐cysteine (10 mmol/L) (65 ± 4% inhibition vs. 61 ± 3% inhibition in the presence of glibenclamide) or NaHS (1 mmol/L) (82 ± 6% inhibition vs. 79 ± 7% inhibition in the presence of glibenclamide). Glibenclamide had no effect on CCh‐induced contraction (329 ± 21 gram‐seconds above basal tension vs. 309 ± 35 gram‐second in the presence of glibenclamide, n = 3), suggesting that inhibition of CCh‐induced contraction by l‐cysteine was not due to activation of KATP channels.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
Inhibition of carbachol‐induced contraction by l‐cysteine and NaHS in colonic smooth muscle. (A) Muscle strips from the rabbit colon were allowed to equilibrate to a passive tension of 1 g for 1 h and then treated with carbachol (carbachol [CCh], 10 μmol/L). The increase in tension (area under the curve [AUC]) in response to CCh from t = 0 min to t = 2 min was calculated as a contraction. In separate experiments, muscle strips were incubated with different concentrations of l‐cysteine (0.1 to 100 mmol/L) or sodium hydrosulfide (NaHS, 0.01 to 10 mmol/L) for 10 min before addition of carbachol. Inhibition of contraction was measured as a percent decrease in AUC in the presence of l‐cysteine or NaHS compared to control response. CCh induced contraction of 369 ± 26 gram‐seconds (AUC) above basal tension. Values are mean ± SEM of four experiments. *P < 0.05; **P < 0.01, significant inhibition of CCh‐induced muscle contraction. (B) Muscle cells isolated from circular muscle layer of rabbit colon were treated with increasing concentrations of CCh (0.01 nmol/L to 1 μmol/L) for 10 min to induce sustained contraction. In some experiments, cells were pretreated with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min and then treated with CCh for 10 min. Muscle cell length was measured by scanning micrometry. Contraction by CCh was calculated as the decrease in muscle cell length from basal cell length of 109 ± 5 μm. Values are mean ± SEM of five experiments. (C) Muscle cells isolated from rabbit colon were treated with CCh (0.1 μmol/L) for 10 min to induce sustained contraction. In some experiments, cells were pretreated with different concentrations of l‐cysteine (0.01 mmol/L to 100 mmol/L) or NaHS (0.01 mmol/L to 100 mmol/L) for 10 min and then treated with CCh for 10 min. Muscle cell length was measured by scanning micrometry. Contraction by CCh was calculated as a percent decrease in muscle cell length from basal cell length (30 ± 3% decrease in cell length from the basal cell length of 109 ± 5 μm). Values are mean ± SEM of seven experiments. *P < 0.05; **P < 0.01, significant inhibition of CCh‐induced muscle contraction.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
## Inhibition of muscle contraction by H2S in isolated muscle cells
*In the **Inhibition of muscle contraction by H2S in isolated muscle cells** section:*
Contractile agonists (e.g., CCh) induce biphasic contraction in muscle cells: a rapid contraction within 30 sec followed by a sustained contraction for 20 min (Murthy et al. 2003a). The effect of l‐cysteine and NaHS on sustained contraction was measured. Muscle cells were pretreated with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) and then treated with different concentrations of CCh (0.01 nmol/L to 1 μmol/L) for 10 min. CCh caused a concentration‐dependent contraction in muscle cells with EC50 of 10−9 ± 3 × 10−10 mol/L. Maximal contraction (32 ± 4% decrease in cell length from control length of 109 ± 5 μm, n = 7) was obtained with 1 μmol/L of CCh. Addition of l‐cysteine or NaHS shifted the contractile response to CCh to the right with EC50 values of 3.5 × 10−8 ± 5 × 10−9 mol/L and 9 × 10−8 ± 7 × 10−9 mol/L, respectively, suggesting inhibition of CCh‐induced contraction by H2S (Fig. 2B).[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
Contraction in response to maximal dose of CCh (1 μmol/L) was inhibited by l‐cysteine or NaHS in a concentration‐dependent fashion (Fig. 2C). Glibenclamide, a KATP channels blocker, was used to test whether the effect of H2S involved activation of KATP channels. Pretreatment with glibenclamide (10 μmol/L) for 10 min did not affect the inhibition of contraction by l‐cysteine (10 mmol/L) (53 ± 2% inhibition alone vs. 52 ± 4% inhibition in the presence of glibenclamide, n = 6) or NaHS (1 mmol/L) (63 ± 3% inhibition alone vs. 60 ± 6% inhibition in the presence of glibenclamide, n = 6). Glibenclamide had no effect on CCh‐induced contraction (31 ± 3% contraction vs. 29 ± 2% contraction in the presence of glibenclamide, n = 4). These results demonstrate that H2S inhibits muscle contraction and the effect of H2S does not involve KATP channel activation.
Control studies showed that inhibition of contraction by levcromakalim (10 μmol/L), a potassium channel activator, was reversed by glibenclamide (52 ± 7% inhibition of contraction alone vs. 8 ± 4% inhibition in the presence of glibenclamide).[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
Inhibition of CCh‐induced contraction by l‐cysteine was blocked by treatment of cells with the CSE inhibitor, dl‐propargylglycine (DL‐PPG) (1 mmol/L) for 10 min (Fig. 3). Treatment of cells with l‐cysteine (10 mmol/L) for 10 min caused significant inhibition (59 ± 3% inhibition) of contraction in response to CCh. The inhibitory effect of l‐cysteine was attenuated (15 ± 8% inhibition) in the presence of DL‐PPG (1 mmol/L). In contrast, inhibition of sustained contraction by NaHS (1 mmol/L) was not affected by DL‐PPG (59 ± 3% inhibition vs. 60 ± 9% inhibition with DL‐PPG), suggesting that the inhibition of muscle contraction by l‐cysteine was mediated by the activation of CSE and generation of H2S.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
Effect of cystathionine‐γ‐lyase (CSE) inhibitor, dl‐propargylglycine (DL‐PPG) on l‐cysteine‐induced inhibition of sustained contraction. Muscle cells isolated from colon were treated with carbachol (CCh) (1 μmol/L) for 10 min in the presence or absence of l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) pretreatment for 10 min. In some experiments, cells were pretreated with CSE inhibitor DL‐PPG (1 mmol/L) for 10 min before the addition l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for another 10 min. Muscle cell length was measured by scanning micrometry. Contraction in response to CCh was calculated as a percent decrease in muscle cell length from control cell length (32 ± 4% decrease in cell length from the basal cell length of 109 ± 6 μm). Values are mean ± SEM of four experiments. **P < 0.01, significant inhibition of CCh‐induced muscle contraction.[](https://www.ncbi.nlm.nih.gov/mesh/C009055)
Pretreatment of cells with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min also inhibited sustained contraction in mouse and human. In mouse colonic muscle cells, contraction (31 ± 3% decrease in cell length) in response to 1 μmol/L CCh was inhibited by l‐cysteine (45 ± 4% inhibition) and NaHS (57 ± 6% inhibition) (Fig. 4A). In human colonic smooth muscle cells, contraction (30 ± 3% decrease in cell length) in response to 1 μmol/L CCh was inhibited by l‐cysteine (34 ± 3% inhibition) and NaHS (47 ± 5% inhibition) (Fig. 4B).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Inhibition of carbachol‐induced contraction by l‐cysteine and NaHS in colonic smooth muscle. Muscle cells isolated from circular muscle layer of the mouse (top panel) and human colon (bottom panel) were pretreated with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min and then treated with carbachol (CCh; 1 μmol/L) for 10 min to induce sustained contraction. Muscle cell length was measured by scanning micrometry. Contraction by CCh was calculated as the decrease in muscle cell length from basal cell length (31 ± 3% decrease in cell length from the basal cell length of 115 ± 7 μm in mouse colon and 30 ± 3% decrease in cell length from the basal cell length of 93 ± 6 μm in the human colon). Values are mean ± SEM of 5–7 experiments. **P < 0.01, significant inhibition of CCh‐induced muscle contraction.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
## Inhibition of CCh‐induced RhoA activity by H2S
*In the **Inhibition of CCh‐induced RhoA activity by H2S** section:*
Previous studies in GI smooth muscle have shown that sustained MLC20 phosphorylation and contraction were mediated by RhoA/Rho kinase pathway (Murthy et al. 2003a), raising the possibility that inhibition of contraction by H2S could be due to inhibition of this pathway.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
RhoA is a small G protein that is bound to GDP in the basal state and GTP in the activated state. Addition of CCh (1 μmol/L) for 10 min caused an increase in the RhoA activity, measured as an increase in the incorporation of GTP into RhoA using an antibody that specifically recognizes RhoA‐GTP. Pretreatment of cells with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min caused inhibition of RhoA activity in response to CCh (Fig. 5A). These results suggest that endogenous and exogenous H2S inhibits RhoA activity in colonic muscle cells. The notion that l‐cysteine exerts its effect via activation of CSE was examined using DL‐PPG.
Pretreatment of cells with DL‐PPG (1 mmol/L) for 10 min reversed the effect of l‐cysteine on RhoA activity in response to CCh (Fig. 5A). DL‐PPG, in contrast, had no effect on NaHS‐induced inhibition of RhoA activity in response to CCh (Fig. 5A).[](https://www.ncbi.nlm.nih.gov/mesh/D006153)
S‐Sulfhydration of RhoA and inhibition of carbachol‐induced RhoA activity by l‐cysteine and NaHS in colonic smooth muscle. (A) Muscle cells isolated from circular layer of rabbit colon were treated with carbachol (CCh) (1 μmol/L) 10 min in the presence or absence of the l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) pretreatment for 10 min. In some experiments, cells were pretreated DL‐PPG (1 mmol/L) for 10 min before the addition of l‐cysteine or NaHS for another 10 min. RhoA activity in response to CCh was measured by incorporation of GTP into RhoA by western blot using an antibody that is specific for RhoA‐GTP. The image depicts representative blot of four separate experiments. Values are mean ± SEM of four experiments. **P < 0.01, significant inhibition of CCh‐induced RhoA activity. (B) Lysates obtained from HEK cells were treated in the presence or absence of NaHS (0.1 or 1 mmol/L) for 15 min and blocked with methyl methanethiosulfonate (MMTS). H2S modified ‐SSH residues are labeled with N‐[6‐(biotinamido) hexyl]‐3′‐(2′‐pyridyldithio)‐propionamide (HPDP‐biotin) and biotinylated proteins were pulled down and analyzed by western blot. Western blot analysis showed the difference in the sulfhydration levels in the control and NaHS‐treated samples. The figure depicts representative blot of three separate experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
## S‐Sulfhydration of RhoA by H2S
*In the **S‐Sulfhydration of RhoA by H2S** section:*
S‐sulfhydration is known to be the primary mechanism through which H2S signals. H2S alters the function of KATP channels via S‐sulfhydration (Mustafa et al. 2009; Wang 2012). Therefore, we examined if H2S alters the activity of RhoA via S‐sulfhydration. S‐sulfhydration of RhoA in response to NaHS (0.1 and 1 mmol/L) was analyzed using a biotin switch assay. Basal sulfhydration of RhoA was not detected in control cells (Fig. 5B).
Addition of NaHS for 15 min, however, caused sulfhydration of RhoA, suggesting that inhibition of RhoA activity by H2S could be due to S‐sulfhydration of RhoA (Fig. 5B).[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
## Inhibition of Rho kinase activity by H2S
*In the **Inhibition of Rho kinase activity by H2S** section:*
Contractile agonists induce an increase in Rho kinase activity in smooth muscle via activation of RhoA. Consistent with the previous studies (Murthy et al. 2003a,b), treatment of smooth muscle cells with CCh for 10 min caused a significant increase in Rho kinase activity in a concentration‐dependent fashion with an EC50 of 5 × 10−9 ± 1 × 10−9 mol/L (Fig. 6A). The concentration‐dependent curve was shifted to the right with EC50 values of 2.5 × 10−8 ± 4 × 10−9 mol/L and 9.8 × 10−8 ± 1.5 × 10−8 mol/L, respectively, by pretreatment of cells with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min, suggesting inhibition of Rho kinase activity by H2S (Fig. 6A).[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
Inhibition of carbachol‐induced Rho kinase activity by l‐cysteine and NaHS in colonic muscle cells. (A) Muscle cells isolated from rabbit colon were treated with different concentrations of carbachol (CCh) for 10 min. In some experiments, cells were pretreated with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min and then treated with CCh for 10 min. Rho kinase activity was measured by immunokinase assay using [32P]ATP. Results are expressed as cpm/mg protein. Values are mean ± SEM of 4–6 experiments. (B) Muscle cells isolated from rabbit colon were treated with CCh (1 μmol/L) for 10 min. In some experiments, cells were pretreated with different concentrations of l‐cysteine (0.01 mmol/L to 100 mmol/L) or NaHS (0.01 mmol/L to 100 mmol/L) for 10 min and then treated with CCh (1 μmol/L) for 10 min. Rho kinase activity was measured by immunokinase assay using [32P]ATP. Results are expressed as a percent inhibition of CCh‐induced Rho kinase activity. Values are mean ± SEM of 4–6 experiments. **P < 0.01, significant inhibition of CCh‐induced Rho kinase activity.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
Pretreatment of cells with increasing concentrations of l‐cysteine or NaHS for 10 min caused inhibition of CCh‐stimulated Rho kinase activity and the inhibitory effect is concentration dependent with an EC50 of 5.6 × 10−5 ± 1.2 × 10−5 mol/L for NaHS and 9.9 × 10−5 ± 2.4 × 10−5 mol/L for l‐cysteine (Fig. 6B). Maximal inhibition was 89 ± 9% with NaHS (10 mmol/L) and 73 ± 7% with l‐cysteine (10 mmol/L). These results suggest that endogenous and exogenous H2S inhibits Rho kinase activity in colonic muscle cells. The kinase assay was specific for Rho kinase. CCh (1 μmol/L)‐stimulated Rho kinase activity was blocked by the addition of Rho kinase inhibitor Y27632 (10 μmol/L) (26,345 ± 3245 cpm/mg protein vs. 3214 ± 452 cpm/mg protein in the presence of Y27632, n = 5).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Inhibition of Rho kinase activity by l‐cysteine is mediated via CSE/H2S
*In the **Inhibition of Rho kinase activity by l‐cysteine is mediated via CSE/H2S** section:*
The involvement of endogenous H2S generation by l‐cysteine to inhibit Rho kinase activity was tested by two methods: (1) by transfection with CSE‐specific siRNA in cultured muscle cells, and (2) by treatment with a selective CSE inhibitor, DL‐PPG in dispersed muscle cells. CCh induced a significant increase in Rho kinase activity in cultured muscle cells (28,819 ± 2312 cpm/mg protein; P < 0.001, n = 5) that is similar to the increase in dispersed muscle cells. Pretreatment of cells with l‐cysteine (10 mmol/L) for 10 min significantly inhibited CCh‐stimulated Rho kinase activity (63 ± 8% inhibition) in cells transfected with control siRNA and the inhibition was blocked in cells transfected with CSE‐specific siRNA (7 ± 1% inhibition, NS) (Fig. 7A), implying activation of CSE by l‐cysteine. Pretreatment of cells with NaHS (1 mmol/L) for 10 min also inhibited CCh‐stimulated Rho kinase activity in cells transfected with control siRNA (66 ± 7% inhibition) and the inhibition, however, was not affected in cells transfected with CSE‐specific siRNA (60 ± 7% inhibition).[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Inhibition of carbachol‐induced Rho kinase activity by l‐cysteine via activation of cystathionine‐γ‐lyase (CSE). (A) Cultured rabbit colonic muscle cells were transfected with control siRNA or CSE‐specific siRNA for 48 h. Cells were treated with carbachol (CCh; 1 μmol/L) for 10 min in the presence or absence of l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) pretreatment for 10 min. Rho kinase activity was measured by immunokinase assay using [32P]ATP. Results are expressed as cpm/mg protein. Values are mean ± SE of 4–6 experiments. **P < 0.001, significant inhibition of CCh‐stimulated Rho kinase activity. Expression of CSE in cells transfected with control siRNA or CSE siRNA was analyzed by western blot (inset). (B) Dispersed muscle cells from rabbit colon were treated with CCh (1 μmol/L) for 10 min in the presence or absence of l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) pretreatment for 10 min. In some experiments, cells were pretreated with CSE inhibitor, DL‐PPG (1 mmol/L) for 10 min before the addition of l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for another 10 min. Rho kinase activity was measured by immunokinase assay using [32P]ATP. Results are expressed as cpm/mg protein. Values are mean ± SEM of 4–6 experiments. **P < 0.001, significant inhibition of CCh‐stimulated Rho kinase activity.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
Similar studies were performed in cells pretreated with DL‐PPG (1 mmol/L) for 10 min before treatment with l‐cysteine or NaHS in dispersed muscle cells. Pretreatment of cells with l‐cysteine (10 mmol/L) for 10 min significantly inhibited CCh‐stimulated Rho kinase activity (53 ± 7% inhibition) and the inhibition was blocked by DL‐PPG (1 mmol/L) (11 ± 5% inhibition, NS) (Fig. 7B). Pretreatment of cells with NaHS (1 mmol/L) for 10 min also significantly inhibited CCh‐stimulated Rho kinase activity in control cells (64 ± 8% inhibition) and the inhibition, however, was not affected by DL‐PPG (65 ± 3% inhibition) (Fig. 7B). These results provide evidence for the involvement of CSE, via generation of H2S, in the inhibition of Rho kinase activity by l‐cysteine.[](https://www.ncbi.nlm.nih.gov/mesh/C009055)
Pretreatment of cells with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min also inhibited Rho kinase activity in response to CCh in mouse and human. In mouse colonic muscle cells, inhibition by l‐cysteine was 43 ± 6% and inhibition by NaHS was 58 ± 5% (Fig. 8A). In human colonic muscle cells, inhibition by l‐cysteine was 40 ± 4% and inhibition by NaHS was 54 ± 6% inhibition (Fig. 8B).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Inhibition of carbachol‐induced Rho kinase activity by l‐cysteine and NaHS in colonic muscle cells. Muscle cells isolated from mouse (upper panel, A) or human (lower panel, B) colon were pretreated with l‐cysteine (10 mmol/L) or NaHS (1 mmol/L) for 10 min and then treated with CCh for 10 min. Rho kinase activity was measured by immunokinase assay using [32P]ATP. Results are expressed as cpm/mg protein. Values are mean ± SEM of 4–6 experiments. **P < 0.01 significant inhibition of CCh‐stimulated Rho kinase activity.[](https://www.ncbi.nlm.nih.gov/mesh/D002217)
## Discussion
*In the **Discussion** section:*
Recent studies have demonstrated that H2S, like nitric oxide (NO) and carbon monoxide (CO), is justified to be defined as a physiological transmitter (Linden et al. 2010; Wang 2012). Like NO and CO, H2S is a small gaseous and freely permeable molecule. The physiological importance of H2S has been underscored by its significance in the regulation of several functions including GI functions (Distrutti et al. 2006; Linden et al. 2008; Hennig and Diener 2009; Xu et al. 2009; Wallace 2010; Strege et al. 2011; Sha et al. 2014; Nalli et al. 2015). H2S is endogenously synthesized via CSE, CBS, and 3‐mercaptopyruvate sulfurtransferase (Wang 2012). Of these three enzymes, CSE and CBS have been well studied.
Both enzymes are dependent on pyridoxal‐5’‐phosphate and use l‐cysteine as a substrate to produce H2S (Wang 2012; Farrugia and Szurszewski 2014). Although the pattern of expression of CSE and CBS in various tissues is largely known, the regulatory mechanisms for expression and activities are not clear. Using exogenous H2S donors (e.g., NaHS) or l‐cysteine (substrate of CSE), pharmacological inhibitors of enzymes that generate H2S (e.g., DL‐PPG) and molecular approaches (siRNA, and CSE/CBS null mice), several functions of H2S have been demonstrated in the cardiovascular system, central nervous system, GI tract, and energy metabolism (Abe and Kimura 1996; Yang et al. 2008; Linden et al. 2010; Wang 2012). Studies from the ablation of CSE or downregulation of CSE highlight the importance of endogenous H2S in the regulation of smooth muscle relaxation (Zhao et al. 2001; Yang et al. 2008). One of the most studied targets of H2S is the KATP channel.
The effect of H2S on vascular muscle is due to the activation of KATP channels, whereas inhibition of KATP channels with glibenclamide blocked the effect of H2S (Zhao et al. 2001; Zhao and Wang 2002; Tang et al. 2005; Mustafa et al. 2011; Wang 2012). The relaxant effect of NaHS in mouse aorta and bronchial rings, however, was not affected by glibenclamide (Kubo et al. 2007a,b). An excitatory action of H2S on bladder contraction involving capsaicin‐sensitive nerves was reported (Pattacchini et al. 2004).[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Although H2S is known to cause inhibition of contraction both in vivo and in vitro in the GI tract, the underlying mechanism of action of H2S is not clear. Targets such as voltage‐dependent Ca2+ channels and Ca2+‐dependent K+ channels are implicated in mediating the effect of H2S in rat colonic smooth muscle cells (Quan et al. 2015). Inhibition of contraction mediated by H2S in rabbit and guinea pig ileum, and in rat longitudinal muscle was not affected by the KATP channel blocker, glibenclamide (Teague et al. 2002; Nagao et al. 2011). NaHS‐induced inhibition of spontaneous contraction in isolated segments of mouse colon and jejunum was unaffected by tetrodotoxin, capsaicin, and N‐nitro‐l‐arginine, but reduced by apamin (Gallego et al. 2008). Tetrodotoxin also had no effect on the relaxation in the distal colon of mouse and human suggesting a direct effect on smooth muscle (Gade et al. 2013; Martinez‐Cutillasa et al. 2015; Quan et al. 2015).
These studies suggest that the underlying mechanism of action of H2S to inhibit contraction varies with the tissue and species. Increase in spontaneous contraction in the presence of CSE inhibitors suggests a role for endogenous production of H2S via CSE in the regulation of GI motility (Martinez‐Cutillasa et al. 2015). In the GI tract, the loci of generation and action of H2S include epithelial cells, enteric neurons, ICCs, and smooth muscle cells.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
In the present study, we investigated the expression of H2S synthesizing enzymes and identified the mechanisms underlying the inhibitory effects of H2S in the smooth muscle from the colon of rabbit, mouse, and human. Our findings include: (1) selective expression of CSE in smooth muscle cells of rabbit, mouse, and human, where it is responsible for H2S production; (2) inhibition of CCh‐induced contraction by l‐cysteine and NaHS in muscle strips and dispersed smooth muscle cells of rabbit and blockade of inhibition by a selective inhibitor of CSE (DL‐PPG), but not by a selective inhibitor of KATP channels (glibenclamide); (3) S‐sulfhydration of RhoA by NaHS and attenuation of CCh‐induced RhoA and Rho kinase activities and muscle contraction by both l‐cysteine and NaHS; (4) blockade of l‐cysteine‐induced inhibition by DL‐PPG or CSE siRNA; and (5) inhibition of Rho kinase activity and muscle contraction by l‐cysteine and NaHS in colonic smooth muscle cells from mouse and human.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
Expression of CSE and CBS are tissue and species specific. In rat colon, mucosal cells, myenteric neurons, and smooth muscle cells were immunoreactive for CBS and CSE (Quan et al. 2015). In rat jejunum, myenteric neurons, but not smooth muscle were immunoreactive for CBS and CSE (Kasparek et al. 2012). In mouse colon, mucosal cells were immunoreactive for CBS and CSE, whereas myenteric neurons were immunoreactive for only CSE (Linden et al. 2008). In the colon of human and guinea pig, submucosal cells and myenteric neurons were immunoreactive for CBS and CSE, whereas only myenteric interstitial cells of Cajal were immunoreactive for CSE (Schicho et al. 2006).
Expression of H2S synthesizing enzymes also varies with the different regions of the GI tract. Expression of CSE is more abundant in the proximal regions (stomach, duodenum, and jejunum), compared to distal regions (ileum and colon) of rat GI tract. In contrast, expression of CBS was low in proximal regions (duodenum and jejunum) and high in the distal regions (ileum and colon) (Martin et al. 2010). Our studies using cultured muscle cells devoid of enteric neurons and ICCs clearly demonstrated selective expression of CSE in smooth muscle cells. This agrees with the previous reports that CBS expression is abundant in central nervous system, and rare in a peripheral system (Bao et al. 1998; Linden et al. 2010;
Wang 2012).[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
l‐Cysteine and/or NaHS were commonly used in vivo and in vitro, in whole organ and nerve‐muscle preparation, to examine the physiological significance of H2S in the regulation of GI motility. Although H2S inhibits contractions, as discussed above, there are considerable differences in the mechanism of action. Interpretation of results was also confounded by the multiple cells types present in the tissue preparation used in these experiments. Our studies showed that both endogenous H2S, generated via CSE, and exogenous H2S inhibit sustained contraction in muscle cells, suggesting a direct effect of both endogenous and exogenous H2S on smooth muscle cells. The concentrations of NaHS and l‐cysteine used in our studies are similar to those used in other studies to induce a response (Distrutti et al. 2006; Gallego et al. 2008; Dhaese and Lefebvre 2009; Gil et al. 2011; Kasparek et al. 2012; Nagao et al. 2012).
These high concentrations reflect concentrations of H2S closer to the target site but not a global tissue concentration due to variations in the metabolism of H2S in different cellular compartments (Levitt et al. 2009; Wang 2012). The recovery of contractile activity after washout of the initial application of l‐cysteine and NaHS was rapid and complete, suggesting that the concentrations used in the study are not toxic. In most of the in vitro studies, a significant effect was produced with a tissue bath H2S concentration of high μmol/L (100 μmol/L) or mmol/L (d'Emmanuele diVilla Bianca et al. 2009; Bucci et al. 2012). These concentrations have been considered to be physiologic, as several reports suggest that the tissue H2S concentration normally ranges from high μM to low mM range (30 μmol/L to >100 μmol/L) (Yang et al. 2012). The free H2S concentrations in GI smooth muscle and the significance of H2S at low μM concentration as a messenger molecule are yet to be determined.
Our studies also showed that the effect of H2S is not dependent on the activation of KATP channels, but dependent on the S‐sulfhydration of RhoA.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
In the present study, we have identified RhoA as S‐sulfhydration target of H2S in mediating the inhibitory effect on muscle contraction. Like S‐nitrosylation by nitric oxide, S‐sulfhydration by H2S appears to be the common post‐translational mechanism to alter the function of proteins (Mustafa et al. 2009; Wang 2012). S‐sulfhydration of several proteins including receptors, ion channels, and enzymes have been described in previous studies (Mustafa et al. 2009; Wang 2012). Initially, it was suggested that S‐sulfhydration always results in the increase of protein activity; however, the inhibitory effect of sulfhydration was also demonstrated (Mustafa et al. 2009; Wang 2012). Previous studies in gastric muscle strips demonstrate regulation of MLC phosphatase activity by H2S (Dhaese and Lefebvre 2009).
In GI smooth muscle, activation of RhoA by contractile agonists causes sustained contraction via phosphorylation of the regulatory subunit of MLCP (MYPT1) and inhibition of MLC phosphatase leading to an increase in MLC20 phosphorylation and muscle contraction (Murthy et al. 2003a; Murthy 2006). H2S exerts an inhibitory effect on RhoA activity causing attenuation of Rho kinase activity and disinhibition of Rho kinase‐mediated MLCP activity, augmentation of MLC20 dephosphorylation, and inhibition of sustained contraction. Blockade of l‐cysteine effect by DL‐PPG in dispersed muscle cells and CSE siRNA in cultured muscle cells provides evidence for the involvement of CSE activation.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
In summary, our studies demonstrated the selective expression of CSE in colonic smooth muscle cells and identified the molecular mechanism by which H2S inhibits muscle contraction via S‐sulfhydration of RhoA and inhibition of RhoA and Rho kinase activities.[](https://www.ncbi.nlm.nih.gov/mesh/D006862)
## Author Contributions
*In the **Author Contributions** section:*
K. S. M., A. N., and H. W. participated in research design; A. N., H. W., S. B., B. B., and K. S. M. conducted experiments; K. S. M. and A. N. performed data analysis; K. S. M., A. N., and S. B. wrote or contributed to the writing of the manuscript. None contributed to new reagents or analytical tools.
## Disclosure
*In the **Disclosure** section:*
None declared.
# References
*In the **References** section:*
|
# Introduction
Are boat transition states likely to occur in Cope rearrangements? A DFT study of the biogenesis of [germacranes](https://www.ncbi.nlm.nih.gov/mesh/D045788)
# Abstract
*In the **Abstract** section:*
It has been proposed that elemanes are biogenetically formed from germacranes by Cope sigmatropic rearrangements. Normally, this reaction [proceeds](https://www.ncbi.nlm.nih.gov/mesh/D012717) through a transition state with[ a chair co](https://www.ncbi.nlm.nih.gov/mesh/D045788)nformation. However, the transformation of schkuhriolide (germacrane) into elemanschkuhriolide (elemane) may occur through a boat transition state due to t[he final conf](https://www.ncbi.nlm.nih.gov/mesh/C432449)ig[uration of](https://www.ncbi.nlm.nih.gov/mesh/D045788) the el[emanschkuhriolide, ](https://www.ncbi.nlm.nih.gov/mesh/D012717)bu[t this ](https://www.ncbi.nlm.nih.gov/mesh/D012717)transition state is questionable due to its high energy. The possible mechanisms o[f this transformati](https://www.ncbi.nlm.nih.gov/mesh/D012717)on were studied in the density functional theory frame. The mechanistic differences between the transformation of (Z,E)-germacranes and (E,E)-germacranes were also studied. We found that (Z,E)-germacranolides are significantly[ more stable than](https://www.ncbi.nlm.nih.gov/mesh/D045788) (E,E[)-germacranolides](https://www.ncbi.nlm.nih.gov/mesh/D045788) and elemanolides. In the specific[ case of schkuhriolid](https://www.ncbi.nlm.nih.gov/mesh/C432449)e, even when the boat transition sta[te is not energetical](https://www.ncbi.nlm.nih.gov/mesh/C432449)ly fa[vored, a pre](https://www.ncbi.nlm.nih.gov/mesh/D012717)vious hemiacetalization lo[wers enough t](https://www.ncbi.nlm.nih.gov/mesh/C432449)he energetic barrier to allow the formation of a very stable elemanolide that is even more stable than its (Z,E)-germacrane.[](https://www.ncbi.nlm.nih.gov/mesh/D012717)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Germacranes are biogenetic precursors of elemanes [–], because germacranes can be easily transformed into elemanes by heating through a Cope rearrangement. In some cases, these transformations are so favorable that it has been mentioned that the observed elemanes are only artifacts produced at the extraction [–]. It is known that 1,5-dienes suffer Cope rearrangements at temperatures between 200 and 300 °C, but some structural changes in the diene, such as the anionic oxy-Cope transformation allows the reactions to happen at temperatures below 0 °C []. The Cope rearrangement is a [3,3]-sigmatropic reaction and in general, occurs through a single transition state (TS), which has, normally, a chair conformation due to the higher energy of the boat conformation [,,–].
In this mechanism, the electron density of the TS is delocalized into the six carbon atoms [–]. However, if the diene contains free radical stabilizing groups, this mechanism could have significant contributions from other mechanisms that involve radical species [,,,–]. Detailed discussions about Cope rearrangements can be found in several studies and reviews that have been published previously [,–]. The confi[guration of](https://www.ncbi.nlm.nih.gov/mesh/D045788) elemanes formed via a Cope re[arrangem](https://www.ncbi.nlm.nih.gov/mesh/D012717)ent from germa[cranolides ](https://www.ncbi.nlm.nih.gov/mesh/D045788)only depends on the configuratio[n of the](https://www.ncbi.nlm.nih.gov/mesh/D012717) most stable germacrane conformer since it is mainly a concerted reaction
[,,]. It is accepted that the conformers that normally carry out a [Cope rea](https://www.ncbi.nlm.nih.gov/mesh/D012717)rrangement are the ones that have crossed double bonds, as they can g[enerate a ](https://www.ncbi.nlm.nih.gov/mesh/D000466)chair TS. The configuration of the final elemanolide is also affected by the substituents in the germac[ranol](https://www.ncbi.nlm.nih.gov/mesh/D000466)ides, the pseudo-equatorial position is preferred over the pseudo-axial position [,–]. These are the factors that dictate that specific germancranes will only rearrange to yield one or potentially two elemanolide configurations.[](https://www.ncbi.nlm.nih.gov/mesh/D002244)
The schkuriolide (1, Scheme 1) is a sesquiterpene lactone, specifically a (Z,E)-germacranolide, named melampolide, that coexists in the same natural source with the elemanschkuhriolide (3), which is an elemanolide with a stereochemistry structurally similar to 1 (C14αH5β). In order to know if 1 and 3 have biogenetic relation, 1 was transformed into 3 by heating 1 for 10 minutes at 200 °C. This suggests that 1 is a biogenetic precursor of 3 []. It is important to mention that 1 suffers a hemiacetalization in addition to a Cope rearrangement to form 3. The non-hemiacetaled compound 3 was found in neither the natural source nor the products of the biomimetic transformation of 1 into 3.
This transformation is very interesting since in order to explain the stereochemistry of elemane 3, a boat-like TS is necessary (path M, Scheme 1) [–]. This is one of the few reported cases of elemane’s biogenetic formations where a boat TS can be proposed instead of the normal chair TS [,–]. In a second proposed mechanism for the transformation of 1 into 3, the (Z,E)-germacranolide isomerizes into (E,E)-germacranolide and in a second step a Cope rearrangement forms the elemane.
In this case a normal chair TS is proposed to generate the correct elemane configuration (path N, Scheme 1) [,–]. It is possible that an enzyme is responsible to allow reactions that happen in the flask at very high temperatures in two ways, stabilizing the transition state, or destabilizing the ground states energy of the reactants. An antibody-catalyzed oxy-Cope reaction has already been described [] as well as a proposed reaction mechanism []. In the study presented in this paper, we performed density functional theory (DFT) calculations of the possible mechanisms for the transformation of 1 into 3, to elucidate which mechanism is more likely and to determine if the Cope TS with a boat conformation during the transformation is energetically favorable.
The study will also help to understand the structural factors that determine the energetic evolution of germacranolides’ Cope transformations.[](https://www.ncbi.nlm.nih.gov/mesh/C432449)
Biogenetic hypothesis for the transformation of schkuhriolide (1) into elemanschkuriolide (3).[](https://www.ncbi.nlm.nih.gov/mesh/C432449)
## Computational methods
*In the **Computational methods** section:*
DFT has been proved to be a good method for the study of reaction mechanisms of natural products' biogenesis and it has been used in many studies [–] and it is the method of choice for pericyclic reactions studies [,]. In particular, third generation hybrid functionals have improved the description of the potential energy surface and produce very reliable results [–]. Our studies in terpene biogenesis show that these hybrid functionals competes successfully with others in the determination of the energetic profile of reaction coordinates []. The third-generation hybrid functional improves the description of the energetic barriers with respect to the popular B3LYP functional [].
Moreover, the B3LYP functional was used in a Cope rearrangement study of several germacranes and it was unable to obtain accurate results when the energy differences between germacranes and elemanes were small [].[](https://www.ncbi.nlm.nih.gov/mesh/D013729)
All calculations were performed with Gaussian 09 []. All the geometries were fully optimized using the DFT hybrid method M06x [], a functional that is very reliability in calculations of activation energies [,]. The 6-31+G(d,p) basis set was used for all calculations.
Diffuse functions in double split valence basis have shown to be more important than a triplet split of the valence basis when reaction energies and activation energies are calculated with DFT []. The stability of the wave functions of all the transition states was checked. An unrestricted wave function was used to calculate the activation energy of the cis/trans isomerization of the (Z,E)-germacranolide. All energies were reported with zero-point energy corrections and all TS geometries have only one imaginary frequency.[](https://www.ncbi.nlm.nih.gov/mesh/C432449)
## Results and Discussion
*In the **Results and Discussion** section:*
Besides the two previously proposed mechanisms (Scheme 1), there are two other possible mechanisms for the transformation of 1 into 3. It is also likely that the hemiacetalization occurs before the Cope rearrangement. Fig. 1 shows the reaction coordinate of these four mechanisms. In the first proposed mechanism (path M, Fig. 1) a conformational transformation of 1 must occur first.
The most stable conformer has chair-boat conformation that according to Samek nomenclature is [15D5,1D14] (1a). This conformer is the one that is present in solution []. Nevertheless, conformer 1a does not have the proper geometry to directly generate the correct stereochemistry of 3. Both C–C bonds next to the C10–C1 double bond of conformer 1a have to rotate to generate the boat-boat conformer (1b, [15D5,1D14]), which is 3.5 kcal/mol less stable than 1a, but conformer 1b has the proper conformation to generate the configuration of 3 (ground state destabilization).
The second step is the Cope rearrangement. The saddle point (TS1b-2) for this process has a high relative energy (47.0 kcal/mol). Thus, the transformation of 1 into 2 through TS1b-2 is unlikely at 200 °C (temperature at which the biogenetic transformation was performed) []. In case of path N, the activation energy of 4 to reach the Cope TS (TS4-2) is 25.9 kcal/mol, and the relative energy of TS4-2 is 35.0 kcal/mol.
The chair TS was, as expected, less energetic than the boat TS. However, before the Cope rearrangement can proceed, the (Z,E)-germacranolide 1, must isomerize to the corresponding (E,E)-germacranolide 4. This process is highly unfavorable, its energetic barrier is about 55.7 kcal/mol, which is very close to the reported activation energies for the ethylene thermal isomerization (≈65 kcal/mol) [–]. Therefore, this high energy TS makes path N and path P unlikely.
It is important to point out that in nature this isomerization of germacranes can be catalyzed by different mechanism. For example, other cis/trans transformations have been biomimetically catalyzed by SeO2 [,–]. The only remaining route for the thermal transformation of 1 into 3 is path O. In this path, the hemiacetalization is the first step.
We used a water molecule to facilitate the proton transfer in this stage. In the experiment, a hydroxylic group of other proximate germacranolide molecule or an actual water molecule could participate as donor and acceptor of the germacranolide proton. In fact, in the solid state 1 cocrystalizes with a water molecule []. The next step in this mechanism is the Cope rearrangement which have a TS (TS5-3) less energetic than the Cope TS without hemiacetal group (TS1b-2).
This could be because the hemiacetalization reduces the transannular distance between C10 and C5 (3.20 Å, 2.91 Å, 2.72 Å and 2.66 Å for 1b, 4, 5 and 6, respectively) that facilitates orbital interactions and bond formation. The chair TS (TS6-3) is still more stable, the energy difference between TS5-3 and TS6-3 is almost the same than in TS1b-2 and TS4-2, but the energy of boat TS (TS5-3) decreases to 38.0 kcal/mol, which is small enough to be overcome at 200 °C. Thus, hemiacetalization lowers the activation energy of the boat Cope TS which allows the reaction to be completed at a temperature significantly lower than the temperature that a standard boat TS would need (≈260 °C) [].[](https://www.ncbi.nlm.nih.gov/mesh/C432449)
Reaction paths M (blue), N (orage), O (yellow) and P (green) for the transformation of 1 into 3. Relative free energies in kcal/mol. The energetic barriers for the hemiacetalization steps are calculated including a water molecule to facilitate the proton transfer.
The hemiacetalization also allows the transformation of a (Z,E)-germacranolide 1 into a elemanolide 3. This is an exception because all the biomimetical transformation of germacranolides into elemanolides reported until now are of (E,E)-germacranolides [,–]. Fig. 1 shows that the elemanolide 2 is less stable than the (Z,E)-germacranolide 1, so it is not possible to obtain 2 from 1 without a transformation that stabilizes 2.
In this special case, the hemiazetalization significantly lowers the energy of the elemanolide what makes the global process spontaneous. In fact, previous studies show that the transformation of (Z,E)-germacranolides with a blocked C6 hydroxy group do not produce the corresponding elemanolide (Scheme 2) []. Contrary, elemane (2) is more stable than the (E,E)-germacranolide. Therefore, a elemanolide can be formed directly from a (E,E)-germacranolide.
Moreover, the (E,E)-germacranolide 4 is even less stable than 1 (9.1 kcal/mol), which explains the lack of published cases for a transformation of an (E,E)-germacranolide into a (Z,E)-germacranolide, but in the opposite direction there are some examples [,,].[](https://www.ncbi.nlm.nih.gov/mesh/C432449)
Similar compounds to melampolide 1 unable to be hemiacetaled.[](https://www.ncbi.nlm.nih.gov/mesh/C432449)
The hemiacetalization by itself does not guarantee the stabilization of an elemane. If compound 1 had to suffer a normal Cope (chair TS), it would generate the C5 epimer of 2 (2’, Fig. 2). This epimer is 2.6 kcal/mol less stable than 1a, so the formation of 2’ from 1, as in case of 2, is thermodynamically forbidden. Epimer 2’ can also produce a hemiacetal (3’) but this compound has a higher energy than 3 by 4.2 kcal/mol; this is due to the C5 propenyl group in 3’ is axially oriented instead of equatorially as in 3. In contrast to 3, the formation of epimer 3’ from 1 is not thermodynamically highly favored.
Therefore, the hemiacetal formation with the right orientation is fundamental to produce an elemanolide more stable than the (Z,E)-germacranolide.[](https://www.ncbi.nlm.nih.gov/mesh/D012717)
Schematic representations of the calculated C5 epimeric structures of 2 and 3. Relative electronic energies in kcal/mol. The energies are relative to 1a.
It has been proposed that the configuration of an elemane depends on the most stable conformation of the germacradiene from which it is derived [,], although this is not a general rule and, in some cases, a conformer with higher energy is the conformer that reacts. To explain this behavior, some authors have proposed that the least energetic conformation of the Cope TS is what determines the elemane configuration [,,]. However, any of these arguments can explain the configuration of 3. Compound 3 has neither the configuration of the most stable conformer of 1 (1a) nor the configuration of the least energetic conformation of the Cope TS (a chair TS that would generate 2’).
Compound 3 comes from conformer 1b which is not the most stable one and from a boat TS that is not the least energetic TS. Then, why does compound 3 have this configuration? The answer is simple, although not obvious; the elemanolide 2 has the right configuration to allow a hemiacetalization that reduced its energy via the formation of a significantly more stable hemiacetaled elemanolide 3. The configuration of 3 is the most stable among all other possible configurations. Therefore, the energy of the different elemane configurations and their possible subsequent rearrangement reactions should also be considered when a prediction of the configuration of an elemane is determined before a Cope rearrangement.[](https://www.ncbi.nlm.nih.gov/mesh/D012717)
Finally, we studied the role the γ-lactone ring plays in the transformation of germacranes into elemanes. Takeda et al. carried out a series of experiments, which proved that γ-lactone rings prevent the Cope rearrangement of (Z,E)-germacranolides when the ring is closed but not when the ring is open []. A study of the Cope rearrangement in open ring (Z,E)-germacranolide 1 (7, Fig. 3) and (E,E)-germacranolide 4 (8, Fig. 3) was done in order to analyze the effect of the lactone ring. Fig.
3 shows that energetic differences between cis and trans isomers do not vary significantly. When the lactone ring is closed, this difference is 9.1 kcal/mol, and when it is open it is 8.9 kcal/mol, so the opening of the lactone ring does not affect in any way the relative stability of the isomers. Another possible explanation of the inhibition of Cope rearrangement by the lactone ring (proposed by Takeda) is that the lactone ring raises the Cope TS energy, as the lactone ring strains the germacrane ring. Contrary to what Takeda predicted, the relative energies of opened lactones, TS7-9 (50.8 kcal/mol) and TS8-9 (39.3 kcal/mol), are higher in comparison with their respectively closed lactone, TS1-2 (47.0 kcal/mol) and TS4-2 (35.0 kcal/mol). Elemanolide 9, product from Cope rearrangements of 7 and 8, has a closer energy to (Z,E)-germacranolide than its closed ring analog, 2.
The difference between 1 and 2 is 4.5 kcal/mol while between 7 and 9 is only 2.1 kcal/mol. Thus, the lactone ring destabilizes the elemanolide, which explains why closed ring (Z,E)-germacranolides cannot carry out a Cope rearrangement contrary to the open ring (Z,E)-germacranolides. Again, the relative stability of an elemane versus that of a germacrane determines the likelihood of the transformation. The conclusion is that the γ-lactone ring in the Cope rearrangement destabilizes the corresponding elemane and it has a little or no effect in the Cope TS. This conclusion could be extrapolated to any 5-membered or smaller rings.
The smaller a ring is the more susceptible it is to the strain generated by a second ring (5-members or smaller) fused to it. Therefore, the impact of the γ-lactone ring on the elemane ring (6-members) is significantly more than on the Cope TS ring (10-members).[](https://www.ncbi.nlm.nih.gov/mesh/D007783)
Reaction paths of the Cope rearrangements of closed (dark blue and orange) and open (red and pink) γ-lactone ring. Relative free energies in kcal/mol.[](https://www.ncbi.nlm.nih.gov/mesh/D007783)
## Conclusion
*In the **Conclusion** section:*
The Cope rearrangement is commonly used to determine the germacrane conformation in solution, since the specialized literature establishes that the elemane configuration is due to the most stable conformer of germacrane. However, this is not always true as in the case studied here, where the product observed has neither the configuration of the most stable conformer nor the configuration of the least energetic conformation of the Cope TS. The configuration of elemane 3 is the most stable configuration of this compound. Therefore, it is also important to consider the energy of the different configurations of an elemane to correctly predict the conformation of a germacrene.[](https://www.ncbi.nlm.nih.gov/mesh/D045788)
Interestingly, the (Z,E)-germacranes are significantly more stable than (E,E)-germacranes. Then, cis/trans isomerization can only happen in one way, (E,E)-germacranes → (Z,E)-germacranes. Moreover, this isomerization cannot be thermally activated because of the high energy of the associated TS. (Z,E)-Germacranolides are also more stable than elemanolides, (Z,E)-germacranolides cannot transform into elemanolides unless there is a subsequent reaction that reduces the energy of the elemanolide, like the hemiacetalization does in the case of 3. The transformation studied herein is possible due to a previous hemiacetalization that reduces the energy of the boat transition state by enforcing a shorter distance between the atoms that form the new C–C bond.
Moreover, the transformation (Z,E)-germacranolide → elemanolide is only possible when the lactone ring is open, since in this case the elemanolide is more stable than the (Z,E)-germacranolide. Contrary to what other authors have proposed, the inability of (Z,E)-germacranes to transform into elemanes via a Cope rearrangement when (Z,E)-germacranes have small rings (lactone) fused is not due to an increase of the activation energy of the Cope rearrangement. The activation energy does not change significantly when the fused ring is open or closed, but when the ring is open, the elemane is more stable. A fused small ring produces a lot of strain in the elemanes. In summary, fused small rings increase significantly the energy of elemanes, but those rings do not significantly modify the energy of germacrane’s Cope TSs.[](https://www.ncbi.nlm.nih.gov/mesh/D045788)
## Supporting Information
*In the **Supporting Information** section:*
|
# Introduction
Organic anion transporters, OAT1 and OAT3, are crucial biopterin transporters involved in bodily distribution of [tetrahydrobiopterin](https://www.ncbi.nlm.nih.gov/mesh/C003402) and exclusion of its excess
# Abstract
*In the **Abstract** section:*
Tetrahydrobiopterin (BH4) is a common coenzyme of phenylalanine-, tyrosine-, and tryptophan hydroxylases, alkylglycerol monooxygenase, and NO synthases (NOS). Sy[nthetic BH4 is used](https://www.ncbi.nlm.nih.gov/mesh/C003402) m[edi](https://www.ncbi.nlm.nih.gov/mesh/C003402)cinally for BH4-responsive phenylketonuria and inherited BH4 deficiency. BH4 supplementation has also drawn attention as a therapy for various NO[S-r](https://www.ncbi.nlm.nih.gov/mesh/C003402)elated cardio-vascular di[sea](https://www.ncbi.nlm.nih.gov/mesh/C003402)ses, but its use has met with limited succ[ess](https://www.ncbi.nlm.nih.gov/mesh/C003402) in decreasin[g B](https://www.ncbi.nlm.nih.gov/mesh/C003402)H2, the oxidized form of BH4. An increase in the BH2/BH4 ratio leads to NOS dysfunction. Previous studies revealed that BH4 supplementation caused a rapid urina[ry ](https://www.ncbi.nlm.nih.gov/mesh/C017226)loss of BH4 accompanied[ by](https://www.ncbi.nlm.nih.gov/mesh/C003402) an increase in the b[loo](https://www.ncbi.nlm.nih.gov/mesh/C017226)d[ BH](https://www.ncbi.nlm.nih.gov/mesh/C003402)2/BH4 ratio and an involvement of probenecid-sensitive but unkno[wn ](https://www.ncbi.nlm.nih.gov/mesh/C003402)transporters was strongly suggested in these pro[ces](https://www.ncbi.nlm.nih.gov/mesh/C003402)ses. Here we show that OAT1 and OAT3 enab[led](https://www.ncbi.nlm.nih.gov/mesh/C017226) [cel](https://www.ncbi.nlm.nih.gov/mesh/C003402)ls to take up BP (BH4 and/or [BH2) in a ](https://www.ncbi.nlm.nih.gov/mesh/D011339)probenecid-sensitive manner using rat kidney slices and transporter-expressing cell systems, LLC-PK1 cells and Xenopus oocytes. Both OAT[1 ](https://www.ncbi.nlm.nih.gov/mesh/D001708)an[d O](https://www.ncbi.nlm.nih.gov/mesh/C003402)AT3 pref[err](https://www.ncbi.nlm.nih.gov/mesh/C017226)ed BH2 [and sepiap](https://www.ncbi.nlm.nih.gov/mesh/D011339)terin as their substrate roughly 5- to 10-fold more than BH4. Administration of probenecid acutely reduced the urinary exclusion of endogenous BP a[cco](https://www.ncbi.nlm.nih.gov/mesh/C017226)mpani[ed by a ris](https://www.ncbi.nlm.nih.gov/mesh/C016727)e in blood BP in vivo. These results indicated that [OAT](https://www.ncbi.nlm.nih.gov/mesh/C003402)1 and OAT3 played cr[ucial role](https://www.ncbi.nlm.nih.gov/mesh/D011339)s: (1) in determining baseline levels of blood BP by [ex](https://www.ncbi.nlm.nih.gov/mesh/D001708)cluding endogenous BP through th[e ](https://www.ncbi.nlm.nih.gov/mesh/D001708)urine, (2) in the rapid distribution to organs of exogenous BH4 and the exclusion to urine of a BH4 excess, particularl[y ](https://www.ncbi.nlm.nih.gov/mesh/D001708)when BH4 was administered[, ](https://www.ncbi.nlm.nih.gov/mesh/D001708)and (3) in scavenging blood BH2 by cellular uptake as the gateway to the [sal](https://www.ncbi.nlm.nih.gov/mesh/C003402)vage pathway of BH4, which reduce[s B](https://www.ncbi.nlm.nih.gov/mesh/C003402)H2 back to BH4.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
(6R)-L-erythro-Tetrahydrobiopterin (BH4) is an essential coenzyme of a group of aromatic amino acid hydroxylases. BH4 is also required by nitric oxide synthases (NOSs) both for enzyme catalysis and for functional dimerization. Further, BH4 is the coenzyme of alkylglycerol monooxygenase which catalyzes irreversible ether lipid metabolism. BH4 is autogenously synthesized in various cells which require this compound as the coenzyme. Inherited BH4 deficiencies are characterized by hyperphenylalaninemia and defective biosynthesis of classic monoamines such as dopamine, noradrenaline, adrenaline, as well as serotonin.
BH4 supplementation ameliorates hyperphenylalaninemia in cases of inherited BH4 deficiency and also benefits patients with BH4-responsive phenylketonuria. BH4 therapy using 6RBH4 has been successful in replacing BH4 except in the brain. Once 6RBH4 is administered to animals, presumably including humans, a certain portion is utilized in replacing innate BH4 and is integrated through endogenous metabolic pathways, similar to the intake of vitamins. The most common use of 6RBH4 to date is as an orphan drug for these inherited diseases. However, like most drugs or supplements, its retention in the body is inefficient.
According to the guidelines for BH4 therapy, the recommended dose is 5–15 mg/kg of 6RBH4 a day, roughly 600 mg for a 60 kg adult patient. These doses are over several hundred times greater than the 0.98 mg lost daily in urinary excretion which is likely close to the amount of BH4 synthesized daily in healthy humans. BH4 replacement has also drawn increased attention with respect to whether it ameliorates NOS dysfunction in the cardio-vascular system. NOS dysfunction is largely caused by an increase in BH2 relative to BH4, a type of oxidative stress, rather [than by a deficiency in BH4. One s](https://www.ncbi.nlm.nih.gov/mesh/C003402)im[ple](https://www.ncbi.nlm.nih.gov/mesh/C003402) hypothesis was that supplying BH4 in an amount exceeding the endogenous BH[4 l](https://www.ncbi.nlm.nih.gov/mesh/C003402)evel would pull the redox balance of BH2 and BH4 in favor of a relative BH4 increase.
This hypothesis was likely based [on ](https://www.ncbi.nlm.nih.gov/mesh/C003402)the assumption that the administered BH4 would accumulate in the cell interio[r keeping i](https://www.ncbi.nlm.nih.gov/mesh/D004987)ts tetrahydro[-fo](https://www.ncbi.nlm.nih.gov/mesh/C003402)rm. This approach to ameliorating cardio-vascular symptoms caused by NOS dysfunction has had limited [suc](https://www.ncbi.nlm.nih.gov/mesh/C003402)cess to date. In recent reports of trials on portal hypertension, the authors concluded that “S[apropterin](https://www.ncbi.nlm.nih.gov/mesh/D015306) (6RBH4·2[HCl) mar](https://www.ncbi.nlm.nih.gov/mesh/D004298)ke[dly increased](https://www.ncbi.nlm.nih.gov/mesh/D009638) t[etrahydrob](https://www.ncbi.nlm.nih.gov/mesh/D004837)iopterin (BH4[) levels,](https://www.ncbi.nlm.nih.gov/mesh/D012701) b[ut ](https://www.ncbi.nlm.nih.gov/mesh/C003402)also levels of its oxidized forms, which may counteract its potential ben[efi](https://www.ncbi.nlm.nih.gov/mesh/C003402)cial effects.”[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
In general, an intracellularly functional and hydrophilic compound such as BH4 might be impermeable to the lipid bilayer of the cell membrane. Needless to say, permeation of BH4 across the cell membrane might require appropriate transporters. Although our knowledge of BH4 transport is incomplete, we attempted to clear some of this ambiguity by characterizing the relevant transporter(s) involved in the BH4 transport system. The work addresses three vital areas, A, B, and C, of which the core processes have been unclear due to a lack of knowledge of the crucial transporter(s) involved. A. Short retention of administered BH4.
The extreme inefficiency seen after BH4 administration was likely brought about by the short retention of BH4 in the body caused by its massive exclusion into urine and feces; about a 90% gross urinary exclusion, which exceeded more than 60% of the dose, and took place within 2 h in rats. The rapid exclusion to urine essentially occurred via renal tubular secretion, namely, by trans-cellular transport across the tubular epithelial cell layer. These processes obviously involved a form of high-capacity transporter activity. The process was most probably mediated by at least two transporters acting in series, one engaging in uptake on the vascular side in a manner sensitive to probenecid (PBC) and the other controlling secretion to the urine on the lumenal side in a manner sensitive to cyclosporine A (CSA). B. Increase in BH2 after BH4 administration.
The exogenous BH4 was delivered throughout the body after systemic oxidation to 7,8BH2, and the BH2 was then reduced back to BH4 by the salvage pathway, resulting in the accumulation of BH4 in target organs such as the liver and kidney. Along this line of exploration, we discovered that ENT1 and ENT2, representatives of equilibrative nucleoside transporter families SLC29A1 and SLC29A2, respectively, were appropriate transporters for providing a gateway for sepiapterin (SP) and 7,8BH2 to enter the BH4 salvage pathway, an essential step in cell-to-cell inter-cellular BH4 redistribution. The ENTs were suspected of being inadequate for such high-capacity transport in the kidney after BH4 administration which represents a highly unnatural event. The putative transporters were distinct from the ENTs which are sensitive to nitrobenzylthioinosine (NBMPR) but less sensitive to PBC. In other words, other transporters must be furnished within the kidney to enable it to proceed with the massive uptake of biopterin species (BP), including BH4 and its oxidized form BH2, which are virtually all derived from the administered 6RBH4.
Consequently, they were secreted by the kidney cells to the urine. C. Systemic BH2 scavenging mechanism. In our latest work on the pharmacokinetics of BH4 administration, PBC-sensitive transporter(s) was/were shown to be the key transporter in removing BH2 from the blood. It was also suggested that the PBC-sensitive transporter played a key role in lowering the BH2/BH4 ratio which would otherwise be raised by BH4 administration.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
We searched for a transporter which was able to uptake BH4 in a PBC-sensitive manner in the kidney. Here, we report that OAT1 and OAT3, representatives of the organic anion transporter families SLC22A6 and SLC22A8, are the plausible biopterin transporters engaging in massive BH4 exclusion in the kidney after BH4 administration. These transporters have been localized at the basolateral membrane of renal tubular epithelium, and they are both sensitive to PBC and have characteristic substrate specificities but with an overlapping preference. Their fundamental features as transporters at the molecular level, such as their substrate selectivity, tissue distribution, and localization of their gene expression, have been extensively studied (for reviews). Accordingly, we first examined whether renal epithelial cells were able to take up BH4 using kidney slices.
We then examined whether rOAT1- and rOAT3-expressing LLC-PK1 cells of renal epithelial origin exhibited specific uptake of BH4, BH2 and SP. We confirmed their preference for BH2 and SP, precursors of the BH4 salvage pathway, using a Xenopus oocyte system expressing the respective transporters. Our finding that OAT1 and OAT3 were well suited to acting as a gateway of the salvage pathway of BH4 biosynthesis prompted us to conclude that these transporters played a crucial role in lowering the BH2/BH4 ratio by scavenging BH2 after BH4 administration. As one would expect, the OAT inhibitor PBC, when administered to healthy rats, raised blood BP levels not by BH4 supplementation but by a reduction in the baseline exclusion of endogenous BP in the urine.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## Materials and methods
*In the **Materials and methods** section:*
(6R)-L-erythro-5,6,7,8-Tetrahydrobiopterin dihydrochloride (6RBH4·2HCl) was donated by Suntory (Asubio Pharma, Kobe, Japan) and sepiapterin (SP: 6-lactoyl-7,8-dihydropterin) and 7,8-dihydrobiopterin (BH2) were purchased from Schircks Laboratories (Jona, Switzerland). Methotrexate (MTX) was purchased from Wako (Osaka). Probenecid (PBC: 4-(dipropylsulfamoyl)benzoic acid), p-aminohippuric acid (PAH), penicillin G (PCG), cimetidine (CIM), and estrone sulfate (ES) were obtained from Sigma-Aldrich (St. Louis, MO). Working solutions of hydrophobic chemicals (100-fold concentration over the final concentration) were usually dissolved in an appropriate amount of DMSO and diluted in isotonic salt solution, and the pH of the medium was made neutral if needed. Collagenase (for Xenopus oocyte defolliculation) was purchased from Wako Pure Chemical Industries (Osaka, Japan).[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## BH4 uptake experiments using rat kidney slices
*In the **BH4 uptake experiments using rat kidney slices** section:*
Rats (SD: Sprague–Dawley) were obtained from Japan SLC (Hamamatsu, Japan). Kidney slices were prepared essentially according to Wedeen and Weiner and BH4 uptake studies were carried out as described in Prof. Sugiyama’s laboratory, Tokyo University, Department of Pharmaceutical Science. In brief, slices (0.3 mm thick) of whole kidneys from male rats were put in an ice-cold oxygenated incubation buffer containing 120 mM NaCl, 16.2 mM KCl, 1 mM CaCl2, 1.2 mM MgSO4, and 10 mM NaH2PO4/Na2HPO4 adjusted to pH 7.5. Two slices, together weighing 10–20 mg, were randomly selected and then incubated in a 12-well plate with 1 mL of oxygenated incubation buffer containing 6RBH4 in the presence of 1 mM dithiothreitol and other compounds. After incubation at 37 °C for 15 min, they were then rinsed with ice-cold buffer, blotted, and weighed.
The slices were soaked in 100 µL of 0.1 M HCl and they were then frozen in liquid nitrogen. Biopterin contents were determined the next day after acid- or alkaline-I2 oxidation as described below.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## PBC administration to rats and collection of blood, urine, and liver and kidney tissues
*In the **PBC administration to rats and collection of blood, urine, and liver and kidney tissues** section:*
The experimental procedure using rats was almost the same as described in the previous paper except that the rats did not receive 6RBH4. In brief, rats (SD, 8–10-week-old males) were loaded with PBC (200 mg/kg) under anesthesia. Blood and urine were collected for BP determination from individual rats at designated times under long-lasting anesthesia on a warm gel pad for 6 h. The 0-time samples were taken from rats before the drug administration. At 6 h after PBC dosing, the rats were sacrificed to dissect the liver and kidney for BP determination.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## Uptake by cells under a monolayer culture in 96-well culture plates
*In the **Uptake by cells under a monolayer culture in 96-well culture plates** section:*
LLC-PK1 cells were a generous gift of Dr. Naohiko Anzai, Kyorin University School of Medicine, and rOat1- and rOat3-transfected LLC-PK1 cells were kindly donated by Dr. Hiroyuki Kusuhara (Graduate School of Pharmaceutical Sciences, University of Tokyo). The OAT-transfected cells were maintained as monolayer cultures in Dulbecco’s modified Eagle’s medium (DMEM, GIBCO® Invitrogen) containing 5% fetal calf serum and G418 (400 µg/mL) at 37 °C in 5% CO2/95% air. The non-transfected cells were maintained similarly but without G418. Cells of this naïve cell line were used as the control of the transfected LLC-PK1 cells.[](https://www.ncbi.nlm.nih.gov/mesh/C010680)
LLC-PK1 cells were plated on a 96-well analytical culture plate (Falcon 3072) and grown to a confluence of 4 × 104 cells/well with 200 µL of the culture medium the day before the experiments. Prior to the uptake experiments, cells were adapted to a “basal culture medium” for 15 min. The “basal culture medium” was a modified Hank’s balanced salt solution which consisted of 137 mM NaCl, 5.37 mM KCl, 0.34 mM Na2HPO4, 0.44 mM KH2PO4, 0.34 mM K2HPO4, 5.5 mM glucose, and 5 mM HEPES, pH 7.4. Most transport experiments were conducted with reagents in the basal medium (100 µL) containing 1 mM dithiothreitol. Removal of the culture medium, leaving cells attached to the substrate, was performed by sucking off the medium with an 18-gauge needle (connected to an aspirator) inserted vertically and lightly touching the bottom of the culture plate.
Additional reagents such as pterin substrate (6RBH4, BH2, or SP) with or without other transporter ligands were introduced into the wells following a rinse and replacement with the new medium. Cellular uptake of particular components was allowed for designated times, and the uptake was terminated by a thorough change of medium without the substrate. Cells were then left in the substrate-free medium for 5 min, and subjected to three repeated rinses with ice-cold Ca2+- and Mg2+-containing phosphate-buffered saline (PBS(+)).[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## Expression of hOAT1 and hOAT3 in Xenopus oocytes
*In the **Expression of hOAT1 and hOAT3 in Xenopus oocytes** section:*
African clawed frogs, Xenopus laevis, were purchased from Hamamatsu Seibutsu Kyozai (Hamamatsu, Japan). The gonads were dissected under ice anesthesia and subjected to collagenase treatment (1 mg/mL, 1 h). Mature oocytes were then subjected to manual defolliculation, essentially according to Bianchi and Driscoll.[](https://www.ncbi.nlm.nih.gov/mesh/D007053)
Vectors hOAT1 and hOAT3 were donated by Dr. A. Anzai, Kyorin University, School of Medicine. They were inserted into pcDNA3.1 (Invitrogen). Complementary RNAs (cRNAs) of hOAT1 and hOAT3 were prepared by in vitro transcription with T7 RNA polymerase in the presence of ribonuclease inhibitor and an RNA cap analog using an mMESSAGE mMACHINE kit (Ambion, Austin, TX). Defolliculated oocytes were injected with 50 ng of the respective cRNAs or the same volume of water as the control and incubated in modified Barth’s solution (82.5 mM NaCl, 2 mM KCl, 1 mM MgCl2, and 5 mM HEPES, pH 7.4) at 19 °C for 2 days.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Transport experiment with Xenopus oocytes
*In the **Transport experiment with Xenopus oocytes** section:*
Five oocytes each were placed in 100 µL of ND96 buffer (96 mM NaCl, 2 mM KCl, 1 mM MgCl2, and 5 mM HEPES, pH 7.4) containing 1 mM dithiothreitol in the wells of a 96-well plate at 25 °C for 60 min. Pterin uptake was initiated by replacing the medium with 100 μL of solution containing the desired concentration of the required ligands. The uptake was terminated at designated times by washing the oocytes three times with ice-cold ND96 buffer followed by the addition of 70 μL of acid-I2 or alkaline-I2 solution, as described below, for biopterin analysis. Subsequently, the oocytes were crushed evenly using a plastic rod with a flat tip (5-mm diameter), and allowed to oxidize for 60 min. They were then mixed with 70 μL of 4% ascorbic acid in 4 M HClO4 and cooled on ice for 1 h. Precipitates were removed by centrifugation. A slight turbidity remained and was removed by filtering the supernatant through a 3 mm cotton ball using a yellow-tipped pipette (Gilson) pressed against the bottom of the plastic tube (600-μL Eppendorf-type).
The clear supernatant was then subjected to HPLC analysis. The endogenous biopterin, 0.020–0.025 pmol/oocyte, was disregarded.[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
## Determinations
*In the **Determinations** section:*
Biopterin was determined essentially according to Fukushima and Nixon as described previously. Sepiapterin was not detectable in the extracts of LLC-PK1 or Xenopus oocytes under ordinary conditions. Even after SP was supplied, it was negligibly small in amount, presumably due to the high endogenous activity of sepiapterin reductase. In this study, therefore, uptake of SP and BH2 by LLC-PK1 cells or Xenopus oocytes was determined indirectly using the amount of biopterin present after the acidic oxidation, i.e., the sum of BH2 and BH4.[](https://www.ncbi.nlm.nih.gov/mesh/D001708)
The BH4 uptake by a given biological sample was expressed as the clearance, using the distribution volume (V d): where[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
In the case of BP uptake by the kidney slices at 15 min in the presence of extracellular BP, [BP](out) (µM), the BP uptake was expressed as:[](https://www.ncbi.nlm.nih.gov/mesh/D001708)
For the BP uptake by LLC-PK1 cells in the presence of extracellular BP, [BP](out) (µM) for example, the BP uptake per hour per well of confluent cells was expressed as:[](https://www.ncbi.nlm.nih.gov/mesh/D001708)
For the Xenopus oocytes adapted to the experimental procedure (5 cells per assay):
## Statistics
*In the **Statistics** section:*
Statistical significance was analyzed by Student’s t-test or Williams’ test. The significance of difference between determinations at different times with individual animal groups was analyzed by a paired t-test. The significance of difference between three groups was analyzed by Holm’s test. All data were statistically analyzed using Pharmaco Basic Ver.15.0.1 (Scientist Co. Ltd., Tokyo).
## Results
*In the **Results** section:*
## BH4 uptake by rat kidney slices
*In the **BH4 uptake by rat kidney slices** section:*
Uptake of 6RBH4 by kidney slices in the presence or absence of typical ligands of organic anion transporters. The kidney slices were prepared as described in “Materials and methods” section. The slices took up 6RBH4 and the uptake was inhibited by OAT ligands. The reagents used were 6RBH4 (10 µM and 3 mM), penicillin G (PCG, 1 mM), probenecid (PBC, 1 mM) and p-aminohippuric acid (PAH, 1 mM). The uptake of BH4 for 15 min was expressed as a portion of the clearance. *P < 0.05, **P < 0.01 (Holm’s test); each point represents the mean ± S.D. (n = 3–7)[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
The rat kidney slices taken from the cortex area contained 2.67 ± 0.78 nmol of BP per mg tissue weight, ca. 4-fold more than the whole kidney average (0.68 ± 0.11 nmol/g, P < 0.01). In the uptake experiment, the BP content of the slices increased almost linearly for more than 20 min, 79.5 ± 32.2 nmol/mg at 15 min in the presence of 10 µM 6RBH4, and the clearance was calculated to be 8.7 ± 3.2 µL/(15 min·mg) as depicted in Fig. 1. At a very high concentration of 6RBH4 (3 mM), the uptake was significantly decreased (P < 0.01), suggesting that the process was saturable with regard to BH4, consistent with the carrier-mediated process but not with physicochemical diffusion. Moreover, the process was inhibited by a group of organic anion transporter ligands; namely, through strong inhibition by PBC (80%, P < 0.01) and moderate inhibition by PCG (40%, P = 0.040), a preferred substrate of OAT3, or PAH (50%, P = 0.024), a preferred substrate of OAT1.[](https://www.ncbi.nlm.nih.gov/mesh/D001708)
Kidney slices are known to be one of the best functional systems in vitro with regard to their characteristic uptake which takes place exclusively at the basolateral membrane due to occlusion at the cut ends of the tubular cross section. Accordingly, BH4 uptake by the slices in this manner was accounted for by the basolateral uptake of tubular epithelium. Furthermore, the 80% suppression by PBC strongly suggested that the relevant transporters involved in the BH4 uptake were mostly the organic anion transporters OAT1 and OAT3 which were reported to participate in the removal of various water-soluble compounds in plasma. Although the involvement of other transporters in BH4 removal was not ruled out, the near exclusive participation of the PBC-sensitive transporters in the removal was suggested, and it might be a prerequisite for the rapid release of these compounds to the urine at proximal tubules in rats as observed previously.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## BH4 uptake by rOat1- or rOat3-transfected LLC-PK1 cells
*In the **BH4 uptake by rOat1- or rOat3-transfected LLC-PK1 cells** section:*
Biopterin uptake by rOat1- and rOat3-expressing LLC-PK1 cells and naïve LLC-PK1 cells. Tetrahydrobiopterin uptake by OAT-expressing or naïve LLC-PK1 cells was examined under a monolayer culture (4 × 104 cells/well, 96-well analytical culture plate). a LLC-PK1 cells transfected with rOat1 (left) or with rOat3 (right) were used. The cells were exposed for 1 h to 50 µM 6RBH4 in the absence (gray bars, control labeled “none”) or in the presence (open bars) of OAT ligands (1 mM each except for methotrexate (MTX) at 80 µM). The rOat1- and rOat3-expressing LLC-PK1 cells took up 6RBH4 and the uptake was inhibited by the ligands of OAT1 and OAT3. The OAT ligands used were probenecid (PBC), estronesulfate (ES), p-aminohippuric acid (PAH), penicillin G (PCG), methotrexate (MTX) and cimetidine (CIM). b rOat1-LLC-PK1 cells (left) or rOat3- LLC-PK1 cells (right) were given 50 µM each of 6RBH4, dihydrobiopterin (BH2) or sepiapterin (SP) in the absence (gray bars) or presence (open bars) of 1 mM PBC for 1 h. The resultant biopterin accumulations of BH2 + BH4 were then compared between those in the absence of PBC vs the presence of PBC, and levels of BH4
vs. BH2 and of BH2
vs SP. c Uptakes of 6RBH4 and BH2 (50 µM each) by naïve LLC-PK1 cells were also compared (left panel). The uptake of 6RBH4 (50 µM) for 1 h (right panel) was analyzed in the absence (gray bar, labeled “none”) or presence (open bars) of 200 µM nitrobenzylthioinosine (NBMPR) or 1 mM PBC. The uptake of the pterins was expressed as a portion of the clearance. *P < 0.05, **P < 0.01 (Holm’s test); ††
P < 0.01 (Student’s t-test); each point represents the mean ± S.D. (n = 5–6)[](https://www.ncbi.nlm.nih.gov/mesh/D001708)
Transport of BH4 by OAT1 and OAT3 and of BH2 and SP was characterized using rOat1- or rOat3-transfected LLC-PK1 cells, a cell line derived from the kidney of a male pig. As shown in Fig. 2a, rOat1-LLC-PK1 and rOat3-LLC-PK1 cells took up 6RBH4, and these processes were inhibited by known ligands of OAT1 and OAT3. The BH4 uptake was strongly inhibited in rOat1-LLC-PK1 cells by the typical ligands of OAT1, PAH and cimetidine. On the other hand, BH4 uptake by the rOat3-LLC-PK1 cells was inhibited by the ligands of OAT3, ES and PCG. These results are consistent with the reported preference of the respective compounds for rOat1 and rOat3.
The respective uptakes of 6RBH4, BH2 and SP by these cells were compared in the presence or absence of PBC (1 mM, Fig. 2b) to ensure that the PBC-sensitive portions of the pterin uptake were distinct from the other processes. Both rOat1-LLC-PK1 and rOat3-LLC-PK1 cells took up BH2 and SP, the precursors of the BH4 salvage pathway, in a PBC-sensitive manner and much more efficiently than their uptake of 6RBH4. The uptake of all three pterins in the presence of PBC was minor and it may have been mediated by transporters other than OAT1 or OAT3. Naïve LLC-PK1 cells were capable of taking up 6RBH4 to a lesser extent than the above transfected cells but they were insensitive to PBC. The naïve cells took up BH2 in preference to BH4 (Fig.
2c, left). The endogenous BH4 uptake by these cells was around 20.7 ± 3.60 nL/(h·4 × 104 cells) and was significantly inhibited by nitrobenzylthioinosine (NBMPR), a typical ENT ligand (P < 0.01) but not by PBC (Fig. 2c, right) suggesting that the uptake was mainly mediated by ENT1, ENT2 as described previously. Hence, our observations of the pterin uptake in the presence or absence of PBC using OAT-transfected LLC-PK1 cells, as shown in Fig. 2b, enabled us to distinguish between the endogenous transporters and those expressed as a result of the transfection.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## Transport of BH4, BH2, and SP by hOAT1- or hOAT3-expressing Xenopus oocytes
*In the **Transport of BH4, BH2, and SP by hOAT1- or hOAT3-expressing Xenopus oocytes** section:*
Uptake of 6RBH4, BH2, and sepiapterin by hOAT1- and hOAT3-expressing Xenopus oocytes. Xenopus oocytes were individually injected (gray bars) with 50 nL of hOAT1 (a) or hOAT3 (b) cRNA (1 ng/nL). As the control (hatched bars), oocytes were injected with 50 nL of distilled water. The oocytes were then allowed to express the respective transporters at 19 °C for 2 days. In the uptake experiment, all pterins were used at 50 µM. All oocytes injected with either cRNA took up significantly more pterins than the control and the uptake was inhibited by OAT ligands (1 mM each). The OAT ligands used were probenecid (PBC), p-aminohippuric acid (PAH), estronesulfate (ES) and penicillin G (PCG). a 6RBH4, BH2, and sepiapterin (SP) were taken up by hOAT1-expressing oocytes (main panel) for 1 h, and the BH2 uptake was analyzed in the absence (gray bar, labeled “none”) or presence (open bars) of OAT1 ligands (upper panel). (b) Uptake of 6RBH4, BH2, and SP by hOAT3-expressing oocytes (main panel) and inhibition of BH4 uptake by OAT3 ligands in the absence (gray bar, labeled “none”) or presence (open bars) of OAT3 ligands (upper panel). The uptake of the pterins was expressed as a portion of the clearance. *P < 0.05, **P < 0.01 (Holm’s test); each point represents the mean ± S.D. (n = 4–9)[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
Xenopus oocytes were separately injected with the cRNA of hOAT1 or hOAT3 and allowed to express the respective transporters at 19 °C for 2 days, as described in Materials and Methods. We observed an increased uptake of 6RBH4, BH2, and SP by all oocytes injected with the respective cRNAs, as shown in Fig. 3. The oocytes injected with hOAT1-cRNA showed the most pronounced uptake of BH2 followed by that of SP (Fig. 3a). Enhancement of 6RBH4 uptake was rather moderate when compared to that of BH2 or SP.
The BH2 uptake was significantly inhibited by PBC (P < 0.01) and PAH (P < 0.01), both typical OAT1 substrates. In hOAT3-expressing oocytes, SP uptake was predominant followed by BH2 uptake (P < 0.01), while BH4 uptake was poorly enhanced, even when compared to BH2 (P < 0.01) (Fig. 3b). Nonetheless, the BH4 uptake was inhibited by OAT3 ligands, suggesting that it was also mediated by the hOAT3 expression product. Considering these results together with those obtained using rOAT-transfected LLC-PK1 cells, OAT1 of either rats or humans mediated the uptake of BH2 better than that of SP, and OAT3 mediated uptake of SP more than that of BH2. The pronounced preference for the dihydropterins 7,8BH2 and SP compared to BH4 seemed to be common to both OAT1 and OAT3 (cf.
Figure 2b).[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## Inhibition of OAT1 and OAT3 by probenecid and a decrease in bodily exclusion of endogenous BP in the rat
*In the **Inhibition of OAT1 and OAT3 by probenecid and a decrease in bodily exclusion of endogenous BP in the rat** section:*
Elevation of blood BP accompanied by a decrease in the urinary loss of endogenous BP with a single dose of probenecid. Rats were given probenecid (PBC, 200 mg/kg, i.p.), and the blood (a) and urine (b) were collected sequentially from individual rats at the indicated times under sustained anesthesia for 6 h, then the kidney and liver (c) were dissected from the same rats at 6 h after the PBC dosing. The 0-time samples were taken from the rats without PBC treatment. The 0-time amounts of BH2 + BH4 (BP, open symbols) were compared with those of PBC-treated rat samples (grey symbols). In a and b, *P < 0.05, **P < 0.01 (“0-time” vs. “PBC-treated”, paired Student’s test), and in c, **P < 0.01, or n.s., no significant difference (“before” vs. “PBC-treated”, Williams’ test). Data are mean ± S.E. (n = 4)[](https://www.ncbi.nlm.nih.gov/mesh/D001708)
In order to confirm that PBC prevents urinary exclusion of endogenous BH4, which was originally synthesized systemically de novo, by inhibiting the representative transporters, OAT1 and OAT3, we compared BP levels before and after PBC treatment in the circulating blood, in urine in the bladder, and in tissues of the kidney and liver. Rats were injected with probenecid (200 mg/kg, i.p.) under sustained anesthesia without BH4 administration. BP levels in the blood were gradually elevated to 1.6- and 1.8-fold within 6 h after PBC administration (Fig. 4a, P < 0.01). With the same rats, significant decrease was observed in the urinary BP contents, normalized with time-matched creatinine; and levels were about 30% less than those of the untreated rats (Fig. 4b, P < 0.05). Moreover, the tissue BP in the kidney of the same rats increased to 4.6-fold over the initial value (Fig. 4c, P < 0.01).
Meanwhile, we did not observe any significant change in endogenous BP in the liver after the PBC treatment.[](https://www.ncbi.nlm.nih.gov/mesh/D011339)
## Discussion
*In the **Discussion** section:*
In earlier studies, we noted a facilitated clearance of BP by the kidney, particularly after 6RBH4 administration, and we demonstrated that it was due to the tubular secretion of exogenous BH4, distinct from removal by renal glomerular filtration. In our in vivo experiments analyzing the effect of PBC on the pharmacokinetics of administered 6RBH4 in rats, we showed that a PBC-sensitive transporter(s) enabled the liver and kidney to play their crucial role in bodily retention of BH4 and tubular secretion to the urine. Organic anion transporters (OATs) are representative PBC-sensitive transporters known to participate in the uptake process in certain tissues such as the kidney and liver and to exhibit a particular localization. The role they play in kidney clearance of various xenobiotics and metabolic wastes has been well studied (reviews). We previously demonstrated that ENT1 and ENT2 were capable of transporting SP, BH2 and BH4, and that both were relevant as a gateway of the BH4 salvage pathway.
ENTs comprise a family of equilibrative transporters that mediate the bidirectional permeation of nucleobases and similar heterocyclic compounds across biological membranes, indicating that BH4 and nucleotides share a common gateway for the salvage pathway. Their near ubiquitous distribution, including in endothelial cells, seemed to be appropriate for their body-wide role in biopterin distribution; however, more tissue-specific localization of vigorous transporters was expected for massive kidney-specific clearing. We therefore looked for other transporters which could play a major role in mediating biopterin permeation and clearance by the kidney. Accordingly, we focused our present search on tubular epithelium in examination of biopterin exclusion.[](https://www.ncbi.nlm.nih.gov/mesh/D001708)
## OAT1 and OAT3 as biopterin transporters
*In the **OAT1 and OAT3 as biopterin transporters** section:*
Our first clue to uncovering the transporters responsible for biopterin uptake came from using kidney slices which took up BH4 and finding that this uptake was strongly inhibited by PBC together with other ligands of OAT1 and/or OAT3. It is also known that uptake by the basolateral side of tubular epithelium can be exclusively elicited in vitro using rat kidney slices. Hence, we hypothesized that the transporters OAT1 and OAT3 were responsible for driving the renal exclusion of biopterin after systemic administration of 6RBH4. In order to prove this hypothesis, we employed an expression system using LLC-PK1 cells transfected with the rat OAT genes rOat1 and rOat3 following a method which established the functionality of these transporters in the uptake required for the exclusion of organic anions, nucleobases and nucleosides. As a result, the ability of these transporters was strongly suggested to mediate the cellular uptake of BH4, BH2, and SP.
One uncertainty in this experimental system was that the naïve LLC-PK1 cells were able to take up BP to a noticeable extent. However, this fraction of BP uptake was thought to be mediated by other transporters such as the ENTs which are rather ubiquitous and essential for proliferation of cells in culture owing to their fundamental role as a gateway of the nucleotide salvage pathway. The ability of OAT1 and OAT3 to transport BH4, BH2 and SP was further confirmed by the uptake experiment using Xenopus oocytes expressing the human OAT genes hOAT1 and hOAT3. Although OAT1 and OAT3 both have a strong ability to transport BH4, BH2, and SP, this does not necessarily infer that they are the proprietary transporters in BP uptake. Instead, we consider that BP in plasma was targeted as a xenobiotic or metabolic waste to be eliminated by the kidney.
The present result does not exclude the possible relevance to BP permeation of other transporters not examined here.[](https://www.ncbi.nlm.nih.gov/mesh/D001708)
## Relevance of OAT1 and OAT3 in systemic BH4 metabolism
*In the **Relevance of OAT1 and OAT3 in systemic BH4 metabolism** section:*
As for the body-wide relevance of OATs in BH4 metabolism, we observed a major movement of administered BH4, including massive accumulation in the liver and rapid exclusion from the kidney. Notably, in the liver and kidney, these processes were both inhibited by prior treatment with PBC. Based on these observations, we considered that the PBC-sensitive transporter(s) played a crucial role in enabling the liver to absorb and retain a large amount of BP and the kidney to take it up and exclude it in the urine. Renal trans-epithelial transport is composed of tandem permeations across the cell membrane, namely, uptake from the vascular side and release from the tubular lumen side, in which the former was strongly prevented by PBC and the latter, by CSA. In the present study, OAT1 and OAT3 were identified as the major PBC-sensitive transporters for BP exclusion.
However, if they act too vigorously to exclude BP, this would raise concern that BP could be thoroughly removed from the body. In addressing this issue, we previously examined BP levels in urine and plasma separately from red blood cells. We noted that most of the plasma BP, whose level rose sharply after 6RBH4 administration, was rapidly and massively excluded in the urine by trans-cellular BP secretion which far exceeded glomerular filtration but only until the point at which the plasma BP had decreased to about 1 nmol/mL, 10-fold the endogenous level, and the level remained higher than the endogenous level for a period of hours. Considering the exclusion dynamics of exogenous BP, these transporters likely play the core role in removing extraordinarily high concentrations of plasma BP as a sort of protective mechanism. Although the transporter ligands CSA and PBC strongly blocked the exclusion of BP across the epithelium to the urine, they raised the blood BP to a very high concentration while they had no effect on the glomerular filtration of BP.
The ligand treatment, therefore, did not improve the efficiency of BP replacement because the inhibition of the trans-cellular BP passage was offset by a more rapid outflow through glomerular filtration due to a compensatory elevation of plasma BP. Despite the “low efficiency” of peripherally administered 6RBH4, this supplement can provide enough BH4 cofactor for pterin-dependent hydroxylases in peripheral organs, such as the liver, in patients with an endogenous BH4 deficiency. 6RBH4 administration immediately elevates plasma BP levels. We consider that an extreme concentration higher than threshold was targeted for biological detoxification by means of the tubular secretion in which OAT1 and OAT3 played the essential role. Hence, the “low efficiency” in BH4 supplementation was a consequence of the blood BP elevation over the critical level, which was 10-fold the endogenous level in the case of rats, in the early period after 6RBH4 administration.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
## Remarks on the relevance of OAT1 and OAT3 for systemic BH2 scavenging after BH4 supplementation
*In the **Remarks on the relevance of OAT1 and OAT3 for systemic BH2 scavenging after BH4 supplementation** section:*
Contrary to general expectations, the administration of 6RBH4 initially caused an increase in BH2 in the circulation. However, the elevated BH2 ratio declined within a short time and dropped even further to a level lower than the initial level. An initial surge in BH2 and a sharp elevation of the BH2/(BH2 + BH4) ratio was first observed in the blood in a BH4 replacement experiment in which mice were intraperitoneally administered 6RBH4, 7,8BH2 or SP. A rather long-lasting BH2 increase was also observed in rat plasma after intravenous administration of 6RBH4, subsequently, the elevated BH2 ratio in the plasma gradually returned to normal over a period of hours. Presumably, the increased plasma BH2 had been removed by its selective uptake and was subsequently converted to BH4.
This consequence was paradoxically illustrated by the finding that treating the rat with PBC prior to 6RBH4 administration strongly stimulated the BH2 increase in the blood and urine while it attenuated the BP increase in the liver and simultaneously retarded urinary BP excretion.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
We have demonstrated in this work that both OAT1 and OAT3 drove the cellular uptake of the dihydro-forms of pterins, BH2 and SP, in roughly a 5- to 10-fold preference over the tetrahydro-form, BH4. The liver and kidney are undoubtedly well furnished with a BH4 salvage pathway as shown by the fact that the tissue BH2 ratio, BH2/(BH2 + BH4), in these organs did not significantly change even after the massive uptake of BH2 from the plasma after 6RBH4 administration. BH2 uptake is a prerequisite of BH2 conversion to BH4 in the liver and of BH2 removal to the urine by the kidney. Furthermore, the strong preference of OAT1 and OAT3 for BH2 over BH4 made it much more efficient for these organs to perform their respective functions. Based on these reports together with the present observation regarding the transporter preference for BH2, we consider that OAT1 and OAT3 work as part of a body-wide machinery for scavenging BH2 and maintaining the redox balance, at least after BH4 administration.
With regard to phenylalanine-, tyrosine-, and tryptophan hydroxylases, the relative increase in BH2 does not interfere with the enzyme activity to any significant extent, however, it does lead to a critical failure in NOS function.[](https://www.ncbi.nlm.nih.gov/mesh/D011622)
Administration of 6RBH4 has been attempted as a translational medicinal approach to improving eNOS dysfunction, especially in cardio-vascular disorders. The dysfunction of eNOS is characterized by uncontrolled uncoupling of O2 reduction producing “reactive oxygen/nitrogen species” (ROS/RNS), and it is believed to be a great risk factor for cardio-vascular disease. Various encouraging experimental studies have been reported, however, these attempts have had limited success in trials to ameliorate human cardio-vascular dysfunction. The binding of BH2 to active NOS causes an uncoupling of the O2 reduction involved in the enzyme reaction. To avoid sustained eNOS dysfunction, the enzyme should not be exposed to a high concentration of BH2 relative to BH4 because the affinity of NOS for BH2 was reported to be not particularly low; the IC50 of BH2 to eNOS was comparable to that of BH4 and the Ki of nNOS to BH2 was only 10-fold higher than to BH4.
Hence, the success of 6RBH4 administration for improving NOS dysfunction might primarily depend on its lowering of the BH2/(BH2 + BH4) ratio rather than on raising the BH4 concentration. This has been noted by many researchers. As two examples, (1) the importance of the BH4 salvage pathway driven by dihydrofolate reductase has been well documented and the scavenging of BH2 was shown to be enabled by vigorous reductase activity. The authors argued that endothelial cells were poorly furnished with dihydrofolate reductase, which made NOS function in these cells extremely vulnerable to BH2 produced in situ in the cell interior. It was also reported that (2) endothelial cells of human origin showed dramatically less dihydrofolate reductase activity compared to cells of other species including cows and mice.
Taken together, endothelial cells did not appear to have the means to scavenge BH2 in the local cell interior. Nonetheless, blood BH2 can be produced either near or at some distance away from local endothelial cells. In addition, endothelial cells readily take up plasma BH2 because of their asymmetrically dense expression of ENT2 on the apical membrane facing the vascular lumen side. As an extreme example, after BH4 administration, eNOS of endothelial cells is vigorously exposed to a high BH2 concentration from the circulating blood. In this context, attenuation of systemic BH2 production or stimulation of BH2 scavenging has a direct impact on the eNOS dysfunction of endothelial cells.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
This work has demonstrated that OAT1 and OAT3 are both involved in systemic BH2 scavenging in terms of their characteristic mediation of BH2 uptake and their known distribution on the vascular side of renal tubular endothelium. Effective BH2 scavenging in the kidney is obviously enabled by the high-capacity uptake of blood BH2 mediated by OAT1 and OAT3. Although OAT3 is the dominant OAT expressed in the liver, other OAT counterparts in this organ, such as OAT2, are also interesting but their roles remain elusive.[](https://www.ncbi.nlm.nih.gov/mesh/C017226)
## Relevance of PBC-sensitive transporters in exclusion of endogenous BP in the urine
*In the **Relevance of PBC-sensitive transporters in exclusion of endogenous BP in the urine** section:*
As a matter of general physiology, we are also interested in the relevance of OAT1 and OAT3 in the homeostasis of BH4 metabolism under ordinary conditions without exogenous BH4 supplementation. Our experiments addressed this issue (Fig. 4) and demonstrated that the administration of PBC to healthy rats significantly decreased BP exclusion in the urine which was accompanied by a rise of BP in the blood, with both levels measured using the same individual rats. Since there are no known reasons for the drug to stimulate de novo BP biosynthesis, the increase in the blood BP was accounted by a kickback from inhibition of the renal secretion of BP by PBC. These results suggested that the loss of BP through the urine strongly influences the endogenous amount of BP in the blood. We also observed that the BP content in the kidney was significantly increased in the presence of PBC.
An almost proportional increase in kidney BP relative to blood BP was even more pronounced in our previous study in which rats were administered “6RBH4 alone” or “6RBH4 + PBC”. The mechanism of this in vivo increase involving local OAT1 and/orOAT3 remains unclear. Since OAT1 and OAT3 are the only known PBC-sensitive transporters which enable BP to permeate across the cell membrane, it is highly plausible that OAT1 and OAT3, if not acting locally in the kidney, are involved in the mechanism controlling the bodily retention of BH4. However, this subject awaits further in vivo exploration.[](https://www.ncbi.nlm.nih.gov/mesh/C003402)
As mentioned at the beginning of the Discussion, ENT1 and ENT2 have been identified as BP transporters. ENTs and OATs are similar in their preference for SP and BH2 as substrates rather than BH4. However, in contrast to the localization of OATs in specific tissues such as in the kidney and liver, ENTs are characterized by their near ubiquitous distribution including in endothelial cells. Generally speaking, it is likely that ENT1 and ENT2 assume their share of the mobilization of BP, including its precursor SP, in the body interior, while OAT1 and OAT3 deal with detoxication in response to a BP excess and act in removal of this pterin from the body.[](https://www.ncbi.nlm.nih.gov/mesh/C016727)
# Compliance with ethical standards
*In the **Compliance with ethical standards** section:*
## Conflicts of interest
*In the **Conflicts of interest** section:*
The authors declare that they have no conflicts of interest.
## Research involving animals and human rights
*In the **Research involving animals and human rights** section:*
The animal experiments were conducted in accordance with the ethical guidelines of the Teikyo University of Science and Technology Animal Experimentation Committee and the guidelines of the Japanese Pharmacological Society.
# References
*In the **References** section:*
|
# Introduction
A [Water](https://www.ncbi.nlm.nih.gov/mesh/D014867)-Bridged [Cysteine-Cysteine](https://www.ncbi.nlm.nih.gov/mesh/C046557) Redox Regulation Mechanism in Bacterial Protein Tyrosine Phosphatases
# Abstract
*In the **Abstract** section:*
Summary
The emergence of multidrug-resistant Mycobacterium tuberculosis (Mtb) strains highlights the need to develop more efficacious and potent drugs. However, this goal is dependent on a comprehensive understanding of Mtb virulence protein effectors at the molecular level. Here, we used a post-expression cysteine (Cys)-to-dehydrolanine (Dha) chemical editing strategy to identify a water-mediated motif that modulate[s access](https://www.ncbi.nlm.nih.gov/mesh/D003545)ib[ili](https://www.ncbi.nlm.nih.gov/mesh/D003545)ty of[ the protein ](https://www.ncbi.nlm.nih.gov/mesh/C015102)ty[ros](https://www.ncbi.nlm.nih.gov/mesh/C015102)ine phosphatase A (PtpA) catalytic pocket.[ Impo](https://www.ncbi.nlm.nih.gov/mesh/D014867)rtantly, this water-mediated Cys-Cys non-covalent motif is also present in the phosphatase SptpA from Staphylococcus aureus, w[hich ](https://www.ncbi.nlm.nih.gov/mesh/D014867)suggests a[ potent](https://www.ncbi.nlm.nih.gov/mesh/C046557)ially preserved structural feature among bacterial tyrosine phosphatases. The identification of this structural water provides insight into the known resistance of Mtb PtpA to the oxidative conditions that prevail within an [infec](https://www.ncbi.nlm.nih.gov/mesh/D014867)ted host macrophage. This strategy could be applied to extend the understanding of the dynamics and function(s) of proteins in their native state and ultimately aid in the design of small-molecule modulators.
Graphical Abstract
Highlights
Chemical editing
Cys-H2O-Cys bridge
Insights on resistance to oxidative inactivation
The Bigger Picture
The emergence of Mycobacterium tuberculosis (Mtb) resistance is a serious threat to public health. However, the quest for more efficient drugs against Mtb is hampered by the lack of a detailed understanding of Mtb virulence protein effectors. Here, we describe the swift modification of select Cys residues in multi-Cys proteins directly through chemistry. New insights into the biochemistry of emerging bacterial drug targets were obtained. We reveal a water Cys-Cys bridging mechanism that offers an explanation for the known resistance of Mtb protein tyrosine phosphatase A (PtpA) to the oxidative conditions that prevail within an infected host macrophage. This water Cys-Cys bridge motif is also found in the phosphatase SptpA from Staphylococcus aureus, suggesting its potential conserved structural role. The rationalization of the unique features of PtpA, an important target for Mtb drug discovery, could now be used in the design of novel small-molecule modulators.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Using a post-expression chemical editing strategy, Bernardes and colleagues have identified a water-mediated Cys-Cys non-covalent motif in bacterial tyrosine phosphatase A (PtpA) from Mycobacterium tuberculosis (Mtb) and Staphylococcus aureus. Importantly, the identification of the Cys-water-Cys bridge provides insight into the known resistance of Mtb PtpA to the oxidative conditions that prevail within an infected host macrophage. This chemical mutagenesis approach could help the understanding of the dynamics and function(s) of proteins in their native state and ultimately aid in the design of small-molecule modulators.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Regioselective Cys Chemical Editing[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
(A) Catalytic mechanism of protein tyrosine phosphatases.
(B) Cys-to-Dha conversion through a bisalkylation elimination reaction.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
(C) Left: conversion of PtpA-Cys53 to Dha with reagent 1. Right: conversion of all Cys residues in PtpA to Dha by treatment with an excess of 1.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Tuberculosis affects millions of people each year and is a leading cause of deaths worldwide. The emergence of multidrug-resistant Mycobacterium tuberculosis (Mtb) strains is linked to the ability of Mtb to overcome host defenses, especially macrophage digestion and overoxidation,, pressuring the long-standing endeavor of disease eradication. Once inside the macrophage vacuole, Mtb circumvents the proteolysis machinery by inhibiting phagosome maturation and its fusion with lysosome. Among others, protein tyrosine phosphatase A (PtpA) is a key player for Mtb survival in this oxidative environment. PtpA is secreted into the macrophage cytosol and interferes directly with phagosome maturation by disrupting key components of the macrophage endocytic pathway.,
However, as macrophages produce reactive oxygen and nitrogen species as a defense mechanism against Mtb, proteins, including PtpA, are likely to be inhibited under oxidative conditions. Protein tyrosine phosphatases (PTPs) contain multiple Cys residues that play a paramount role regulating signaling pathways (Figure 1A)., , The formation of a disulfide bridge between the catalytic Cys and a backdoor Cys residue located within the catalytic pocket is a structural feature that can finely control the redox mechanism of PTPs. However, such regulating mechanism(s) that delay oxidative inactivation remain elusive for PtpA.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
Typically, the relationship between amino acid sequence and protein activity and function is determined through site-directed mutagenesis. However, this technique is restricted to the introduction of 20 canonical amino acid building blocks. On the other hand, semi-synthesis of proteins via native chemical ligation, followed by refolding is limited to simpler proteins. Alternatively, site-selective chemical mutagenesis offers an expeditious and elegant means of studying native, folded proteins by post-expression installation of non-canonical amino acid residues., , , These could allow activity and functional studies per se or act as chemical tags for subsequent functionalization. Here, we describe our efforts to leverage site-selective post-expression mutagenesis by using non-canonical amino acids in order to understand the interplay between multiple Cys residues and their role in redox regulation mechanisms displayed by bacterial PTPs.
Using Mtb PtpA, Yersinia enterocolitica tyrosine phosphatase YopH, and Staphylococcus aureus tyrosine phosphatase SptpA as examples of PTPs and a post-expression chemical editing method for converting Cys to dehydroalanine (Dha), we obtained exquisite regioselectivity. Unexpectedly, we unveiled that the protein's solvation state regulates its reactivity toward modification by the chemical editing reagent. Our findings illustrate the importance of a structural water and the reactivity of the non-catalytic Cys53 residue as a protection mechanism against catalytic oxidation in PtpA. Indeed, the role of water molecules and water networks is central to understanding the hydrophobic effect, protein function, and molecular recognition in general., , , All together, our data offer a rationale for Cys oxidation mechanisms by xenobiotic species and offer insights into new biology that can be used for designing innovative antimicrobial PTP-targeting chemical probes and therapeutic agents.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## Results and Discussion
*In the **Results and Discussion** section:*
## Regioselective Cys53 Modification Indicates Its Enhanced Reactivity and Its Role as an Oxidant Scavenger
*In the **Regioselective Cys53 Modification Indicates Its Enhanced Reactivity and Its Role as an Oxidant Scavenger** section:*
In contrast to current approaches that rely on extensive protein sequence remodeling, we investigated PtpA function and dynamics by post-expression conversion of Cys to Dha by using the editing reagent α,α′-di-bromo-adipyl(bis)amide, 1 (Figure 1B). Dha provides an ideal mutation to study and understand protein dynamics because of its small size and possible use as a tag for functionalization or as a spectroscopic probe (C–H stretching).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Chemical Mutation of PtpA and Its Effect on Catalytic Activity
(A) Deconvoluted ESI-MS spectra overlay of native and Cys-to-Dha-modified PtpA isoforms. Reactions were carried out with 1 in NaH2PO4 buffer (pH 8.0, 50 mM) for 1 hr at 37°C at the indicated concentrations. Mass peak assignments: 19,924 Da, Cys11/16/53; 19,889 Da, Cys11/16/Dha53; 19,820 Da, Dha11/16/53.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
(B) Dha53-containing peptide 40-VTSAGTGNWHVGSDhaADER-57 obtained upon treatment of wild-type PtpA with 15 mM of reagent 1.[](https://www.ncbi.nlm.nih.gov/mesh/C015102)
(C) Catalytic activity profile of PtpA and isoforms, as assessed by p-nitrophenylphosphate hydrolysis over time.[](https://www.ncbi.nlm.nih.gov/mesh/C008644)
(D) Circular dichroism spectra of PtpA, site-directed single-mutant C53A, and chemical mutant Dha53. Protein samples were concentrated to 10 μM in 25 mM NH4HCO3 (pH 7.4).[](https://www.ncbi.nlm.nih.gov/mesh/C015102)
(E) Atomic fluctuation (Cα) analysis of PtpA wild-type (left) and triple-mutant (right) obtained from 500 ns MD simulations. The data correspond to the average structure of both molecules throughout the simulations.
(F) H2O2 inactivation profile of PtpA, Dha53, and site-directed mutant C53A, as assessed by p-nitrophenylphosphate hydrolysis over time. PtpA and Dha53 mutant (0.5 μM) were pre-incubated with 100 μM H2O2 for 1 min, then 20 mM p-nitrophenylphosphate was added to the reaction and absorbance of the released p-nitrophenolate was monitored at 410 nm over 10 min. The data represent mean ± SD of three independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D006861)
We reacted Mtb PtpA and its alanine (Ala) and serine (Ser) mutants with 1 at varying concentrations (5–60 mM) at pH 8.0 and 37°C for 1 hr (Figures 1C and S1). Inspection of the deconvoluted electrospray ionization mass spectrometry (ESI-MS) spectra for the wild-type and mutant PtpA counterparts revealed a puzzling profile. In the wild-type PtpA, noticeable chemical mutation of one Cys residue was obtained at 5 mM of 1, and its full conversion was achieved at 15 mM of modifying reagent (m/z = 19,889 Da; Figures 2A and S1). No further Cys-to-Dha modifications were prominently identifiable with concentrations of 1 below 30 mM. However, at concentrations of 1 equal to 30 or 60 mM, simultaneous, yet incomplete modification of Cys53, Cys16, and Cys11 occurred as ascertained from the identification of two distinct mutant PtpA sub-populations (m/z = 19,820 and 19,887 Da; Figures 2A and S1). Interestingly, modification of the homologous Y. enterocolitica YopH (Figure S2), which contains five free Cys residues, similarly preceded regioselectively at a single position (Figures S2D and S2E).
To the best of our knowledge, these examples represent the first regioselective modifications of a single Cys on a native, multiple-containing Cys protein. Tight control of pH, time of incubation, and concentration of 1 is required to achieve regioselective modification. For example, prolonged incubation times with 15 mM of compound 1 also resulted in simultaneous and incomplete modification of Cys53, Cys11, and Cys16 (Figure S3).[](https://www.ncbi.nlm.nih.gov/mesh/D000409)
Regioselective Cys Modification Is Regulated by PtpA Structural Features[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
(A) 2D radial pair distribution function (2D RDF) computed after a 500 ns molecular dynamics (MD) simulation, suggests an H-bridged Cys11-Cys16 interaction. “Hotter” color suggests higher probability of water molecules. PtpA inset with predicted water molecule positions in catalytic cleft was computed after 500 ns and imaged with PyMOL (Schrödinger LLC).[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
(B) Overlay of 100 equidistant protein structure snapshots from 500 ns of MD simulations (gray cartoon). The inset shows the conformational flexibility of the Cys residues over the entire simulation length, overlayed with a single snapshot of the X-ray structure protein. Cys53 exhibits high conformational flexibility, whereas Cys11 and Cys16 adopt more fixed conformations (PDB: 1U2P).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
(C) Covalent docking to both the X-ray PtpA and 500 ns MD relaxed PtpA structures. More negative docking scores indicate a better fit of the covalently bound ligand.
(D) X-ray structure of PtpA, highlighting the water network at the catalytic cleft.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
(E) WaterMap predictions of spatially localized waters, i.e., high solvent density, within 6 Å of Cys11, Cys16, and Cys53. The remainder of the protein is shown as a green cartoon with a transparent gray molecular surface. Empty areas in the images are fully solvated during the simulations but without significantly enhanced solvent density in relation to bulk water. Cys11 and Cys16 are within the vicinity of multiple high-density water locations, whereas Cys53 is located on the protein surface with few bound water molecules nearby. High-density water positions are displayed in spherical representation, and color corresponds to the free energy in relation to bulk water (green, ΔG ≤ −2 kcal/mol; brown, ΔG ≥ +2 kcal/mol). Images were created with Maestro v10.5 (Schrödinger LLC).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
(F) Second-derivative FTIR spectrum of PtpA in the 3,100−2,700 cm−1 region measured at pH 8.0. The samples were hydrated with H2O (blue) or H218O (red). Labeled frequencies correspond to the water O–H stretching vibrations.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
(G) Second-derivative FTIR spectrum of native PtpA and Dha53 mutant in the 3,100−2,700 cm−1 region measured at pH 8.0. The samples were hydrated with H2O. Labeled frequency corresponds to the missing water O–H stretching vibration after Dha.[](https://www.ncbi.nlm.nih.gov/mesh/C015102)
(H) Second-derivative FTIR spectrum of native PtpA and triple mutant C11/16/53Dha in the 3,100−2,700 cm−1 region measured at pH 8.0. The samples were hydrated with H2O. Labeled frequencies correspond to the missing water O–H stretching vibrations after complete Dha installation.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
Kinetic Parameters of PtpA and the Chemically Derived Mutants PtpADha53 and PtpADha53Dha16Dha11
The absence of stoichiometric correlation in the Cys-to-Dha chemical mutation prompted our curiosity and deeper exploration to understand which factors determined the observed regioselectivity. To address this question, we engineered and expressed single Cys-to-Ser and single, double, and triple Cys-to-Ala PtpA mutants in order to gain insights into which of the Cys residues were preferentially modified. The bisalkylation elimination reaction with the single C53A PtpA mutant only proceeded at 60 mM of 1 (Figure S4). Conversely, conversion of Cys53 to Dha on the double C11A/C16A mutant resulted in complete Cys-to-Dha conversion at 15 mM of 1 (Figures S4–S7), suggesting favored Cys53 modification and fully in line with our previous data. Tryptic digestion (Figure 2B) of different PtpA populations led to 40-VTSAGTGNWHVGS(X)ADER-57 fragments, which corroborated preferential Dha installation at position 53 after tandem mass spectrometry (MS/MS) analyses (Figures 3B and S8–S10).
Critically, the Dha53 PtpA mutant displays an identical pH-dependent activity profile (Figure S12), kinetic parameters, and melting temperature to the wild-type counterpart (Figures 2C and S12–S14 and Table 1), providing a solid rationale for further bioorthogonal point of functionalization in PtpA. For example, Michael addition of β-mercaptoethanol to Dha53 was readily achieved on the engineered Dha53 PtpA (Figures S9 and S10). Circular dichroism spectra for the wild-type PtpA, Ala site-directed, and chemically mutated species show identical folding, with the exception of C11A/C16A, C11/16/C53A, and Dha11/16/53A (Figures 2D and S15; Matiollo et al.). C11/16/53/A and Dha11/16/53 displayed a pronounced loss of α-helical content (Figure S15, green dotted line). Molecular dynamics (MD) simulations performed on the C11/16/53A mutant corroborated a higher degree of flexibility than with the wild-type protein (Figure 2E). These data support the absence of a protein-fold-promoted Cys modification upon installation of Dha53 and significant fold changes upon full Cys mutation.
Likewise, the non-catalytic Cys259 residue was preferentially modified in YopH as established by MS/MS analyses (Figures S3D and S3E).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
It had been previously established that S-nitrosylation plays a paramount role in the dynamic post-translational regulation of several proteins., , In particular, S-nitrosylation of PtpA with S-nitrosoglutathione (GSNO) follows a pattern identical to that of Dha53 installation because it occurs exclusively at Cys53 (Table 1 and Figures S10 and S11), in contrast to the preferred catalytic Cys oxidation in Ptp1B. However, unlike Dha53 installation, S-nitrosylation of Cys53 partially suppresses the activity of PtpA (Table 1; Ecco et al.). Both the C53A and the Dha53 mutant are more prone to H2O2 inactivation than the wild-type counterpart (Figure 2F). Moreover, Cys53 is the first residue to be overoxidized after H2O2 incubation (Figure 2F). This pattern corroborates the critical role of Cys53 as a modulator of the PtpA redox state. Albeit intriguing, a molecular mechanism for selective S-oxidation is not yet known to date, despite its potential implications on chemical biology and drug discovery.[](https://www.ncbi.nlm.nih.gov/mesh/D026422)
## Water-Mediated Interplay between the Catalytic Cys11 and Backdoor Cys16 Modulates Redox Regulation
*In the **Water-Mediated Interplay between the Catalytic Cys11 and Backdoor Cys16 Modulates Redox Regulation** section:*
We investigated the reaction products obtained from the double C16A/C53A mutant. Chemical mutation of Cys11 occurred at concentrations of 1 as low as 15 mM and complete conversion occurred at 30 mM. Surprisingly, a different outcome was observed for the C11A/C53A mutant, i.e., no conversion of Cys16 to Dha was attained at 30 mM of editing reagent (Figure S7), advocating for an interplay between the catalytic Cys11 and backdoor Cys16 residues (Figures S5–S7) because these residues are modified simultaneously in the wild-type protein. Consequently, our data suggest intricate regulating factors for Cys modification, which were further corroborated by multiple thiol titrations with Ellman's reagent (Figure S16 and Table S1). Finally, mutation of the catalytic Cys11 residue substantially decreases PtpA functional activity (Table 1). The formation of a disulfide bridge as an explanation for the intricate interplay between Cys11 and Cys16 was dismissed because all the reactions were conducted under reducing conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
With the chemical mutagenesis data in hand, we endeavored to explore this elusive mechanism by using a combination of biophysical and computational approaches. Protein-bound water molecules have been increasingly recognized as key in the modulation of protein structure, folding, and dynamics., Just recently, a water-mediated allosteric network was reported to govern activation of Aurora kinase A. We thus hypothesized the existence of a water-mediated Cys-Cys non-covalent bridge regulating the active-site dynamics, and consequently Cys-to-Dha modification and GSNO-mediated oxidation. To address this question, we performed a series of MD simulations up to 500 ns with the apo structure of PtpA (PDB: 1U2P ). Despite evidence of a pKa of ca. 5 for the catalytic Cys, and having performed the Cys-modifying reactions at pH 8, we probed the different ionization states of all relevant residues in the catalytic cleft and computed 2D radial pair distribution functions (2D RDF).
The simulation data distinctly support that a water molecule bridges the ionizable Cys residues for 99% of the simulation time, with the average residence time for a water molecule of ca. 100 ps. It is thus feasible that such an H-bridge network can account for the observed regioselective PtpA modifications (Figure 3A). Over 500 ns, the PtpA structure was found to be very stable, such that the Cα-root-mean-square deviation (RMSD) from the X-ray structure converged to ca. 1.5 Å after 100 ns (average RMSD of 1.52 Å over the remaining simulation trajectory).
For Cys11 and Cys16, side-chain root RMS fluctuations of 0.9 and 1.4 Å, respectively, were found. A larger value of 1.9 Å was found for Cys53, indicating its higher conformational flexibility and exploration of diverse solvent-exposed surface conformations (Figure 3B). In contrast, Cys11 and Cys16 showed conformations buried into the catalytic pocket during the course of the simulation.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
In order to directly assess each Cys residue's propensity for undergoing chemical reactions, we modeled the nucleophilic substitution reaction of 1 with Cys. As protein structures, both the X-ray crystal structure and the protein structure at the end of the 500 ns MD simulation were used, and all three stereoisomers of reagent 1 (R,R; S,S; and meso-R,S) were used equally as ligands. We conducted covalent docking by using both fast and thorough protocols to sample changes in protein structure. In the fast-docking mode, docked poses were found only for Cys53. Conversely, the more thorough docking mode did find covalently docked poses of 1 for all three Cys residues.
More poses and better docking scores were obtained for Cys53 (Figure 3C). Thus, the docking predictions support the hypothesis that Cys53 is more solvent exposed, flexible, and sterically accessible, i.e. a preferred reaction partner with 1.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Motivated by the potential role of water molecules, together with the presence of water molecules in the catalytic cleft of the crystalized PtpA (Figure 3D), we performed MD simulations coupled with statistical thermodynamic analysis to assess the location and energetics of structural waters. We used the program WaterMap,, which combines a short (2 ns) MD simulation with solvent clustering and thermodynamic analysis by using inhomogeneous solvation theory., This approach has been used to characterize the energetics of water molecules at the surface of proteins and explain selectivity between highly related protein binding sites,, , binding kinetics, and the role of water networks in entropy and/or enthalpy compensation., Cys53 is predicted to have little tightly bound water structure around it, in line with its higher reactivity. On the other hand, the analysis revealed multiple stable water positions in the proximity of Cys11 and Cys16 (Figure 3E).
Both energetically favorable and unfavorable water molecules were found within that solvated pocket, advocating for kinetic and thermodynamic barriers regulating reactant entry and pocket desolvation.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
Fourier transformed infrared spectroscopy (FTIR) is a powerful tool in the structural biology of proteins. Strong absorbance from bulk water often limits the assignment of important structural waters; insights on structural (and internal) water clusters and their hydrogen-bonding networks are obtainable in the 3,700–2,700 cm−1 spectral region in certain experimental conditions., , , , , Although studies at the single-water-molecule level are not possible by FTIR alone, when it’s used in combination with site-directed mutagenesis or the post-expression chemical mutagenesis strategy discussed here, FTIR can aid in the assessment of water-molecule orientation in structural water hydrogen-bonding networks., ,[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
To further validate our WaterMap data, we analyzed and compared O–H vibrational energies of native Mtb PtpA, Cys-to-Ala, Cys-to-Ser, and Dha mutants (Dha53 and Dha53/11/16). We highlighted two regions of interest in the native Mtb PtpA spectrum that warranted further investigation (bands at 2,936 and 2,895 cm−1) because they appeared in the structural water region (3,100–2,700 cm−1) and undergo a wavenumber shift and/or a change in absorbance upon hydration with H218O (Figure 3F). The change of these bands upon isotope exchange suggests that they originate from the O–H stretching of water molecules, and not from overlapping N–H or C–H stretches., The band at 2,936 cm−1 was altered in both the triple Dha mutant (Dha11/16/53; Figures 3H and S17) and to a lesser extent in the triple Cys-to-Ala mutant (C11A/C16A/C53A; Figure 3F), suggesting that this band might represent a hydrogen-bonded water molecule oriented toward Cys11 and Cys16 in the active site of PtpA. In the case of single and double mutants, changes at 2,936 cm−1 were not significant, in agreement with the fact that only the triple mutation is able to disrupt the water-molecule hydrogen network in the active site. We detected other changes that can be related to water molecules interacting with residues outside the active site.
In particular, the Dha53 and C53A mutants showed changes at 2,895 cm−1 (Figures 3G, 3B, and S17). Changes in these regions might correspond to a structural change around the Cys53 residue not related to the active site. The mutant C16A/C53A (with only Cys11; Figures 3C and S17) again showed a change in the region 2,895 cm−1. However, the mutant C11A/C16A (with only Cys53; Figures 3D and S17) did not show clear changes in this region, within error. As a further control, we designed and produced the mutant C11S because Ser might also form stable hydrogen bonds.
The FTIR spectra of the C11S mutant displayed an almost identical pattern to that observed for the wild-type PtpA (Figure S17), which further corroborates the ability of the mutant to replace the Cys11 and still maintain the water-bridging network between Ser11 and Cys16. Furthermore, we also performed 500 ns MD simulations on mutants C11S, C16S, and the double mutant, observing the clear existence of bridging water molecules between residues 11 and 16 (Figure S18). This result is in good agreement with the FTIR data of mutant C11S. In addition, the atomic fluctuation study on these mutants confirms that the 3D structure is only slightly modified by mutation, highlighting the crucial role of the water pockets on the global structure of these proteins. These observations fully confirm our WaterMap calculations and provide strong experimental evidence for the location of structural waters within the PtpA catalytic pocket, i.e., the hydrogen bond network is dependent on the presence of Cys11 and Cys16.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Our biophysical analysis also revealed differences in the folding of the site-directed mutant C11/16/53A and the chemical mutant Dha11/16/53 (Figure S15). It is clear that modification of Cys11 and Cys16 disrupts PtpA structure, and because of this disruption, it is likely that the structural waters within the PtpA catalytic cleft are displaced. It is useful to consider, however, that the structural changes that led to the displacement of the water-bridging motif appear entirely unalike in the Dha and the site-directed mutant. In the site-directed C11/16/53A, as the protein is translated it acquires disrupted folding, which is induced by the absence of Cys11, Cys16, and Cys53; therefore, the water-bridging mechanism is not formed. On the other hand, in the Dha11/16/53 mutant, it is feasible to hypothesize that because of the harsh conditions (Figure S2; 24 hr incubation at 37°C and saturated compound concentrations, i.e., 60 mM of 1), the catalytic pocket topology was disrupted, the structural waters were displaced, and hence simultaneous Cys11 and Cys16 modification was allowed.
In these conditions, we were not able to achieve complete conversion of Dha11/16/53 (Figure S2). Two protein populations in the MS spectra (50/50 ratio, single Dha53 and triple Dha11/16/53) were persistent during the longer incubation times tested. Nevertheless, the full Dha11/16/53 conversion was possible after forced disruption of the protein structure, induced by the significant changes we made in the pH and compound concentrations (data not shown). We assume that the water-bridging motif is well stabilized and structurally tight, to the point that only a part of the protein population undergoes the full chemical modification in the designed conditions we tested. This observation offers a credible explanation for the absence of stoichiometric correlation in the Cys-to-Dha chemical mutation, given that the water-bridging motif ultimately mediates Cys11 and Cys16 accessibility and reactivity.
Moreover, it is reasonable to assume that this water motif can also mediate the access of nitrosative and oxidative species to the catalytic pocket, consequently preventing the overoxidation of Cys11. Such features are consistent with the known resistance of Mtb PtpA to the oxidative conditions that prevail within an infected host macrophage.[](https://www.ncbi.nlm.nih.gov/mesh/C015102)
## The Cys-Cys Water-Bridging Motif Is Conserved among Phosphatases with Structurally Similar Catalytic Clefts
*In the **The Cys-Cys Water-Bridging Motif Is Conserved among Phosphatases with Structurally Similar Catalytic Clefts** section:*
The Catalytic Pocket Is Highly Conserved among the Protein Phosphatase Family
(A) Superposition of selected active-site residues and waters of PtpA from M. tuberculosis (PDB: 1U2P) in gray, YwlE arginine phosphatase from G. stearothermophilus (PDB: 4PIC) in yellow, tyrosine phosphatase AmsI from E. amylovora (PDB: 4D74) in cyan, and TT1001 protein from T. thermophilus HB8 (PDB: 2CWD) in gray.
(B) The active site from the Mtb PtpA crystal structure.
(C–F) The active sites from (C) S. aureus, (D) G. stearothermophilus, (E) E. amylovora, and (F) T. thermophilus HB8. The protein residues are drawn in stick representation, and conserved water molecules are drawn as spheres. Atom colors are gray for carbon, blue for nitrogen, red for oxygen, yellow for sulfur, bright orange for selenium, and orange for phosphorus.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
Motivated by the presence of a Cys-Cys water-bridging motif in the catalytic pocket of Mtb PtpA, we decided to investigate whether this mechanism is conserved among other bacterial PTPs. Several structures of PTPs have been determined either by X-ray crystallography or solution nuclear magnetic resonance. However, a large number of these have different structures that are dependent on their crystal form or ligand. Using the FATCAT algorithm operating in rigid mode, we found that the structures of phosphatases from Vibrio cholera O395 (PDB: 4LRQ ), Entamoeba histolytica (PDB: 3IDO ), S. aureus (PDB: 3ROF ), Thermus thermophilus HB8 (PDB: 2CWD), and arginine phosphatases from Erwinia amylovora (PDB: 4D74 ) and Geobacillus stearothermophilus (PDB: 4PIC) shared the highest structural similarity. Next, we performed searches on PDBeFold by using chain A of the Mtb PtpA structure as a query (PDB: 1U2P ); we found 19 top hits with alignments (Figure S19) sharing 27%–42% sequence identity and with an RMSD of Cα atomic coordinates varying from 1.23 to 1.66 Å.
The Mtb PtpA belongs to the low-molecular-weight PTPase family in which the catalytic pocket is highly conserved, with a signature sequence of (H/V)CX5R(S/T)., , This prompted us to investigate if the water molecules found in the Mtb PtpA catalytic pocket (W171, W182, and W212, in PDB: 1U2P) were also present in the X-ray structures of other PTPs. Importantly, a superposition analysis of the closest 3D structures shows the presence of these water molecules in the active-site cavity. As demonstrated by our structural analysis data (Figure 4), a water-molecule network comprises an important allosteric arrangement that stabilizes the catalytic pocket of bacterial PTPs. A water molecule has also been invoked to play a role in the reaction mechanism of an arginine phosphatase from E. amylovora. Moreover, it has also been hypothesized as a mechanism of oxidative regulation in Ptp1B from S. aureus that involves the reversible oxidation of the catalytic Cys to a sulfenate, thus suggesting a potential role of a water molecule.
Nevertheless, the observation of a conserved water network is not observed in all homologous PTPs, given that in some X-ray structures these water molecules are most likely displaced by ligands found within the active site., ,[](https://www.ncbi.nlm.nih.gov/mesh/C046557)
Averaged Second-Derivative FTIR Spectra of PtpA and SptpA in the 3,100−2,700 cm−1 Region Measured at pH 8.0
The samples were hydrated with H2O or H218O.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
(A) FTIR spectra of native Mtb PtpA. Labeled blue frequencies correspond to the water O–H stretching vibrations.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
(B) FTIR spectra of native SptpA. Labeled red frequencies correspond to the water O–H stretching vibrations.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
All spectra represent an average of three replicates from three independent experiments. The width of the line indicates ± standard error of the mean.
To confirm whether the Cys-Cys water-bridging motif is conserved among bacterial PTPs, we chose the SptpA protein from S. aureus. This phosphatase adopts the general architecture of the low-molecular-weight PTPase family, displaying an α/β fold with a central four-stranded parallel β sheet providing the scaffold for the active site. Importantly, the catalytic Cys8 as well as the backdoor Cys13 in SptpA from S. aureus are structurally identical to the catalytic Cys11 and backdoor Cys16 in Mtb PtpA. Next, using FTIR, we analyzed whether absorbances indicating a Cys-Cys water-bridging motif for Mtb PtpA were also present for SptpA from S. aureus. Similarly, the native SptpA also yielded a broad IR absorbance spectrum with five main bands (2,936 cm−1, 2,917 cm−1, 2,905 cm−1, 2,895 cm−1, and 2,852 cm−1) that underwent a change in the corresponding vibrational energy upon hydration with H218O (Figure 5).
Importantly, the IR band at 2,936 cm−1 was found in the spectra of both phosphatases, indicating the presence of a water molecule. This observation points toward a conserved Cys-Cys water-bridging motif among bacterial PTPs. In addition, the absorbance spectra (Figure S20) of both proteins showed similar peaks in the region 3,000–2,800 cm−1. Importantly, this region can potentially be used to spectroscopically probe the catalytic pocket of similar phosphatases.[](https://www.ncbi.nlm.nih.gov/mesh/C046557)
## Conclusions
*In the **Conclusions** section:*
Using a robust post-expression mutagenesis approach, we have demonstrated that the non-catalytic residues Cys53 in PtpA and Cys259 in YopH are the most reactive Cys residues in phosphatases of clinically relevant bacteria. Although steric hindrance is likely to play a role in the observed regioselective modification, we confirmed a water-mediated structural motif that modulates the interplay between the catalytic Cys11 and the backdoor Cys16 at a molecular level in Mtb PtpA. Such structural motif is also found in the phosphatase SptpA from S. aureus, which indicates that the mechanism might actually be conserved among phosphatases that share structural identity in the catalytic cleft. This hitherto unknown regulation mechanism is key for protein structure and function and sharply contrasts with the well-established disulfide bridge paradigm. Significantly, this mechanism also provides a molecular rationale for selective PtpA Cys53 oxidation by GSNO and H2O2 and insights into new biology and host-pathogen interaction in PtpA resistance given that all of the Cys residues might work synergistically in an elegant interplay to protect the protein against the harsh macrophage environment. Considering that drug-target biology assessment and validation is a critical preliminary step toward the development of innovative therapeutics, our strategy provides a broadly applicable platform in chemical biology and molecular medicine to aid in the understanding of native protein dynamics.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Experimental Procedures
*In the **Experimental Procedures** section:*
Full experimental procedures are provided in the Supplemental Information.
## Author Contributions
*In the **Author Contributions** section:*
G.J.L.B. conceived the study. G.J.L.B. and H.T. supervised the study. J.B.B. performed protein expression, purification and modification, and biophysical characterization experiments. L.D. and L.A.R. performed protein expression and purification. T.R. and F.C. performed molecular dynamic simulations.
O.B. analyzed protein-modification reactions. J.B.B., F.A.A., and L.D. performed FTIR experiments. M.C.M. performed the structural alignments. T.B.S. and W.S. conducted the WaterMap and covalent docking calculations and analysis. T.R., J.B.B., and G.J.L.B. wrote the manuscript with contributions from all authors.
# References and Notes
*In the **References and Notes** section:*
## Supplemental Information
*In the **Supplemental Information** section:*
Supplemental Information includes Supplemental Experimental Procedures, protein sequences, 32 figures, and 1 table and can be found with this article online at http://dx.doi.org/10.1016/j.chempr.2017.07.009.
|
# Introduction
Structure and Functional Analysis of Promoters from Two Liver Isoforms of CPT I in Grass Carp Ctenopharyngodon idella
# Abstract
*In the **Abstract** section:*
Carnitine palmitoyltransferase I (CPT I) is a key enzyme involved in the regulation of lipid metabolism and fatty acid β-oxidation. To understand the transcriptional mechanism of CPT Iα1b and CPT Iα2a gene[s, we](https://www.ncbi.nlm.nih.gov/mesh/D008055) cloned the 2695[-bp and 26](https://www.ncbi.nlm.nih.gov/mesh/D005227)31-bp regions of CPT Iα1b and CPT Iα2a promoters of grass carp (Ctenopharyngodon idella), respectively, and explored the structure and functional characteristics of these promoters. CPT Iα1b had two transcription start sites (TSSs), while CPT Iα2a had only one TSS. DNase I foot printing showed that the CPT Iα1b promoter was AT-rich and TATA-less, and mediated basal transcription through an initiator (INR)-independent mechanism. Bioinformatics analysis indicated that specificity protein 1 (Sp1) and nuclear factor Y (NF-Y) played potential important roles in driving basal expression of CPT Iα2a gene. In HepG2 and HEK293 cells, progressive deletion analysis indicated that several regions contained cis-elements controlling the transcription of the CPT Iα1b and CPT Iα2a genes. Moreover, some transcription factors, such as thyroid hormone receptor (TR), hepatocyte nuclear factor 4 (HNF4) and peroxisome proliferator-activated receptor (PPAR) family, were all identified on the CPT Iα1b and CPT Iα2a promoters. The TRα binding sites were only identified on CPT Iα1b promoter, while TRβ binding sites were only identified on CPT Iα2a promoter, suggesting that the transcription of CPT Iα1b and CPT Iα2a was regulated by a different mechanism. Site-mutation and electrophoretic mobility-shift assay (EMSA) revealed that fenofibrate-induced PPARα activation did not bind with predicted PPARα binding sites of CPT I promoters. Additionally,[ PPARα was ](https://www.ncbi.nlm.nih.gov/mesh/D011345)not the only member of PPAR family regulating CPT I expression, and PPARγ also regulated the CPT I expression. All of these results provided new insights into the mechanisms for transcriptional regulation of CPT I genes in fish.
## 1. Introduction
*In the **1. Introduction** section:*
Lipids are the major sources of metabolic energy in fish. Body lipid composition results from the balance among deposition of dietary lipids, de novo synthesis of fatty acids and oxidation of fatty acids. While the relations between food intake and lipid deposition as well as nutritional control of fatty acid synthesis are well documented, fatty acid catabolism has received little attention. The β-oxidation of fatty acids plays a critical role in the production of energy, and most oxidation occurs in the mitochondria. Carnitine palmitoyltransferase I (EC.2.3.1.21; CPT I), located in outer membranes of mitochondria, controls the flux through β-oxidation and is the main regulatory enzyme of fatty acid oxidation.
The studies about the structure and transcriptional regulation of CPT I gene are useful for the understanding of the β-oxidation in fish. In mammals, three CPT I isoforms encoded by distinct genes have been discovered: a liver isoform (CPT Iα), a muscle isoform (CPT Iβ), and a brain isoform (CPT Ic). In fish, however, due to fish-specific genomic duplication event, various CPT I isoforms have been cloned. For example, three α-copies and one β-copy of CPT I was obtained in yellow catfish Pelteobagrus fulvidraco and seven complete CPT I cDNA sequences (CPT Iα1a-1a, CPT Iα1a-1b, CPT Iα1a-1c, CPT Iα1a-2, CPT I α2a, CPT Iα2b1a, CPT Iβ) and a partial cDNA sequence (CPT Iα2b1b) were cloned in goby Synechogobius hasta. In grass carp, the complete cDNA sequences of three CPT Iα genes (CPT Iα1a, CPT Iα1b and CPT Iα2a) and one CPT Iβ gene isoforms have successfully been cloned.
Although these isoforms of CPT I gene can express CPT I protein which catalyzes the same reaction, they have different properties. For example, McGarry and Brown pointed out that mammalian CPT Iβ had a much lower IC50 and higher Km for carnitine than CPT Iα (from). Lineage- and species-specific genome duplication events can lead to increased diversity in protein regulation and function. At present, while the characteristics of CPT I gene and struct[ure pr](https://www.ncbi.nlm.nih.gov/mesh/D008055)ediction as well as its enzyme kinetics are well document[ed in](https://www.ncbi.nlm.nih.gov/mesh/D008055) fish, mechanisms involving the transcriptional regulations of CPT[ I gen](https://www.ncbi.nlm.nih.gov/mesh/D008055)e received no attention[.](https://www.ncbi.nlm.nih.gov/mesh/D005227)/www.ncbi.nlm.nih.gov/mesh/D005227)
Considering the importance of CPT I in regulating fatty acid oxidation, it is very important and meaningful to explore the regulatory mechanism of CPT I mRNA expression. At present, most studies on the mRNA expression and/or activity of CPT I isoforms in fish involve the response to either dietary or hormonal treatments. However, expression of eukaryotic genes is controlled at the level of transcription initiation. Promoters, which contain cis-acting sequences bound by a wide variety of regulatory factors, control the expression of individual genes. Therefore, it is very important to analyze the structure and function of CPT I promoter, which helps to understand the regulatory mechanism of CPT I itself.
At present, the promoter of the CPT Iα gene has been obtained only in mammals, but not in fish. The present study hypothesizes that significant differences exist in structure and function of CPT I promoters between fish and mammals.[](https://www.ncbi.nlm.nih.gov/mesh/D005227)
Lipid metabolism is closely controlled by diverse regulatory systems involving many transcription factors. Peroxisome proliferator-activated receptors (PPARs), which belong to ligand-dependent transcription factors, regulate the expression of various genes involved in lipid metabolism. Among the PPAR family member, PPARα plays crucial roles in the catabolism of fatty acids by increasing the expression of key lipolytic enzymes (also CPT I). Studies demonstrated that the PPARα mRNA expression was positively correlated to CPT I mRNA expression. Further investigation indicated that PPARα stimulated through a peroxisome proliferator-responsive element (PPRE) in the first and second intron of the human and rat CPT Iα genes, respectively.
PPARγ, involved in the regulation of lipogenesis and lipid storage, preferentially control the transcription of genes in triglyceride synthesis. In an earlier study, Chen et al. found that mRNA expression of PPARγ was positively correlated with CPT I expression, suggesting a potential regulation of PPARγ on CPT I expression. At present, although several evidences suggested that CPT Iα was a target gene for PPAR, a lack of knowledge regarding the DNA sequence responsible for this predicted regulatory mechanism has left this a controversial issue. Thus, considering the importance of PPARs in lipid metabolism, it is very important to explore the regulation of CPT I expression by PPARs.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
Grass carp (Ctenopharyngodon idella) was an important herbivorous freshwater fish widely farmed all over the world because of its good taste and high market price. Its aquaculture yield amounted to 6 million metric tons in China in 2016. In some countries of European and Northern America, grass carp were used to control aquatic plants because of their aggressive feeding on vegetation. At present, grass carp is considered a good model for the study of lipid metabolism because it stores excess fat in liver and adipose tissues under intensive aquaculture. Recently, the draft genome of the grass carp has been released, which is considered a convenient tool for identifying genomic structure of genes involved in lipid metabolism.
In the present study, we characterized CPT Iα1b and CPT Iα2a promoters in grass carp. Their transcriptional regulation by peroxisome proliferators was also explored. These studies will provide new insights into the transcriptional regulatory mechanism of CPT I genes in fish.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
## 2. Results
*In the **2. Results** section:*
Studies indicated that, compared with other isoforms, mRNA levels of CPT Iα1b and CPT 1α2a were predominant in the liver. Therefore, CPT Iα1b and CPT 1α2a were considered as liver isoforms. To investigate their transcriptional regulatory mechanism, for the first time, we cloned the sequences of promoters of the two liver isoforms (CPT Iα1b and CPT Iα2a), and explored their functional characteristics in fish.
## 2.1. Identification of Transcription Start Site (TSS)
*In the **2.1. Identification of Transcription Start Site (TSS)** section:*
In the present study, the 2695 bp of CPT Iα1b promoter and 2631 bp of CPT Iα2a promoter were cloned and submitted to an online transcription factor database (MatInspector) for sequence analysis. RNA ligase-mediated rapid amplification of 5′ cDNA ends (RLM-5′RACE) was performed to identify the TSS of CPT Iα1b and CPT Iα2a promoters. This amplification generated two different TSSs of CPT Iα1b which approximately corresponded to the alternative 5′ splice variants of CPT Iα1b mRNA, and one TSS of CPT Iα2a without alternative 5′ variant. The first nucleotide of the CPT Iα1b gene, mapped to the most upstream position from the grass carp liver cDNA library, was arbitrarily designated as +1′ and the alternative 5′ splicing site was designated as +1 (Figure 1A,B). The first nucleotide of the CPT Iα2a gene was designated as +1 (Figure 1C).[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
## 2.2. DNase I Foot Printing Assay of Core Promoter of CPT Iα1b
*In the **2.2. DNase I Foot Printing Assay of Core Promoter of CPT Iα1b** section:*
Figure 2A showed the core region of CPT Iα1b promoter from −268 bp to +37 bp containing transcription start site (TSS1). Predicted TATA-box was located between 148 bp and 167 bp of the FAM-labeled fragment, and the electropherograms around this region presented similar peak patterns between control group (0 µg nuclear proteins, 20 µg bovine serum albumin, BSA) and DNase I digested group (10 µg nuclear proteins, 10 µg BSA). In contrast, the region between 290 bp and 360 bp presented different peak patterns between control group and DNase I digested group, where the initiator (INR) was located. Figure 2B showed the core region of CPT Iα1b promoter from −581 bp to −236 bp containing alternative transcription start site (TSS2). Predicted TATA-box on this fragment was located between 327 bp and 343 bp of the FAM-labeled fragment, and the electropherograms on this region were similar between control group (0 µg nuclear proteins, 20 µg BSA) and DNase I digested group (10 µg nuclear proteins, 10 µg BSA).
In contrast, the different peak patterns were discovered at the region between 285 bp and 310 bp, where the INR was located. Taken together, these indicated that the INR on the promoter was sufficient for the transcription initiation of CPT Iα1b gene.
## 2.3. Sequence Analysis of the CPT Iα1b and CPT Iα2a Promoters
*In the **2.3. Sequence Analysis of the CPT Iα1b and CPT Iα2a Promoters** section:*
Several putative core promoter elements close to the TSS on the CPT Iα1b promoter, including two TATA-box (TBP) located from −160 bp to −176 bp and from −293 bp to −309 bp, and two initiator (INR) located at −2 bp to +10 bp (TSS1) and −333 bp to −343 bp (TSS2), were identified (Figure 3). Meanwhile, on the core region of CPT Iα2a promoter, three CCAAT-box (NF-Y) were identified, located at −46 bp to −60 bp, −146 bp to −160 bp and −165 bp to −179 bp, respectively. Besides, two GGGCGG-box (Sp1), located at −13 bp to −29 bp and −127 bp to −143 bp, were also identified on the core promoter of CPT Iα2a (Figure 4). Some relevant TFBSs of CPT Iα1b and CPT Iα2a were presented in Figure 3 and Figure 4. There were two thyroid hormone receptor α (TRα) binding sites on the CPT Iα1b promoter at the position −1070 bp to −1094 bp and −2067 bp to −2091 bp, and three thyroid hormone receptor β (TRβ) binding sites on the CPT Iα2a promoter, at the position −39 bp to −63 bp, −1103 bp to −1127 bp and −1331 bp to −1355 bp, respectively.
In addition, we discovered one HNF4 binding site on the CPT Iα1b promoter, located at −2379 bp to −2403 bp, one HNF4 binding site on the CPT Iα2a promoter, located at the position −406 bp to −430 bp, and one HNF4α binding site on the CPT Iα2a promoter, located at the position −2587 bp to −2611 bp. Moreover, analysis using MatInspector database revealed two PPAR binding sites on the CPT Iα1b promoter and four PPAR binding sites on the CPT Iα2a promoter. Among these sites, one PPARα/RXR binding site located at the position −1814 bp to −1836 bp and one PPARγ binding site located at the position −1719 bp to −1741 bp were predicted on the CPT Iα1b promoter. Meanwhile, there were four important binding sites of transcriptional factors on the CPT Iα2a promoter, distributed at the position −1939 bp to −1961 bp (PPARα/RXR binding site), −1179 bp to −1201 bp (PPARγ binding site), −1104 bp to −1136 bp (PPARγ binding site) and −1044 bp to −1066 bp (PPARγ binding site).
## 2.4. Deletion Assay of the CPT Iα1b and CPT Iα2a Promoter
*In the **2.4. Deletion Assay of the CPT Iα1b and CPT Iα2a Promoter** section:*
Deletion analysis of CPT Iα1b and CPT Iα2a promoters was presented in Figure 5. The reporter activity for each serial deletion was compared with the activity of pGl3-basic vector, and the pGl3-basic was chosen as the baseline. Figure 5A showed the result of deletion assay of the CPT Iα1b promoter sequence from −2695 bp to −86 bp in HepG2 cells. Deletion of the region from −2276 bp to −2695 bp significantly increased the relative luciferase activity of the promoter. Subsequent deletion to −1716 bp significantly decreased the relative luciferase activity.
Deletion of the sequence from −581 bp to −1716 bp showed no significant effect, whereas deletion of the sequence from −581 bp to −86 bp significantly decreased the relative luciferase activity. Figure 5B showed the result of deletion assay of the CPT Iα1b promoter in HEK293 cells. Deletion of the sequence from −2695 bp to −2276 bp significantly increased the relative luciferase activity, and the sequence deletion from −2276 bp to −1716 bp significantly reduced the activity. Subsequent deletion to −581 bp presented no significant effects on the relative luciferase activity. The sequence deletion from −581 bp to −86 bp significantly decreased the relative luciferase activity.
Figure 5C presented the result of deletion assay of the CPT Iα2a promoter sequence from −2631 bp to −97 bp in HepG2 cells. The relative luciferase activity of CPT Iα2a promoter showed no significant difference from −2631 bp to −1646 bp. Deletion of the sequence from −1646 bp to −1304 bp significantly increased the relative luciferase activity. Subsequent deletion to −848 bp presented no significant effects on the relative luciferase activity. Deletion of the sequence of −848 bp to −428 bp and −428 bp to −97 bp significantly decreased the relative luciferase activity.
Figure 5D showed the result of deletion assay of CPT Iα2a promoter in HEK293 cells. Deletion of the sequence from −2631 bp to −1165 bp presented no significant effects on the relative luciferase activity. All of these sequence deletions from −1165 bp to −848 bp, −848 bp to −428 bp and −428 bp to −97 bp significantly decreased the luciferase activity.
## 2.5. Site-Mutation Analysis of PPAR Binding Sites
*In the **2.5. Site-Mutation Analysis of PPAR Binding Sites** section:*
Site-mutation analysis was used to evaluate the contribution of each PPAR binding site to the basal expression of the grass carp CPT Iα1b and CPT Iα2a genes in HepG2 cells (Figure 6). The disruption of the −1814/−1836 PPARα binding site did not change the relative luciferase activity against the wild-type pGl3-CPTIα1b-2276, and disruption of the −1814/−1836 PPARα binding site did not influence the fenofibrate-induced change of luciferase activity, indicating that −1814/−1836 PPARα binding site did not contribute to the transcriptional response of CPTIα1b gene to fenofibrate (Figure 6A). The disruption of the −1939/−1961 PPARα binding site significantly up-regulated the relative luciferase activity against the wild-type pGl3-CPTIα2a-2041. In contrast, disruption of the −1939/−1961 PPARα binding site did not influence the fenofibrate-induced change of luciferase activity induced, suggesting that the −1939/−1961 sequence did not contribute to the transcriptional response of CPTIα2a gene to fenofibrate. We also disrupted each PPARγ binding site by site-directed mutagenesis in the context of the pGl3-CPTIα1b-2276 and pGl3-CPTIα2a-1304 vectors, respectively; meantime, three double mutants and one triple mutant of PPARγ binding site were produced on the pGl3-CPTIα2a-1304 vector (Figure 6B).
The disruption of the −1719/−1741 PPARγ binding site did not change the relative luciferase activity against the wild-type pGl3-CPTIα1b-2276, and disruption of the −1719/−1741 PPARγ binding site did not influence the pioglitazone-induced change of luciferase activity, suggesting that −1719/−1741 PPARγ binding site did not contribute to the transcriptional response of CPTIα1b gene to pioglitazone. Disruptions of the PPARγ binding sites on pGl3-CPTIα2a-1304 vectors showed that the −1044/−1066 PPARγ binding site up-regulated relative luciferase activity against the pGl3-CPTIα2a-1304. Other mutant vectors, including double and triple mutant of PPARγ binding sites, presented no significant difference in luciferase activities against the wild-type pGl3-CPTIα2a-1304, indicating that the −1044/−1066 PPARγ binding site possibly played a negative regulatory role in CPTIα2a transcription. In addition, disruption of the −1719/−1741 PPARγ binding site reduced the luciferase activity induced by pioglitazone, and disruption of −1719/−1741 PPARγ binding site along with either −1179/−1201 PPARγ binding site or −1104/−1136 PPARγ binding site also reduced the luciferase activity induced by pioglitazone, suggesting that −1719/−1741 PPARγ binding site contributed to the transcriptional response of CPTIα2a to pioglitazone. Taken together, these results indicated that PPARα could not regulate the transcription of CPT Iα1b and CPT Iα2a at their predicted binding sites, and the transcription of the grass carp CPT Iα2a gene expression could be controlled by PPARγ.[](https://www.ncbi.nlm.nih.gov/mesh/D011345)
## 2.6. EMSA of Each PPAR Binding Sequence
*In the **2.6. EMSA of Each PPAR Binding Sequence** section:*
Having demonstrated that the putative PPAR binding site was important for the transcriptional activities of CPT Iα1b and CPT Iα2a genes, we next examined whether PPARs could bind to this site directly. We used EMSA assay to confirm this mechanism (Figure 7). Two close weak bands were observed at the −1814/−1836 PPARα binding sequence of CPT Iα1b promoter, and neither a 100-fold excess unlabeled probe nor a 100-fold excess unlabeled point-mutated probe could compete out the labeled probe, indicating that this sequence was not bound by PPARα (Figure 7A). Only the free probe band was discovered at the −1719/−1741 PPARγ binding sequence of CPT Iα1b promoter (Figure 7B), suggesting that this sequence was not bound by any transcriptional factors. A strong band close to a weak band was observed at the −1939/−1961 PPARα binding sequence of CPT Iα2a promoter, and neither a 100-fold excess unlabeled probe nor a 100-fold excess unlabeled point-mutated probe could compete out the labeled probe, indicating that this sequence was not bound by PPARα (Figure 7C).
Similarly, Figure 7D and E also indicated that the −1179/−1201 and −1104/−1136 PPARγ binding sequences of CPT Iα2a promoter were not bound by PPARγ. Only the sequence corresponding to the −1104/−1066 PPARγ binding site of the CPT Iα2a promoter could bind with proteins from HepG2 nuclear extract (NP) and be disrupted by a 100-fold excess of unlabeled wild-type, and restored by a point-mutant probe (Figure 7F), confirming that −1044/−1066 PPARγ binding site of the CPT Iα2a promoter could react with PPARγ.
## 3. Discussion
*In the **3. Discussion** section:*
The reaction catalyzed by CPT I is a rate-controlling step in the pathway of LCFA β-oxidation. Currently, five isoforms of CPT I genes (CPT Iα1a, CPT Iα1b, CPT Iα2a, CPT Iα2b and CPT Iβ) were identified in grass carp (C. idella). Moreover, these studies indicated that, compared with other isoforms, mRNA levels of CPT Iα1b and CPT 1α2a were predominant in the liver. Therefore, CPT Iα1b and CPT 1α2a were considered as the liver isoform. To investigate their transcriptional regulatory mechanism, for the first time, we cloned the sequences of CPT Iα1b and CPT Iα2a promoters in fish, and explored their functional characteristics.
In the present study, we found two TSSs of CPT Iα1b corresponding to the alternative 5′ splice variants of CPT Iα1b mRNA. Studies suggested that alternative TSSs usually occurred in the proximal promoter of genes lacking TATA and CCAAT boxes. Batarseh et al. pointed out that multiple TSSs were typically TATA-less and they were located within CpG islands. Park et al. found that the rat L-CPT I (CPT Iα) promoter was GC rich and TATA-less and had an alternative transcription initiation. However, our present study found some variations in TSSs of the CPT Iα1b promoter in grass carp.
Grass carp CPT Iα1b promoter was AT-rich and contained two TATA elements without canonical CpG islands, but DNase I foot printing assay showed that both TATA elements were not protected from DNase I digestion, whereas the INR, which encompassed the TSS, was protected from DNase I digestion. These phenomena indicated that the basal transcription of the CPT Iα1b gene required the INR to position the basal transcription machinery. In agreement with our study, Smale and Kadonaga pointed out that the INR was located at the TSS and it was independent of, or in synergy with the TATA box. Thus, our results suggested that the basal transcription of the CPT Iα1b gene might be mediated through an INR-independent mechanism.
For CPT Iα2a gene, the present study indicated that the core promoter of grass carp CPT Iα2a was GC-rich and did not contain a TATA box.
In agreement with rat CPT Iα gene, our study indicated that the proximal promoter region of CPT Iα2a contained several Sp1 and NF-Y binding sites, whereas only one transcription initiation was identified on the promoter. Steffen et al. pointed out that the Sp1, Sp3 and NF-Y factors played major roles in driving basal expression of rat CPT Iα gene. Sp1, a ubiquitously expressed prototypic C2H2-type zinc finger protein, can activate or repress transcription after physiological and pathological stimuli. Studies demonstrated that multiple Sp1 binding sites were a common feature of TATA-less promoters. Moreover, the Sp1 can bind GC-rich motifs and regulate the expression of genes via protein-protein interactions or interplay with other transcription factors and/or components of the transcriptional machinery.
NF-Y, one of the major transcriptional factors binding to the CCAAT box, may interact with Sp1 to regulate transcription of various genes. In agreement with these studies, the present study indicated that Sp1 and NFY factors were identified on the core region of CPT Iα2a promoter in a similar manner, indicating a similar transcription initiation for CPT Iα2a transcription. Taken together, our study indicated that transcription initiation of the CPT Iα1b and CPT Iα2a genes presented different mechanisms, suggesting that the expression of two genes from grass carp was induced by different transcriptional initiation.
Identification of TFBSs is very important for deciphering the mechanisms of gene regulation. To better understand the regulation of CPT Iα1b and CPT Iα2a at the transcriptional level, we functionally characterized the CPT Iα1b and CPT Iα2a promoters of grass carp. The present study identified a cluster of TFBS, such as TR, HNF4 and PPAR family, on the grass carp CPT Iα1b and CPT Iα2a promoters. Similarly, Jackson-Hayes et al. showed that the rat CPT Iα gene had a thyroid hormone response element (TRE) which was required for the thyroid hormone receptor (TR) binding. In the present study, several TREs were also observed on the grass carp CPT Iα1b and CPT Iα2a promoters.
Interestingly, our study found that CPT Iα1b and CPT Iα2a promoters were bound by different isoforms of TRs. The two TREs on the CPT Iα1b promoter were only for TRα binding, and the three TREs on the CPT Iα2a promoter were only for TRβ binding. In fish, TRα and TRβ were expressed at different developmental stages, suggesting their functional differentiation. Additionally, TRα and TRβ were differentially regulated by systemic thyroid status in fish. Thus, these studies strongly suggested that the transcriptions of CPT Iα1b and CPT Iα2a genes were regulated through different mechanism in the liver.[](https://www.ncbi.nlm.nih.gov/mesh/D013963)
In the present study, deletion analysis indicated that several regions of the promoters contained a potential cis-active element(s) which enhanced/decreased transcriptional activities of the grass carp CPT Iα1b and CPT Iα2a genes. Furthermore, the regions of the CPT Iα1b and CPT Iα2a promoters presented different reporter activities in HepG2 and HEK293 cells. Obviously, the enhancing/decreasing reporter activity indicated the existence of potential positive/negative regulators on the regions, respectively. For the promoter of CPT Iα1b gene, we found that, compared to the region between −1716 and 1379 bp, the luciferase activity between the region from −2276 bp to −1716 bp significantly increased in HepG2 and HEK293 cells. Interestingly, we noticed that the −2067/−2091 TRα binding site, −1939/−1961 PPARα binding site and −1719/−1741 PPARγ binding site were located at the region from −2276 bp to −1716 bp, which was reported to correlated with CPT I expression in mammals.
For the promoter of CPT Iα2a gene, we discovered that the luciferase activity increased from the TSS to −848 bp in HepG2 cells, whereas the activity increased from TSS to −1165 bp in HEK293 cells, indicating a different regulation at the region from −848 bp to −1165 bp between the two kinds of cell lines. In the meantime, we found the −1103/−1127 TRβ binding site, −1044/−1066 PPARγ binding site and −1104/−1126 PPARγ binding site were located in this region. Besides, the luciferase activity declined at the region from −1304 bp to −1646 bp in HepG2 cells, but not in HEK293 cells, and this region contained the −1331/−1355 TRβ binding site. Considering different TREs on CPT Iα1b and CPT Iα2a promoter, TR enhanced the promoter activity of corresponding genes and might play important roles in regulating the expression of CPT Iα2a in different tissues. Moreover, we also discovered that the luciferase activity declined in the region from −2695 bp to −2276 bp of CPT Iα1b and the region from −1304 bp to −1646 bp of CPT Iα2a promoter in HepG2 cells.
Obviously, some negative regulators binding on these regions regulated CPT Iα1b and CPT Iα2a expression. In addition, studies suggested that the luciferase activity declined in the deletion region from more that −6000 bp to −1653 bp on the promoter of rat CPT Iα. However, the present study indicated that the transcription activity increased on the upstream of the promoters of grass carp CPT Iα1b and CPT Iα2a, suggesting that positive regulators existed on the upstream region of CPT Iα1b and CPT Iα2a promoters. Thus, it appears that the regulation of CPT Iα1b and CPT Iα2a transcription was more complicated in fish than mammals.
PPARs are key transcriptional factors which mediate the regulation of many enzymes related with lipid metabolism. Studies suggested that CPT I was one of the target genes of PPARα. In our study, the activities of CPT Iα1b and CPT Iα2a promoters were induced by fenofibrate, PPARα agonist. However, site-directed mutagenesis and EMSA analysis indicated that CPT Iα1b and CPT Iα2a were not regulated through those predicted PPARα binding sites. Accordingly, the reporter activities were up-regulated by fenofibrate, indicating that other potential PPARα binding sites or other related factors existed on the promoter.
For instance, studies suggested that PPARα-induced activation of CPT Iα gene was enhanced in a ligand-dependent manner by PGC-1, and PGC-1, as a co-activator, can activate gene transcription through HNF4α. Additionally, studies established the necessity of the first intron in the transcriptional regulation of the CPT Iα gene. Taken together, the induction of CPT Iα1b and CPT Iα2a by fenofibrate may involve several nuclear factors and/or other promoter regions of the gene.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
PPARγ is one of transcriptional factors which plays important roles in lipogenesis. The present study indicated that transcription of grass carp CPT I was regulated by PPARγ. Pioglitazone, the agonist of PPARγ, could increase the activity of grass carp CPT Iα2a promoter, and site-mutation on the −1044/−1066 PPARγ binding site reduced the activity. Moreover, EMSA assay confirmed that the sequence at −1044 bp to −1066 bp was a functional binding locus. Similarly, Gilde et al. found that overexpression of PPARγ in cardiomyocytes was accompanied by basal and ligand-activated transcription of the CPT I promoter.
Patsouris et al. reported that PPARγ compensated for PPARα by mediating the HFD-induced up-regulation of PPARα target genes involved in fatty acid oxidation in PPARα-null mice. Moreover, studies suggested that the addition of a classical agonist ligand promoted the dissociation of the co-repressor and the binding of co-activator proteins resulting in an enhancement in the basal transcriptional level of specific genes. Stanley found that the PPARγ-directed pioglitazone enhanced the affinity for co-activators and decreased the affinity for co-repressor on PPARγ, indicating that PPARγ possibly activated gene transcription by causing the dissociation of co-repressors and promoting the association of co-activator proteins. Similarly, the present study indicated that site-directed mutagenesis on the −1044/−1066 PPARγ binding site possibly decreased the activity of CPT Iα2a promoter, and pioglitazone-induced activation of PPARγ could up-regulate the activity of CPT Iα2a promoter. These evidences indicated that PPARγ probably played an important role in regulating CPT Iα2a transcription and compensated for PPARα-induced expression of lipolytic genes in fish.[](https://www.ncbi.nlm.nih.gov/mesh/D000077205)
In summary, the 2695-bp CPT Iα1b and 2631-bp CPT Iα2a promoters in grass carp had been cloned and characterized. The promoters of CPT Iα1b and CPT Iα2a genes showed the different structures in their core regions. Several putative TFBSs had been predicted in their promoter regions. Analysis of 5′ deletion mutants presented the regulatory characteristics of CPT Iα1b and CPT Iα2a promoters. Fenofibrate activated the activities of CPT Iα1b and CPT Iα2a promoters.
PPARγ played an important role in regulating CPT I expression. The present study provided new insights into the regulatory mechanisms of liver isoforms of CPT I genes in fish.[](https://www.ncbi.nlm.nih.gov/mesh/D011345)
## 4. Materials and Methods
*In the **4. Materials and Methods** section:*
## 4.1. Experimental Animals and Cells
*In the **4.1. Experimental Animals and Cells** section:*
Juvenile grass carp was obtained from Hubei Honghu Fisheries Farm (Jingzhou, Hubei Province, China). HepG2 and HEK293 cell lines were obtained from the Cell Resource Center in Fishery College of Huazhong Agricultural University. We ensured that the experiments were performed in accordance with the experimental protocols of Huazhong Agricultural University (HZAU) and approved by the ethics committee of HZAU (identification code: Fish-2015-0324, Date: 24 March 2015).
## 4.2. Rapid Amplification of 5′ cDNA Ends (5′ RACE)
*In the **4.2. Rapid Amplification of 5′ cDNA Ends (5′ RACE)** section:*
The TSSs of CPT Iα1b and CPT Iα2a genes were determined using the GeneRacer Kit (Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s instructions. Briefly, total RNA was isolated from the liver tissue using TRIzol Reagent (Invitrogen), and then the 5′-ready cDNA libraries were prepared using reverse transcription kit (Invitrogen). Nested PCR was performed using a commercial nested 5′ primer (Invitrogen) in combination with a reverse gene-specific primer complementary to CPT Iα1b and CPT Iα2a genes. The PCR reactions were performed using TaKaRa PrimeSTAR® HS DNA Polymerase kit (TaKaRa, Otsu, Japan) under the following PCR conditions: pre-incubation at 94 °C for 3 min, 30 cycles of 94 °C for 15 s, 60 °C for 30 s and 72 °C for 1 min. Amplified PCR products were gel-purified and subcloned into pMD19-T for sequencing (Tsingke, Wuhan, China).[](https://www.ncbi.nlm.nih.gov/mesh/C411644)
## 4.3. Cloning of Promoter and Plasmid Construction
*In the **4.3. Cloning of Promoter and Plasmid Construction** section:*
Based on the published draft genome of grass carp, we cloned the sequences of CPT Iα1b and CPT Iα2a promoters. Genomic DNA was extracted from grass carp tail fins using a commercial kit (Omega, Norcross, GA, USA). For amplification of the CPT Iα1b and CPT Iα2a promoter sequences, specific primers with overlapping sequence were designed and listed in Table 1. For the generation of the luciferase reporter constructs, the PCR product and pGl3-Basic vectors (Promega, Madison, WI, USA) were purified and digested using corresponding endonucleases, and then products were ligated using ClonExpress™ II One Step Cloning Kit (Vazyme, Piscataway, NJ, USA). According to the distance from its TSS, the plasmids were named as pGl3-CPTIα1b-2695 and pGl3-CPTIα2a-2632, respectively.
Plasmids pGl3-CPTIα1b-2276, pGl3-CPTIα1b-1716, pGl3-CPTIα1b-1073, pGl3-CPTIα1b-581, pGl3-CPTIα1b-86, pGl3-CPTIα2a-2041, pGl3-CPTIα2a-1646, pGl3-CPTIα2a-1304, pGl3-CPTIα2a-1165, pGl3-CPTIα2a-848, pGl3-CPTIα2a-428 and pGl3-CPTIα2a-97, which contained unidirectional deletions of the promoter regions, were generated with the Erase-a-Base system (Promega) using templates of pGl3-CPTIα1b-2695 and pGl3-CPTIα2a-2632, respectively. The PCR reactions were performed using TaKaRa PrimeSTAR® HS DNA Polymerase kit (TaKaRa) as mentioned above. All plasmids were sequenced in a commercial company (Tsingke).
## 4.4. Sequence Analysis
*In the **4.4. Sequence Analysis** section:*
For sequence analysis of promoters of CPT Iα1b and CPT Iα2a genes in grass carp, putative TFBSs were predicted by MatInspector online (). Nucleotide sequences of CPT Iα1b and CPT Iα2a promoters were compared with DNA sequences present in the GenBank database () and the UCSC Genome Browser ().[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
## 4.5. DNase I Foot Printing Assay
*In the **4.5. DNase I Foot Printing Assay** section:*
DNase I foot printing assays were performed based on the method of Zianni et al.. In brief, 303-bp and 346-bp proximal regions of CPT Iα1b promoter, which contained two TSSs (TSS1 and TSS2, respectively), were PCR amplified and cloned into pMD-19T vector (TaKaRa). Then, the amplicons were used as the template for further preparation of fluorescent 6-carboxy-fluorescein (FAM)-labeled probes with M13F and M13R-FAM to label the coding strand. After agarose gel electrophoresis, the FAM-labeled probes were purified by Gel Extraction Kit (Omega, USA) and quantified with NanoDrop 2000 (Thermo, Waltham, MA, USA). 10 µg of proteins extracted from HepG2 cell lines were incubated with 500 ng of probes in the same binding buffer based on Zianni et al..
DNase I digestion was performed for 3 min at room temperature and then terminated by the addition of DNase I stop solution (Promega). Digested samples were precipitated with alcohol and then analyzed with the 3730 DNA Analyzer in the commercial company (Tsingke).[](https://www.ncbi.nlm.nih.gov/mesh/C024098)
## 4.6. Transfections and Luciferase Assays
*In the **4.6. Transfections and Luciferase Assays** section:*
HepG2 and HEK293 cells were cultured in DMEM medium supplemented with 10% (v/v) heat-inactivated FBS (Invitrogen) and 2 mM l-glutamine in a humidified atmosphere with 5% CO2 at 37 °C. Prior to the transient transfection, HepG2 or HEK293 cells were seeded in 24-well cell culture plate at a density of 1.2 × 105 and cultured for 24 h to reach 70–80% convergence. Plasmids were transiently transfected into HepG2 or HEK293 cells using Lipofectamine™ 2000 (Invitrogen) following the manufacture’s protocol. All reporter plasmids were used in equimolar amounts in Opti-MEM (Invitrogen), and they were co-transfected with 35 ng pRL-TK as the control. After 4 h, the transfection medium was replaced by 10% FBS-DMEM.
Then, with 24-h incubation, cells were harvested to assay the relative luciferase activity by Dual-Luciferase Reporter Assay System (Promega). The relative luciferase activity was presented as the ratio of firefly luciferase to renilla luciferase. Results were normalized to the control reporter pGl3-Basic. All experiments were performed in triplicates and measured at least three times.[](https://www.ncbi.nlm.nih.gov/mesh/D005973)
## 4.7. Site-Mutation Analysis of PPAR Binding Sites on the Grass Carp CPT Iα1b and CPT Iα2a Promoters
*In the **4.7. Site-Mutation Analysis of PPAR Binding Sites on the Grass Carp CPT Iα1b and CPT Iα2a Promoters** section:*
To identify the corresponding PPAR binding sites on the grass carp CPT Iα1b and CPT Iα2a promoters, we performed site-directed mutagenesis according to the manufacture instruction of QuickChange II Site-Directed Mutagenesis Kit (Vazyme). pGl3-CPTIα1b-2276, pGl3-CPTIα2a-2041 and pGl3-CPTIα2a-1304 were used as templates. The mutagenesis primers were listed in Table 1, and the PCR reactions were performed as mentioned above. These mutant constructs were named as 1bMut-PPAR1, 1bMut-PPAR2, 2aMut-PPAR1, 2aMut-PPAR2, 2aMut-PPAR3, 2aMut-PPAR4, 2aMut-2PPAR1, 2aMut-2PPAR2, 2aMut-2PPAR3 and 2aMut-3PPAR, respectively. Then the constructs and pRL-TK were co-transfected into HepG2 cell lines using LipofectamineTM 2000 following the manufacture’s protocol.
After 4 h, the transfection medium was replaced by 10% FBS-DMEM with 50 μM fenofibrate or 10 μM pioglitazone. After 24-h incubation, cells were harvested to assay the luciferase activity according to the procedure above.[](https://www.ncbi.nlm.nih.gov/mesh/C086724)
## 4.8. Electrophoretic Mobility-Shift Assay (EMSA)
*In the **4.8. Electrophoretic Mobility-Shift Assay (EMSA)** section:*
EMSA was performed to confirm the functional PPAR binding sites of the promoters. Proteins for EMSA were extracted from HepG2 cell lines. Nuclear and cytoplasmic extracts were prepared based on the methods of Read et al.. Protein concentrations were determined by the BCA method. These extracts were stored at −20 °C until analyzed. Each oligonucleotide duplex of PPAR binding sites was incubated with 10 µg nuclear extracts at room temperature according to the instruction of LightShift™ Chemiluminescent EMSA Kit (Invitrogen), and each unlabeled probe was pre-incubated 10 min prior to the addition of biotin-labeled probe.
The reaction was allowed to proceed for 30 min after the addition of biotin-labeled probe at room temperature, and then were detected by electrophoresis on 6% native polyacrylamide gels. Competition analyses were performed by using 100-fold excess of unlabeled oligonucleotide duplex with or without the mutation. All the oligonucleotide sequences of EMSA were listed in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
## 4.9. Statistical Analysis
*In the **4.9. Statistical Analysis** section:*
Results were presented as mean ± SEM (standard errors of means) in at least three independent biological experiments. Prior to statistical analysis, all data were tested for normality of distribution using the Kolmogornov-Smirnov test. Differences between wild types and drug-treated groups were compared using the Student’s t test. Difference was considered significant at p < 0.05. All statistical analyses were performed using the SPSS10.0 for Windows (SPSS, Michigan Avenue, Chicago, IL, USA).
# Supplementary Materials
*In the **Supplementary Materials** section:*
# Author Contributions
*In the **Author Contributions** section:*
Zhi Luo and Yi-Huan Xu designed the experiment; Yi-Huan Xu conducted the experiment with the help of Kun Wu, Yao-Fang Fan, Wen-Jing You and Li-Han Zhang; Yi-Huan Xu, Kun Wu and Zhi Luo analyzed the data; Yi-Huan Xu drafted the manuscript and Zhi Luo revised the manuscript. All the authors read and approved the manuscript.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflicts of interest with the contents of this article.
# Abbreviations
*In the **Abbreviations** section:*
CPT I carnitine palmitoyltransferase I FAM fluorescent 6-carboxy-fluorescein L-CPT liver carnitine palmitoyltransferase LCFA long-chain fatty acid M-CPT muscle carnitine palmitoyltransferase PCR polymerase chain reaction PPAR peroxisome proliferator-activated receptor PPRE peroxisome proliferator-responsive element RLM-5′RACE RNA ligase-mediated rapid amplification of 5′ cDNA ends RXR retinoid X receptor TFBS transcription factor binding site TR thyroid hormone receptor TRE thyroid hormone response elements[](https://www.ncbi.nlm.nih.gov/mesh/C024098)
# References
*In the **References** section:*
A map of the first two exons in the CPT Iα1b and CPT Iα2a genes was shown. Exons were denoted by black rectangles, introns by a fold line and transcriptional direction (5′-3′) by an arrow line. The initiation codon (ATG) in exon 2 represented the start site of protein translation. Numbers were relative to the distance from transcription start site (+1). (A) structure of transcription start site (TSS) of CPT Iα1b gene (B) structure of alternative splicing transcription start site (TSS’) of CPT Iα1b gene (C) structure of transcription start site of CPT Iα2a gene.
DNase I foot printing assay of proximal promoter of CPT Iα1b. (A) 303-bp proximal promoter region of CPT Iα1b (B) 346-bp proximal promoter region of CPT Iα1b. The sequence used for FAM-labeled probe was presented, based on the result of DNase I foot printing. Putative binding sequence was underlined and italicized with labels. Capital letters indicate the coding sequence of proximal promoter region of CPT Iα1b, and lowcase letters indicate the partial sequence of pMD-19T vector. The primer sequences used for DNase I foot printing assay M13F and M13R-FAM were labeled by arrows.
Nucleotide sequence of grass carp CPT Iα1b promoter. +1 denoted the transcription start site (TSS1) obtained from RLM-5′RACE experiment. TSS2 presented another transcription start site (−346, TSS′). Numbers indicated the distance from TSS1. The highlighted sequences show putative transcription factor binding sites.[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
Nucleotide sequence of grass carp CPT Iα2a promoter. +1 denotes the TSS obtained from RLM-5′RACE experiment. Numbers present distance from TSS. The highlighted sequences show putative transcription factor binding sites.[](https://www.ncbi.nlm.nih.gov/mesh/D009711)
5′ Unidirectional deletion analysis of the CPT Iα1b and CPT Iα2a promoter regions for grass carp. Schematic diagrams of truncated promoters were shown at the left panel. The corresponding luciferase reporter assay results were shown in the right panel. Promoter activity of constructs is presented with the values of relative light unit. A series of plasmids containing 5′ unidirectional deletions of the CPT Iα1b promoter region fused in frame to the luciferase gene were transfected into HepG2 cells (A) and HEK293 cells (B), and a series of plasmids containing 5′ unidirectional deletions of the CPT Iα2a promoter region were transfected into HepG2 cells (C) and HEK293 cells (D). Values represent the ratio between firefly and renilla luciferase activities, normalized to the control plasmid pGl3-Basic. Results were expressed as the mean ± SEM of three independent experiments (Student’s t-test, * p < 0.05).
Analysis of putative PPAR binding sites by site-directed mutagenesis. Site-mutation constructs are presented in the left panel. Promoter activity of constructs is presented in the middle. Promoter activity treated with agonist was presented in the right panel. (A) site-mutations of PPARα binding sites on pGl3-CPTIα1b-2276 and pGl3-CPTIα2a-2041 vectors (B) site-mutation of PPARγ binding sites on pGl3-CPTIα1b-2276 and pGl3-CPTIα2a-1304 vectors. Values represent the ratio between firefly and renilla luciferase activities, normalized to the control plasmid pGL3-Basic. Bars are the mean ± SEM of three independent experiments (Student’s t-test, * p < 0.05).
Electrophoretic mobility-shift assay (EMSA) of putative PPAR binding sequences. The 5′-biotin labeled double-stranded oligomers were incubated with HepG2 nuclear extract (NP). A 100-fold excess of the competitor and Mutative competitor oligomers was added to the competition and mutant competition assay, respectively. The oligonucleotide sequences are given in Table 1. (A) PPARα/RXR binding sequence located at −1814 bp to −1836 bp of CPT Iα1b promoter (B) PPARγ binding sequence located at −1719 bp to −1741 bp of CPT Iα1b promoter (C) PPARα/RXR binding sequence located at −1939 bp to −1961 bp of CPT Iα2a promoter (D) PPARγ binding sequence located at −1179 bp to −1201 bp of CPT Iα2a promoter (E) PPARγ binding sequence located at −1104 bp to −1136 bp of CPT Iα2a promoter (F) PPARγ binding sequence located at −1044 bp to −1066 bp of CPT Iα2a promoter.[](https://www.ncbi.nlm.nih.gov/mesh/D001710)
Primers used in the experiments.
|
# Introduction
Protective effects of [Mentha piperita L. leaf essential oil](https://www.ncbi.nlm.nih.gov/mesh/C015424) against [CCl4](https://www.ncbi.nlm.nih.gov/mesh/D002251) induced hepatic oxidative damage and renal failure in rats
# Abstract
*In the **Abstract** section:*
Background
Mentha piperita L. is a flowering plant belonging to the Lamiaceae family. Mentha plants constitute one of the main valuable sources of essential oil used in foods and for medicinal purposes.[](https://www.ncbi.nlm.nih.gov/mesh/D009822)
Methods
The present study aimed to investigate the composition and in vitro antioxidant activity of Mentha piperita leaf essential oil (MpEO). A single dose of CCl4 was used to induce oxidative stress in rats, which was demonstrated by a significant rise of serum enzyme markers. MpEO was administrated for 7 consecutive days (5, 15, 40 mg/kg body weight) to Wistar rats prior to CCl4 treatment and the effects on serum alanine aminotransferas[e (ALT), aspartate aminotransferas](https://www.ncbi.nlm.nih.gov/mesh/C015424)e [(AST](https://www.ncbi.nlm.nih.gov/mesh/C015424)), alkaline phosphat[ase ](https://www.ncbi.nlm.nih.gov/mesh/D002251)(ALP), lactate dehydrogenase (LDH), and γ -glutamyl transpeptidase (γ-GT) levels, as well as the liver and kidney su[pero](https://www.ncbi.nlm.nih.gov/mesh/C015424)xide dismutase (SOD), catalase (CAT) and glutathione peroxidase (GPx) activity and thiobarbituri[c ac](https://www.ncbi.nlm.nih.gov/mesh/D002251)id reactive substances (TBARS) levels were evaluated. In addition, histopathological examinations of livers and kidneys was performed.[](https://www.ncbi.nlm.nih.gov/mesh/D017392)
Results
The in vitro antioxidant activity of MpEO was lower than that of silymarin. Pretreatment of animals with MpEO at a dose of 5 mg/kg did not have a significant effect on ALT, AST, ALP, LDH, γGT, urea or creatinine levels in CCl4-induced stress. Whereas pretreatment with MpEO at doses of 15 and 40 mg/kg prior to CCl4, significantly reduced stress parameters (ALT, AST, ALP, LDH, γGT, urea and creatinine) compared to the CCl4-only group. Moreover, a significant reduction in hepatic and kidney lipid peroxidation (TBARS) and an increase in antioxidant enzymes SOD, CAT and GPx was also observed after treatment with MpEO (40 mg/kg) compared to CCl4-treated rats. Furthermore, pretreatment with MpEO at 40 mg/kg can also markedly ameliorate the histopathological hepatic and kidney lesions induced by administration of CCl4.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
Conclusions
We could demonstrate with this study that MpEO protects liver and kidney from CCl4-induced oxidative stress and thus substantiate the beneficial effects attributed traditionally to this plant.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Background
*In the **Background** section:*
Reactive oxygen species (ROSs) are various forms of activated oxygen. A disproportion of the reactive oxygen species and the absence of their scavenge systems in cells lead to oxidative stress and increases the risk of several human chronic diseases []. ROS contributes to the development of various diseases such as diabetes, atherosclerosis, cancer, neurodegenerative diseases, liver cirrhosis and the aging process []. The liver plays a central role in the maintenance of systemic lipid homeostasis and is especially susceptible to ROS damage.
CCl4 is now of greatest concern as an environmental contaminant []. It was reported that CCl4 was one of the most commonly used toxins in the experimental study of liver diseases []. Abraham et al. [] showed that the nephrotoxic effects of CCl4 were also associated with free radical production.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
To prevent the damage caused by ROS, living organisms have developed an antioxidant defense system that includes the presence of non-enzymatic antioxidants and enzymes such as catalase (CAT), superoxide dismutase (SOD) and glutathione peroxidase (GPx) []. It has been anticipated that in addition to these natural antioxidants, other synthetic or natural ROS scavengers may reduce the incidence of free radical-mediated diseases. The use of antioxidants in the prevention and cure of various diseases is intensifying, and there is considerable interest in the study of the antioxidant activities of molecules such as plant polyphenolic and carotenoid components [, ].
Antioxidants appear to act against disease processes by increasing the levels of endogenous antioxidant enzymes and decreasing lipid peroxidation [].[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
A number of studies showed that various herbal extracts could protect liver and kidney against CCl4-induced oxidative stress by inhibiting lipid peroxidation and enhancing antioxidant enzyme activity []. Silymarin, a flavonolignan mixture of milk thistle (Silybum marianum), is one such important herbal hepatoprotective drug. Silymarin exhibits hepatoprotective effects by altering cytoplasmic membrane architecture and, in turn, preventing the penetration of hepatotoxic substances, such as carbon tetrachloride (CCl4), thioacetamide and D-galactosamine [].[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
The well-known and widely used peppermint (Mentha piperita L.) (Lamiaceae) is a cultivated natural hybrid of Mentha aquatica L. (water mint) and Mentha spicata L. (spearmint). Although a native genus of the Mediterranean region, it is cultivated all over the world for its use in flavor, fragrance, medicinal, and pharmaceutical applications. Peppermint oil is one of the most widely produced and consumed essential oils [, ]. Besides its uses in food, herbal tea preparations, and confectioneries, the medicinal uses of mint, which date back to ancient times, include carminative, anti-inflammatory, antispasmodic, antiemetic, diaphoretic, analgesic, stimulant, emmenagogue, and anticatharrhal application.
It is also used against nausea, bronchitis, flatulence, anorexia, ulcerative colitis, and liver complaints. Mint essential oils are generally used externally for antipruritic, astringent, rubefacient, antiseptic, and antimicrobial purposes, and for treating neuralgia, myalgia, headaches, and migraines [, ].[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
From the experimental and clinical studies performed on Mentha piperita leaf essential oil (MpEO), it seems that most of its pharmacological actions are due to its antioxidant activity which is mainly due to its ability to scavenge free radicals and/or inhibit lipid peroxidation [, ]. Antioxidants are substances that delay or prevent the oxidation of inter- or intra-cellular oxidizable substrates from oxidative stress. In this study, we report the chemical composition and antioxidant effects of MpEO in several in vitro systems (DPPH and superoxide scavenging activities). Besides, we are interested in determining the possible protective effects of MpEO against oxidative damage of the liver and kidney following an intraperitoneal administration of CCl4, by assessing the oxidative stress profile and some serum biochemical parameters.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Methods
*In the **Methods** section:*
## Plant material
*In the **Plant material** section:*
Fresh leaves of M. piperita L. samples were harvested from the local market at Sfax (Tunisia) (N: 34.4426°, E: 10.4537°) during the vegetative stage in June 2013. The samples were identified and authenticated by a senior botanist, Pr. Ferjani Ben Abdallah, at the Faculty of Science of Sfax, University of Sfax (Tunisia). From 50 individual M. piperita L. plants each, a total of 80–100 leaves (≈ 12 cm2 in size) were randomly collected from the base to the apex. The fresh leaves were mixed and immediately dried in the shade away from light at room temperature.
After drying, the samples were grounded to a fine powder that was used for the extraction of essential oil.[](https://www.ncbi.nlm.nih.gov/mesh/D009822)
## Essential oil preparation
*In the **Essential oil preparation** section:*
MpEO was extracted by the steam distillation method. A mass of 3 kg of dry plant material was hydrodistillated for 2 h in a Clevenger-type apparatus. The recovered (0.47%) essential oil was dried with anhydrous Na2SO4, and stored at 4 °C.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Mentha piperita essential oil composition
*In the **Mentha piperita essential oil composition** section:*
MpEO compositional analysis of the volatile constituents was performed on a Hewlett-Packard gas chromatograph GC: 5890 series II. The fused HP-Innowax capillary column (polyethylene glycol, 30 m, 0.25 μm, ID, 0.25 mm film thickness) was directly connected to the mass spectrometer. Nitrogen was used as a carrier gas at a flow rate of 1.2 ml/min. Oven temperature was initially set at 50 °C (1 min) and gradually raised to 250 °C (5 min) at 7 °C/min. The temperatures of the injection port and detector were maintained at 250 and 280 °C, respectively.
The mass spectrometer was operated (full scan-mode) in the EI-mode at 70 eV.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Component identification
*In the **Component identification** section:*
The essential oil components were identified based on their mass spectra and computer matching with the data available in the Wiley 275 library (Wiley, New York).[](https://www.ncbi.nlm.nih.gov/mesh/D009822)
## In vitro antioxidant activities test
*In the **In vitro antioxidant activities test** section:*
The antioxidant activity of the MpEO was determined by two methods and compared with the activity of silymarin, a standardized extract of the milk thistle seeds that containes a mixture of flavonolignans. Silymarin has a number of potential mechanisms including chemoprotective effects from environmental toxins and anti-inflammatory activity and is used as a drug.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Measurement of free radical-scavenging action
*In the **Measurement of free radical-scavenging action** section:*
2,2-Diphenyl picrylhydrazyl (DPPH) free radicals scavenging activity was assessed according to Blois [], with a slight modification. Different concentrations of the MpEO and silymarin (5–100 μg/ml) were mixed with 1 ml of 0.1 mM DPPH in ethanol solution and 450 μL of 50 mM Tris-HCl buffer (pH 7.4) was added. The solution was incubated at 37 °C for 30 min and the reduction of DPPH free radicals was measured by reading the absorbance at ʎ = 517 nm. Silymarin was used as reference standard.
The activity is given as % DPPH scavenging and calculated according to the following equation:[](https://www.ncbi.nlm.nih.gov/mesh/C004931)
The antioxidant activity of MpEO is expressed as IC50, defined as the concentration of MpEO required to cause a 50% decrease in initial DPPH concentration. Each sample was analyzed six times.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Scavenging of superoxide anion
*In the **Scavenging of superoxide anion** section:*
The influence of MpEO on the generation of superoxide anion was measured according to the method described by Yen & Chen, 1995 []. Superoxide anion was generated in a non-enzymatic system and determined by spectrophotometric measurement for the reduction of nitroblue tetrazolium. The reaction mixture, which contained 100 μL of essential oil in ethanol, 800 μL of 1 M phosphate buffer (pH 7.4), 400 μL of distilled water, 100 μL of 0.1 M Na4EDTA, 100 μL of 1.5 mM NBT and 50 μL of 0.12 mM riboflavin was incubated at ambient temperature for 5 min, and the color was read at ʎ = 560 nm against blank samples.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
Where blank OD is the absorbance of the control reaction and sample OD is the absorbance in the presence of MpEO. The IC50 was calculated from the plot of the inhibition percentage against the essential oil concentration. Each sample was analyzed six times.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## In vivo antioxidant properties
*In the **In vivo antioxidant properties** section:*
## Animal
*In the **Animal** section:*
Male Wistar rats, weighing about 200–220 g, were purchased from the Central Pharmacy of Tunisia (SIPHAT, Tunisia). They were housed at 22 ± 3 °C with light/dark periods of 12 h and a minimum relative humidity of 40%. The animals had free access to commercial pellet diet (SICO, Sfax, Tunisia) and water ad libitum. The general guidelines for the use and care of living animals in scientific investigations were followed []. The handling of the animals was approved by the Tunisian Ethical Committee for the Care and Use of laboratory animals.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Experimental design
*In the **Experimental design** section:*
After acclimatizing to the laboratory conditions for 1 week, 70 rats were divided into 7 groups of 10 animals and treated for 7 days as follow []:
The rats of group 1 served as normal control and received saline orally daily for 7 days and were injected with 1 ml/kg BW of just olive oil (the solvent of CCl4) on the 7 day. The rats of group 2 served as CCl4-hepato and renotoxicity control and were received saline orally daily for 7 days and were injected with 1 ml/kg BW of CCl4 and olive oil mixture on the 7 day (a single intraperitoneal injection). The CCl4 dose was selected according to the reference dose for chronic oral exposure (RFD) as recommended for CCl4 (CASRN 56–23-5) [].[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
The rats of group 3 were pretreated orally seven times with a dose of 50 mg/kg BW of reference drug silymarin with an interval of 24 h [].[](https://www.ncbi.nlm.nih.gov/mesh/D012838)
The rats of groups 4, 5, 6 and 7 were pretreated orally seven times with doses of 5, 15 and 40 mg MpEO /kg BW, respectively with an interval of 24 h [].[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
After pretreatment with either silymarin or MpEO for 7 days, the rats of groups 3, 4, 5 and 6 received a single intraperitoneal injection of CCl4 (1 ml/kg BW) on the 7 day.[](https://www.ncbi.nlm.nih.gov/mesh/D012838)
Rats were killed 24 h after vehicle or CCl4 single injection. The animals in the different groups were killed by cervical decapitation to avoid stress conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
## Sample collection
*In the **Sample collection** section:*
Serum was prepared by centrifugation (1500×g, 15 min, 4 °C; Beckman-Coulter, Marseille, France) and stored at −80 °C for further biochemical assays. The liver and kidney tissues were immediately removed and dissected over ice-cold glass slides and a part was homogenized (10% w/v) with an Ultra Turrax homogenizer in ice-cold, 1.15% KCl-0.01 M sodium, potassium phosphate buffer. Homogenates were centrifuged at 10000×g for 20 min at 4 °C. The resulting supernatants were used for immediate lipid peroxidation and protein oxidation determination. Homogenate aliquots were stored at −80 °C for further biochemical assays.
Other parts of these livers and kidney tissues were fixed in 10% formaldehyde solution and processed for paraffin sectioning and histological studies.[](https://www.ncbi.nlm.nih.gov/mesh/D011189)
## Biochemical assays
*In the **Biochemical assays** section:*
## Biochemical markers in plasma
*In the **Biochemical markers in plasma** section:*
Plasma levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), γ-glutamyl transpeptidase (γ-GT), cholesterol (TC), triglycerides (TG), low-density lipoprotein (LDL), high-density lipoprotein (HDL), creatinine and urea rates were measured in plasma samples by standardized enzymatic procedures using commercial kits from (Biolabo, Maizy, France) on an automatic biochemistry analyzer (Vitalab Flexor E, Diamond Diagnostics, Holliston, MA).[](https://www.ncbi.nlm.nih.gov/mesh/D002784)
## Protein quantification
*In the **Protein quantification** section:*
Protein content in liver and kidney tissues were determined according to the method of Lowry et al. [] using bovine serum albumin as a standard.
## Lipid peroxidation
*In the **Lipid peroxidation** section:*
Malondialdehyde concentrations (marker for lipid peroxidation) in liver and kidney tissues were determined spectrophotometrically according to Draper & Hadley []. Briefly, an aliquot of liver and kidney extracts supernatant was mixed with 1 ml of 5% trichloroacetic acid and centrifuged at 2500×g for 10 min. One ml of thiobarbituric acid reagent (0.67%) was added to 500 μl of supernatant and heated at 90 °C for 15 min. The mixture was cooled and the absorbance measured at 532 nm using a spectrophotometer (Jenway UV-6305, Essex, England). The malondialdehyde values were calculated using 1,1,3,3-tetraethoxypropane as standard and expressed as nmol of malondialdehyde/mg of protein.[](https://www.ncbi.nlm.nih.gov/mesh/D008315)
## Determination antioxidant enzyme activities in liver and kidney tissue
*In the **Determination antioxidant enzyme activities in liver and kidney tissue** section:*
Catalase (CAT) activity was measured according to Aebi []. A total of 20 μL tissue homogenate (about 1.5 mg proteins) was added to 1 ml phosphate buffer (0.1 M, pH 7) containing 100 mM H2O2. Rate of H2O2 decomposition was followed by measuring the decrease in absorbance at 240 nm for 1 min. The enzyme activity was calculated using an extinction coefficient of 0.043 mM−1 cm−1 and expressed in international units (I.U.), i.e. in μmol H2O2 destroyed/min/ mg protein, at 25 °C.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
Superoxide dismutase (SOD) activity was estimated according to Beyer and Fridovich []. The reaction mixture contained 50 mM of tissue homogenates in potassium phosphate buffer (pH 7.8), 0.1 mM EDTA, 13 mM L-methionine, 2 mM riboflavin and 75 mM nitro blue tetrazolium (NBT). The developed blue color in the reaction was measured at 560 nm. Units of SOD activity were expressed as the amount of enzyme required to inhibit the reduction of NBT by 50% and the activity was expressed as units/mg of protein, at 25 °C.
Glutathione peroxidase (GPx) activity was measured by the procedure of Flohe and Gunzler []. One milliliter of reaction mixture containing 0.3 ml of phosphate buffer (0.1 M, pH 7.4), 0.2 ml of 2 mM glutathione (GSH), 0.1 ml of sodium azide (10 mM), 0.1 ml of H2O2 (1 mM) and 0.3 ml of liver and kidney supernatant were prepared. After incubation at 37 °C for 15 min, the reaction was terminated by adding 0.5 ml 5% TCA. Tubes were centrifuged at 1500×g for 10 min and the supernatant was collected.
To 0.1 ml of this reaction supernatant, 0.2 ml of (0.1 M pH 7.4) and 0.7 mL of 5,5 dithiobis-(2-Nitrobenzoic acid) (DTNB, 0.4 mg/ml) were added. After mixing, absorbance was recorded at 420 nm and the enzyme activity was calculated as μmol of GSH oxidized/min/mg protein.[](https://www.ncbi.nlm.nih.gov/mesh/C013216)
## Histopathological studies
*In the **Histopathological studies** section:*
At the time of sacrifice, the liver and kidney tissues were removed and fixed in 10% formaldehyde solution and washed. The tissues were dehydrated in increasing gradient of ethanol, finally cleared in toluene and embedded in molten paraffin wax. Sections were cut at 4–5 μm thickness and stained with hematoxylin and eosin (H&E). The slides were photographed with an Olympus UTU1X-2 camera connected to an Olympus CX41 microscope (Tokyo, Japan).[](https://www.ncbi.nlm.nih.gov/mesh/D005557)
The histological damage in liver was quantified by measuring the index of tissue large numbers of inflammatory cells such as lymphocytes together with hepatic sinusoidal inflammation, hepatocyte necrosis and devastating liver architecture. Moreover, the histological damage in kidney was quantified by measuring the index of tissue the glomerular and tubular necrosis. To evaluate the severity of lesions, the degree of liver and kidney damage was graded according to a zero to four-point scoring system [], where 0 indicates no damage, I indicates slight damage (1–25%), II indicates discrete damage (26–50%), III indicates moderate damage (51–75%) and IV indicates severe damage (76–100%).The tabulation of data and the statistical analysis were made in accordance with the number of animals with established scores. All the parameters were quantified by a single observer who was not aware of the treatment groups.
## Statistical analysis
*In the **Statistical analysis** section:*
All values are expressed as mean ± SE for continues variables or as median with inter quartile range [25%, 75%] where appropriate. The results were analyzed by One-Way Analysis of Variance (ANOVA) followed by Tukey test for multiple comparisons using SPSS for Windows (version. 12) or ANOVA-on-ranks with Dunn’s correction. Differences were considered significant at p < 0.05.
## Results
*In the **Results** section:*
## Chemical constitution of Mentha piperita L. leaf essential oil
*In the **Chemical constitution of Mentha piperita L. leaf essential oil** section:*
Chemical composition (%) of leaves essential oil from Tunisian M.piperita as identified by GC/MS analysis[](https://www.ncbi.nlm.nih.gov/mesh/D009822)
Chemical composition of MpEO was determined by GC/MS analysis. The compounds, their percentages as well as their retention indices are listed in Table 1. MpEO is a mixture with 26 compounds representing 98.17% of the total oil composition. The most abundant chemicals categories for MpEO are oxygenated monoterpenes (79.50%), followed by monoterpene hydrocarbons (16.23%) and sesquiterpene hydrocarbons (2.44%). The major components of MpEO are menthol (33.59%) and iso-menthone (33.00%).
In lower amounts we found a variety of compounds including limonene (8.00%), piperitone (3.20%), 1,8-cineole (2.80%), linalool (2.64%), iso-pulegol (2.40%), caryophyllene (1.95%) and pulegone (1.60%).[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Essential oil antioxidant activity
*In the **Essential oil antioxidant activity** section:*
MpEO effects and positive controls on the in vitro free radical(DPPH and superoxide)[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
The antioxidant activity of MpEO was compared to that of silymarin, a well-known antioxidant, using two different assays, namely DPPH and superoxide oxygen radicals inhibition, the results are reported in Table 2. DPPH showed for MpEO an IC50 value around 3 times higher than the one recorded for silymarin indicating that antioxidant activity of MpEO was lower than that of silymarin.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Serum biochemical parameters
*In the **Serum biochemical parameters** section:*
Effects of CCl4, MpEO and their combination MpEO/CCl4 on hepatic markers in serum of control and experimental rats[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
Effects of CCl4,MpEO and their combination MpEO/CCl4 on lipid profile in serum of control and experimental rats[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
Effects of CCl4, MpEO and their combination MpEO/CCl4 on kidney markers in serum of control and experimental rats[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
The results of biochemical indicators of liver and kidney function are summarized in Tables 3, 4 and 5. The administration of CCl4 caused severe hepato and reno-toxicity in the treated rats, as evidenced by the significant elevations of serum ALT, AST, ALP, LDH, γGT, total cholesterol, triglycerides, LDL urea and creatinine levels, while HDL level was decreased compared to control animals.[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
Pretreatment with the MpEO at doses of 15 or 40 mg/kg significantly reduced levels of ALT, AST, ALP, LDH, γGT, total cholesterol, triglycerides, LDL urea and creatinine and increased the level of HDL compared to the CCl4 group. It is worth noting that the treatment with 5 mg/kg MpEO did not induce any significant changes in the biochemical parameters (ALT, AST, ALP, LDH, γGT, total cholesterol, triglycerides, LDL, urea, creatinine or HDL) when compared to the CCl4 group. Treatment of rats with only MpEO (40 mg/kg BW) did not result in significant alterations in biochemical parameters compared to control rats.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
Pretreatment with silymarin (50 mg/kg), used as positive control, significantly decreased the elevated levels of ALT, AST, ALP, LDH, γGT, total cholesterol, triglycerides, LDL urea and creatinine and increased of HDL level as compared to CCl4 group. Its effect was comparable in reducing of liver and kidney damage induced by CCl4 with that observed for the highest dose of MpEO (40 mg/kg).[](https://www.ncbi.nlm.nih.gov/mesh/D012838)
## Effects on lipid peroxidation
*In the **Effects on lipid peroxidation** section:*
(
)
a Effects of CCl4, MpEO and their combinations MpEO/CCl4 on hepatic TBARS of control (Con) and experimental rats. Con, control group; mod, CCl4-model group; SL/CCl4, silymarin 50 mg/kg + CCl4; MpEO/CCl4 5 mg/kg + CCl4 group; MpEO/CCl4 15 mg/kg + CCl4 group; MpEO/CCl4 40 mg/kg + CCl4 group; MpEO 40 mg/kg group. Values are mean ± SEM for ten rats in each group. CCl4, MpEO, MpEO/CCl4 treated groups vs control group; *p < 0.05, **p < 0.01, *** p < 0.001, CCl4 group vs (MpEO/CCl4) group; #p < 0.05, ##p < 0.01, ###p < 0.001. b. Effects of CCl4, MpEO and their combinations MpEO/CCl4 on kidney TBARS of control (Con) and experimental rats. Con, control group; mod, CCl4-model group; SL/CCl4, silymarin 50 mg/kg + CCl4; MpEO/CCl4 5 mg/kg + CCl4 group; MpEO/CCl4 15 mg/kg + CCl4 group; MpEO/CCl4 40 mg/kg + CCl4 group; MpEO 40 mg/kg. Values are mean ± SEM for ten rats in each group. CCl4, MpEO, MpEO/CCl4 treated groups vs control group; *p < 0.05, **p < 0.01, *** p < 0.001, CCl4 group vs (MpEO/CCl4) group; #p < 0.05, ##p < 0.01, ###p < 0.001[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
TBARS level is widely used as a marker for free radical mediated lipid peroxidation injury. We determined TBARS levels in the liver and kidney tissues of the investigated animals and our results are shown in Fig. 1a, b. The levels of TBARS were significantly increased in both liver and kidney tissues of CCl4-treated animals when compared to control untreated rats.[](https://www.ncbi.nlm.nih.gov/mesh/D017392)
Pre-treatment with the MpEO at doses of 15 and 40 mg/kg BW significantly reduced levels of TBARS in liver and kidney tissues as compared to CCl4 group. There was a dose effect; treatment with MpEO at 5 mg/kg BW did not induce any significant decrease in the levels of TBARS in liver and kidney as compared to CCl4 group. When rats were treated with only MpEO (40 mg/kg BW), no significant differences in the TBARS values was observed compared to control rat. Pretreatment with silymarin (50 mg/kg) significantly decreased the elevated levels of TBARS in both liver and kidney compared to CCl4 control. Moreover, the effect of silymarin (50 mg/kg) in attenuation of TBARS levels in liver and kidney was comparable with highest dose of MpEO (40 mg/kg).[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
## Effects on antioxidant enzymes
*In the **Effects on antioxidant enzymes** section:*
Effects of CCl4, MpEO and their combination MpEO/CCl4 on the activities of enzymatic antioxidants in liver of control and experimental rats[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
Effects of CCl4, MpEO and their combination MpEO/CCl4 on the activities of enzymatic antioxidants in kidney of control and experimental rats[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
Results presented in Tables 6 and 7 showed a significant decrease in the levels of CAT, SOD, GPx in liver and kidney in CCl4-treated group when compared to control group. The decrease in hepatic and kidney CAT, SOD and GPx levels induced by CCl4 injection were significantly restored (elevated) in the MpEO and silymarin groups, and this effect was more pronounced with the increase of essential oil concentration. Pretreatment with MpEO at doses of 15 and 40 mg/kg significantly increased levels of hepatic and kidney CAT, SOD and GPx as compared to CCl4 group. It is worth noting that the treatment with MpEO at a dose 5 mg/kg did not induce any significant increase in the levels of hepatic and kidney CAT, SOD and GPx as compared to CCl4 group. No significant differences in the values were observed in rats treated with MpEO only (40 mg/kg) compared to control rat values.
Moreover, the effect of silymarin (50 mg/kg) was comparable in attenuation of levels of hepatic and kidney CAT, SOD and GPx with highest dose of MpEO (40 mg/kg).[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
## Histopathological findings
*In the **Histopathological findings** section:*
Effect of MpEO on CCl4-induced liver damage. a Control group; (b) CCl4-model group showing marked inflammatory cells, necrosis and reduced lesions of necrosis; (c) Silymarin 50 mg/kg + CCl4 group; (d) MpEO 5 mg/kg + CCl4 group (e); MpEO 15 mg/kg + CCl4 group; (f) MpEO 40 mg/kg + CCl4 group; (g) MpEO 40 mg/kg group. Hematoxylin/eosin staining; magnification ×400. : Marked inflammatory cells; : Necrosis cells; : Reduced lesions of necrosis; : Regeneration area; : Prominent nucleolus; : Mild inflammatory cells[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
Grades of inflammatory cells and cellular necrosis in rat liver
The normal liver architecture was observed in liver histology of control group (Fig. 2a). Large numbers of inflammatory cells such as lymphocytes together with hepatic sinusoidal inflammation, hepatocyte necrosis and devastating liver architecture were observed in the CCl4 group (Fig. 2b). However, pretreatment with MpEO (40 mg/kg, Fig. 2f) can remarkably ameliorate the histopathological hepatic lesions induced by administration of CCl4. MpEO 15 mg/kg showed very few inflammatory cells along with prominent nucleolus (Fig. 2e).
The highest dose of MpEO (40 mg/kg) and silymarin (50 mg/kg) significantly attenuated the damaged liver depicting marked focal regenerative changes which are illustrated by presence of actively dividing cells with a prominent nucleolus (Fig. 2f). In addition, silymarin at the dose of 50 mg/kg has shown to produce hepatoprotection evidenced by area of regeneration and dark nucleus (Fig. 2c). The histological pattern was almost normal in rats treated with MpEO oil alone. By analyzing the histopathological scoring attributed to the liver tissues it is possible to note that the highest dose of MpEO (40 mg/kg) or silymarin (50 mg/kg) pretreatment just conferred good protection on CCl4-induced liver damage (Table 8).[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
Effect of MpEO on CCl4-induced kidney damage. a Control group; (b) CCl4-model group showing some nephrotoxic lesions, as evidenced by the glomerular and tubular necrosis; (c) Silymarin 50 mg/kg + CCl4 group; (d) MpEO 5 mg/kg + CCl4 group (e); MpEO 15 mg/kg + CCl4 group; (f) MpEO 40 mg/kg + CCl4 group; (g) MpEO 40 mg/kg group. Hematoxylin/eosin staining; magnification ×400. : glomerular necrosis; : necrosis in epithelial cells of the proximal tubules[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
Grades of glomerular and epithelial cells of the proximal tubules necrosis in rat kidney
Kidney sections of normal histological appearance (Fig. 3a) and the CCl4 control group showed some nephrotoxic lesions, as evidenced by the glomerular and tubular necrosis (Fig. 3b). However, pretreatment with MpEO (40 mg/kg, Fig. 3f) can remarkably ameliorate the histopathological kidney lesions induced by administration of CCl4 (Fig. 3f). In addition, silymarin at the dose of 50 mg/kg has shown to produce renoprotection evidenced by amelioration the histopathological kidney lesions induced by injection of CCl4 (Fig. 3c). The histological pattern in kidney was almost normal in rats treated with MpEO alone.
By analyzing the histopathological scoring attributed to the kidney tissues it is possible to note that the highest dose of MpEO (40 mg/kg) or silymarin (50 mg/kg) pretreatment just conferred good protection on CCl4-induced kidney damage (Table 9).[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
## Discussion
*In the **Discussion** section:*
Chemical composition of the essential oil obtained from MpEO was determined by GC-MS analysis. The compounds, their percentages as well as the retention indices are listed in Table 1. The essential oil is a complex mixture with 26 compounds representing 98.17% of the total oil composition. The major component of the essential oil is menthol (33.59%) followed by iso-menthone (33.00%). In lower amounts we found a variety of compounds including limonene (8.00%), piperitone (3.20%), 1,8-cineole (2.80%), linalool (2.64%), iso-pulegol (2.40%), caryophyllene (1.95%) and pulegone (1.60%).
The obtained results are in accordance with previous studies of M. piperita oils from Turkey, Spain (Barcelona), Norway and Poland that also had menthone and menthol as their most important components [–]. On the other hand, the composition of the essential oil from Iran is totally different, with α-terpinene (19.70%), isomenthone (10.30%), trans-carveol (14.50%), pipertitinone oxide (19.30%) and β-caryophyllene (7.60%) as the major compounds, and also the oil from the Girona region (Spain) is different, where limonene and 1.8-cineole, eucalyptol are the main compounds (33.37% and 30.75%, respectively) [, ].[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
These studies showed variable chemotypes of M. piperita L. extracts with various major oil components. Differences in chemical composition observed for essential oils is likely related to abiotic factors such as soil type and climate specific regions of provenance samples and geographical factors []. Furthermore, menthol and iso-menthone, found at relatively high concentrations in the MpEO used in the present study, have been reported to exhibit anti-inflammatory activity [], making MpEO use a promising candidate against oxidative damage of the liver and kidney following an intraperitoneal administration of CCl4.
To be noted: working with natural extracts, the antioxidant activity is considered to be primary related to the major active compounds in the essential oil such as menthol and its derivatives []. However the antioxidant activity could also come from a minor compound interacting in a synergistic or antagonistic way, to create an effective system against free radicals [, ], this has to be realized when evaluating different MpEO preparations.[](https://www.ncbi.nlm.nih.gov/mesh/D009821)
Liver injury after CCl4 exposure is characterized by the elevated levels of serum hepatic marker enzymes indicating the cellular leakage and loss of functional integrity of hepatic membrane architecture. High levels of ALT, AST, ALP, LDH and γGT activities are sensitive indicators of liver cell injury and are most helpful in recognizing hepatic diseases []. CCl4-treated rats show increased activities of these enzymes, reflecting damage to the liver cells or changes in the cell membrane permeability leading to leakage of enzymes from cells to the circulation []. In the present study increased levels of serum hepatic markers suggested that an extensive liver injury was occasioned by CCl4 due to increased lipid peroxidation which had the ability to cause membrane damage. It is now generally accepted that CCl4 hepatotoxicity is the result of reductive dehalogenation, which is catalyzed by its specific isoenzyme of cytochrome P450 2E1, and which forms the highly reactive free radical.
Hence, the suppression of P450 2E1 could result in reduced levels of reactive metabolites, and thus decreased tissue damage [].[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
The liver plays a fundamental role in the metabolism of lipids. Injection of CCl4 caused a significant increase in the triglyceride, total cholesterol, and LDL levels and decrease in HDL level. Increase in the cholesterol levels might be due to the increased esterification of fatty acids, inhibition of fatty acid β-oxidation, and decreased excretion of cellular lipids []. CCl4 stimulates the transfer of acetate into liver cells (probably by increasing access to acetate) and leads to an increase in cholesterol synthesis. It also increases the synthesis of fatty acids and triglyceride from acetate and enhances lipid esterification [].
The accumulation of triglyceride in liver might occur due to the inhibition of lysosomal lipase activity and VLDL secretion [].[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
The administration of CCl4 induced also renal toxicity evidenced by an elevation of serum creatinine and urea [, ]. These pathological changes can also be attributed to damages touching the structural integrity of nephrons [], which is consistent with reports confirming that the level of serum creatinine increases only if at least half of the kidney nephrons are already damaged [].[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
Treatment of rats with MpEO prior to CCl4 exposure resulted in a dramatically protective effect against acute hepato and renotoxicity and oxidative stress, which was further also confirmed by the hepatic histopathological examinations. The stimulation of hepatic regeneration makes the liver more resistant to damage by the toxin []. Treatment with silymarin (50 mg/kg) or MpEO (40 mg/kg) significantly decreased the elevated levels of ALT, AST, ALP, LDH, γGT, total cholesterol, triglycerides, LDL urea and creatinine and increased of HDL level as compared to CCl4 group. Pharmacological studies have shown that essential oil derived from various plant materials possesses anti-inflammatory activities [, ] Knowing that sesquiterpenes have excellent anti-inflammatory activities [], the anti-inflammatory activity of M. piperita L. leaf essential oil could be partly explained by the presence of sesquiterpenes, such as spathulenol, cadinene, caryophyllene and caryophyllene oxide.
The ethanolic extract of parsley leaves also showed significant anti-inflammatory [] and antioxidant activities [, ] which may contribute to its hepatoprotective action. Furthermore, menthol and iso-menthone, found at relatively high concentrations in the MpEO used in the present study, have been reported to exhibit anti-inflammatory activity [].[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
In the present study, the two fold increase in the TBARS levels and reduce activity of SOD, CAT and GPx observed in liver and kidney homogenate of CCl4-intoxicated rats. Silymarin significantly reversed CCl4-induced TBARS levels elevation but values obtained with MpEO at the highest dose was comparable in attenuation of TBARS levels in liver and kidney. Silymarin reversed CCl4-induced SOD, CAT and GPx activities decrease but values obtained with MpEO at the highest dose (40 mg/kg) was comparable in attenuation of SOD, CAT and GPx activities in liver and kidney.[](https://www.ncbi.nlm.nih.gov/mesh/D017392)
These results suggested that MpEO could exert its antioxidant and/or radical scavenging activities thus preventing the formation of the carbon free radicals originated from CCl4 metabolism as well as ROS and peroxidation products. This hypothesis is supported by the recent findings on the in vitro antiradical and antioxidative activities of MpEO []. Previous studies showed that menthol and its derivatives were the major compounds responsible for antioxidant activity of MpEO [].[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
In the present study, the rats of group 2 served as CCl4-hepato and renotoxicity control and rats of groups 3, 4, 5 and 6 were injected with 1 ml/kg BW of CCl4 and olive oil mixture on day seven (a single intraperitoneal injection). It has already been shown that a single dose of CCl4 initiates lipid peroxidation [–] that results in the disruption of cellular and organelle membrane integrity and subsequent leakage of cellular contents into the blood [, ]. CCl4 is further well known to induce fibrosis of the hepatic tissue that may further progress to cirrhosis if the stimuli persists [–]. Thus, a single CCl4 injection in mice can be used as an attractive and highly reproducible model of liver regeneration after toxic injury. The first appearance of histological fibrosis and scarring fibers is usually observed after repeated CCl4 treatment for 2 to 3 weeks, depending on the dosage and mouse strains used [].
In the present work, the hepatic histoarchitecture of the CCl4-treated rats resulted large numbers of inflammatory cells such as lymphocytes along with hepatic sinusoidal inflammation, hepatocyte necrosis and devastating liver architecture The highest dose of MpEO (40 mg/kg) or silymarin (50 mg/kg) significantly attenuated the damaged liver depicting marked focal regenerative changes which are illustrated by presence of actively dividing cells with a prominent nucleolus. The administration of MpEO reducing the histological alterations in liver provoked by CCl4 was quite noticeable. In fact, the histological changes seen in the kidney of rats treated with CCl4 were characterized by some nephrotoxic lesions, as evidenced by the glomerular and tubular necrosis. Our results confirmed previous findings of Ozturk et al. [] who had found degenerative changes in kidney of rats exposed to CCl4.
The results suggest that MpEO treatment prior to CCl4 intoxication could prevent the CCl4-induced alterations in kidney tissues of treated animals.[](https://www.ncbi.nlm.nih.gov/mesh/D002251)
## Conclusions
*In the **Conclusions** section:*
The contents of MpEO not only protect the integrity of plasma membrane but, at the same time, increased the regenerative and reparative capacity of the liver and kidney. These results suggest that the compound present in MpEO has hepatorenal protective effects against CCl4 induced oxidative stress in rats. Further investigations are essential to elucidate the precise mechanism of active agents of MpEO protection against CCl4-induced hepatotoxicity and nephrotoxicity and it has to be tested against other biological parameters.[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
# Abbreviations
*In the **Abbreviations** section:*
BHT[](https://www.ncbi.nlm.nih.gov/mesh/D002084)
Butylated hydroxytoluene[](https://www.ncbi.nlm.nih.gov/mesh/D002084)
CAT
Catalase
DPPH[](https://www.ncbi.nlm.nih.gov/mesh/C004931)
1-Diphenyl-2-picrylhydrazyl[](https://www.ncbi.nlm.nih.gov/mesh/C004931)
DTNB[](https://www.ncbi.nlm.nih.gov/mesh/D004228)
5,5′-Dithio-bis(2-nitrobenzoic acid)[](https://www.ncbi.nlm.nih.gov/mesh/D004228)
EDTA[](https://www.ncbi.nlm.nih.gov/mesh/D004492)
Ethylenediaminetetraacetic acid[](https://www.ncbi.nlm.nih.gov/mesh/D004492)
GPX
Glutathione peroxidase
GSH[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
Glutathione[](https://www.ncbi.nlm.nih.gov/mesh/D005978)
M. piperita
Mentha piperita
MpEO[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
Mentha piperita essential oil[](https://www.ncbi.nlm.nih.gov/mesh/C015424)
NBT[](https://www.ncbi.nlm.nih.gov/mesh/D009580)
Nitroblue Tetrazolium[](https://www.ncbi.nlm.nih.gov/mesh/D009580)
ROS[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
Reactive oxygen species[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
SD
Standard deviation
SOD
Superoxide dismutase
TBA[](https://www.ncbi.nlm.nih.gov/mesh/C029684)
Thiobarbituric Acid[](https://www.ncbi.nlm.nih.gov/mesh/C029684)
TBARS[](https://www.ncbi.nlm.nih.gov/mesh/D017392)
Thiobarbituric acid reactive substances[](https://www.ncbi.nlm.nih.gov/mesh/D017392)
TBS[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
Tris Buffered Saline[](https://www.ncbi.nlm.nih.gov/mesh/D012965)
TCA[](https://www.ncbi.nlm.nih.gov/mesh/D014238)
Trichloroacetic Acid[](https://www.ncbi.nlm.nih.gov/mesh/D014238)
TRIS[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
Trishydroxymethyl aminomethane[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
# Authors’ contributions
*In the **Authors’ contributions** section:*
KB, ABH, KA, JvP, FMA, TR and AE designed and wrote the paper. All authors have read and approved the final manuscript.
## Ethics approval and consent to participate
*In the **Ethics approval and consent to participate** section:*
Not applicable.
## Consent for publication
*In the **Consent for publication** section:*
Not applicable.
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interests.
## Publisher’s Note
*In the **Publisher’s Note** section:*
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
# References
*In the **References** section:*
|
# Introduction
Enabling high-mobility, ambipolar charge-transport in a [DPP-benzotriazole copolymer](https://www.ncbi.nlm.nih.gov/mesh/C012771) by side-chain engineering†
# Abstract
*In the **Abstract** section:*
In this article we discuss the synthesis of four new low band-gap co-polymers based on the diketopyrrolopyrrole (DPP) and benzotriazole (BTZ) monomer unit.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
In this article we discuss the synthesis of four new low band-gap co-polymers based on the diketopyrrolopyrrole (DPP) and benzotriazole (BTZ) monomer unit. We demonstrate that the BTZ unit allows for additional solubilizing side-chains on the co-monomer and show that the introduction of a linear side-chain on the DPP-unit leads to an [increase in](https://www.ncbi.nlm.nih.gov/mesh/D011108) thin-film ord[er and charge-carrie](https://www.ncbi.nlm.nih.gov/mesh/D011758)r [mob](https://www.ncbi.nlm.nih.gov/mesh/D011758)ility [if a sufficie](https://www.ncbi.nlm.nih.gov/mesh/C012771)nt[ly ](https://www.ncbi.nlm.nih.gov/mesh/C012771)solubilizing, branched, side chain is at[tac](https://www.ncbi.nlm.nih.gov/mesh/C012771)hed to the BTZ. We compare two different synthetic routes, direct arylation and Suzuki-polycondensation, by a direct comparison of p[oly](https://www.ncbi.nlm.nih.gov/mesh/D011758)mers obtained via the two routes and show that direct arylation produces polymers with lower electrical performance which we attribute to a higher[ de](https://www.ncbi.nlm.nih.gov/mesh/C012771)nsity of chain Furthermore we demonstrate that a polymer utilizing this design motif and synthesized via Suzuki-polyc[ondensat](https://www.ncbi.nlm.nih.gov/mesh/D011108)ion ((l-C18)-DPP-(b-C17)-BTZ) exhibits exceptionally high and near ba[lanced a](https://www.ncbi.nlm.nih.gov/mesh/D011108)verage electron and hole mobilities >2 cm2 V–1 s–1 which are among the highest, robustly extracted mobility values re[ported ](https://www.ncbi.nlm.nih.gov/mesh/D011108)for DPP copolymers in a top-gate configuration to date. Our results demons[trate clearly that line](https://www.ncbi.nlm.nih.gov/mesh/C012771)ar side chain substitution of the DPP unit together with co-monomers that allow for the use of sufficiently long or branched solubilizing side chains can be an attractive design[ motif for sol](https://www.ncbi.nlm.nih.gov/mesh/D011108)ution processable, high mobility DPP copolymers.[](https://www.ncbi.nlm.nih.gov/mesh/D011758)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Low-bandgap, donor–acceptor conjugated copolymers are being intensely researched for applications in organic field-effect-transistors (OFETs) and solar cells. They have led to the development of polymer-based OFETs with high field-effect-mobilities exceeding 1 cm2 V–1 s–1 which outperform amorphous-Si (a-Si:H) thin film transistors and could enable new applications for OFETs, including logic circuit applications that require fast switching and current-driven flexible displays based on organic light-emitting diodes., For logic circuit applications a complementary circuit design based on p- and n-type FETs is beneficial to enable low-power consumption and achieve adequate noise margins for larger-scale integration. Polymers that are intrinsically ambipolar, i.e. capable of operating in either p-type or n-type operation mode, facilitate fabrication of such circuits, either using ambipolar or genuinely complementary logic configurations. In the latter case a single polymer semiconductor is used to realize both the p-type and the n-type devices, but the device configuration is engineered to suppress either electron or hole transport, respectively, through the use of appropriate source-drain contacts or gate dielectrics.
Ambipolar polymer semiconductors have reached exceptionally high performances with p- and n-type mobilities exceeding 2 cm2 Vs–1 and simple ambipolar circuits, such as ring oscillators with oscillation frequencies of up to 380 kHz and simple logic gates have been demonstrated.–[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
Some of the highest performing ambipolar donor–acceptor polymers known today are based on the diketopyrrolopyrrole (DPP) electron accepting unit. FET mobilities up to 12 cm2 V–1 s–1 for hole- and up to 6.3 cm2 V–1 s–1 for electron-transport have been claimed,,– although in these ambipolar materials there is ongoing debate about overestimation of mobility values, when mobilities are extracted from a narrow gate voltage range near the onset of hole accumulation, in which the hole concentration might be enhanced by residual electrons in the channel. The DPP unit is usually flanked by side units like phenylene, thiophene or selenophene. Units based on five-membered rings, such as thiophene or selenophene, tend to exhibit less backbone torsion and improved backbone co-planarity resulting in improved OFET performance compared to six-membered rings. Recently Nielsen and co-workers published a very extensive review comparing more than 80 different DPP containing polymers and summarizing the effects of different side chain substitution, co-monomers and processing conditions on FET performance.
In most polymers containing the DPP unit, long, branched side-chains are attached to the DPP-core to endow the polymers with solution processability. It is generally understood that the side chain substitution exerts an important influence on the packing and backbone conformation of conjugated polymers in the solid state and may even dictate the ability of the polymer to form an ordered microstructure. Therefore choosing the right chemical type (e.g. alkyl, fluoroalkyl, …), structure (e.g. linear or branched) and length of side-chains is an important part of conjugated polymer design and can lead to major differences in electrical performance of a semiconducting polymer. The deliberate choice of side-chains to enable good electrical performance of a conjugated polymer is referred to as side-chain engineering and has gained in importance in the last few years leading to an increasing number of publications on this topic. It has been found in a number of conjugated polymers that the introduction of branched instead of linear alkyl side-chains improves polymer solubility but is often detrimental for efficient charge-transport as e.g. shown by Osaka et al.
and Guo et al. for two series of polymers in organic field-effect transistors. In their respective works both groups found decreased charge-carrier mobilities for polymers containing branched side-chains compared to the linear side-chain substituted variants and both authors attributed the decreased transistor performance to an increase in side-chain bulkiness of branched alkyl-chains preventing efficient charge-transport through the polymer film. However by examining a series of DPP-based polymers and varying the attachment density of a branched 2-octyldodecyl side-chain Zhang et al. found that branched side-chains can also improve charge-carrier mobility by facilitating a favourable polymer backbone alignment in the thin film. Based on these results other groups studied the effect of branched side-chains on polymer semiconductors by evaluating the influence of branching point position, as well as side chain type in various DPP-based polymers.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
However despite the increasing amount of publications on this topic no universal set of rules for the efficient design of side-chains in DPP-based conjugated polymers exists up to date. Therefore further studies on the effects of side-chain substitution on the electrical properties and morphology in a wider range of DPP copolymers are urgently needed. In this publication we therefore discuss the effects of linear and branched side-chains and their positioning on thin-film morphology and electrical performance in a novel dithienyl-DPP based polymer obtained by copolymerization with the weakly electron-deficient benzotriazole (BTZ) monomers (Scheme 1). The polymer allows for facile substitution with either linear or branched side chains on both the DPP-core and the BTZ co-monomer unit. To the best of our knowledge we report in this work the first DPP-based polymer FET with a mobility >2 cm2 V–1 s–1 in which the DPP unit is substituted with a linear side chain.
In fact, among the wide range of DPP polymers tested in our group under comparable conditions, for which mobilities were extracted in a robust manner and in a top-gate device architecture, this mobility is among the highest we have observed for this class of polymers. It is higher, for example, than the value we routinely observe for DPP-T-TT (), one of the most widely studied and best performing DPP-copolymers.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
Synthesis of the four DPP-benzotriazole copolymers investigated in this study.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
## Experimental details
*In the **Experimental details** section:*
## Materials
*In the **Materials** section:*
K2CO3, Li2CO3, Pd(PPh3)4, Pd(OAc)2, phenylboronic acid, bromobenzene were purchased from Sigma-Aldrich or Acros and used as received. Toluene (anhydrous) and dimethylacetamide (DMAc) were received from Sigma-Aldrich and used without further purification. Monomers 2,5-di(2-octyldodecyl)-3,6-bis-(thiophenyl)-1,4-diketopyrrolo[3,4-c]pyrrole and 2,5-dihexadecyl-3,6-bis-(thiophenyl)-1,4-diketo-pyrrolo[3,4-c]pyrrole were prepared according to the literature procedures. The synthesis of the other monomers is given in the ESI.†[](https://www.ncbi.nlm.nih.gov/mesh/C037593)
## Polymer synthesis and characterisation
*In the **Polymer synthesis and characterisation** section:*
## (l-C18)-DPP-(b-C17)-BTZ polymer prepared by a Suzuki polycondensation method
*In the **(l-C18)-DPP-(b-C17)-BTZ polymer prepared by a Suzuki polycondensation method** section:*
To a Schlenk tube were added dipotassium (b-C17)-BTZ bis(trifluoroborate) 2 (0.200 g, 0.351 mmol), (l-C18)-DPP dibromide 1 (0.338 g, 0.351 mmol), Pd(PPh3)4 (0.02 g, 0.018 mmol), Li2CO3 (0.400 g), K2CO3 (0.400 g) and Bu4NBr (0.011 g, 0.035 mmol). , The reaction mixture was then purged with N2. Toluene (N2 bubbled, 6 cm3) was added to the Schlenk tube and the resulting mixture was stirred at 90 °C for 5 min. Water (N2 bubbled, 4 cm3) was added to the tube at 90 °C. After the heterogeneous mixture was stirred for 4 h at 90 °C, a solution of phenylboronic acid (0.043 g, 0.35 mmol) dissolved in toluene (N2 bubbled, 2 cm3) was added. The reaction mixture was stirred for another 1 h at 90 °C.
Bromobenzene (0.11 g, 0.70 mmol) was added to the mixture. After the reaction mixture was stirred for further 2 h, it was allowed to cool down to room temperature and precipitated into stirring MeOH (200 cm3). The precipitates were filtered, washed with MeOH, and dried under reduced pressure. For purification of the polymer, the dried precipitate was dissolved in CHCl3 (300 cm3) and a pipetteful amount of ammonium hydroxide was added to the solution. It was then passed through a short plug of silica gel with flushing with a copious amount of CHCl3 and concentrated under reduced pressure.
The concentrated polymer solution was dropped into stirring MeOH (200 cm3). The precipitated polymer was filtered, washed in a Soxhlet extraction apparatus with acetone for 48 h and dried under reduced pressure again to give the copolymer (l-C18)-DPP-(b-C17)-BTZ as a dark brown solid (0.26 g).[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
GPC (chlorobenzene, 80 °C): Mn = 63 000 g mol–1, Mw = 204 000 g mol–1, PDI = 3.2. Tdecomp. (onset) = 390 °C. Found: C, 75.6; H, 10.2; N, 6.0; S, 5.5%.
(C73H117N5S2O2)n requires: C, 75.5; H, 10.2; N, 6.0; S, 5.5%.[](https://www.ncbi.nlm.nih.gov/mesh/C031294)
## (l-C18)-DPP-(l-C8)-BTz polymer prepared by a Suzuki polymerisation method
*In the **(l-C18)-DPP-(l-C8)-BTz polymer prepared by a Suzuki polymerisation method** section:*
Following the procedure above (l-C18)-DPP dibromide 1 (0.217 g, 0.226 mmol), dipotassium(l-C8)-BTZ bis(trifluoroborate) 3 (0.100 g, 0.226 mmol), Pd(PPh3)4 (0.013 g, 0.011 mmol), Li2CO3 (0.200 g), K2CO3 (0.200 g), Bu4NBr (0.0073 g, 0.023 mmol) in toluene (3 cm3) were reacted at 90 °C for 5 min, after which water (2 cm3) was added. Following reaction for 4 h, the mixture was endcapped as above using degassed solutions of phenylboronic acid (0.028 g, 0.23 mmol) in toluene (1 cm3) and then bromobenzene (0.071 g, 0.45 mmol). The polymer as isolated as above to give the polymer (l-C18)-DPP-(l-C8)-BTZ as a dark blue solid (0.11 g):[](https://www.ncbi.nlm.nih.gov/mesh/D011758)
GPC (chlorobenzene, 80 °C): Mn = 11 000 g mol–1, Mw = 55 000 g mol–1, PDI = 4.0. Tdecomp. (onset) = 413 °C. Found: C, 72.7; H, 9.4; N, 6.5; S, 6.1%.
(C64H99N5S2O2)n requires: C, 74.3; H, 9.6; N, 6.8; S, 6.2%.[](https://www.ncbi.nlm.nih.gov/mesh/C031294)
## (b-C20)-DPP-(l-C8)-BTZ polymer prepared by direct arylation method
*In the **(b-C20)-DPP-(l-C8)-BTZ polymer prepared by direct arylation method** section:*
A mixture of (b-C20)-DPP 4 (0.146 g, 0.170 mmol), (l-C8)-BTZ dibromide 5 (0.0661 g, 0.170 mmol), Pd(OAc)2 (0.0016 g, 0.007 mmol), and K2CO3 (0.0345 g, 0.25 mmol) in anhydrous DMAc (5 cm3) was purged with Ar stream and stirred for 72 h at 110 °C under Ar. After cooling to room temperature, the reaction mixture was precipitated into acidic MeOH and stirred for another 30 min. The resulting precipitates were filtered, washed with MeOH and purified by Soxhlet extraction with MeOH, acetone and hexane before collecting the higher molecular weight CHCl3 fraction. To the CHCl3 extract was added an aqueous solution of sodium diethyldithiocarbamate (ca. 1 g per 100 cm3) and the mixture was heated to 60 °C with vigorous stirring for 2 h. After cooling to room temperature, the layers were separated and the organic fraction was washed with water (2 × 150 cm3), and concentrated under reduced pressure.
The resulting residue was dissolved in a minimum amount of CHCl3 and added dropwise to a vigorously stirred MeOH (250 cm3). The precipitates were filtered using a 0.45 μm PTFE filter and dried under reduced pressure to afford the desired polymer. A final purification by preparative GPC afforded the polymer as a dark powder (0.083 g, 42%).[](https://www.ncbi.nlm.nih.gov/mesh/D011758)
GPC (chlorobenzene, 80 °C): Mn = 93 000 g mol–1Mw = 142 000 g mol–1, PDI = 1.5. Found: C, 74.1; H, 9.4; N, 6.1%. (C68H107N5S2O2)n requires: C, 74.9; H, 9.9; N, 6.4%.[](https://www.ncbi.nlm.nih.gov/mesh/C031294)
## (l-C16)-DPP-(l-C8)-BTZ prepared by direct arylation method
*In the **(l-C16)-DPP-(l-C8)-BTZ prepared by direct arylation method** section:*
Using a similar procedure to above, 4,7-dibromo-2-octyl-2,1,3-benzotriazole (66.1 mg, 0.17 mmol), 2,5-dihexadecyl-3,6-bis-(thiophenyl)-1,4-diketopyrrolo[3,4-c]pyrrole (127.4 mg, 0.17 mmol), Pd(OAc)2 (1.6 mg, 0.007 mmol), and K2CO3 (34.5 mg, 0.25 mmol) in anhydrous dimethylacetamide DMAc (5 cm3) were reacted for 72 h at 110 °C. Following purification as above l-PDPPBTz was isolated as a dark powder. Yield: 77 mg, 47%.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
GPC (chlorobenzene, 80 °C): Mn = 27 600 g mol–1; Mw = 54 800 g mol–1; PDI = 2.0. Found: C, 72.9; H, 8.7; N, 6.7%. (C60H89N5O2S2)n requires: C, 73.8; H, 9.2; N, 7.2%.[](https://www.ncbi.nlm.nih.gov/mesh/C031294)
## Device fabrication and characterisation
*In the **Device fabrication and characterisation** section:*
Bottom-contact top-gate field-effect transistors were fabricated with poly(methyl methacrylate) (PMMA) as polymer gate dielectric. 20 nm Au source and drain contacts with a 4 nm Cr adhesion layer were constructed photolithographically on Corning 1737 alkali-free borosilicate glass substrates. The pre-structured substrates were then transferred into a nitrogen glovebox for all subsequent processing steps. All polymer semiconductor films with the exception of (l-C18)-DPP-(l-C8)-BTZ were deposited on top of the electrodes by spin-coating from a 9 mg mL–1 solution in anhydrous chlorobenzene yielding a layer thickness of 40–60 nm. As (l-C18)-DPP-(l-C8)-BTZ showed a very low solubility in chlorobenzene it was spin-coated from 1,2-dichlorobenzene using the same parameters as for the other polymers and yielding a comparable polymer film thickness.
Afterwards the polymer films were annealed at the designated temperature for one hour before cooling to room temperature. The PMMA polymer gate dielectric film (from Polymer Source, electronic-grade, syndiotactic PMMA) was spin-coated from a 50 mg mL–1 solution in anhydrous n-butylacetate and subsequently annealed at 90 °C for 30 minutes resulting in a 500 nm thick dielectric layer. Finally Au gate electrodes were evaporated thermally through a shadow mask to complete the devices. Due to the evaporator being located in ambient atmosphere the samples were taken out of the glovebox and mounted in the evaporator in ambient air prior to evaporation.[](https://www.ncbi.nlm.nih.gov/mesh/D019904)
Electrical characterization of all devices was performed with an Agilent 4155B Semiconductor Parameter Analyzer inside a nitrogen glovebox. Saturated field-effect mobilities (μsat) were determined from the slope of ID1/2 over VG in the last 10 V of the transfer curves (μh: VG = –50 to –60 V, μe: VG = 50 to 60 V) by using the following equation:μsat = (dID1/2/dVG)22L/(WCins)with L and W being the transistor channel length and width and Cins being the gate dielectric capacitance. Due to a non-negligible gate-voltage dependency of charge-carrier mobilities, extracted mobility values were slightly lower if a larger gate-voltage region was used to determine mobility. However even if almost the whole transistor operating range (e.g. for μh: VG = –60 to –30 V) was used for mobility extraction, extracted mobilities were still 70–90% of the values extracted in the last 10 V of the transfer curve and ranged up to 2 cm2 V–1 s–1 for the best devices. This shows that the used method for mobility extraction is robust and does not yield unrealistically high mobility values.
For measurements at low temperatures a Desert Cryogenics low temperature probe station was used. Measurements at low temperatures were carried out by cooling the devices to low temperatures and reheating them afterwards. Measurement points were taken in both directions (during cooling and heating cycle).[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
## Thin-film structural characterization
*In the **Thin-film structural characterization** section:*
Synchrotron-based grazing-incidence wide-angle X-ray scattering (GIWAXS) measurements of (l-C16)-DPP-(l-C8)-BTZ and (b-C20)-DPP-(l-C8)-BTZ were performed at the small/wide angle X-ray scattering beamline at the Australian Synchrotron. 14 keV photons were used with the 2D diffraction patterns recorded on an MAR-165 CCD detector. A grazing incidence angle of 0.1°, close to the films critical angle, was employed causing the X-rays to penetrate the entire film and therefore probe the films bulk structure. The sample-to-detector distance was calibrated using a silver behenate standard. Data acquisition times of 60 s were used with no evidence found for beam damage when comparing data taken at shorter and longer acquisition times.
GIWAXS data were analysed using the software SAXS15ID version 3299. X-ray diffraction data are expressed as a function of the scattering vector Q that has a magnitude of 4π/λ sin θ, where θ is half the scattering angle and λ is the wavelength of the incident radiation.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
Additional GIWAXS measurements of (l-C18)-DPP-(b-C17)-BTZ and (l-C18)-DPP-(l-C8)-BTZ were obtained at DTU Energy, Risø Campus using Cu Kα radiation (λ = 1.542 Å), generated by a rotating anode and focused using a 1D multilayer. The data were recorded on photostimulable image plates at a distance of 121 mm from the sample. A grazing incidence angle of 0.18° was employed, just below the critical angle of the Si wafer substrates, thus maximizing the diffracted signal from the polymer film. GIWAXS data were analysed using the software SIMDIFFRACTION.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
## Results and discussion
*In the **Results and discussion** section:*
## Polymer design and synthesis
*In the **Polymer design and synthesis** section:*
Co-polymerisation of the DPP-unit with another electron-deficient co-monomer has been explored previously to control the optical and electronic properties of such polymers. Strong acceptor units like benzothiadiazole (BT) or benzobisthiadiazole (BBT) have been used and have led to high mobility ambipolar polymer semiconductors with very low band gaps (Eg < 0.7 eV) and high ambipolar charge-carrier mobilities slightly exceeding 1 cm2 V–1 s–1. Despite their impressive performance, DPP polymers containing non-alkylated co-monomers are often poorly soluble even in chlorinated solvents and their solutions form gels when stored at room temperature. This constitutes important issues for their practical use. In this publication we investigate benzotriazole (BTZ) as an attractive alternative, weakly electron-deficient co-monomer unit.
Compared to co-monomers like BT or BBT, BTZ offers additional solubility due to the possibility of alkylating the middle nitrogen atom. This allows us to explore the use of linear – as opposed to standard branched – side chains on the main DPP unit whilst maintaining a sufficient level of solubility and therefore enables the evaluation of the influence of side-chain type and positioning on FET performance in DPP-polymers.[](https://www.ncbi.nlm.nih.gov/mesh/D011758)
We therefore synthesized four different modifications of a dithienyl-DPP-benzotriazole polymer with different linear and branched side chains on both the main DPP-core and the benzotriazole unit (see Scheme 1 for polymer chemical structures) to evaluate the influence of side-chain length, type and positioning on electrical performance. Despite the previous use of benzotriazole containing copolymers in organic photovoltaics and OFETs, to our knowledge this is the first demonstration of high performing OFETs based on DPP copolymers with a benzotriazole co-monomer.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
In order to access the four DPP-based polymers, we had to adopt two different synthetic protocols based on Pd catalysed cross-coupling, namely Suzuki polycondensation and direct arylation polymerisation (DARP). The first two polymers, (l-C18)-DPP-(b-C17)-BTZ and (l-C18)-DPP-(l-C8)-BTZ, were prepared by the Suzuki polycondensation reactions in which dipotassium 2-alkylbenzotriazolyl bis(trifluoroborate)s 2 and 3 were employed as stable alternatives to the boronic acid or boron ester of 2-alkylbenzotriazole. An interesting point in the synthesis of dipotassium (l-C8)-BTZ bis(trifluoroborate) 3 is the difficulty in transforming dibromo-2-octylbenzotriazole into the latter through a sequence of lithiation employing tBuLi, substitution reaction with isopropoxydioxaborolane and fluorination using an aqueous KHF2 solution. Only a tiny amount of impure monomer 3 was recovered after recrystallization, which totally differs from the case of dipotassium (b-C17)-BTZ bis(trifluoroborate) 2. It seems that the dibromo-2-octylbenzotriazole decomposes in the presence of tBuLi.
Several consecutive, failures with modified reaction conditions made us look at alternative protocols, one of which was the Suzuki–Miyaura borylation using bis(pinacolato)diboron and Pd(dppf)Cl2 catalyst. A sequence of borylation, conversion to the bis(trifluoroborate) 3 and crystallization from a mixture of acetonitrile and water provided the monomer 3 suitable for the synthesis of (l-C18)-DPP-(l-C8)-BTZ. Detailed synthetic procedures of relevant monomers are described in the ESI.†[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
Another important point in the synthesis of polymers is the previously reported modification of the polymerisation conditions. For the polymerisation of dibromobenzothiadiazole and dipotassium (b-C17)-BTZ bis(trifluoroborate) 2, successful Suzuki polycondensations were possible employing a mixture of Et4NOH and LiOH. However, the same reagents were not effective for promoting condensation reactions with (l-C18)-DPP dibromide 1. Only a small amount of precipitate was recovered by filtration and the solid had low molecular weight. We therefore carefully looked into the polymerisation conditions, and came to suspect an incompatibility of the hydrolysable amide functional groups in the DPP unit with the strongly basic conditions induced by Et4NOH and LiOH.
Switching to a less basic reagent mixture of K2CO3 and Li2CO3 was thus regarded as a viable option, which allowed us to acquire the desired polymers, (l-C18)-DPP-(b-C17)-BTZ and (l-C18)-DPP-(l-C8)-BTZ.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
The challenges in obtaining the boronic acid derivatives of monomer 3 discussed above prompted the investigation of alternative polymerization routes. In particular we were interested in methods which could utilize the readily available dibromo-2-octylbenzotriazole monomer 5 directly. Direct arylation polymerization (DARP) was of particular interest since organometallic coupling groups do not need to be introduced onto the DPP or BTz monomers, potentially facilitating synthesis and scale-up. However the selectivity of the polymerization when there are multiple aromatic C–H bonds present is a concern in some cases, and there have also been reports of significant homo-coupling of both monomers depending on the reaction conditions., This is a potentially significant limitation, since such defects would be chemically bound to the polymer, and could not be removed.
However it is challenging to directly identify such defects in conjugated polymers, particularly with polymers which show strong aggregation in solution such as DPP based materials, due to both the low number of potential defects present and also the poor resolution of NMR based techniques for such aggregated materials. Here by preparing similar straight chain polymers, namely (l-C16)-DPP-(l-C8)-BTZ and (l-C18)-DPP-(l-C8)-BTZ, by the DARP and Suzuki routes we have an indirect method to compare the two polymerization routes, by investigating the thin film morphology and device performance of the two polymers.[](https://www.ncbi.nlm.nih.gov/mesh/D001897)
Thus a co-polymer containing all linear sidechains, (l-C16)-DPP-(l-C8)-BTZ as well as a branched derivative (b-C20)-DPP-(l-C8)-BTZ in which the branching sidechain was on the DPP unit were prepared. Both polymers were obtained using the phosphine free conditions described by Scherf and co-workers utilizing Pd(OAc)2 as a catalyst with potassium carbonate in DMAc. Following purification by solvent extraction and preparative GPC, the two polymers were isolated in yields of ∼40%.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
The molecular weights of all four polymers were measured by gel permeation chromatography in chlorobenzene at 80 °C against polystyrene standards, and the results are shown in Table 1. All four polymers show relatively high average molecular weights (Mw) with broadly comparable values obtained for both polymerization methods. It is notable that both sets of branched polymers have considerably higher molecular weight than the linear polymers made by the same polymerization chemistry, which we ascribe to the improved solubility of the branched sidechain polymers. This helps to keep the growing polymer chain in solution during the polymerization process. Attempts to characterize the polymers by 1H NMR were not very informative due to the broad, poorly resolved peaks observed.
This is a common issue for relatively rigid low band gap polymers which show pronounced aggregation in solution.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
aDetermined by GPC.
bDetermined from onset of optical absorption.
Properties of the synthesized polymers. Number and weight average molecular weight (Mn/Mw), polydispersity index (PDI), maximum absorption wavelength in chlorobenzene solution and in thin-films (λmax,solution/λmax,film) estimated optical band gap (Eg) and average extracted hole and electron mobility in saturation at optimized film annealing temperature (μh/μe)[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
## Polymer properties
*In the **Polymer properties** section:*
The UV-VIS absorption characteristics of the different DPP-BTZ variations in thin-film and in chlorobenzene solution are shown in Fig. 1. The band gaps of all polymers described in this work are estimated from the onset of the low energy absorption peak in the solid state and are shown together with other physical data in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
(a) UV-VIS absorption spectra of (l-C18)-DPP-(b-C17)-BTZ (black) and (l-C18)-DPP-(l-C8)-BTZ (red) measured in chlorobenzene solution (dashed lines) and thin-films (solid lines). (b) UV-VIS absorption spectra of (l-C16)-DPP-(l-C8)-BTZ (red) and (b-C20)-DPP-(l-C8)-BTZ (black) measured in chlorobenzene solution (dashed lines) and thin-films (solid lines).[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
All four polymers exhibit a double peak in their absorption spectra commonly observed in many DPP based polymers and a broad absorption band ranging from 500 up to 1040 nm.,[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
A comparison between (l-C16)-DPP-(l-C8)-BTZ and (b-C20)-DPP-(l-C8)-BTZ shows that the replacement of a branched side chain at the DPP unit with a longer linear side chain results in a major change in absorption behaviour. The replacement leads to an increase in intensity of the 0-1 absorption peak at 777 nm while the 0-0 absorption peak shifts from 840 to 849 nm and decreases in intensity. It also leads to a slightly lower band gap of (l-C16)-DPP-(l-C8)-BTZ (Eg = 1.28 eV) compared to (b-C20)-DPP-(l-C8)-BTZ (Eg = 1.30 eV) which is indicative of a longer conjugation length. In both polymers the solution and thin film spectra look alike suggesting similar polymer conformations in the solution and solid states.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
Comparing the two linear sidechain polymers made by different polymerisation routes (l-C18)-DPP-(l-C8)-BTZ and (l-C16)-DPP-(l-C8)-BTZ, both polymers show similar positions of the two main peaks in solution (861 and 761 nm) and in thin-films (860 and 770 nm). However, we detect a change in the onset of absorption with the (l-C16)-DPP-(l-C8)-BTZ polymer prepared by direct arylation exhibiting a less well defined transition and a smaller band gap of 1.28 eV compared to 1.32 eV for the (l-C18)-DPP-(l-C8)-BTZ polymer prepared by Suzuki polycondensation. These changes are similar to those recently identified by Janssen et al. and are most likely the result of homo-coupling defects during the DARP polymerization. Such homo couplings would likely act as traps.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
The most interesting behaviour is exhibited by (l-C18)-DPP-(b-C17)-BTZ, which has a branched side-chain attached to the BTZ unit together with a slightly longer linear side-chain on the DPP unit. It exhibits the best resolved fine structure of the absorption bands and the largest 0-0/0-1 peak ratio indicating a small reorganisation energy. Together with the small band gap value of 1.25 eV and a considerable red shift of the main absorption maximum from 858 to 875 nm upon thin-film formation this suggests that the most planar, ordered backbone conformation and the longest conjugation length among the four polymers is present in (l-C18)-DPP-(b-C17)-BTZ. This is fully consistent with the FET performance reported below.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
## Thin-film microstructure
*In the **Thin-film microstructure** section:*
The 2D GIWAXS patterns of films of all four polymers annealed at temperatures to achieve optimum charge-transport are presented in Fig. 2. The scattering pattern of (b-C20)-DPP-(l-C8)-BTZ (see Fig. 2c) is essentially featureless indicating a lack of long-range periodic order and a near amorphous microstructure. The ring at Q ∼ 0.38 Å–1 is due to scattering from a Kapton window used in the experiment.
In contrast, films of (l-C16)-DPP-(l-C8)-BTZ (see Fig. 2d) exhibit a series of (h00) scattering peaks oriented along Qz corresponding to a semicrystalline microstructure with edge-on orientated lamellae. The (100) peak is located at Qz ∼ 0.27 Å–1 corresponding to a lamellar stacking distance of ∼23.2 Å. A faint (010) stacking peak is observed at Qxy ∼ 1.75 Å–1, corresponding to a π–π stacking distance of ∼3.6 Å. GIWAXS measurements of (l-C18)-DPP-(l-C8)-BTZ (see Fig. 3b) show a comparable semicrystalline thin-film structure as (l-C16)-DPP-(l-C8)-BTZ with the first three orders of the (h00) scattering peaks along the Qz-axis and a π–π stacking peak at Qxy ∼ 1.75 Å–1. The position of the (100) peak at Qz ∼ 0.24 Å–1 corresponds to a lamellar stacking distance of ∼25.7 Å.
This slightly longer stacking distance is consistent with the difference in side-chain length between the two polymers. However (l-C18)-DPP-(l-C8)-BTZ is more ordered as three diffraction orders along Qz are clearly visible compared to two peaks in (l-C16)-DPP-(l-C8)-BTZ with a weak indication of the 3rd order peak. Furthermore, the pi-stacking peak is better defined for this polymer. The ring at 1.4 Å–1 is due to packing of disordered side-chains.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
2D GIWAXS patterns of (a) (l-C18)-DPP-(b-C17)-BTZ film annealed at 110 °C, (b) (l-C18)-DPP-(l-C8)-BTZ film annealed at 110 °C, (c) (b-C20)-DPP-(l-C8)-BTZ annealed at 200 °C and (d) (l-C16)-DPP-(l-C8)-BTZ also annealed at 200 °C. Intensities are shown on a false color log scale.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
Transfer characteristics of ambipolar transistors (L = 20 μm, W = 1000 μm) based on (a) (l-C18)-DPP-(b-C17)-BTZ annealed at 110 °C, (b) (l-C18)-DPP-(l-C8)-BTZ annealed at 110 °C, (c) (b-C20)-DPP-(l-C8)-BTZ annealed at 200 °C and (d) (l-C16)-DPP-(l-C8)-BTZ annealed at 200 °C.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
Obviously films of (l-C16)-DPP-(l-C8)-BTZ and of (l-C18)-DPP-(l-C8)-BTZ which both incorporate a linear side-chain on the DPP-unit are more ordered than films of the (b-C20)-DPP-(l-C8)-BTZ polymer with a branched side-chain on the DPP unit. This is not entirely unexpected and is consistent with results from Tamayo et al. who studied film-order in films formed of DPP molecules and observed a similar increase in film disorder when incorporating branched side-chains at the DPP-core. This is likely because linear side-chains on the DPP-unit do not interfere with aggregation as much as branched side-chains which provide a higher level of steric hindrance.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
A different behaviour is observed when the branched side-chain is not incorporated on the DPP-unit but on the BTZ unit instead. As the GIWAXS data in Fig. 2a clearly shows, thin-films of (l-C18)-DPP-(b-C17)-BTZ are highly ordered. They exhibit clear semicrystalline order but with a different texture as evidenced by the occurrence of the h00 peaks along the Qxy direction, i.e. a face-on orientation with the π-stacking peak centred along the Qz axis (surface normal). The 100 peak is located at Qxy ∼ 0.22 Å–1, corresponding to a lamellar packing distance of ∼28.6 Å, and second and third order diffraction peaks at higher Qxy values.
Peaks at the same positions are present in the out-of-plane direction as well but show reduced intensity compared to the in-plane peaks. This likely corresponds to a minority component of edge-on oriented crystallites. The weak ring at q ∼ 1.3 Å–1 (∼4.8 Å) is most likely caused by packing of disordered side-chains. These results indicate a distribution of face-on and edge-on oriented crystallites with a preferred face-on crystal orientation and a high overall degree of out-of-plane π–π stacking with short (∼3.7 Å) π–π stacking distance.[](https://www.ncbi.nlm.nih.gov/mesh/D011758)
Obviously the positioning of a long branched side-chain at the BTZ unit in comparison to a linear side-chain leads to a significant enhancement in thin-film order. This is especially surprising as the comparison of (l-C16)-DPP-(l-C8)-BTZ and (b-C20)-DPP-(l-C8)-BTZ shows that a branched side-chain at the main DPP-core significantly hinders the formation of an ordered film in comparison to a linear side-chain. The strong tendency of the DPP-units to form aggregated structures is widely known and was already shown for oligothiophene DPP small molecules by Tamayo et al. who demonstrated that the DPP units stack in a coplanar fashion and form columnar stacks. It is not unreasonable to assume that the same stacking mechanism of DPP-units is present in DPP-containing polymers but can be hindered by bulky side-chains.
Clearly, the addition of a branched side-chain at the BTZ unit not only improves solubility but also does not hinder stacking of DPP units which is evident from the high degree of π–π stacking present in thin-films of (l-C18)-DPP-(b-C17)-BTZ. Furthermore it is interesting to note that a high degree of thin-film order is already present at low annealing temperatures of 110 °C and can be further improved using higher annealing temperatures up to 300 °C (see Fig. S3†).[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
## Charge-transport and ambipolar transistors
*In the **Charge-transport and ambipolar transistors** section:*
Fig. 3 shows representative transfer characteristics of bottom-contact top-gated field-effect transistors based on the four versions of DPP-BTZ polymer. As dielectric a 500 nm thick PMMA layer was used to prevent electron trapping at the semiconductor–dielectric interface and therefore enable n-type or ambipolar charge-transport.,, All transistors exhibit ambipolar transfer characteristics with low hysteresis and in the case of (l-C16)-DPP-(l-C8)-BTZ even balanced electron and hole transport. The extracted saturated hole- and electron-mobilities of all four polymers as a function of temperature are shown in Fig. 4.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
Temperature dependence of saturated hole (circles) and electron (diamonds) mobility of (l-C18)-DPP-(b-C17)-BTZ (green), (l-C18)-DPP-(l-C8)-BTZ (orange), (l-C16)-DPP-(l-C8)-BTZ (blue) and (b-C20)-DPP-(l-C8)-BTZ (red).[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
In general (l-C16)-DPP-(l-C8)-BTZ and (b-C20)-DPP-(l-C8)-BTZ which were synthesized by direct arylation show lower transistor performances compared to (l-C18)-DPP-(b-C17)-BTZ and (l-C18)-DPP-(l-C8)-BTZ which were synthesized by Suzuki polycondensation. Both polymers exhibit a similar trend of higher annealing temperatures resulting in increased saturated field-effect mobilities.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
The maximum extracted values for hole and electron mobilities were obtained after annealing at 300 °C and amount to 0.075 and 0.073 cm2 V–1 s–1 for (l-C16)-DPP-(l-C8)-BTZ and 0.011 and 0.018 cm2 V–1 s–1 for (b-C20)-DPP-(l-C8)-BTZ respectively. The almost seven-fold increase of charge carrier mobilities from (b-C20)-DPP-(l-C8)-BTZ to (l-C16)-DPP-(l-C8)-BTZ is consistent with the increase in thin-film order as observed in the GIWAXS data. The linear side-chains used for (l-C16)-DPP-(l-C8)-BTZ lead to a better π–π stacking and improved charge-transport.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
As (l-C16)-DPP-(l-C8)-BTZ and (l-C18)-DPP-(l-C8)-BTZ are almost identical in terms of their chemical structure and mainly differ in terms of synthesis method (DARP vs. Suzuki polycondensation), molecular weight (see Table 1) and a small difference in side-chain length, it is interesting to draw a direct comparison between the transistor performance of the two polymers. (l-C16)-DPP-(l-C8)-BTZ has a slightly higher molecular weight than (l-C18)-DPP-(l-C8)-BTZ but exhibits lower saturated electron and hole mobilities of 0.075 cm2 V–1 s–1 and 0.073 cm2 V–1 s–1 compared to 0.48 cm2 V–1 s–1 and 0.31 cm2 V–1 s–1 for (l-C18)-DPP-(lC8)-BTZ. Since in most polymer systems mobility is found to increase with molecular weight rather than decrease, the difference in device performance between these two polymers is most likely a reflection of the different synthesis routes used. The direct arylation method has been reported to incorporate homo-coupling defects into the chain, and such defects are expected to affect device performance adversely., Note that evidence for such homo-coupling defects was found in the UV-VIS spectra above.
The increase in mobility is consistent with the higher degree of thin-film order as seen in the GIWAXS data.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
As might have been anticipated from the high degree of microstructural order evident in GIWAXS, transistors based on (l-C18)-DPP-(b-C17)-BTZ show the highest FET performance (see Fig. 3a) with high Ion/Ioff ratios of 106 to 107 at low source-drain voltages (VD = –5 V), low subthreshold swing values of 1.5 V dec–1 and high average saturated hole mobilities of 2.4 cm2 V–1 s–1 with maximum values for hole mobilities ranging up to 2.8 cm2 V–1 s–1. Electron injection from the Au contacts into the (l-C18)-DPP-(b-C17)-BTZ polymer films seems to be slightly hindered as evident from the lack of transfer curve symmetry, especially at low gate voltages (VD = –5 V) and by the high relative voltage difference between drain and gate electrode needed to induce n-type transport. It was not possible to operate the transistors in the electron transport regime while using positive source-drain voltages because the high positive gate voltages needed to observe n-type transport using positive VD usually lead to a breakdown of the polymer gate dielectric. Nevertheless transistors based on (l-C18)-DPP-(b-C17)-BTZ showed high average saturated electron mobilities around 1.5 cm2 V–1 s–1 with maximum electron mobilities ranging up to 2.4 cm2 V–1 s–1.
The mobilities achieved with (l-C18)-DPP-(b-C17)-BTZ for either holes and electrons are amongst the highest values published for DPP-containing polymer semiconductors in top-gated structures.,,, Even though mobilities exceeding 3 cm2 V–1 s–1 have been published for devices using bottom-gated structures, these devices usually exhibit severe non-idealities in their device characteristics which manifest as kinks in the ID1/2 over VG plot. These kinks lead to an unusual gate-voltage dependency of mobilities with very high mobilities in a small region at low VG and much lower mobilities at higher VG when the transistor is switched on completely.,, As pointed out in the experimental section our transistors do not show this behaviour and the extracted mobilities are valid in a large region of the transfer curve. As obvious from the transport data, the introduction of a branched side-chain at the BTZ unit as in (l-C18)-DPP-(b-C17)-BTZ improves charge-transport significantly compared to the linear side-chain in (l-C18)-DPP-(l-C8)-BTZ. However it is still surprising that the introduction of a single branched side-chain on the BTZ unit leads to such a dramatic improvement in charge-carrier mobility compared to a linear side-chain. A possible explanation for this is the planarization of the polymer backbone by a more favourable side-chain arrangement leading to an improvement in thin-film order, a more planar backbone conformation with longer conjugation lengths and a reduced reorganization energy.
This is consistent with the absorption spectrum of (l-C18)-DPP-(b-C17)-BTZ showing a high 0-0/0-1 absorption peak intensity ratio which is indicative for a low reorganization energy and long conjugation lengths.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
Low reorganization energies caused by different structural arrangement due to differences in side-chain packing in (l-C18)-DPP-(b-C17)-BTZ and (l-C18)-DPP-(l-C8)-BTZ can also explain the difference in annealing dependence of the saturation mobilities of the two polymers compared to (l-C16)-DPP-(l-C8)-BTZ and (b-C20)-DPP-(l-C8)-BTZ. Due to higher degrees of film order and therefore low-reorganization energies already being present at low annealing temperatures in (l-C18)-DPP-(b-C17)-BTZ and (l-C18)-DPP-(l-C8)-BTZ higher annealing temperatures do not lead to large enhancements in charge-carrier mobility in these polymers. Even though (l-C18)-DPP-(l-C8)-BTZ and (l-C16)-DPP-(l-C8)-BTZ mainly differ in molecular weight, a small difference in side-chain length and the used synthesis method, the polymers do not show the same annealing behavior. If annealed at temperatures higher than 110 °C the hole mobility of (l-C18)-DPP-(l-C8)-BTZ decreases while the electron mobility steadily increases up to annealing temperatures of 300 °C. (l-C18)-DPP-(b-C17)-BTZ essentially shows the same annealing behavior as (l-C18)-DPP-(l-C8)-BTZ, even though the increase in electron mobility is not as pronounced. In comparison (l-C16)-DPP-(l-C8)-BTZ and (b-C20)-DPP-(l-C8)-BTZ both show a significant increase in saturated hole and electron mobility when annealed at temperatures of up to 300 °C which could be caused by the need to eliminate specific trapping sites and/or to enhance thin-film ordering and therefore lower reorganization energy by annealing at high temperatures for these polymer synthesized by DARP.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
As shown by GIWAXS measurement of films of (l-C18)-DPP-(bC17)-BTZ at different annealing temperatures (see Fig. S3†) the semicrystalline order in (l-C18)-DPP-(b-C17)-BTZ increases when annealed at higher temperatures. However no related increase in hole mobilities can be observed suggesting that the film morphology at 110 °C annealing temperature is already ordered enough to enable the highest possible degree of charge-transport performance in spin-coated films of (l-C18)-DPP-(b-C17)-BTZ.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
To gain a deeper insight into the charge transport of the studied polymers, temperature dependent measurements were performed on transistors annealed at temperatures ensuring optimal transistor performance (i.e. 300 °C for (l-C16)-DPP-(l-C8)-BTZ and (b-C20)-DPP-(l-C8)-BTZ and 110 °C for polymers (l-C18)-DPP-(b-C17)-BTZ and (l-C18)-DPP-(l-C8)-BTZ).[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
Fig. S4† shows the dependence of the extracted saturated charge-carrier mobilities of (l-C16)-DPP-(l-C8)-BTZ and (b-C20)-DPP-(l-C8)-BTZ as a function of measurement temperature. The carrier mobilities show a simple Arrhenius-like, temperature activated behavior as demonstrated by the linear dependence of mobility on a 1/T scale. Activation energies for holes and electrons of 135 meV and 141 meV for (l-C16)-DPP-(l-C8)-BTZ and 157 meV and 214 meV for (b-C20)-DPP-(l-C8)-BTZ can be extracted. The decreased activation energy for charge-carrier hopping in (l-C16)-DPP-(l-C8)-BTZ in comparison with (b-C20)-DPP-(l-C8)-BTZ is consistent with the increased charge-carrier mobilities.
As expected from the increased mobility and the higher degree of structural order, the activation energy for charge-transport in (l-C18)-DPP-(l-C8)-BTZ with values of 100 meV for holes and electrons is even smaller than for (l-C16)-DPP-(l-C8)-BTZ.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
In contrast to (l-C18)-DPP-(l-C8)-BTZ, the low-temperature charge transport measurements of (l-C18)-DPP-(b-C17)-BTZ do not show a simple Arrhenius behavior with just one slope. The Arrhenius plot of saturated hole-mobility over 1/T shown in Fig. 5 shows two clearly distinguishable regions with different slopes in the temperature regime ranging from 300 down to 120 K with a change of temperature behavior at around 250 K for (l-C18)-DPP-(b-C17)-BTZ. In the high temperature region above 250 K a high activation energy of 180 meV is extracted, while in the low temperature region below 250 K a much lower activation energy value of around 65 meV is observed suggesting different charge-transport mechanisms in the two regimes. The charge-transport activation energy value of 180 meV extracted from the high temperature regime is unusually high for a high mobility polymer, as other high mobility polymers like PBTTT, DPP-BT or PSeDPP-BT usually show much lower activation energies in the range of 50–90 meV.,,[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
Arrhenius plot of temperature dependent p-type field effect mobility of (l-C18)-DPP-(b-C17)-BTZ (squares) and (l-C18)-DPP-(l-C8)-BTZ (circles). The blue, the red and the orange curves show linear fits for the activation energy extraction in the high and low temperature region (above and below 250 K) for (l-C18)-DPP-(b-C17)-BTZ (red/blue) and across the whole temperature range for (l-C18)-DPP-(l-C8)-BTZ (orange). All measurements were taken in forward and reverse direction to.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
A similar temperature dependent transport behavior with two distinct activation energy regimes was already proposed by Street et al. in 2005 for semicrystalline films of PQT-12. In this publication Street et al. explained the two different activation energy regions by the activation of charge transport through the disordered grain boundary regions at higher temperatures leading to an increase in activation energy. Recently Noriega et al. extended this model to explain the exceptionally high charge-carrier mobility values of poorly ordered polymers like DPP-BT by intrachain transport along long polymer tie-chains bridging the disordered regions between small semicrystalline domains. A comparable model based on percolation transport model can be used to explain the observed temperature dependence of charge-transport activation energy in (l-C18)-DPP-(b-C17)-BTZ.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
We propose that charge-transport in (l-C18)-DPP-(b-C17)-BTZ at temperatures below 250 K is dominated by a few percolation paths leading through more ordered regions of the film and mainly containing low-activation energy hops. At temperatures above 250 K the higher thermal energy enables additional transport paths to occur through somewhat more disordered regions of the polymer thin film increasing the measurable activation energy to values close to activation energies usually observed in purely amorphous polymers like regioregular-amorphous-P3HT. Interestingly the same behaviour is not observed in (l-C18)-DPP-(l-C8)-BTZ suggesting that in this polymer the difference in energy between the few percolation paths that allow charge transport at low temperature and the transport paths through more disordered regions is too large for the latter to become thermally accessible near room temperature.[](https://www.ncbi.nlm.nih.gov/mesh/C012771)
## Conclusions
*In the **Conclusions** section:*
We prepared four new low band-gap DPP-based co-polymers with BTZ to allow for additional solubilizing side-chains on the BTZ co-monomer. We used two different synthetic routes, direct arylation and Suzuki-polycondensation, and found from a comparison of polymers with only linear side chains made by the two routes that direct arylation produces a higher density of chain defects which manifest themselves as low band gap trap sites in UVVIS spectroscopy and lead to a lower electrical performance. We showed that the introduction of a linear side-chain on the DPP-unit leads to an increase in thin-film order and charge-carrier mobility provided that the BTZ co-monomer is substituted with a sufficiently solubilizing branched side chain. (l-C18)-DPP-(b-C17)-BTZ exhibits near balanced electron and hole mobilities; the average hole mobility of 2.4 cm2 V–1 s–1 is exceptionally high, in fact it is one of the highest, robustly extracted values reported for a DPP copolymer in a top-gate configuration to date. Our results demonstrate clearly that linear side chain substitution of the DPP unit is an attractive route for improving the charge transport properties of DPP copolymers.
It requires the use of co-monomers however, that allow attachment of sufficiently long or branched side chains to retain sufficient solubility for solution processing. The experimental study presented here demonstrates very clearly that for a given conjugated polymer backbone judicious selection of the side chain substitution is crucial to ensure optimum device performance.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
One of the challenges from a practical design point of view is that it is difficult to predict from mere reasoning or simple molecular modelling how a particular side chain substitution is likely to influence the polymer microstructure in the solid state. A microscopic understanding of the sensitive dependence of microstructure on side chain substitution generally requires sophisticated molecular dynamics modelling that goes beyond the scope of the present work. , However, as such detailed calculations will be performed in the future on more high mobility polymers, such as the ones reported here, general molecular design guidelines for side chain substitution are likely to emerge. Our work suggests that copolymers with linear side chains on the DPP unit and branched side chains on the co-monomer may be an attractive design motif for high mobility DPP copolymers.[](https://www.ncbi.nlm.nih.gov/mesh/D011108)
## Supplementary Material
*In the **Supplementary Material** section:*
|
# Introduction
Regulated Mesenchymal Stem Cells Mediated Colon Cancer Therapy Assessed by Reporter Gene Based Optical Imaging
# Abstract
*In the **Abstract** section:*
Colorectal cancer is the most common cancer in both men and women and the second most common cause of cancer-related deaths. Suicide gene-based therapy with suicide gene-transduced mesenchymal stem cells (MSCs) is a promising therapeutic strategy. A tetracycline-controlled Tet-On inducible system used to regulate gene expression may be a useful tool for gene-[based therap](https://www.ncbi.nlm.nih.gov/mesh/D013752)ies. The aim[ of](https://www.ncbi.nlm.nih.gov/mesh/D013752) this study was to develop therapeutic MSCs with a suicide gene that is induced by an artificial stimulus, to validate therapeutic gene expression, and to monitor the MSC therapy for colon cancer using optical molecular imaging. For our study, we designed the Tet-On system using a retroviral vector and developed a response plasmid RetroX-TRE (tetracycline response elem[ent](https://www.ncbi.nlm.nih.gov/mesh/D013752)) expressing a mutant form of herpes simplex virus thymidine kinase (HSV1-sr39TK) [with dual re](https://www.ncbi.nlm.nih.gov/mesh/D013752)porters (eGFP-Fluc2). Bone marrow-derived MSCs were transduced using a RetroX-Tet3G (Clontech, CA, USA) regulatory plasmid and RetroX-TRE-HSV1-sr39TK-eGFP-IRES-Fluc2, for a system with a Tet-On (MSC-Tet-TK/Fluc2 or MSC-Tet-TK) or without a Tet-On (MSC-TK/Fluc2 or MSC-TK) function. Suicide gene eng[ine](https://www.ncbi.nlm.nih.gov/mesh/D013752)ered MSCs[ we](https://www.ncbi.nlm.nih.gov/mesh/D013752)re co-cultured wi[th ](https://www.ncbi.nlm.nih.gov/mesh/D013752)colon cancer cells[ (C](https://www.ncbi.nlm.nih.gov/mesh/D013752)T26/Rluc) in the presence of the prodrug ganciclovir (GCV) after stimulation with or without doxycycline (DOX). Treatment efficiency was monitored by as[sessing Rlu](https://www.ncbi.nlm.nih.gov/mesh/D015774)c [(CT](https://www.ncbi.nlm.nih.gov/mesh/D015774)26/Rluc) and Fluc (MSC-Tet-TK and MS[C-TK) activ](https://www.ncbi.nlm.nih.gov/mesh/D004318)it[y u](https://www.ncbi.nlm.nih.gov/mesh/D004318)sing optical imaging. The bystander effect of therapeutic MSCs was confirmed in CT[26/](https://www.ncbi.nlm.nih.gov/mesh/D013752)Rluc cells after GCV treatment. Rluc activity in CT26/Rluc cells decreased significantly with GCV treatment of DOX(+) cells (p <[ 0.](https://www.ncbi.nlm.nih.gov/mesh/D015774)05 and 0.01) whereas no significant changes were observed in DOX(−) cells.[ In](https://www.ncbi.nlm.nih.gov/mesh/D015774) addition, Flu[c a](https://www.ncbi.nlm.nih.gov/mesh/D004318)ctivity in also decreased significantly with DOX(+) MSC-Tet-TK cells, but no s[ign](https://www.ncbi.nlm.nih.gov/mesh/D004318)al was observed in DOX(−) cells. In addition, an MSC-TK bystander effect wa[s a](https://www.ncbi.nlm.nih.gov/mesh/D004318)lso conf[irm](https://www.ncbi.nlm.nih.gov/mesh/D013752)ed. We assessed therapy with this system [in ](https://www.ncbi.nlm.nih.gov/mesh/D004318)a colon cancer xenograft model (CT26/Rluc). We successfully transduced cells and developed a Tet-On system with the suicide gene HSV1-sr39TK. Our results confirmed the therapeutic efficiency of a suicide [gen](https://www.ncbi.nlm.nih.gov/mesh/D013752)e with the Tet-On system for colon cancer. In addition, our results provide an innovative therapeutic approach using the T[et-](https://www.ncbi.nlm.nih.gov/mesh/D013752)On system to eradicate tumors by administration of MSC-Tet-TK cells with DOX and GCV.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 1. Introduction
*In the **1. Introduction** section:*
Colorectal cancer (CRC) is the third leading cause of cancer death in men and women in the United States and the second most common cause of cancer-related deaths in both men and women world-wide. With regular screening, colon cancer can be detected early, when treatment is most effective. When detected and treated at a later stage, recurrence of the cancer is more common. Therapeutic strategies are available for treatment of CRC; however, the five-year survival rate is still only about 65% (Available online: ). Therefore, new therapeutic strategies are urgently needed to improve the prognosis of CRC.
In this regard, mesenchymal stem cells (MSCs) carrying therapeutic genes might be one therapeutic option.
MSCs are multipotent cells that preferentially reside in perivascular niches in all human tissues and organs, including adipose tissue, bone marrow, heart and lung tissue, and neonatal tissues including amniotic membranes, umbilical cords, and placentas. The application of MSCs to drug therapy extends beyond tumor and immune suppression, and because of their tumor-tropic and migratory properties, MSCs have been engineered to express pro-apoptotic, anti-proliferative, and anti-angiogenic factors for localized and systemic treatment of diseases. As such, MSCs have become important drug delivery candidates, generating optimism that therapies with greater efficacy can be developed. Of particular note, MSCs have been used in gene therapy, and experimental studies are being conducted to determine their therapeutic potential in cardiovascular and neurodegenerative disorders.
One of the main advantages of using MSCs as a drug delivery system is their ability to self-renew, allowing them to act as “self-maintaining drug-delivery vehicles”. Studies have been performed on the delivery of therapeutic genes by MSCs that capitalize on MSC’s self-proliferation and differentiation capacities and that have employed MSCs transduced with therapeutic genes such as tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL), a trans-membrane protein that causes selective apoptosis in tumor cells.
Gene-directed enzyme prodrug therapy (GDEPT), also known as suicide gene therapy, takes advantage of the properties of MSCs and, combined with a prodrug, offers additional promise as a therapy. This combined system relies on a vector to deliver a transgene and prodrug to the target (i.e., tumor) cell. The suicide gene is transcribed and translated into an active enzyme in the MSCs, which in turn converts the prodrug to its cytotoxic form leading to the death of the MSCs, as well as to the death of nearby cancer cells through a mechanism called the bystander effect, eventually resulting in tumor regression. The most commonly used suicide gene-based therapy incorporates herpes simplex virus type 1-thymidine kinase (HSV1-TK) and the drug ganciclovir (GCV), which is converted into the cytotoxic GCV triphosphate.[](https://www.ncbi.nlm.nih.gov/mesh/D015774)
In order to be of use in patients, therapeutic gene-based systems with enhanced safety are required. The expression of a transduced therapeutic gene needs to be controlled to prevent adverse side effects caused by inappropriate expression of the gene. A tetracycline-inducible system (Tet-system) has been used to control long-term transgene expression in small animal models. This inducible system is regulated by the presence or absence of doxycycline (DOX) (Tet-On or Tet-Off systems). The suicide gene HSV1-sr39TK combined with a Tet-system may therefore be a useful tool for treating cancer with reduced side effects that are commonly related to gene-based therapy.
In the current study, MSCs with DOX-inducible HSV1-sr39TK were created and their bystander cytotoxic potential was assessed in colon cancer cells, both in an in vitro and in vivo model, using an optical molecular imaging system.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 2. Results
*In the **2. Results** section:*
## 2.1. Characterization of MSC-Tet-TK and MSC-TK
*In the **2.1. Characterization of MSC-Tet-TK and MSC-TK** section:*
The transduced MSC-Tet-TK and MSC-TK cells were created by retroviral transfection, and eGFP-positive cells were sorted by FACS Aria III. Cells with an eGFP concentration of more than 50% were used for further study. The expression of MSC-TK cell eGFP and MSC-Tet-TK cell eGFP after induction with DOX (2 µg/mL) for 24 h (Figure 1A) was confirmed by confocal microscopy image. Morphology of MSC-Tet-TK cells are heterogeneous and not uniform, because MSCs are a heterogeneous cell population. MSC-Tet-TK cells with eGFP positive revealed more rod-like morphology than others and it might be related to the genetic alteration.
Some of MSC-TK cells revealed higher eGFP signals by higher eGFP expression, because protein synthesis can vary with individual cells by different copy numbers of the transduced gene. The Fluc activity of MSC-TK cells increased with increasing cell numbers (Figure 1B, R2 = 0.90). The BLI signal in MSC-Tet-TK cells increased 10, 15, 18, 22, 25, and 29-fold with increasing concentrations of DOX (0.0625 to 2 µg/mL), whereas no signal was observed in DOX(−) cells (Figure 1C). Furthermore, we found that Fluc activity increased 17, 19, 20, 20, 15, and 7-fold with increasing DOX treatment (0.5 to 4 µg/mL) for 48 h, and then decreased at 8 and 16 µg/mL DOX (Figure S1). Therefore, in this study, we elected to use 2 µg/mL DOX for transgene activation.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 2.2. 3H-PCV Uptake Assay
*In the **2.2. 3H-PCV Uptake Assay** section:*
In order to confirm the functional activity of transduced cells, we performed a 3H-PCV uptake assay. The suicide gene HSV1-sr39TK converts PCV into penciclovir monophosphate, and, subsequently, cellular monophosphate kinases and diphosphate kinases convert penciclovir monophosphate to the triphosphate form. The accumulation of penciclovir inside the cells indirectly indicates the functional efficiency of the HSV1-sr39TK enzyme expressed in cells. MSC-TK and DOX(+) MSC-Tet-TK cells showed higher 3H-PCV uptake compared to naive MSCs and DOX(−) MSC-Tet-TK cells (Figure 1D).[](https://www.ncbi.nlm.nih.gov/mesh/C053539)
## 2.3. Characterization of CT26/Rluc Cells
*In the **2.3. Characterization of CT26/Rluc Cells** section:*
The Rluc activities of stably transduced CT26/Rluc and parent CT26 cells were determined by BLI. CT26/Rluc activity increased with increasing cell number (Figure S2A, R2 = 0.91). mCherry expression was also confirmed by fluorescence microscopy (Figure S2B).
## 2.4. Cell Viability of MSCs after Treatment with DOX and GCV
*In the **2.4. Cell Viability of MSCs after Treatment with DOX and GCV** section:*
To confirm drug (GCV and DOX) effects in naive MSCs, we analyzed cell viability. We found that GCV did not affect naive MSC viability (Figure S3A) after 48 h of treatment. Since we used a DOX-inducible gene system for gene activation, we also assessed the effect of DOX on the viability of naïve MSCs. We found that MSC cell viability was reduced significantly at 8 and 16 µg/mL DOX (p < 0.05) (Figure S3B).[](https://www.ncbi.nlm.nih.gov/mesh/D015774)
## 2.5. Fluc Activity of Suicide Gene-Transduced MSCs after Treatment with GCV
*In the **2.5. Fluc Activity of Suicide Gene-Transduced MSCs after Treatment with GCV** section:*
The relative Fluc activity of MSC-Tet-TK cells decreased 56, 50, 43, 34, 28, and 22% in DOX(+) MSC-Tet-TK cells treated with increasing concentrations of GCV (0.25, 0.5, 1, 2, 4, or 8 µM, respectively). In contrast, the DOX(−) group did not show any detectable Fluc signal. In addition, the relative Fluc activity of MSC-TK cells also decreased 62, 54, 52, 45, 41, and 37% with increasing concentrations of GCV (Figure 2). Therefore, in this study, we successfully developed MSCs with a Tet-On system (MSC-Tet-TK), and confirmed the induced expression of Fluc in the presence of DOX, as well as the cytotoxic effect of GCV.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 2.6. Bystander Effects on Colon Cancer Cells with Suicide Gene Expressed by Engineered MSCs
*In the **2.6. Bystander Effects on Colon Cancer Cells with Suicide Gene Expressed by Engineered MSCs** section:*
The therapeutic effect of MSC-Tet-TK and MSC-TK on colon cancer cells was analyzed. To assess this, we initially co-cultured naïve MSCs with CT26/Rluc cells and treated them with GCV for 48 h to assess the effect of GCV on naive MSCs. The Rluc activity was not changed by GCV treatment, confirming that GCV has no effect on naive MSCs (Figure 3A). Further, to evaluate the bystander effect, we co-cultured either MSC-Tet-TK or MSC-TK cells separately with CT26/Rluc cells at a 1:1 ratio, and increasing concentrations of GCV were administered (0.125 to 1 µM), with or without prior DOX induction. The relative Rluc activity of CT26/Rluc cells significantly declined 69, 49, 39, and 35% (p < 0.01) with increasing concentrations of GCV (Figure 3B) in DOX(+) MSC-Tet-TK cells co-cultured with CT26/Rluc cells.
However, the relative Rluc activity of CT26/Ruc cells did not decrease significantly (104, 104, 101, and 98%) with increasing GCV concentrations (Figure 3B) in DOX(−) MSC-Tet-TK cells co-cultured with CT26/Rluc cells. In addition, the relative Fluc activity of MSC-Tet-TK cells significantly decreased 61, 54, 48, and 46% with GCV (0.125 to 1 µM respectively), demonstrating the function of the Tet-On HSV1-sr39TK/GCV suicide system in DOX(+) cells. In contrast, there was no Fluc signal observed in the DOX(−) MSC-Tet-TK cells (Figure S4).
In addition, the therapeutic effect of MSC-TK cells was also monitored in CT26/Rluc cells, and the Rluc activity was found to be decreased by 86, 67, 42, and 35%, respectively (p < 0.05, 0.01), after 48 h of GCV treatment (Figure 3B). The relative Fluc activity of MSC-TK cells also significantly decreased 98, 76, 68, and 43% after 48 h of GCV treatment (Figure S4). Therefore, these studies confirmed the bystander effect of MSCs expressing the suicide gene (MSC-Tet-TK or MSC-TK) in an in vitro colon cancer cell line.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 2.7. Effects of Suicide Gene-Transduced MSCs on Colon Cancer in a Mouse Model
*In the **2.7. Effects of Suicide Gene-Transduced MSCs on Colon Cancer in a Mouse Model** section:*
To determine the effect of suicide gene-transplanted MSCs on colon cancer growth, Rluc imaging was used to monitor tumor progression. After suicide gene-transduced MSC transplantation and GCV treatment, we found, by monitoring CT26/Rluc activity using BLI, that tumor growth was inhibited in the GCV+DOX-treated MSC-Tet-TK group (Figure 4A) compared to that in the DOX-only treatment group. The MSC-TK+GCV-treated group also exhibited significantly decreased Rluc activity compared with that of the MSC-TK group (Figure 4B). We also monitored the survival of MSC-Tet-TK cells on day 1 by measuring Fluc activity. The Fluc activity on day 1 was very high in mice injected with either cell type (MSC-Tet-TK and MSC-TK), but declined after administration of GCV compared to mice injected with the cells alone (Figure S6A,B).
Tumor weights were significantly (p < 0.05) reduced in the MSC-Tet-TK+DOX+GCV-treated group compared with those in the MSC-Tet-TK+DOX group (Figure 4C). The body weight of the treatment group was lower than untreated group, but it was not statistically significant (data not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D015774)
## 2.8. In Vivo Tumor Apoptosis Analysis
*In the **2.8. In Vivo Tumor Apoptosis Analysis** section:*
To further confirm the antitumor effect of transduced MSC-mediated apoptosis in vivo, we utilized a tumor apoptosis assay. The ApopTag Peroxidase In Situ Apoptosis Detection Kit confirmed in vivo tumor apoptosis in the transduced MSC and GCV treatment groups; the effect was enhanced by the application of GCV. This result was confirmed by TUNEL staining. Therefore, our results suggest that suicide gene-transduced MSCs with (MSC-Tet-TK) or without the Tet-On system (MSC-TK) inhibit tumor growth by inducing cell apoptosis in vivo (Figure 5A,B).[](https://www.ncbi.nlm.nih.gov/mesh/D015774)
## 3. Discussion
*In the **3. Discussion** section:*
In this current study, we successfully established an “on” or “off” switch for the suicide gene HSV1-sr39TK MSCs also co-expressing a Fluc2 reporter gene, and we assessed the therapeutic effect of this system, both with and without the Tet-On function, in a pre-established tumor using optical imaging. For this study, we used a retroviral vector to stably transfect cells and then selected the transduced MSCs for further use. The FACS-sorted immortalized MSCs expressing HSV1-sr39TK exhibited in vitro susceptibility to GCV. In the present study, we selected the transduced MSCs (MSC-Tet-TK and MSC-TK) after induction with or without DOX and confirmed the cell-killing effect of GCV following DOX induction. We also showed that suicide gene-transduced MSCs played an effective antitumor role in a human colon cancer xenograft model by inducing cell apoptosis in xenograft tumor cells.
The colon cancer cells used in this study had previously been transduced with mCherry-Rluc using lentiviral particles. The MSCs were introduced into the tumor by direct injection. The MSCs and CT26 tumor cells could each easily be monitored by measuring their respective Fluc and Rluc activities by bioluminescent imaging (BLI) using the substrates d-Luciferin and coelenterazine, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
The clinical safety of traditional constitutively overexpressed therapeutic genes is often compromised by side effects related to the failure to control the timing and level of gene expression, which are critical for the functions of genes. The Tet system can be used to induce gene expression and transcription, and to reversibly turn on or off expression in the presence of an inducing factor such as tetracycline or doxycycline (DOX). In this study, we used the “Tet-On 3G” protein, which is considerably more sensitive to DOX than the native Tet-On system and provides for much higher expression of the HSV1-sr39TK suicide gene. We assessed the effect of the pro-drug GCV in the genetically transduced MSCs (either with or without the Tet-On system, namely MSC-Tet-TK and MSC-TK cells, respectively), using BLI. The results of our study showed that these suicide gene-expressing MSCs responded to the prodrug in a dose dependent manner (Figure 2).[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
Next, we studied the consequence of prodrug conversion to its cytotoxic form inside the suicide gene-transduced MSCs by assessing the ability of these MSCs to kill neighboring cancer cells (the so-called bystander effect). The results showed that MSCs either with (MSC-Tet-TK) or without (MSC-TK) the DOX system significantly induced cell death in co-cultures with colon cancer cells (CT26/Rluc) following treatment with increasing concentrations of GCV, as evidenced by decreases in Rluc imaging and relative Rluc activity in the co-cultured colon cancer cells (Figure 3B). Importantly, GCV did not have any effect on Rluc activity in co-cultures of colon cancer cells (CT26/Rluc) with naive MSC (Figure 3A). MSCs expressing the suicide gene HSV1-TK that are sensitive to GCV can then facilitate the permeation of phosphorylated GCV to neighboring cells that do not express the suicide gene, a phenomenon known as the bystander effect that has been observed in U-87 brain cancer cells in the presence of GCV. Gap junctions may exist between MSCs and the colon cancer cells used here because the bystander effect induced by GCV triphosphate has been shown to be dependent on active transport via gap junction intercellular molecules.
This is also supported by the findings of another study by Matuskova et al., who demonstrated the formation of gap junctions between MSCs and cancer cells after co-culture. Our data, therefore, confirmed the bystander effect in co-cultures of MSCs with colon cancer cells in vitro.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
One advantage of our system is that therapeutic gene expression was linked with expression of a reporter gene Fluc2, meaning that expression of the therapeutic gene could be monitored simply by measuring Fluc expression. Therefore, the combination of the inducible Tet system and the fluorescent reporter gene allow for very easy control of the expression of the therapeutic gene, and this might be helpful in reducing side effects related to the expression of the therapeutic gene. In addition, the molecular imaging approach used in this study is ideally suited to evaluate in vivo tumor progression and regression in the evaluation of new cancer therapies.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
Kucerova et al. successively applied retrovirus-transfected with cytosine deaminase (CD) into human adipose tissue MSCs for treating colon cancer. Higashi et al. developed a successful cancer vaccine strategy using combined IL18 and HSV-TK suicide gene therapy. However, there is no specific study for the treatment of colon cancer with the suicide gene of thymidine kinase. MSCs expressing thymidine kinase suicide gene has been used for treating glioblastoma, breast, and melanoma. Apart from other studies, the main advantage of the present study is monitoring the therapeutic gene expression which was controlled by inducible Tet On system.
Suicide gene based therapy can induce side effects to the normal tissues, however, if the suicide gene is delivered only to the targeted cancer cells, it does not impose any harm onto healthy cells. Targeting of cancer by using MSCs as the delivery agent and the inducible system which can regulate the transgene expression, are able to decrease toxicities of the therapy to normal tissues. This inducible system might be useful for treating other cancer types by using suicide gene or other therapeutic genes such as TRAIL, interferons, etc.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
The prodrug’s bystander effect is thought to be mediated by the number of MSCs that come into contact with tumor cells, and this may therefore be an important factor in the therapeutic outcome of suicide gene-based therapies using MSCs. In the current study, we injected all suicide gene-transduced MSCs intratumorally, which guaranteed that all tumors received the MSCs. To validate the viability of MSCs after delivery to the tumors, all mice were imaged using BLI to measure Fluc activity for MSC-Tet-TK (1 mg/kg body weight DOX) and without the DOX system MSC-TK (approximately 1 × 106) were injected intra-tumorally after 24 h of DOX induction (or without induction). After treatment with GCV for 5 days and daily DOX injection to ensure stable transgene activation in in the MSC-Tet-TK cells (but not the MSC-TK cells), all of the mice were imaged using BLI again to monitor Fluc activity in order to determine whether the MSCs responded to the GCV (Figure S6A,B). To measure the effect of GCV on the viability of suicide gene-transduced MSCs, a second dose of suicide gene-transduced MSCs was injected on day 6.
For the group of animals treated with MSCs having the Tet On system (MSC-Tet-TK) these were treated with DOX and GCV and compared to a group of animals also treated with MSC-Tet-TKs but only treated with DOX. Similarly, for the animals treated with MSCs not having the DOX system (MSC-TK) these were treated with GCV and compared to animals treated with MSC-TKs only.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
Tumor growth in vivo during treatment with MSC-Tet-TK cells was measured using BLI to assess the Rluc activity of the CT26/Rluc cells on days 0, 6, and 13. Rluc activity was found to be decreased at days 6 and 13 in the MSC-Tet-TK+DOX+GCV treatment group relative to the MSC-Tet-TK+DOX (i.e., without GCV) treatment group (Figure 4A). The therapeutic efficiency of MSC-TK cells in this same colon cancer model was also studied. We also observed that Rluc activity was decreased on days 6 and 13 in the MSC-TK+GCV treatment group compared to the MSC-TK only treatment group (Figure 4B). Based on the data, we have successfully created MSCs with a Tet-On system (MSC-Tet-TK) and confirmed their therapeutic effect after induction with DOX in a colon cancer xenograft model.
Because colon cancer cells (CT26/Rluc) are known to form fast growing tumors; we elected to sacrifice the mice at day 13, after the second injection of cells and subsequent DOX and GCV treatments. After sacrifice, the tumors were excised, weighed, and the tumor tissue processed for a TUNEL assay. In MSC-Tet-TK treated animals, tumor weights decreased significantly in the DOX(+) GCV treatment group compared to the DOX group alone. Similarly, in MSC-TK treated animals there was a significant reduction in tumor weight relative to the MSC-TK group (Figure 4C). Leng et al. have shown similar data, namely that human umbilical cord-derived MSCs (hUC-MSCs), engineered to express HSV1-TK and injected into tumor-bearing mice, enhanced the therapeutic effect of GCV treatment.
Administration of these hUC-MSCs suppressed the angiogenesis pathway, thereby inhibiting the migration and proliferation of MDA-MB-231 cells. In addition, Amano et al. have reported an in vivo experiment where intracranial C6 tumors in Sprague-Dawley rats were injected intratumorally with MSC-TK cells and through the bystander effect produced significant tumor growth suppression and prolonged survival time following GCV administration.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
Apoptosis is widely thought to be the mechanism through which most cancer therapies induce tumor cell death. The TUNEL assay has been used extensively to assess apoptosis in vivo. In the current study, we used TUNEL staining and found an increased number of apoptotic cells in the tumors from mice treated with MSC-Tet-TK cells +DOX+GCV compared to the tumors from mice treated with MSC-Tet-TK cells +DOX (Figure 5A). Similarly, tumors from mice treated with MSC-TK cells +GCV showed an increase in apoptotic cells compared to tumors from mice treated with MSC-TK cells only (Figure 5B). These results confirm that, in mice bearing a xenograft colon cancer tumor injected with MSCs expressing the HSV1-sr39TK suicide gene (either inducible by DOX or expressed constitutively), GCV enhances apoptotic cell death in the tumor xenograft.
Ryu et al. similarly studied the antitumor activity of MSCs-TK treated with valproic acid and found that this activity enhanced apoptotic cell death in intracranial gliomas since the tissue sections from TUNEL-stained glioma-bearing mice showed a significant increase in the number of apoptotic cells.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
In conclusion, we have successfully produced a Tet-On-inducible therapeutic gene system linked with an optical reporter gene system, which allowed non-invasive monitoring of therapeutic gene expression both in vitro and in vivo. With this innovative reporter gene coupled Tet-On system, we can control expression of the therapeutic gene and minimize the potential for adverse impacts related to gene-based therapies. Experimental studies using various therapeutic genes such as interferon-β, interleukin-23, interleukin-2, cytosine deaminase, and TNF-related apoptosis-inducing ligand have been reported. Tet system-based noninvasive molecular imaging, a reproducible and controllable imaging tool, could also be used with other therapeutic genes. In order for MSC-based gene therapies to be introduced into the clinic, suicide gene-expressing MSCs must be examined for viability and their inability to promote tumors in vivo.
The present study, however, supports the usefulness of MSCs with an inducible suicide gene for the treatment of colon cancer.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 4. Materials and Methods
*In the **4. Materials and Methods** section:*
## 4.1. Chemicals
*In the **4.1. Chemicals** section:*
pRetroX-TRE3G and pRetroX-Tet3G plasmids, tetracycline-free fetal bovine serum (FBS), and DOX were purchased from Clontech (Mountain View, CA, USA). Gentamicin, a CaPO4 transfection kit, and mouse bone marrow-derived MSCs were purchased from Invitrogen (Carlsbad, CA, USA). MSCs were cultured in Dulbecco’s modified Eagle’s medium (DMEM)-F12 (HyClone, Logan, UT, USA) supplemented with 10% FBS (Hyclone), 1× GlutaMAX (Invitrogen), and 1% gentamicin (Gibco-BRL Life Technologies, Gaithersburg, MD, USA). The mouse colon cancer cell line CT26 (American Tissue Culture Collection (Manassas, VA, USA) was grown in DMEM medium supplemented with 10% FBS and 1% penicillin/streptomycin solution (HyClone).[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 4.2. Retroviral Transduction of MSCs
*In the **4.2. Retroviral Transduction of MSCs** section:*
The therapeutic suicide gene (HSV1-sr39TK; a mutant of HSV1-TK) and eGFP (enhanced green fluorescent protein) were linked to create a fusion protein. Firefly luciferase (Fluc2) was inserted after the internal ribosome entry site (IRES) sequence in the pIRES vector and downstream of the HSV1-sr39TK-eGFP fusion protein to create the vector pIRES-HSV1-sr39TK-eGFP-IRES-Fluc2. From this plasmid, the HSV1-sr39TK-eGFP-IRES-Fluc2 sequence was removed and it was inserted after the pTRE3G region into the pRetroX-TRE3G response plasmid. This plasmid encodes a modified Tet-On advanced transactivator protein called Tet-On 3G, which has markedly increased sensitivity to DOX. The inducible promoter pTRE3G consists of seven repeats of a 19-bp Tet operator sequence that results in very low basal expression and high maximal expression of the transduced protein after induction.
DOX can activate Tet-On 3G and bind specifically to pTRE3G to promote transcription. The retrovirus in pRetroX-TRE-HSV1-sr39TK-eGFP-IRES-Fluc2 and pRetro-Tet-On 3G was isolated from Gryphon E cells (Allele Biotechnology, San Diego, CA, USA) after separately transfecting the cells using the CaPO4 method. Two days after an overnight transfection and a medium change, the medium was collected, filtered through a 0.45 µm filter, and then concentrated using an Amicon Ultra 15 centrifugal filter (Merck Millipore, Burlington, MA, USA). MSCs transduced with a 1:1 dilution of Tet-On3G and Retro-HSV1-sr39TK-eGFP-IRES-Fluc2 were used to make Tet-On MSCs (MSC-Tet-TK/Fluc2 or MSC-Tet-TK) or MSCs without a Tet system (MSC-TK/Fluc2 or MSC-TK). MSC-Tet-TK and MSC-TK cells were sorted based on expression of eGFP.
MSC-Tet-TK cells were treated with DOX at 2 µg/mL for 24 h and then sorted. The sorted cells were screened by optical imaging and confocal microscopy.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 4.3. Lentiviral Transduction of Colon Cancer Cells (CT26)
*In the **4.3. Lentiviral Transduction of Colon Cancer Cells (CT26)** section:*
CT26 cells were transduced with lentiviral particles expressing mCherry-Rluc (Renilla luciferase) under the control of the cytomegalovirus (CMV) promoter (Genecopoeia, Rockville, MD, USA). Positive mCherry cells were selected, and stable clones were selected using Fluorescence activated cell sorting (FACS) Aria III (BD Biosciences, San Jose, CA, USA). The stable cells, named CT26/Rluc, were screened by optical imaging and fluorescence microscopy.
## 4.4. Optical Imaging
*In the **4.4. Optical Imaging** section:*
Bioluminescent imaging (BLI) was performed using an IVIS Lumina II Imaging System (Perkin Elmer, Waltham, MA, USA). Fluc2 and Rluc activity and the response to therapy was monitored in transfected cells using d-Luciferin and coelenterazine as substrates, respectively. CT26/Rluc and the Fluc activity of MSC-Tet-TK and MSC-TK, respectively, were used to monitor tumor development, with d-Luciferin (150 mg/kg) injected intraperitoneally into mice and after 5 min to evaluate Fluc2 expression and coelenterazine (3 mg/kg) injected intravenously into mice to assess Rluc expression. After injection of coelenterazine, mice were imaged immediately with the IVIS Lumina II Imaging System.[](https://www.ncbi.nlm.nih.gov/mesh/C532924)
## 4.5. 3H-Penciclovir (PCV) Uptake Assay
*In the **4.5. 3H-Penciclovir (PCV) Uptake Assay** section:*
To confirm HSV1-sr39TK functional activity, we performed a 3H-penciclovir uptake assays as described in Sekar et al. 2012. The transduced MSCs, MSC-TK, and MSC-Tet-TK cells, along with naive MSCs were separately plated (2 × 105 cells/well) in 6-well plates, and after 24 h, cells were induced with or without DOX for a further 24 h and then incubated with 1 μCi of 3H-penciclovir (Moravek Biochemicals, La Brea, CA, USA) for 1, 2, or 4 h at 37 °C in an atmosphere containing 5% CO2. The cells were then washed twice with ice-cold PBS, lysed in 0.1 mL of 0.1% sodium dodecyl sulfate (SDS), and analyzed in a scintillation counter with 10 mL of scintillation fluid. Activity was measured as counts per minute (CPM) and analyzed based on the CPM of cell lysates/CPM of medium/µg of protein.[](https://www.ncbi.nlm.nih.gov/mesh/C053539)
## 4.6. CCK8 Assay of Cell Viability
*In the **4.6. CCK8 Assay of Cell Viability** section:*
To determine the effect of DOX and GCV on naive cell viability, we performed a CCK8 assay. MSCs, (5 × 103 cells) were seeded into 96-well plates. After incubation overnight, the cells were treated with various doses of DOX and GCV for 48 h. After 48 h, CCK-8 (Dojindo, Kumamoto, Japan) was applied for 1 h, and the developed color was read at 450 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D004318)
## 4.7. Bystander Effect
*In the **4.7. Bystander Effect** section:*
To assess the therapeutic efficiency of the suicide gene-transduced MSC-mediated bystander effect, MSC-Tet-TK and MSC-TK cells were separately mixed at a 1:1 ratio with CT26/Rluc cells in a 48-well plate and then induced with or without DOX for 24 h. After 24 h, various concentrations of GCV were applied for a further 48 h. The fate of cells was monitored by BLI of Rluc and Fluc, using IVIS imaging.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
## 4.8. Tumor Model
*In the **4.8. Tumor Model** section:*
Six-week-old female BALB/c white mice purchased from Central Lab. Animal Inc., Seoul, Korea and mice were housed under standard laboratory conditions. All experimental procedures were reviewed and approved by the Kyungpook National University (KNU-2012-43, KNU-2016-0095, 24 June 2016) Animal Care and Use Committee and performed in accordance with the Guiding Principles for the Care and Use of Laboratory Animals. The mice were injected in the right flank with 1 × 106 CT26/Rluc cells in PBS.
One week later, Rluc activity was measured, and mice were randomly separated into groups. For MSC-Tet-TK, the following groups were compared: (1) the MSC-Tet-TK+DOX group and (2) the MSC-Tet-TK+DOX+GCV group. For MSC-TK, the groups were (1) the MSC-TK group and (2) MSC-TK+GCV group. Mice in both groups were injected intra-tumorally with 1 × 106 transduced MSCs (MSC-Tet-TK and MSC-TK), and the day the experiment began was considered day 0. A schematic of the experimental plan is shown in Figure S5.
Twenty-four hours after injection of the transduced MSCs, the treatment group mice were intraperitoneally injected with GCV (30 mg/kg/day) on days 1–5. A second dose of MSCs was injected on Day 6, and cells were treated with DOX and GCV on days 7–12. At the end of experiment (day 13), mice were sacrificed and tumors were excised, weighed, fixed with 10% buffered formalin (Sigma, St. Louis, MO 63103, USA), and processed for a TUNEL (Terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labeling) assay. Tumor development was evaluated using the BLI of Rluc on days 0, 6, and 13.
The transduced MSCs were tracked by BLI of Fluc on days 1 and 5.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## 4.9. TUNEL Staining for Ex Vivo Tumor
*In the **4.9. TUNEL Staining for Ex Vivo Tumor** section:*
To confirm apoptosis in mice in the GCV-treated group, we performed an in vivo apoptosis assay using an ApopTag Peroxidase In Situ Apoptosis Detection Kit (Millipore, Burlington, MA, USA). This assay identifies apoptotic cells in situ by labeling and detecting DNA strand breaks by the TUNEL method. The results are visualized using bright-field microscopy.[](https://www.ncbi.nlm.nih.gov/mesh/D015774)
## 4.10. Statistical Analysis
*In the **4.10. Statistical Analysis** section:*
For this study, the data are expressed as mean ± standard deviation (SD). Data from the experimental groups were analyzed by t-test using GraphPad Prism 5 software version 5.01 (GraphPad Software, Inc., La Jolla, CA, USA). A P-value of less than 0.05 was considered statistically significant.
## 5. Conclusions
*In the **5. Conclusions** section:*
The results of the current study support the use of MSCs with an inducible suicide gene system for the treatment of colon cancer. This system may open new avenues for the use of other inducible suicide gene systems with different levels of bystander effects in preclinical trials.
# Supplementary Materials
*In the **Supplementary Materials** section:*
The following are available online at .
# Author Contributions
*In the **Author Contributions** section:*
Senthilkumar Kalimuthu (first author) conceived, designed, and performed the experiments, analyzed and interpreted the data, and drafted the manuscript. Byeong-Cheol Ahn (corresponding author) contributed to the conception and design of experiments and revision of the manuscript. Liya Zhu, Ji Min Oh, Ho Won Lee, Prakash Gangadaran, Ramya Lakshmi Rajendran, Se Hwan Baek, and Yong Hyun Jeon performed the experiments. Shin Young Jeong, Sang-Woo Lee, and Jaetae Lee contributed reagents, materials, and/or analytical tools for the work.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
Transduction of mesenchymal stem cells (MSCs) with triple fusion (TF) reporter genes. (A) Enhanced green fluorescent protein (eGFP) confocal microscopy images of MSC-TK, MSC-Tet-TK plus DOX (2 µg/mL, 24 h) or doxycycline(−) (DOX(−)), and parental MSC cells. (B) Bioluminescent imaging (BLI) of Fluc activity in transduced MSC-TK cells at different cell densities. (C) Fluc activity of transduced MSCs (MSC-Tet-TK) after DOX treatment for 24 h assessed by bioluminescent imaging (BLI) and quantitation of this Fluc activity. (D) 3H penciclovir (PCV) uptake assay. 3H-Penciclovir (PCV) uptake of parental MSCs, DOX(−) and DOX(+) MSC-Tet-TK, and MSC-TK cells for 1, 2, and 4 h. p/s, photons/second.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
Fluc activity of MSC-Tet-TK and MSC-TK cells after ganciclovir (GCV) treatment for 48 h. Fluc activity was measured by bioluminescent imaging (BLI) imaging, and the quantitation for MSC-Tet-TK and MSC-TK cells is shown in the right hand panel. Values obtained from three individual experiment are expressed as the mean ± standard deviation (SD), ** p < 0.01, *** p < 0.001 (by Student’s t test). p/s, photons/second.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
Bystander effect of MSC-Tet-TK and MSC-TK cells. (A) Rluc activity in co-cultures (1:1) of naive MSCs and CT26/Rluc cells treated with the indicated concentrations of GCV for 48 h. (B) BLI images of the Rluc activity and quantitation data of CT26/Rluc in co-cultures (1:1) of MSC-TK or MSC-Tet-TK cells in the absence or presence of doxycycline (DOX(−) and DOX 2 μg/mL respectively). Three individual experiment values are expressed as the mean ± standard deviation (SD), * p < 0.05, ** p < 0.01, *** p < 0.001 (by Student’s t test). p/s, photons/second.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
In vivo therapeutic effect of MSC-Tet-TK and MSC-TK cells on inhibiting colon tumor growth. (A) Renilla luciferase (Rluc) imaging of colon cancer cells (CT26/Rluc) in mice treated with either MSC-TK or MSC-Tet-TK cells with or without concurrent GCV treatment. BLI images were taken on days 0, 6, and 13 in five individual mice; (B) Quantitative analysis of the data shown in (A). (C) Tumor weights assessed at study end. Bioluminescence activity is shown in photons/second (p/s). * p < 0.05 compared separately to MSC-Tet-TK (GCV−) and MSC-TK (GCV−).[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
In vivo therapeutic effect of the MSC-Tet-TK and MSC-TK cells to induce colon tumor apoptosis in the presence of GCV. (A) Terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labeling (TUNEL) assay for tumors taken from mice treated with MSC-Tet-TK DOX cells in the presence and absence of GCV. (B) TUNEL assay for tumors taken from mice treated with MSC-TK cells in the presence and absence of GCV. Apoptosis of cells was detected in the ganciclovir-treated group using the ApopTag peroxidase in situ apoptosis detection method.[](https://www.ncbi.nlm.nih.gov/mesh/D013752)
|
# Introduction
Folding and binding pathways of BH3-only proteins are encoded within their intrinsically disordered sequence, not templated by partner proteins
# Abstract
*In the **Abstract** section:*
Intrinsically disordered regions are present in one-third of eukaryotic proteins and are overrepresented in cellular processes such as signaling, suggesting that intrinsically disordered proteins (IDPs) may have a functional advantage over folded proteins. Upon interacting with a partner macromolecule, a subset of IDPs can fold and bind to form a well-defined three-dimensional conformation. For example, disordered BH3-only proteins bind promiscuously to a large number of homologous BCL-2 family proteins, where they fold to a helical structure in a groove on the BCL-2–like protein surface. As two protein chains are involved in the folding reaction, and the structure is only formed in the presence of the partner macromolecule, this raises the question of where the folding information is encoded. Here, we examine these coupled folding and binding reactions to determine which component determines the folding and binding pathway. Using Φ value analysis to compare transition state interactions between the disordered BH3-only proteins PUMA and BID and the folded BCL-2–like proteins A1 and MCL-1, we found that, even though the BH3-only protein is disordered in isolation and requires a stabilizing partner to fold, its folding and binding pathway is encoded in the IDP itself; the reaction is not templated by the folded partner. We suggest that, by encoding both its transition state and level of residual structure, an IDP can evolve a specific kinetic profile, which could be a crucial functional advantage of disorder.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Folded proteins contain a plethora of information; three-dimensional conformations, folding pathways, (un)folding rates, stability, and function are all encoded within the amino acid chain (). Unlike their folded counterparts, the primary sequences of intrinsically disordered proteins (IDPs)5 encode a lack of well-defined three-dimensional structure (). Instead, they exist as an ensemble of conformations. When binding to a partner macromolecule, a subset of IDPs can transition from this conformational ensemble to a stable, well-defined structure (, ). Final conformations of IDPs in these folding-upon-binding reactions can differ depending on the partner protein.
For example, the disordered C-terminal domain of p53 can fold and bind as a strand (), a helix (), or a coil () when interacting with Sir2, S100B(ββ), and cyclin A, respectively. In other cases, an IDP can form essentially the same structure when it binds promiscuously to a number of different partners, as when disordered BH3-only proteins bind to BCL-2 family proteins to form a single helix (). The bound conformation may be transiently sampled by the free IDP but only becomes significantly populated in the presence of the stabilizing partner macromolecule. There are two extreme possibilities for the folding of the IDP in this situation. Either the IDP could encode the folding pathway (that is, determine the order of events), or the reaction could be templated by the partner protein ().[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Φ Value analysis is a method that probes the folding pathway by comparing the interactions a residue makes in the reaction transition state and the final, bound state (). This is achieved through shortening the side chain of the amino acid, thus deleting contacts formed, and monitoring the impact on the stability and kinetics of the reaction. This method has been a useful tool, enabling the folding of families of homologous (well-folded) proteins to be compared (), and it has also been used to investigate IDP–protein interactions (). We previously used Φ value analysis to investigate the interaction between the intrinsically disordered BH3-only protein PUMA and the folded BCL-2–like protein MCL-1 (). Here we investigate the binding of PUMA to A1, another folded BCL-2–like protein, and show that the pathway for folding and binding is conserved.
Based on this observation, we hypothesized that the folding and binding pathway is encoded within the IDP sequence, and we performed a Φ value analysis for BID, another disordered BH3-only protein, when folding and binding to A1 and MCL-1 to test our assumptions. As predicted, BID displayed a conserved pattern of Φ values when binding to A1 and MCL-1, which differed from the pattern observed for PUMA. We therefore conclude that the folding and binding pathways for intrinsically disordered BH3-only proteins are encoded within the sequence of the IDP, not templated by the partner protein.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## Results
*In the **Results** section:*
## Comparison of MCL-1 and A1
*In the **Comparison of MCL-1 and A1** section:*
We have shown previously that PUMA has a relatively early transition state when folding and binding to MCL-1, with few inter- or intramolecular interactions formed (). Residues toward the N terminus of PUMA display slightly higher Φ values, indicating that weak structure formation in this region stabilizes the transition state. To determine whether these transition state interactions are templated by the partner protein, we investigated the interaction of PUMA with A1, another folded BCL-2–like protein. Upon binding A1, PUMA folds into a single contiguous α-helix, forming a structure that closely resembles that of PUMA bound to MCL-1 (Fig. 1). Although the bound conformations are homologous, sequence alignment of the MCL-1 and A1 constructs showed that they share a total sequence identity of only 26% (Fig. 1 and Fig.
S1A). Further analysis of the structures of PUMA bound to MCL-1 () and A1 () indicated that the binding interface shared a greater degree of identity (52%) because of a conserved seven-residue stretch (NWGRIVT). Six of these residues (NWGRVT) make contact with the C-terminal region of PUMA (Fig. 2 and Fig. S2). Importantly, the N-terminal region of PUMA, which had the highest Φ values when binding MCL-1 (), contacts residues with different chemical properties in A1. Furthermore, the presence of Glu-78 and Glu-80 in A1 (Val-234 and Lys-236 in MCL-1) (Fig. S3) introduces a negative charge into the binding grove (), whereas the electrostatic potential of the MCL-1 grove is positive ().
A1 therefore provides a good comparison with MCL-1, as PUMA folds to the same structure, but in the presence of different stabilizing amino acids with different physiochemical properties. If it were the binding partner that templates the folding pathway, then we might expect the pattern of Φ values to be different for PUMA binding to A1 and MCL-1.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Structural homology and sequence identity of complexes between BH3-only and BCL-2–like proteins. The IDPs PUMA (blue) and BID (yellow) fold to a single contiguous α-helix upon binding either the folded partner protein MCL-1 (gray) or A1 (green). Sequence identities were produced from alignments of mouse protein constructs. Root mean square deviation (RMSD) values were obtained from backbone atom structural alignments using PyMOL. N indicates the N terminus of each protein.
The N-terminal region of PUMA contacts residues with different chemical properties when bound to MCL-1 and A1. The bound structure of PUMA and noncontacting residues in MCL-1 and A1 are shown in light gray. Residues that contact PUMA were determined using PyMOL (PDB codes 2ROC and 2VOF), assuming the minimal cutoff (0.001 Å2) to account for all possible contacts between PUMA and its partners. Contacting residues are highlighted in the bound structure according to their conservation as follows: identical residues in red, similar in orange (e.g. both hydrophobic, same charges), and different in blue (e.g. opposite charges; one hydrophobic, one polar) (determined from the alignment of the MCL-1 and A1 sequences using Clustal Omega (, )).
## The folding pathway for the IDP PUMA is conserved when binding to different partners
*In the **The folding pathway for the IDP PUMA is conserved when binding to different partners** section:*
As for the previous Φ value analysis (), PUMA mutations were chosen that probed both the interfacial contacts and the amount of helicity in the reaction transition state (Fig. 3A). Binding was monitored by following the change in fluorescence of TAMRA, a fluorescent dye that was conjugated to the N terminus of PUMA, upon binding to A1 (Fig. S4). Ten of the 12 mutations resulted in a change in Gibbs free energy of binding of >0.6 kcal mol−1 (Fig. 3B), which is typically considered the minimum required to calculate Φ values ().
Analysis of all alanine and glycine mutants demonstrated that changes in koff, rather than kon, were predominantly responsible for differences in binding affinity (Fig. 3C and Table S1). From the linear free energy plot, the gradient of ln(kon) versus ln(Kd) provided a Leffler α value of 0.18 ± 0.04, indicating that the transition state was structurally early. However, although this gives a global picture of the interactions formed in the transition state that is similar to PUMA binding MCL-1 (α = 0.1 ± 0.04) (), it can hide the detail provided by a residue-level technique such as Φ value analysis. We therefore employed this method to determine which PUMA residues were forming interactions with A1 in the transition state.
A remarkable similarity in Φ values was observed for PUMA binding A1 and PUMA binding MCL-1 (Fig. 3D). Both the pattern of higher Φ values at the N terminus and the absolute values were conserved when binding to either partner protein.[](https://www.ncbi.nlm.nih.gov/mesh/C437523)
Mutations to the IDP PUMA predominantly alter the affinity for A1 through modulating koff.
A, the position of PUMA (blue) mutations are indicated in orange. Mutations were either designed to probe interface interactions (sticks) with A1 (green) through shortening the side chain to alanine or intramolecular helicity (spheres) via alanine-to-glycine mutations. The N terminus of PUMA is shown on the left. B, binding destabilization upon probing PUMA interface interactions and helicity by mutation. C, linear free energy plot showing kon (open circles) and koff (filled circles) against Kd for every PUMA mutant binding A1. Kd values were calculated from the ratio of the kinetic rate constants (koff/kon), as the affinity was generally too tight (<1 nm) to measure using equilibrium binding experiments. Some mutations shifted the affinity to a regime that could be measured at equilibrium (Fig. S7). In these cases, the determined Kd values compared well with the values calculated from the ratio of the rate constants (Table S4). D, comparison of Φ values obtained for PUMA binding to its folded partners, A1 and MCL-1. Error bars represent propagated errors.[](https://www.ncbi.nlm.nih.gov/mesh/D000409)
## Choice of mutations for the IDP BID
*In the **Choice of mutations for the IDP BID** section:*
The resemblance of the reaction transition states indicated that the folding and binding pathway is encoded within the IDP (PUMA). To test the generality of this hypothesis, we investigated the binding of BID to A1 and MCL-1. The BH3-only region of BID is another intrinsically disordered protein (Fig. S5) that folds upon binding, forming a structure that is homologous to bound PUMA (Fig. 1). Because of the low sequence identity between BID and PUMA (14%) (Fig. S1B), the Φ values for BID should differ to those observed for PUMA if the IDP encodes the folding and binding pathway.
BID residues that matched the position of the probed residues in PUMA were chosen for analysis. Five of the interface mutations (Fig. 4A) destabilized the complex with both A1 and MCL-1 by more than 0.6 kcal mol−1 (Tables S2 and S3) and were used to determine Φ values. Mutations designed to probe the transition state helicity of BID did not achieve this destabilization cutoff in both A1 and MCL-1 (Tables S2 and S3). A comparative analysis of the transition states of BID binding to A1 and MCL-1 was therefore produced using only interface Φ values.
The folding and binding pathway is encoded within the IDP rather than the stabilizing partner.
A, interface residues (sticks) in comparable helical turns were investigated for both the IDPs, PUMA (blue), and BID (yellow). The N terminus of each peptide is shown on the left. B, comparison of the Φ values obtained for the interface residues in BID (top panel) and PUMA (bottom panel) when binding to either A1 or MCL-1. C, correlation of Φ values obtained for the IDP when binding to MCL-1 and A1. Together with the IDP interface data, the Φ values from helix-probing mutations in PUMA that destabilized both the A1 and MCL-1 complex by >0.6 kcal mol−1 are included in the correlation plot. Error bars represent the propagated errors.
## The folding and binding pathway is encoded within the IDP
*In the **The folding and binding pathway is encoded within the IDP** section:*
All of the calculated interface Φ values were low (<0.2), indicating that few interactions were formed in the transition state. Unlike PUMA, which had higher Φ values at the N terminus, the pattern for BID indicated that interface interactions in the central section were stabilizing the transition state (Fig. 4B). The distinct pattern was replicated for BID binding to both A1 and MCL-1, supporting the original hypothesis that the transition state is encoded within the sequence of the IDP. When the data for both IDPs were combined, correlation of the Φ values obtained for the IDP with each partner protein (Fig. 4C) illustrated that the folding and binding information was contained within the IDP, not the partner protein (Fig. S6).
## Discussion
*In the **Discussion** section:*
IDPs are typically overrepresented in cell signaling (), a process that can comprise complex networks of many protein partners. The BCL-2 family is a model system for folding and binding reactions at the heart of these signaling networks. For example, PUMA and BID are disordered BH3-only proteins that can fold upon binding and initiate apoptosis through interacting with multiple partner proteins (, ).
Although Φ value analysis has been extensively used to characterize the folding pathway and mechanism of folded proteins, only a handful have been performed on other IDP-coupled folding and binding systems (). Generally, these have demonstrated transition states that occur early along reaction coordinates, with few interactions formed (). Our data for disordered BH3-only proteins binding BCL-2–like partners are consistent with this view of relatively unstructured transition states for IDPs that fold to simple conformations upon binding.
The N-terminal region of PUMA displays higher Φ values than the C-terminal end when binding to MCL-1, indicating that this region is important in stabilizing the reaction transition state (). Given that IDPs can fold to different structures upon interacting with different partner proteins (), it is easy to assume that the partner templates the folding reaction (). If this is the case, then changing the partner protein to which PUMA should alter the transition state interactions. We therefore chose to investigate the binding of PUMA to a partner protein that had different interactions at the N terminus of PUMA and altered physiochemical properties of the binding groove (, ). In contrast to the templating hypothesis, we found that the transition state remained unchanged, indicating that the IDP (PUMA) encodes its transition state.
This conclusion was supported by a Φ value analysis of the disordered BH3-only protein BID, which displayed a different pattern of Φ value than PUMA that was replicated when binding to either MCL-1 or A1. Interestingly, simulations of the unbound structure of PUMA indicate that it displays a greater degree of helicity toward the N terminus (), whereas the CD analysis in this work indicated that BID is largely disordered when unbound. This is consistent with both transition states; PUMA has a slightly later transition state with higher Φ values at the N terminus, perhaps suggesting that the unbound structure of the IDP has an influence on the transition state interactions.
IDPs have the potential to encode both their level of residual structure and, as we show here, the structure of the transition states for coupled folding and binding; this provides an opportunity to evolve specific kinetic profiles. This could be of crucial importance in cell signaling processes, where disorder is conserved and abundant (, ), as responses to stimuli may have to occur quickly (e.g. activation of a cell surface receptor), or may need to be decisive and relatively irreversible (e.g. stimulation of apoptosis). Changing the residual structure or encoded transition state provides an accessible method for evolution to tune the lifetimes of these complexes, which may be one explanation for the evolutionary conservation of disorder.
## Experimental procedures
*In the **Experimental procedures** section:*
## Buffers
*In the **Buffers** section:*
All biophysical experiments were carried out in 50 mm sodium phosphate, 0.05% (v/v) Tween 20 (pH 7.0).[](https://www.ncbi.nlm.nih.gov/mesh/C018279)
## Proteins and peptides
*In the **Proteins and peptides** section:*
BID (mouse, residues 76–110, UniProt P70444) and PUMA (mouse, residues 127–161, UniProt Q99ML1) peptides were synthesized by Biomatik. To reduce the oligomerization propensity, WT PUMA contained an M144A mutation, as described previously (). A fluorescent dye, 5-carboxytetramethylrhodamine (TAMRA), was conjugated at the N terminus. The peptides were reconstituted in biophysical buffer and filtered (0.22 μm). Stock solutions were frozen in liquid N2 and stored at −80 °C.
Expression and purification protocols for recombinantly produced MCL-1 (mouse, 152–308 residues, UniProt P97287), A1 (mouse, residues 1–152, UniProt Q07440), and PUMA are described in the supplemental Experimental Procedures.[](https://www.ncbi.nlm.nih.gov/mesh/C437523)
## Circular dichroism
*In the **Circular dichroism** section:*
CD scans were acquired in an Applied Photophysics Chirascan using 1 cm (for PUMA) and 2 mm (for BID) path length cuvettes. Estimates for percentage helicity were calculated using the mean ellipticity at 222 nm and the method of Muñoz and Serrano (). Scans were performed at multiple concentrations to check for the presence of oligomers (PUMA, 0.25–1 μm; BID, 5 and 10 μm).
## Binding kinetics
*In the **Binding kinetics** section:*
Association kinetics were monitored using SX18 and SX20 fluorescence stopped-flow spectrometers (Applied Photophysics) by following the TAMRA dye fluorescence, with excitation at 555 nm and emission recorded above 570 nm. Experiments were done at 25 °C, and a minimum of 15 fluorescence traces were collected and averaged before analysis. Data collected before the deadtime of mixing (1 ms) were removed. Pseudo-first-order conditions were adopted, with the concentration of the folded partner at least 10-fold higher than that of the peptide (). For each peptide concentration, the traces were averaged and fit to a single exponential equation to extract the observed rate constant (kobs) of reaching the new equilibrium.
The gradient from the straight line fit of kobs versus the concentration of partner protein was used to determine kon.[](https://www.ncbi.nlm.nih.gov/mesh/C437523)
Dissociation kinetics were monitored either using Applied Photophysics SX18 and SX20 stopped-flow spectrometers (kobs > 0.03 s−1) or a Varian Cary Eclipse spectrophotometer (kobs < 0.03 s−1) at 25 °C. A preformed complex of peptide and partner protein (PUMA–A1, BID–MCL-1, and BID–A1) was mixed with various concentrations of unlabeled PUMA peptide (used as a competitor). The change in fluorescence upon formation of the new equilibrium was monitored and fit to a single exponential function to determine kobs. With sufficient excess of the competitor, kobs represents koff, and the concentration independent rate constants were averaged to ascertain koff.
## Equilibrium binding
*In the **Equilibrium binding** section:*
Equilibrium dissociation constants (Kd) were measured at 25 °C by fluorescence anisotropy using a Cary Eclipse spectrophotometer (Varian). The TAMRA fluorophore was excited at 555 nm, and the emission was recorded at 575 nm. A detailed description of the data analysis used to extract the Kd is included in the supplemental Experimental procedures.[](https://www.ncbi.nlm.nih.gov/mesh/C437523)
## Data analysis
*In the **Data analysis** section:*
Data were analyzed using Kaleidagraph. The figures were prepared using Kaleidagraph and PyMOL. Errors were propagated using standard methods.
## Author contributions
*In the **Author contributions** section:*
M. D. C. and J. C.
conceptualization; M. D. C., C. A. T.
F. M., and Q. R. B. data curation; M.
D. C., C. A. T. F.
M., and Q. R. B. formal analysis; M. D.
C. methodology; M. D. C., C. A.
T. F. M., and J. C. writing-original draft; M.
D. C., C. A. T. F.
M., Q. R. B., and J. C. writing-review and editing; J.
C. supervision; J. C. funding acquisition; J. C.
project administration.
## Supplementary Material
*In the **Supplementary Material** section:*
The authors declare that they have no conflicts of interest with the contents of this article.
This article contains Figs. S1–S4, Tables S1–S7, Experimental Procedures, and References.
IDP
intrinsically disordered protein
TAMRA[](https://www.ncbi.nlm.nih.gov/mesh/C437523)
5-carboxytetramethylrhodamine.[](https://www.ncbi.nlm.nih.gov/mesh/C437523)
The abbreviations used are:
# References
*In the **References** section:*
|
# Introduction
A WD40 Repeat Protein from Camellia sinensis Regulates [Anthocyanin](https://www.ncbi.nlm.nih.gov/mesh/D000872) and [Proanthocyanidin](https://www.ncbi.nlm.nih.gov/mesh/D044945) Accumulation through the Formation of MYB–bHLH–WD40 Ternary Complexes
# Abstract
*In the **Abstract** section:*
Flavan-3-ols and oligomeric proanthocyanidins (PAs) are the main nutritional polyphenols in green tea (Camellia sinensis), which provide numerous benefits to [human health](https://www.ncbi.nlm.nih.gov/mesh/C404987). To date, the r[egulatory mechani](https://www.ncbi.nlm.nih.gov/mesh/D044945)sm[ of](https://www.ncbi.nlm.nih.gov/mesh/D044945) flavan-3-ol biosynthesis i[n green tea](https://www.ncbi.nlm.nih.gov/mesh/D059808) remains open to study. Herein, we report the characterization of a C. sinensis tryptophan-aspartic acid repeat protein ([CsWD40) tha](https://www.ncbi.nlm.nih.gov/mesh/C404987)t interacts with myeloblastosis (MYB) and basic helix-loop-helix (bHLH) transcription factors (TFs) to regulate the biosynthesis of flavan-3-ols. Full length CsWD40 cDNA was cloned from leaves and was deduced to encode 342 amino acids. An in vitro yeast two-hybrid assay demonstrated that C[sWD40 intera](https://www.ncbi.nlm.nih.gov/mesh/C404987)cted with two bHLH TFs (CsGL3 and CsTT8) and two MYB TFs (CsAN2 and CsMYB5e). T[he overexpr](https://www.ncbi.nlm.nih.gov/mesh/D000596)ession of CsWD40 in Arabidopsis thaliana transparent testa glabra 1 (ttg1) restored normal trichome and seed coat development. Ectopic expression of CsWD40 alone in tobacco resulted in a significant increase in the anthocyanins of transgenic petals. CsWD40 was then coexpressed with CsMYB5e in tobacco plants to increase levels of both anthocyanins and PAs. Furthermore, ge[ne expressio](https://www.ncbi.nlm.nih.gov/mesh/D000872)n analysis revealed that CsWD40 expression in tea plants could be induced by several abiotic stresses. Taken [together, th](https://www.ncbi.nlm.nih.gov/mesh/D000872)ese d[ata](https://www.ncbi.nlm.nih.gov/mesh/D044945) provide solid evidence that CsWD40 partners with bHLH and MYB TFs to form ternary WBM complexes to regulate anthocyanin, PA biosynthesis, and trichome development.[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
## 1. Introduction
*In the **1. Introduction** section:*
Flavonoids are widely distributed and ubiquitous secondary metabolites in the plant kingdom []. Flavonoid compounds have been confirmed to be involved in various important physiological functions in plants, such as seed germination, protection from ultraviolet radiation, and pathogenic microorganism defense erosion [,,]. More importantly, flavonoids have health benefits to human beings, such as antioxidative, antihypertensive, anti-inflammatory, antiaging, and insulin-sensitizing activities [,,].[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
In plants, chalcones, flavones, flavonols, flavan-3-ols, anthocyanins, and proanthocyanidins (PAs, also called condensed tannins) are the primary subgroups of flavonoids []. Their biosynthetic pathways have been extensively studied over the past three decades [,,]. The main pathway enzymes involved in flavonoid biosynthesis include chalcone synthase (CHS), chalconeisomerase (CHI), flavanone 3-hydroxylase (F3H), flavonoid 3′-hydroxylase (F3′H), flavonoid 3′,5′-hydroxylase (F3′5′H), flavonol synthase (FLS), dihydroflavonol 4-reductase (DFR), anthocyanidin synthase (ANS), and anthocyanidin reductase (ANR) [].[](https://www.ncbi.nlm.nih.gov/mesh/D047188)
In addition, regulation of the flavonoid pathway has been studied in-depth. To date, many transcription factors (TFs) have been identified from different plant species. Three TF families, WD40, basic helix-loop-helix, and MYB (v-myb avian myeloblastosis viral oncogene homolog), have been extensively studied to elucidate the regulation mechanism.[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
Numerous past studies have demonstrated that these three families form MBW (MYB-bHLH-WD40) ternary complexes to regulate the biosynthesis of anthocyanins and PA in all investigated plants, and the development of root hairs and trichomes in some plants [,,,,]. For example, in Arabidopsis, four MBW complexes, PAP1(MYB75)–TT8/GL3–TTG1(WD40), PAP2(MYB90)–TT8/GL3–TTG1, MYB113(PAP3)–TT8/GL3–TTG1, and PAP4(MYB114)–TT8/GL3–TTG1 have been shown to activate the expression of late anthocyanin biosynthetic genes, such as AtDFR, AtANS/LDOX, and UF3GT [,,]. The MBW complex TT2 (MYB123)–TT8/GL3–TTG1 regulated the expression of AtDFR, AtANS/LDOX, AtANR/BANYULS (BAN), glutathione-S-transferase (TT19), and a MATE (multidrug and toxic compound extrusion) transporter-encoding (TT12) involved in the PA biosynthesis [,].[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
In all of the MBW complexes, the WD40 repeat protein is localized in the center of ternary structure [,]. Unlike the bHLH and MYB members, WD40 has been shown to functionally enhance complex activation rather than directly participate in the recognition of the target gene promoter []. To date, several other WD40 repeat proteins that are orthologs to Arabidopsis TTG1 have been reported in a few of plant species, such as Perilla frutescens, petunia (Petunia hybrida), cotton (Gossypium hirsutum), and maize (Zea mays).
It is interesting that the gene encoding WD40 has been shown to have only one copy in the same species [,,].
To date, the biosynthetic pathways of flavan-3-ols and PAs have gained intensive studies for human health benefits. Recently, the genome of green tea has been sequenced from different cultivars. These studies provide important knowledge to enhance tea industries for improved health benefits. However, the regulation mechanism of the flavonoid pathway in green tea remains to be elucidated. Herein, we report the cloning and functional analysis of cDNA encoding a WD40 repeat protein from green tea, namely CsWD40.
Transgenic analysis, genetic complementation, and protein–protein interactions were performed to characterize that CsWD40 partner with R2R3-MYB and bHLH to form complexes which regulate the biosynthesis of anthocyanins, flavan-3-ols, and PAs. All data are fundamental to comprehensively elucidate the regulation of the green tea flavonoid pathway.[](https://www.ncbi.nlm.nih.gov/mesh/C404987)
## 2. Results
*In the **2. Results** section:*
## 2.1. Molecular Cloning and Sequence Analysis of CsWD40
*In the **2.1. Molecular Cloning and Sequence Analysis of CsWD40** section:*
The full-length cDNA of CsWD40 was obtained from tender leaves of the local cultivar C. sinensis Nongkangzao’ through rapid amplification of cDNA ends polymerase chain reaction (RACE-PCR). The 1029-bp open reading frame (ORF) of CsWD40 codes a protein with 342 amino acid residues. Alignment with CsWD40 proteins from other species revealed that four WD40 repeat domains are highly conserved among all WD40 repeat proteins (Figure S1). CsWD40 shared 79.62% and 77.17% identity with MdTTG1 (GU173813) from Malus domestica and PgTTG1 (HQ199314) from Punicagranatum, respectively (Figure S1). MdTTG1 and PgTTG1 TFs have been reported to regulate anthocyanin biosynthesis in M. domestica and P. granatum plants, respectively [,].
Therefore, their orthologous gene CsWD40 was predicted to be a TF that regulates anthocyanin biosynthesis in tea plants.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
## 2.2. CsWD40 Interacts with MYB and bHLH TFs
*In the **2.2. CsWD40 Interacts with MYB and bHLH TFs** section:*
The bHLH type TF genes CsTT8 and CsGL3 in C. sinensis are the orthologs of AtTT8 and AtGL3 in Arabidopsis thaliana, respectively. The MYB type TF genes CsAN2 and CsMYB5e are the orthologs of AtMYB75 in A. thaliana and MtMYB5 in M. truncatula, respectively. AtTT8, AtGL3, and AtTT2 have been demonstrated to form MBW complexes with WD40 proteins; these complexes regulate the phenylpropanoid pathway []. Therefore, CsWD40, CsMYB5e, CsTT8, and CsGL3 were selected to determine whether interactions occur among them in the yeast two-hybrid system.[](https://www.ncbi.nlm.nih.gov/mesh/D010666)
The yeast two-hybrid test results revealed that CsWD40 specifically interacted with the bHLH TFs CsTT8 and CsGL3 (Figure 1B). Moreover, both CsAN2 and CsMYB5e could physically interact with CsWD40 (Figure 1C). CsWD40 exhibited no autoactivation. The results indicated that CsGL3, CsTT8, CsAN2, CsMYB5e, and CsWD40 may be involved in jointly regulating the flavonoid pathway in tea plants.[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
## 2.3. CsWD40 Complements the Arabidopsis ttg1 Deficient Phenotype
*In the **2.3. CsWD40 Complements the Arabidopsis ttg1 Deficient Phenotype** section:*
In A. thaliana, TTG1 not only regulates anthocyanin synthesis but is also involved in trichome organogenesis, seed coat pigment synthesis, root hair development, and regulation and control of negative cotyledon hypocotyl stomatal cell movement [,].[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
To determine whether CsWD40 is a functional ortholog of TTG1 that can restore the deficient phenotypes in the Arabidopsis ttg1 mutant, the ORF of CsWD40, under the control of the 35S promoter, was transformed into the mutant. The surface of the ttg1 mutant leaves appeared smooth and there was no trichome on them. The seed coats lost the pigment store and appeared yellow in comparison with the wild type. The ttg1 mutant lost red pigmentation in the stem of seedlings even in those induced by 6% sucrose. The results showed that overexpression of CsWD40 fully complemented the trichome deficiency and pigmentation phenotypes in the stems of the ttg1 mutant (Figure 2).
The seed coat color of the transgenic mutant recovered partially (Figure 2).[](https://www.ncbi.nlm.nih.gov/mesh/D013395)
## 2.4. Anthocyanin and PA Accumulation in the Flowers of CsWD40-Overexpressing Tobacco Plants
*In the **2.4. Anthocyanin and PA Accumulation in the Flowers of CsWD40-Overexpressing Tobacco Plants** section:*
To ascertain the putative function of CsWD40, a tobacco gene transformation system was used for ectopic expression experiments. The petals of T1- and T2-generation CsWD40-overexpressing tobacco plants showed a deeper pink color than those of the control (G28) (Figure 3A). The expression level of CsWD40 was positively associated with anthocyanin content (Figure 3A,B). This result indicates that CsWD40 is involved in anthocyanin biosynthesis. Pang reported that MtWD40 in Medicago truncatula participates in PA accumulation
[]. Therefore, in this study, PA content in the flowers of transgenic tobacco plants was determined. No blue color was developed after a reaction with the DMACA (7-DIMETHYLAMINOCOUMARIN-4-ACETICACID) reagent for transgenic and wild-type tobacco plants (Figure 3C), indicating that CsWD40 overexpression did not affect PA biosynthesis in the flowers of transgenic tobacco plants.[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
## 2.5. Analysis of Expression of Genes Involved in Flavonoid Biosynthesis in the Flowers of CsWD40-Overexpressing Tobacco Plants
*In the **2.5. Analysis of Expression of Genes Involved in Flavonoid Biosynthesis in the Flowers of CsWD40-Overexpressing Tobacco Plants** section:*
qRT-PCR was performed to analyze the expression of genes involved in the flavonoid biosynthetic pathway in transgenic tobacco petals. The results showed that the key structural genes, NtCHS, NtF3′H, NtDFR, and NtANS were upregulated significantly (Figure 4B). Furthermore, the transcription factor genes NtAN2 and NtAN1b were expressed highly in all transgenic tobacco petals (Figure 4B). NtAN2 and NtAN1b were clearly activated in anthocyanin biosynthesis in tobacco []. These results indicate that CsWD40 acts as a positive participator and upregulates key structural and transcript factor genes involved in the anthocyanin pathway in transgenic tobacco plants.[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
## 2.6. Anthocyanin and PA Content in CsWD40- and CsMYB5e-Overexpressing Tobacco Plants
*In the **2.6. Anthocyanin and PA Content in CsWD40- and CsMYB5e-Overexpressing Tobacco Plants** section:*
The aforementioned yeast two-hybrid test results suggested that CsWD40 could positively interact with CsMYB5e. To validate the synergistic effects of CsMYB5e and CsWD40 function, we generated CsWD40- and CsMYB5e-overexpressing tobacco plants. We then cross-pollinated CsWD40 and CsMYB5e transgenic tobacco plants in both ♀ × ♂ and ♂ × ♀ directions to produce co-overexpressing transgenic tobacco.
Eleven CsWD40♀CsMYB5e♂ and nine CsWD40♂CsMYB5e♀ transgenic plants were verified by RT-PCR. There was no phenotypical difference between the flowers of CsWD40♀CsMYB5e♂ and CsWD40♂CsMYB5e♀ transgenic plants (Figure S2).
CsWD40 expression resulted in significantly increased anthocyanin content in the flowers of transgenic tobacco plants but did not affect PA content (Figure 5A,B). The extract solution of petals from the flowers of CsMYB5e lines turned blue after a reaction with DMACA (Figure 5B), indicating high PA accumulation in CsMYB5e-overexpressing tobacco plants, but slight changes in anthocyanin content in CsMYB5e tobacco plants than in control plants (Figure 5).[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
The data showed that anthocyanin and PA accumulation in CsWD40♀CsMYB5e♂ transgenic lines were markedly higher than those in control plants (Figure 5). In the CsWD40♀CsMYB5e♂ transgenic plants, high CsWD40 expression and low CsMYB5e expression resulted in marked anthocyanin accumulation but slight PA accumulation. The highest PA accumulation was detected in lines with high CsWD40 and CsMYB5e expression. The petals of CsMYB5e and CsWD40 co-overexpressing transgenic tobacco plants exhibited higher levels of anthocyanin than those of lines overexpressing CsMYB5e alone (Figure 5, Figure S2), but anthocyanin levels were lower in the co-overexpressing tobacco plants than in the tobacco plants expressing CsWD40 alone (Figure 5A).[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
## 2.7. Expression Analysis of Genes Involved in Flavonoid Biosynthesis in CsWD40 and CsMYB5e Transgenic Tobacco
*In the **2.7. Expression Analysis of Genes Involved in Flavonoid Biosynthesis in CsWD40 and CsMYB5e Transgenic Tobacco** section:*
Gene expression profiling using qRT-PCR analysis was completed to understand the effects of overexpression of CsWD40 and CsMYB5e on flavonoid pathway genes in transgenic tobacco plants. In CsMYB5e alone transgenic flowers, expression levels of NtCHS and NtANR increased more than two-fold (Figure 6A). In CsWD40 (♀) and CsMYB5e (♂) coupled expression transgenic lines, the expression levels NtANS and NtANR increased more than four-fold and five-fold, respectively (Figure 6B). In addition, the expression levels of NtLAR and NtDFR significantly increased. The expression of other genes was either slightly increased or similar between transgenic and wild type flowers.[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
## 2.8. Expression Patterns of CsWD40 in Tea Leaves under Different Abiotic Stresses
*In the **2.8. Expression Patterns of CsWD40 in Tea Leaves under Different Abiotic Stresses** section:*
TaWD40D overexpression enhanced the tolerance of Triticum aestivum plants to salt, mannitol, and abscisic acid (ABA) stresses, indicating that WD40 proteins may be involved in the response of plants to environmental stresses []. Therefore, we analyzed the expression patterns of CsWD40 induced by different abiotic stresses, including sucrose (Suc), ABA, mannitol (Man), sodium chloride (NaCl), salicylic acid (SA), and jasmonic acid (JA) stresses. The results showed that compared to the control, the expression level of the CsWD40 gene increased nearly 2-fold under ABA stress and approximately 6-fold under sucrose stress (Figure 7A). Under mannitol, NaCl, SA, and JA stresses, the expression level of the CsWD40 gene varied slightly in comparison with the control (Figure 7A). To determine whether CsWD40 expression is related to temperature changes, we compared CsWD40 expression under low (10 °C) and high (50 °C) temperatures (Figure 7B).
Our data indicated no significant difference was observed in the expression level of CsWD40 under different temperatures.[](https://www.ncbi.nlm.nih.gov/mesh/D012492)
## 3. Discussion
*In the **3. Discussion** section:*
Flavonoid compounds are widely distributed and are ubiquitous secondary metabolites, mainly existing in the root, leaf, fruit, and skin of plants ranging from spermatophytes to mosses [,]. A complex comprising an R2R3-MYB TF, a bHLH domain protein, and a WD40 repeat protein regulates the production of anthocyanin and PA in seed coats, roots, leaves, and fruit [,,,]. This complex also controls the formation of root hairs and trichomes on aerial tissues in some, but not all, plants [].[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
WD40 repeat proteins are widely present in plants, animals, and unicellular eukaryotes such as fungi and slime molds []. WD40 repeat proteins comprise a superfamily of proteins with a β-propeller structure. The core region of WD40 repeat proteins contains 40 amino acid residues, including histidine–glycine and aspartic acid–tryptophane dipeptides. The conserved motifs can be arranged in 4–16 tandems in the same protein.
Eight such random repeats are present in the β-subunit of G protein in higher eukaryotic organs.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Plant flavonoid-related WD proteins are grouped into the 4 WD-repeat (4WDR) subfamily of the TTG (Transparent teata glabra) family. The functions of several WD40 repeat proteins have been reported in petunia (Petunia hybrida), Perilla frutescens, cotton (Gossypium hirsutum), and maize (Zea mays) [,,]. In A. thaliana, only one WD40 protein (TTG1), involved in anthocyanin biosynthesis and trichome formation, has been reported []. The ortholog (MtWD40-1) of AtTTG1 is necessary for tissue-specific anthocyanin and PA biosynthesis in M. truncatula
[]. Two WD repeat genes (GhTTG1 and GhTTG2) were found in G. hirsutum [], but only GhTTG1 could restore trichome formation in the Arabidopsis ttg1 mutant and complemented the anthocyanin deficiency in the white-flowered Matthiola incana ttg1 mutant.[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
A total of 195 candidate WD genes were isolated from the transcriptome data sets of C. sinensis []. However, only one WD gene was confirmed to be the ortholog of AtTTG1 in tea plants. As shown in Figure 3, the petals of CsWD40 transgenic tobacco plants accumulated a large amount of anthocyanins, but no observable PA was found in these plants. In the Arabidopsis ttg1 mutant, CsWD40 not only complemented trichome deficiency but also restored seed coat color. The results implied that CsWD40 improved PA production in the seeds of the ttg1 mutant.
This result also implied the different roles of CsWD40 in transgenic tobacco and Arabidopsis plants. The cause was probably the difference in the dominant flavonoid biosynthesized in the petals of tobacco plants and the seeds of Arabidopsis plants. Anthocyanin is mainly biosynthesized in the petals of tobacco plants [], whereas PA is mainly biosynthesized in the seeds of Arabidopsis plants [].[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
Besides, not only the structural genes, NtCHS, NtF3′H, NtDFR, and NtANS, but also the transcript factor genes, NtAN2 and NtANb1, were upregulated significantly in CsWD40 transgenic tobacco petals (Figure 5B). A potential reason for this is that CsWD40 improves the stability of some MBW complexes or promotes the formation of some MBW complexes. These complexes may regulate the gene expression of some transcription factors. For example, TT8 expression was found to be directly regulated by TT8 itself through a positive feedback regulatory loop involving redundant MBW complexes [,].
TT8 promoter activity is itself was partially regulated by TT1 [].
The anthocyanin and PA biosynthetic pathways are regulated by different MBW complexes. Different MYB TFs play critical roles in the regulation of the complexes []. For example, TT2 (ATMYB123) in subgroup 5 mainly regulates PA biosynthesis in Arabidopsis. PAP1 (ATMYB75), in subgroup 6, significantly promotes anthocyanin accumulation in Arabidopsis [].
In our study, CsWD40 could interact with CsAN2 (the ortholog of PAP1) and CsMYB5e (the ortholog of TT2) in the yeast two-hybrid system (Figure 1). In our previous study, CsAN2 markedly increased anthocyanin biosynthesis in CsAN2-overexpressing tobacco plants []. In CsMYB5e transgenic tobacco plants, the PA pathway was markedly upregulated []. Expression levels of both anthocyanin and PA biosynthesis related genes were upregulated to varying degrees in the petals of CsWD40 and CsMYB5e co-overexpressing transgenic tobacco plants.
CsWD40 and CsMYB5e co-overexpressing transgenic tobacco plants tended to show less anthocyanin production than transgenic tobacco plants expressing CsWD40 alone. This might be because CsWD40 promoted the regulation effects of CsMYB5e, resulting in higher PA biosynthesis than anthocyan in biosynthesis. In our previous studies, CsANR overexpression reduced anthocyanin production and promoted PA accumulation in the petals of tobacco plants []. In the Arabidopsis mutant of BAN (loss of function of ANR), a high level of anthocyanin accumulated in the seed coat. Hybridization and yeast two-hybrid assays further confirmed that CsWD40 played a crucial role in the regulation of anthocyanin and PA biosynthesis through the formation of MBW ternary complexes
[].[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
In a previous study, the ectopic overexpression of TaWD40D (T. aestivum L.) in Arabidopsis greatly increased the tolerance of the plants to ABA (Abscisic acid), salt, and osmotic stresses during seed germination and seedling development []. In this study, the transcript levels of the CsWD40 gene were significantly upregulated under sucrose and ABA stresses, but no response was found to mannitol, NaCl, and SA (Salicylicacid) stresses under different temperatures. Our previous study showed that under ABA and sucrose stresses the most flavonoid was accumulated and the related gene expression increased []. This result suggests CsWD40 promotes the accumulation of products of the flavonoid biosynthetic pathway, which led to stress tolerance in tea plants.[](https://www.ncbi.nlm.nih.gov/mesh/D000040)
## 4. Materials and Methods
*In the **4. Materials and Methods** section:*
## 4.1. Plant Materials and Growth Conditions
*In the **4.1. Plant Materials and Growth Conditions** section:*
Numerous samples from different organs (including young sprout, young stem, and tender root) of the tea plants growing on the grounds of the research station at Anhui Agricultural University were obtained during the growth period in early spring. All samples were immediately frozen in liquid nitrogen and were stored at −80 °C for the present investigation.[](https://www.ncbi.nlm.nih.gov/mesh/D009584)
Wild-type tobacco “G28” plants were used in genetic transformation studies. Tobacco plants were grown in a controlled environment chamber at a constant temperature of 28 ± 3 °C and a 12/12-h (light/dark) photoperiod with a light intensity of 150–200 μmol·m–2·s–1.
A. thaliana (Ecotype Columbia) ttg1 mutant seeds were purchased from the Arabidopsis Seed Bank (). The methods used for seed germination and plant growth for A. thaliana were the same as those used for tobacco []. For genetic transformation, A. thaliana plants were grown in a green house. The temperature, light intensity, and photoperiod were 22 ± 2 °C, 50 µmol·m−2·s−1, and 16/8 h (light/dark), respectively.
In order to induce the anthocyanin accumulation in seedlings of Arabidopsis thaliana, Arabidopsis seeds were surface sterilized using 70% alcohol for 1 min followed by 3 times of washing with sterile distilled water, then treated with 1.5% sodium hypochlorite (10% Clorox), (SC Johnson Wax, Racine County, WI, USA) for 8 min followed by 6 times of washing with sterile distilled water. Seeds were germinated on MS medium with 6% sucrose. Seedlings were cultured for five days after germination and their phenotypes were observed under microscope.[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
## 4.2. Cloning CsWD40 and CsMYB5e
*In the **4.2. Cloning CsWD40 and CsMYB5e** section:*
Total RNA was isolated from plants using the RNAiso-mate for Plant Tissue Kit (Takara, Dalian, China), according to the manufacturer’s protocol. cDNAs were reverse transcribed using the PrimeScript® RT Reagent Kit (Takara, Dalian, China), according to the manufacturer’s protocol. The specific conditions of PCR were as follows: 98 °C for 30 s; followed by 30 cycles of 98 °C for 10 s and 60 °C for 20 s; 72 °C for 30 s; and then a 10-min extension step at 72 °C. The cDNAs of CsGL3, CsTT8, CsAN2, and CsMYB5e were provided by Tong Li and Xiaolan Jiang in our university.
## 4.3. Extraction and Analysis of Anthocyanins and PAs
*In the **4.3. Extraction and Analysis of Anthocyanins and PAs** section:*
To extract anthocyanins and PAs, the samples were ground in liquid nitrogen and extracted with extraction solution (0.5% HCl:80% methanol:19.5% water) by vortexing and then sonicating for 30 min at low temperature (4 °C). The samples were centrifuged at 5000× g for 20 min, and the residues were re-extracted twice. The supernatants were combined and diluted with extraction solution to 2 mL. Absorbance of the aqueous phase was measured at 530 nm. The extraction method for PA is described in the article by Jiang [].[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
The total PA content was determined spectrophotometrically at 640 nm after a reaction with the DMACA reagent (0.2% (w/v) DMACA in methanol-3 N HCl), with (−)-epicatechin serving as the standard [].[](https://www.ncbi.nlm.nih.gov/mesh/D044945)
## 4.4. Agrobacterium-Mediated Transformation of Arabidopsis and Tobacco
*In the **4.4. Agrobacterium-Mediated Transformation of Arabidopsis and Tobacco** section:*
The ORFs of CsWD40 were cloned into a binary vector pDONR207 by using the Gateway® Cloning System (Invitrogen, Carlsbad, CA, USA). The primers used are listed in Table S1. The vectors were constructed and the genetic transformation of tobacco was performed according to the method of Li et al. []. The phenotypes of transgenic plants growing in the controlled environment chamber were recorded to characterize the effect of transgene overexpression on growth.
## 4.5. Hybridization of Different Transgenic Tobacco Lines
*In the **4.5. Hybridization of Different Transgenic Tobacco Lines** section:*
The T1-generation plants of CsWD40 and CsMYB5e transgenic tobacco plants were chosen for research. Transgenic tobacco plants separately overexpressing CsMYB5e and CsWD40 were crossed in both ♀ × ♂ and ♂ × ♀ directions.
The pollinated ovaries were collected until the seeds ripened. The next generation of plants was grown for subsequent experiments after PCR verification.
## 4.6. Abiotic Stress Treatment of Tea Shoots
*In the **4.6. Abiotic Stress Treatment of Tea Shoots** section:*
For abiotic stress treatment, tea shoots sprouting approximately 10 cm were cultivated in 90 mM sucrose, 200 mM mannitol, 50 mM NaCl, 10 µM JA, 100 µM ABA, and 20 mM SA for 12 h. All samples were exposed to treatments at 22 °C, with a light intensity of 150–200 µmol·m−2·s−1. Control shoots were cultivated in deionized water. For low- and high-temperature treatment, the shoots were exposed to treatments at 10 and 50 °C for 1 and 6 h, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D013395)
## 4.7. Yeast Two-Hybrid Assay
*In the **4.7. Yeast Two-Hybrid Assay** section:*
The plasmids pGADT7 (Clontech Laboratories, Inc., Terra Bella Ave, Mountain View, CA, USA) and pGBKT7 (Clontech Laboratories, Inc., Terra Bella Ave, Mountain View, CA, USA), which contained the GAL4 activation domain and the GAL4 DNA-binding domain, were used for yeast two-hybrid experiments.
To analyze protein interaction, the ORFs of CsWD40, CsGL3, CsTT8, CsAN2, and CsMYB5e were cloned into the pGADT7 (Clontech Laboratories, Inc., Terra Bella Ave, Mountain View, CA, USA) and pGBKT7 vectors through pfu DNA PCR (Thermo Scientific, Waltham, MA, USA). This cloning was performed using the Matchmaker™ Gold Yeast Two-Hybrid System (Clontech), according to the manufacturer’s instructions described in Clontech Yeast Protocol Handbook. The primers with leading and tailing EcoRI and BamHI restriction enzyme sites were used and are listed in Table S1.
BD-CsWD40, CsGL3, CsTT8, CsAN2, or CsMYB5e and AD-CsWD40, CsGL3, CsTT8, CsAN2, or CsMYB5e were cotransformed into the yeast strain Y2HGold (Clontech) using the PEG/LiAC method described in the Clontech Yeast Protocol Handbook. These transformed colonies were tested on synthetic dropout (SD) medium with X-α-Gal lacking leucine, tryptophan, histidine, and adenine (SD/-Ade/-His/-Leu/-Trp).[](https://www.ncbi.nlm.nih.gov/mesh/D011092)
## 4.8. Quantitative Real-Time PCR
*In the **4.8. Quantitative Real-Time PCR** section:*
RNA was extracted from various tissues and was quantified spectrophotometrically (NANODROP 2000, Thermo Scientific). Reverse transcription of RNA into cDNA was performed using 2 μL of 5× PrimeScript RT Master Mix (Takara) and 500 ng of RNA in a reaction volume of 10 µL.
In addition, cDNA was diluted to 25% (v/v) with deionized water before being used as the template. Quantitative real-time PCR was performed in a reaction volume of 20 µL containing 10 µL of SYBR Green PCR Master Mix (Takara), 1.1 µL of cDNA, and 0.8 µL of forward and reverse primers (10 μM). The PCR cycling parameters used were as follows: 95 °C for 30 s and 40 cycles of 95 °C for 5 s, 30 s at 60 °C, and 30 s at 72 °C, followed by melting curve analysis from 55 to 95 °C. The transcription abundance was normalized to the transcription abundance of the control gene and was calculated from three technical replicates. The gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH, Accession No. GE651107) in tea plants and ribosomal protein L25 (RPL25, Accession No. L18908) in tobacco were used as control genes in qPCR analysis.
Their gene primers were listed in Table S1. The relative expression level was calculated using a previously described method [].[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
# Supplementary Materials
*In the **Supplementary Materials** section:*
Supplementary materials can be found at .
# Author Contributions
*In the **Author Contributions** section:*
Y.L. and X.J. conceived and designed the experiments; Y.L. and H.H. performed the experiments and analyzed the data; P.W., W.C., and X.D. contributed reagents/materials/analysis tools; L.G. and T.X. drafted the manuscript. All authors read and approved the final manuscript.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# Abbreviations
*In the **Abbreviations** section:*
ANR Anthocyanidinreductase ANS Anthocyanidin synthase CHS Chalcone synthase DFR Dihydroflavonolreductase DMACA Dimethylaminocinnamaldehyde F3H Flavanone 3-hydroxylase F3′H Flavonoid 3-hydroxylase FLS Flavonol synthase MBW MYB-bHLH-WD40 Pas Proanthocyanidins ttg1 Transparent testa glabra 1[](https://www.ncbi.nlm.nih.gov/mesh/C018523)
# References
*In the **References** section:*
Protein interactions between CsWD40 and other transcript factors from tea plants by two-hybrid system. (A) A model pattern of the MYB–bHLH–WD40 ternary complex. (B) Protein–protein interactions between CsWD40 and bHLH transcript factors (CsTT8 and CsGL3) from tea plants by two-hybrid system. (C) Protein–protein interactions between CsWD40 and MYB transcript factors (CsAN2 and CsMYB5e) from tea plants by two-hybrid system.
CsWD40 complements the phenotypes of Arabidopsis ttg1 mutant. (A) Leaf trichome occurrence. Bar = 0.5mm (B) Seed coat pigmentation. (C) Trichome seedling.
Identification of the CsWD40 function in transgenic tobacco. (A) Analysis of CsWD40 transcription levels in the flowers by semiquantitative PCR. (B) Relative content of anthocyanin in the flowers of CsWD40 overexpressing tobacco. (C) Relative soluble proanthocyanidin content in the flowers of CsWD40 overexpressing tobacco. All data are the means of three biological replicates, and the error bars represent the standard deviation of three replicates. Statistical significance was analyzed using ANOVA software (ANOVA ALL MAC VERSION 2.0, Thomas Hanson, OR, USA). Means followed by the same letter are not significantly different (p > 0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
Expression of genes involved in flavonoid biosynthesis in the flowers of CsWD40-overexpressing tobacco plants. (A) A schematic diagram of flavonoid biosynthetic pathway. (B) Expression profiles of genes in flavonoid pathway in flowers of transgenic CsWD40 tobacco lines. CHS, chalcone synthase; F3H, flavanone 3-hydroxylase; F3′H, flavonoid 3-hydroxylase; DFR, dihydroflavonol reductase; ANS, anthocyanidin synthase; ANR, anthocyanidin reductase; FLS, flavonol synthase; AN2, N. tabacum Anthocyanin 2; AN1a, N. tabacum Anthocyanin 1a; AN1b, N. tabacum Anthocyanin 1b.[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
Anthocyanin and proanthocyanidin accumulation variated in petals of co-overexpressing CsWD40 and CsMYB5e transgenic tobacco. (A) The relative content of anthocyanin was calculated on the record of absorbance at 530 nm. (B) The relative content of proanthocyanidin was calculated on the record of absorbance at 640 nm after reaction with DMACA regent. (C) RT–PCR determination of the CsWD40 and CsMYB5e expression levels in CsWD40, CsMYB5e, and CsWD40♀CsMYB5e♂ transgenic tobacco flowers. All the data were present based on three biological and technical repeats.[](https://www.ncbi.nlm.nih.gov/mesh/D000872)
(A) Relative expression of flavonoid biosynthetic pathway genes in CsMYB5e overexpression tobacco petals. (B) Relative expression of flavonoid biosynthetic pathway genes in CsWD40 and CsMYB5e overexpressing tobacco petals. CHS, chalcone synthase; F3H, flavanone 3-hydroxylase; F3′H, flavonoid 3-hydroxylase; DFR, dihydroflavonol reductase; ANS, anthocyanidin synthase; ANR, anthocyanidin reductase; FLS, flavonol synthase.[](https://www.ncbi.nlm.nih.gov/mesh/D005419)
Effects of different abiotic treatments on the expression levels of CsWD40 in tea leaves. (A) The transcript levels of CsWD40 under mannitol (MAN), sodium chloride (NaCl), salicylic acid (SA), abscisic acid (ABA), jasmonic acid (JA), and sucrose (Suc) stresses. (B) The transcript levels of CsWD40 under different temperatures. The asterisks indicate the significant level (n = 3, * p < 0.05, *** p < 0.001) based on a Tukey’s honestly significant difference test.[](https://www.ncbi.nlm.nih.gov/mesh/D008353)
|
# Introduction
Downregulation of survivin expression and concomitant induction of apoptosis by [celecoxib](https://www.ncbi.nlm.nih.gov/mesh/D000068579) and its non-cyclooxygenase-2-inhibitory analog, [dimethyl-celecoxib](https://www.ncbi.nlm.nih.gov/mesh/C506698) ([DMC](https://www.ncbi.nlm.nih.gov/mesh/C506698)), in tumor cells in vitro and in vivo
# Abstract
*In the **Abstract** section:*
Background
2,5-Dimethyl-celecoxib (DMC) is a close structural analog of the selective cyclooxygenase-2 (COX-2) inhibitor celecoxib (Celebrex®) that lacks COX-2-inhibitory function. However, despite its inability to block C[OX-2 activity, DMC is ](https://www.ncbi.nlm.nih.gov/mesh/C506698)ab[le ](https://www.ncbi.nlm.nih.gov/mesh/C506698)to potently mimic the anti-tumor effects of celecoxib in vitro and in vivo, indicat[ing that ](https://www.ncbi.nlm.nih.gov/mesh/D000068579)bo[th of th](https://www.ncbi.nlm.nih.gov/mesh/D000068579)ese drugs are able to involve targets other than COX-2 to exert their recognized cytotoxic effect[s. ](https://www.ncbi.nlm.nih.gov/mesh/C506698)However, the molecular components that are involved i[n mediati](https://www.ncbi.nlm.nih.gov/mesh/D000068579)ng these drugs' apoptosis-stimulatory consequences are incompletely understood.
Results
We present evidence that celecoxib and DMC are able to down-regulate the expression of survivin, an anti-apoptotic protein that is highly expressed in tumor cells and known to confer resistance of such cells to anti-cancer treatments. Suppression of survivin is specific to these two drugs, as other coxibs (valdecoxib, rofecoxib) or traditional NSAIDs (flurbiprofen, indomethacin, sulindac) do not affect survivin expression at similar concentrations. The extent of survivin down-regulation by celecoxib and DMC in different tumor cell lines is somewhat variable, but closely correlates with the degree of drug-induced growth inhibition and apoptosis. When combined with irinotecan, a widely used anticancer drug, celecoxib and DMC greatly enhance the cytotoxic effects of this drug, in keeping with a model that suppression [of surviv](https://www.ncbi.nlm.nih.gov/mesh/D000068579)in ma[y b](https://www.ncbi.nlm.nih.gov/mesh/C506698)e beneficial to sensitize cancer cells to chemotherapy. Remarkably, these effects are not restricted to in vitro conditions, but also take place in tumors from drug-treated animals, where both drugs similarly repress survivin, induce apoptosis, and inhibit tumor gro[wth in viv](https://www.ncbi.nlm.nih.gov/mesh/C406224)o.[](https://www.ncbi.nlm.nih.gov/mesh/C116926)
Conclusion
In consideration of survivin's recognized role as a custodian of tumor cell survival, our results suggest that celecoxib and DMC might exert their cytotoxic anti-tumor effects at least in part via the down-regulation of survivin – in a manner that does not require the inhibition of cyclooxygenase-2. Because inhibition of COX-2 appears to be negligible, it might be worthwhile to further evaluate DMC's potential as a non-coxib alternative to celecoxib for anti-cancer purposes.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Nonsteroidal anti-inflammatory drugs (NSAIDs) have long been implicated in the treatment or prevention of various types of cancer. The biochemical mechanism generally ascribed to this effect is the inhibition of cyclooxygenase (COX) enzymes, which catalyze the initial step in prostaglandin synthesis [1-3]. The traditional NSAIDs, such as flurbiprofen, indomethacin, or sulindac, are able to inhibit both COX-1 and COX-2 enzymes, while new generation drugs, such as celecoxib (Celebrex®), valdecoxib (Bextra®), or rofecoxib (Vioxx®), inhibit only COX-2. Due to their more selective function, these latter drugs, referred to as coxibs, initially had promised to offer the therapeutic benefit of traditional NSAIDs with less of the associated side effects [4-7]; however, this expectation has come under intense scrutiny and has generated considerable controversy in the recent past [8-10].[](https://www.ncbi.nlm.nih.gov/mesh/D011453)
Celecoxib is widely prescribed under the trade name Celebrex® for relief of symptoms of osteoarthritis and rheumatoid arthritis and was also approved as an adjunct to standard care for patients with familial adenomatous polyposis (FAP). It is suspected that this drug might be useful for the prevention and treatment of colorectal and possibly other types of cancer, and several clinical trials are ongoing to confirm this expectation. In addition, celecoxib has demonstrated potent anti-cancer activity in various animal tumor models in the laboratory [11-17]. Despite these promising results, however, the underlying molecular mechanisms by which celecoxib exerts its anti-tumor potential are not completely understood, in particular because of numerous reports describing potent anti-proliferative and pro-apoptotic effects of this drug in the absence of any apparent involvement of COX-2
[18-24].[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
In order to investigate the COX-2 independent anti-tumor mechanisms of celecoxib in greater detail, we and others have generated close structural analogs of this compound that lack the ability to inhibit COX-2 activity [25-28]. One such analog is 2,5-dimethyl-celecoxib (DMC), a compound that was first developed in the laboratory of Ching-Shih Chen at Ohio State University [26,28]. Intriguingly, despite its inability to inhibit COX-2, DMC is able to faithfully mimic – without exception – all of celecoxib's numerous anti-tumor effects that have been investigated so far, including the reduction of neovascularization and the inhibition of experimental tumor growth in various in vivo tumor models [21,25,26,28-32].
Therefore, DMC appears to be well suited for studies intended to illuminate the COX-2 independent anti-tumor effects of celecoxib [33].[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
Because celecoxib and DMC are potent inducers of apoptosis, we investigated their effects on survivin, which is a member of the inhibitor of apoptosis (IAP) family of proteins that has been implicated in the control of cell division and apoptosis [34]. Survivin's function in mitosis is to preserve the mitotic apparatus and to allow normal mitotic progression, whereas its anti-apoptotic function is executed via its ability to prevent caspase activation. The protein is usually not expressed in differentiated normal adult tissues, but is elevated in the majority of human cancers, with very high levels generally being predictive of tumor progression and poor prognosis. In addition, survivin appears to be involved in tumor cell resistance to some anticancer agents and ionizing radiation (for detailed references, see reviews
[35-37].[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
As the above-described characteristics established survivin as a potential target for anticancer therapy, we investigated whether the expression of this anti-apoptotic protein could be restrained by celecoxib and DMC. Here we report that both drugs are able to down-regulate survivin expression and induce apoptosis in numerous tumor cell lines. These effects are not restricted to in vitro conditions, but also take place in drug-treated animals in vivo, where both drugs repress survivin and induce apoptosis in xenograft tumor tissue. Thus, in consideration of survivin's recognized role as a guardian of tumor cell survival, our results suggest that celecoxib and DMC might exert their cytotoxic anti-tumor effects at least in part via the down-regulation of survivin. Because DMC lacks COX-2 inhibitory function, these anti-tumor effects appear to take place without the involvement of celecoxib's well-known target, cyclooxygenase-2.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Results
*In the **Results** section:*
## Celecoxib and DMC down-regulate survivin protein levels
*In the **Celecoxib and DMC down-regulate survivin protein levels** section:*
To determine whether celecoxib and DMC would be able to affect survivin expression in a variety of human tumor types, we treated a collection of derived cell lines with either drug in vitro. Because it had been established earlier that DMC is generally more potent than celecoxib, we used 30 and 50 μM of DMC, and 40 and 60 μM of celecoxib. As shown in Figure 1, both drugs were able to down-regulate survivin expression in all cell lines investigated, which included cells derived from glioblastoma, lymphoma, multiple myeloma, and carcinoma of the breast, colon, and prostate. Consistent with earlier studies on other targets, DMC exerted stronger effects than celecoxib and caused a more potent down-regulation of survivin. Although this effect was observed in all cell types, the overall magnitude of down-regulation varied between individual cell lines; for example, whereas Raji lymphoma, T98G glioblastoma, and T47D breast carcinoma cells displayed a very strong down-regulation of survivin, LN229 glioblastoma, MCF7 breast carcinoma, and HCT116 colon carcinoma showed a weaker response at the same concentrations.
However, further increased concentrations of these two drugs invariably led to complete downregulation of survivin expression in all cell lines examined, i.e., 60–70 μM DMC or 70–80 μM celecoxib completely suppressed survivin expression, which was accompanied by severe cytotoxicity (not shown).[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
Celecoxib and DMC decrease levels of survivin protein in various cancer cell lines. Several different cancer cell lines were cultured in the presence of celecoxib (Cxb) and DMC for 48 hours as indicated. Total cellular lysates were prepared and analyzed by Western blot analysis with specific antibodies to survivin. As a control for equal loading, all blots were also analyzed with antibodies to actin (only two of these control blots are shown at the bottom). The tumor type of each cell line is indicated on the right.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Down-regulation of survivin is independent of p53
*In the **Down-regulation of survivin is independent of p53** section:*
Because the above results indicated a certain cell type-specific sensitivity with regards to the down-regulation of survivin, we comparatively analyzed several relevant parameters in these cell lines. As it has been shown earlier that the status of the tumor suppressor p53 might influence basal levels of survivin expression [38,39], we investigated whether there was a correlation of p53 status with the basal and/or the differential drug-reduced levels of survivin. As can be seen in Figure 2A, the basal level expression of survivin, i.e., the cellular amount of survivin protein in the absence of drug treatment, varied greatly among the various tumor cell lines. However, overall there was no obvious correlation between this variation of basal level expression and the efficacy of drug-induced repression (compare to Figure 1). But when the mutational status of the p53 gene in these cell lines was investigated from data of the published literature (presented at the top of Figure 2A), and was compared among cell lines of the same tumor type, it appeared that the presence of mutant p53 exerted a small, yet noticeable influence on the efficacy of survivin down-regulation by DMC and celecoxib in some of the cells.
For example, in the pair of breast carcinoma cell lines MCF7 (p53 wt) and T47D (p53 mut), T47D displayed a higher basal level (Figure 2) and stronger down-regulation of survivin than MCF7 (Figure 1). The same held true among the various glioblastoma cell lines we investigated: T98G and U251 (both p53 mut) displayed higher basal levels and a somewhat stronger down-regulation of survivin than U87 and LN229 (both p53 wt). Similarly, the colon carcinoma pair HCT116 (p53 wt) and DLD-1 (p53 mut) followed this pattern as well, although in this case the difference was less pronounced.[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
Basal level expression of survivin and Cox-2 proteins in various cancer cell lines and effect of p53 and p21. In (A), the various cancer cell lines were cultured in the absence of any drug treatment, harvested in log phase, and analyzed by Western blot analysis with antibodies to survivin, cycloxygenase-2 (Cox-2), and actin (as a loading control). In addition, the p53 status of each line (as reported in a variety of reports) is indicated (wt: wild type; m: mutant). (Note that in LN229 cells, wt p53 function is retained, despite a mutation in the coding sequence.) In (B), three variants of HCT116 colon carcinoma cells were treated with celecoxib (Cxb) or DMC and analyzed by Western blot analysis for survivin levels and actin (as a loading control; only one representative panel is shown). The top panel shows results with HCT116 cells that harbor wild type alleles of the p53 and p21 genes; the second panel is from cells with disrupted p53 alleles (p53-/-); the third panel is from cells lacking p21 (p21-/-).[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
However, the correlation between p53 status and basal and drug-reduced survivin levels did not hold true in all cell lines. For example, the pair of prostate carcinoma cell lines, MIA-PaCa-2 and Bx-PC-3, displayed a noticeable difference in their basal levels of survivin and in their response to the drugs, even though these cells both harbor mutant p53. Therefore, in order to distinguish whether the observed differential drug responses were indeed related to p53, or rather were an expression of the general genetic heterogeneity of these aneuploid tumor cells, we used an HCT116 colon carcinoma cell line where the p53 gene (or one of its crucial target genes, the cyclin-dependent kinase inhibitor p21Waf1, which was found to mediate p53's repression of survivin [40]) was disrupted by targeted homologous recombination [41,42]. As shown in Figure 2B, inactivation of p53 resulted in a minor reduction of drug effects, whereas inactivation of p21 had no effect.
Thus, taken together, we conclude that p53 does not play a major role in the observed differential down-regulation of survivin by celecoxib or DMC.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Down-regulation of survivin is independent of cyclooxygenase-2
*In the **Down-regulation of survivin is independent of cyclooxygenase-2** section:*
Another parameter we decided to analyze in the various tumor cell lines was cyclooxygenase-2 (COX-2). Although the use of DMC, which does not inhibit COX-2, already indicated that this enzyme quite likely played no role in the observed drug effects, we determined the levels of COX-2 protein and investigated whether they would correlate with the sensitivity of these cells to DMC and/or celecoxib. The amount of COX-2 protein was established by Western blot analysis and is shown in Figure 2A. However, when compared to the data presented in Figure 1, we found that cell lines with elevated levels of COX-2 (U87, LN229, Bx-PC-3) did not consistently differ in their extent of survivin down-regulation as compared to cell lines lacking COX-2 (Raji, RPMI/8226, HCT116, MIA-PaCa-2). Thus, as expected, no correlation between COX-2 expression and the degree of survivin down-regulation by DMC or celecoxib was found.[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
The lack of COX-2 involvement was further confirmed by comparing the effects of DMC and celecoxib to other established inhibitors of this enzyme. For instance, flurbiprofen, indomethacin, and sulindac are traditional NSAIDs that inhibit both COX-1 and COX-2, whereas valdecoxib and rofecoxib are coxibs that selectively inhibit only COX-2. When two different tumor cell lines were treated with various concentrations of the above inhibitors, no effect on survivin expression was observed, even at concentrations of up to 100 μM (Figure 3, bottom part), which are more than double the effective concentrations of celecoxib and DMC. Thus, the significant down-regulation of survivin by DMC and celecoxib could not be achieved by comparable concentrations of other COX-2 inhibitors, clearly arguing against an involvement of COX-2 in these processes. In addition, none of these other COX-2 inhibitors was able to substantially impinge on cell growth and survival of these cells (Figure 3, top part), nor were these compounds able to induce apoptosis at these concentrations (not shown).
Thus, the differential effects of DMC, celecoxib, and other coxibs and traditional NSAIDs indicated a correlation between the effects on survivin expression and cell survival or death.[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
Downregulation of survivin is specific to celecoxib and DMC and correlates with reduced survival. U251 glioblastoma or BxPc-3 pancreatic carcinoma cells were cultured in the presence of DMC, various non-steroidal anti-inflammatory drugs (NSAIDs), or solvent DMSO alone, at the concentrations indicated. Cell growth and survival was determined by standard MTT assay (top part of figure). In parallel, total cellular lysates were prepared and analyzed by Western blot analysis with specific antibodies to survivin or to actin as a loading control (bottom part of figure).[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Down-regulation of survivin involves transcriptional repression
*In the **Down-regulation of survivin involves transcriptional repression** section:*
We had shown earlier that celecoxib and DMC are able to inhibit the expression of two key cell cycle-regulatory genes, cyclin A and cyclin B, at the transcriptional level [20,25]. To determine whether survivin expression was similarly affected by these drugs, we generated cells that were stably transfected with luciferase reporter constructs under the control of the survivin promoter. Two different constructs were used; one contained 6270 bp of upstream promoter sequences of the survivin gene, the other only 230 bp. As shown in Figure 4, the activity of both of these constructs was similarly inhibited by DMC and celecoxib (not shown for celecoxib), indicating that these drugs were able to impinge on survivin transcription. As controls, we used a reporter construct under the control of the cyclin B promoter, which, as expected, was down-regulated by DMC as well; however, a luciferase construct under the control of the cytomegaloviral (CMV) promoter was not affected, indicating that DMC (and celecoxib) did not block transcription indiscriminately.
Thus, we conclude that, in addition to cyclin A and cyclin B, survivin represents yet another target of these drugs that is affected at the transcriptional level.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
DMC decreases the activity of the survivin promoter. Mass cultures of LN229 cells stably transfected with various luciferase reporter constructs under the control of either the survivin promoter (-6270Surv and -230Surv), the cyclin B promoter, or the cytomegalovirus (CMV) promoter, were treated with different concentrations of DMC for 36 hours. Thereafter, cellular lysates were analyzed for luciferase activity. For each reporter construct, basal level activity in the absence of drug at 36 hours was set to 100%. Shown is the mean (± SD; n = 3) luciferase activity from one experiment, which was repeated twice with similar results.[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
## Down-regulation of survivin correlates with increased apoptosis
*In the **Down-regulation of survivin correlates with increased apoptosis** section:*
Because survivin has a recognized role as an inhibitor of apoptosis, we next investigated whether and how the observed down-regulation of survivin by DMC would relate to the known ability of this drug to induce apoptosis. We used several different representative cell lines (U251, T98G, and LN229 glioblastoma; BxPc-3 and MIA PaCa-2 pancreatic carcinoma) with differing sensitivities to DMC, and comparatively analyzed their response to 30 and 50 μM DMC. As shown in Figure 5, U251, T98G, and BxPc-3 cells responded quite sensitively; these cells displayed a potent down-regulation of survivin, and at the same time strongly increased apoptosis in combination with greatly reduced survival. On the other hand, at these same concentrations of DMC, LN229 and MIA PaCa-2 cells exhibited only a minor down-regulation of survivin, which correlated with marginally increased apoptosis and a much weaker effect on overall cell survival (Figure 5). Thus, the magnitude of survivin down-regulation caused by DMC closely correlated with the extent of apoptosis and with the degree of short-term growth and survival (as determined by MTT assay), as well as long-term survival (as determined by colony forming ability) of these cells.[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
Downregulation of survivin by celecoxib and DMC correlates with increased apoptosis and reduced cell growth and survival. The three glioblastoma cell lines U251, T98G, and LN229 (A), or the two pancreatic carcinoma cell lines BxPc-3 and MIA PaCa-2 (B), were treated with 30 or 50 μM DMC or remained untreated for 48 hours. The effects on cell growth/survival and on cell death were determined by various assays. The panels labeled Number of Colonies display the results from a colony forming assay, where the number of surviving cells able to spawn a colony of newly grown cells was determined; in this assay, the colonies of adherent cells were stained and visualized with methylene blue two weeks after drug treatment and were counted. The panels labeled % Cell Growth and Survival show the results from MTT assays performed at the end of the 48 hour drug treatment period. The panels labeled % Apoptotic Cells present the percentage of cells undergoing apoptosis as revealed by the TUNEL assay after 48 hours of drug treatment. At the bottom of each series of panels in A and B, the level of survivin protein at the end of drug treatment is shown, as determined by Western blot analysis with specific antibodies. Western blots for actin are also shown (as a loading control).[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Celecoxib and DMC enhance cell killing by CPT-11
*In the **Celecoxib and DMC enhance cell killing by CPT-11** section:*
With the use of the U251 and LN229 glioblastoma cell lines, we next investigated whether DMC would be able to synergize with other chemotherapeutic drugs to achieve increased tumor cell killing. For this purpose, we used irinotecan (CPT-11) and temozolomide as two representative drugs that are commonly used for the treatment of high-grade brain tumors [43] and determined tumor cell survival with the use of the colony forming assay. Intriguingly, while DMC dramatically increased the cytotoxicity of CPT-11, no such enhancing effect was observed in combination with temozolomide (Figure 6). Furthermore, the outcome was the same in both cell lines, U251 and LN229, which are known to differ in the status of their p53 and PTEN tumor suppressor genes [44,45] (and probably a few other genes as well).
Thus, while this result established that DMC is able to cause substantial chemosensitization of glioblastoma cells with different genetic backgrounds, it also revealed that this effect apparently does not take place indiscriminantly with any type of anticancer drug.[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
Combination drug effects of DMC with CPT-11 or temozolomide. U251 and LN229 glioblastoma cells were treated with DMC, CPT-11, and temozolomide (TMZ) either alone or in combination as indicated for 48 hours. The percentage of surviving cells was established by the conventional colony forming assay, where the number of surviving cells able to spawn a colony of newly grown cells was determined two weeks after drug treatment. Shown are the results from one experiment performed in triplicate, which was repeated several times with very similar results.[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
## Celecoxib and DMC down-regulate survivin and induce apoptosis in vivo
*In the **Celecoxib and DMC down-regulate survivin and induce apoptosis in vivo** section:*
Finally, we investigated whether the effects of DMC and celecoxib on survivin expression would also take place in vivo. For this purpose, we used a xenograft nude mouse tumor model with subcutaneously implanted glioblastoma cells. After palpable tumors had developed, the animals received chow supplemented with either celecoxib, DMC, or no drug (control group). As shown in Figure 7, the group of animals that were treated with either celecoxib or DMC displayed significantly (p < .01 and p < .003, respectively) reduced tumor growth as compared to the group of untreated animals, which was in keeping with similar results published with the use of prostate carcinoma and Burkitt's lymphoma xenograft mouse tumor models [21,25].[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
Inhibition of tumor growth by celecoxib and DMC in vivo. Nude mice were implanted subcutaneously with U87 glioblastoma cells, and two weeks later received daily chow supplemented with celecoxib, DMC, or no drug. Shown here is the increase in tumor volume over time (mean ± SD; n = 8). At the end of the experiment, the difference in mean tumor volume between the non-treated groups and the groups receiving celecoxib or DMC was statistically significant (p < .01 and p < .003, respectively). Shown are two independent experiments that were performed at different times with different batches of U87 cells and different shipments of animals; therefore, a direct comparison between animals that received celecoxib and animals that received DMC is not possible.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
Representative tumors were collected from the animals and analyzed by immunohistochemistry for survivin expression and with the TUNEL assay for the presence of apoptotic cell death. Typical results from the staining of numerous tumor sections are presented in Figure 8 (bottom half). For comparative purposes, we also performed the same type of analysis on glioblastoma cells cultured and treated with drugs in vitro (see top half of Figure 8). Under in vitro conditions, and in keeping with the results shown further above, celecoxib and DMC caused substantial reduction of survivin expression, and at the same time, increased levels of apoptotic cell death (Figure 8, top). Tumor tissue obtained from control (non-drug treated) animals stained strongly positive for survivin protein, and at the same time, was apparently negative for the presence of apoptotic cell death (Figure 8, bottom).
In contrast, tumor tissue from drug-treated animals displayed drastically reduced levels of survivin, to the point where not a single positive cell could be found in tumors from DMC-treated animals. Concomitantly, the tumor tissue from drug-treated animals stained clearly positive for the presence of apoptotic cell death (Figure 8, bottom). Thus, in agreement with the findings obtained in vitro, we found that in vivo as well, both drugs were able to suppress survivin expression and concomitantly induce apoptosis in tumor tissue.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
Downregulation of survivin by celecoxib and DMC correlates with increased apoptosis in vitro and in vivo. Top half: U87 glioblastoma cells were treated with celecoxib (Cxb) or DMC for 48 hours in vitro; thereafter, cytospins were performed and the cells were subjected to immunohistochemical analysis of survivin protein levels and, in parallel, TUNEL assay for apoptotic cell death. Bottom half: tumor sections from animals described in Figure 7 were analyzed by immunohistochemistry for survivin expression and by TUNEL assay for apoptotic cell death. In all cases, representative sections are shown. Small black rectangles denote enlarged areas of the same photograph shown below. Arrows indicate examples of TUNEL-positive, i.e., apoptotic, cells.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Discussion
*In the **Discussion** section:*
The selective COX-2 inhibitor celecoxib appears to hold promise for the treatment and prevention of colorectal cancer and possibly for other cancers as well. Because COX-2 is an oncogene [46] and over-expressed in a large number of tumors, it is generally thought that the COX-2-inhibitory function of celecoxib is critical for its anti-tumor property [4,47-49]. However, several recent studies [19,21-24,27,50], including from our laboratory [20,51], have indicated that celecoxib might be unique among the class of coxibs because this particular compound appears to be able to also suppress tumor formation in the absence of COX-2 involvement. For example, all coxibs completely inhibit COX-2 at very low micromolar concentrations in cell culture; yet only celecoxib causes efficient growth arrest and induction of apoptosis at low concentrations – an effect that is furthermore independent of the amount, or even the presence, of intracellular COX-2 (i.e., it takes place even in cells that lack COX-2 protein) [20,23,26,30,50,52-54].
Additional strong support for COX-2-independent anti-tumor effects of celecoxib has come from the use of its close structural analog, 2,5-dimethyl-celecoxib (DMC) (.)[33], which lacks COX-2 inhibitory function, yet was shown to faithfully mimic the anti-tumor effects of celecoxib in various experimental systems, including the reduction of neovascularization and the inhibition of experimental tumor growth in prostate carcinoma and Burkitt's lymphoma xenograft mouse tumor models [21,25,26,28-32].[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
The underlying mechanisms of celecoxib's (and DMC's) COX-2 independent anti-tumor effects are not completely understood, although several non-COX-2 targets have been described that are affected by these two drugs in vitro and in vivo [21,25-28,31,32]. In the present report, we demonstrate that survivin, a protein that is critically involved in the regulation of mitosis and the protection of cells from apoptosis, is potently down-regulated by celecoxib and by DMC in all tumor cell lines examined. This effect appears to be independent of any involvement of COX-2, as indicated by three observations: (i) both drugs down-regulate survivin even in cells that do not express detectable amounts of COX-2 (Figure 2A); (ii) none of the other COX inhibitors tested, including the coxibs rofecoxib (Vioxx) and valdecoxib (Bextra), are able to impinge on survivin expression (Figure 3); (iii) DMC does not inhibit COX-2, yet potently down-regulates survivin as well.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
There are a few reports from other groups [55-58] indicating that, in addition to celecoxib, some other NSAIDs appear to be able to reduce survivin expression, and these findings could be viewed as being discrepant to ours. However, much higher concentrations were required; for example, Zhang et al. [58] applied 200 μM of sulindac, and Lin et al. [57] used 300 μM of etodolac to impact survivin expression. Compared to our results presented here, these reports further emphasize our observation that celecoxib and DMC are unique in that these two drugs are able to suppress survivin expression at significantly lower concentrations than other NSAIDs.
Furthermore, studies with the use of non-small cell lung cancer (NSCLC) cell lines have indicated that increased COX-2 activity might contribute to the stabilization of survivin in these cells [59,60]. While these reports indicate a role of COX-2 in the expression of survivin, it appears that this observation cannot be generalized, as we have not observed a correlation between COX-2 activity and the expression levels of survivin in the various tumor cells lines used in our study (Figure 2).[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
The potent down-regulation of survivin by celecoxib and DMC, but not by other COX inhibitors, is reminiscent of earlier reports demonstrating that only celecoxib and DMC, but not other COX inhibitors, are able to efficiently induce apoptosis at comparatively low concentrations [21,25,26,28]. This correlation suggests that survivin might be an important mediator of the cell death-inducing function of celecoxib and DMC. Indeed, when we compared the kinetics of survivin down-regulation with the resulting increase in apoptosis in two cell lines with varying sensitivities to DMC (Figure 5), we noticed a very close correlation between the degree of survivin down-regulation and the induction of apoptosis. In these cases, stronger down-regulation of survivin by DMC was associated with substantially more efficient induction of apoptosis.
These results are also consistent with our observation (Figure 3) that those NSAIDs that did not affect survivin expression (rofecoxib, valdecoxib, flurbiprofen, and others) also did not impinge on cell growth and survival and did not induce apoptosis.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
In addition to survivin, there are several other intracellular proteins that are known to restrain cell death when highly expressed, such as, for example, Bcl-2, Bcl-xL, c-IAP2, XIAP, and FLIP, which also have been found overexpressed in some tumors [61]. While our study did not investigate the potential contribution of these components, studies by others have excluded the involvement of Bcl-2, Bcl-xL, Bax, Bad, or Bak in the apoptosis-stimulating mechanisms of celecoxib and several of its derivatives, and instead provided evidence that these drugs appear to function via the disruption of the mitochondrial membrane potential [62]. This latter observation is of particular relevance, as it has been demonstrated that suppression of survivin expression by RNA interference causes loss of mitochondrial membrane potential and spontaneous apoptosis [63].
Taken together, these data consistently support our view that the observed down-regulation of survivin by celecoxib and DMC might constitute an important step in the induction of apoptotic cell death by these drugs.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
Considering the well-known function of survivin as an inhibitor of caspases and, consequently, as an anti-apoptotic protein [35,64], it is not surprising that down-regulation of this protein by celecoxib and DMC is associated with increased cell death. It has been shown in several other experimental systems that the down-regulation of survivin expression, for example by antisense or siRNA approaches [65], results in elevated "basal level" apoptosis and, perhaps more importantly, causes substantially increased sensitivity of such tumor cells to killing by chemotherapeutic drugs or ionizing radiation (for examples, see [66-71]). From these earlier results, one might expect that the down-regulation of survivin by celecoxib or DMC should sensitize these cells to other cancer drugs. We tested this assumption with two widely used anticancer drugs, CPT-11 (irinotecan; Camptosar®) and temozolomide (Temodar®). Intriguingly, while DMC vastly increased cell killing by CPT-11, no such enhancing effect was observed after co-treatment with temozolomide.
Thus, while these results establish proof-of-principle that DMC can substantially enhance tumor cell killing by other anticancer drugs, this obervation cannot be generalized and certainly deserves further study. In this context, it should be noted that celecoxib has been shown previously to enhance the anti-tumor efficacy of CPT-11 in a xenograft mouse model in vivo [16], and a Phase II study revealed encouraging activity of this drug combination among heavily pretreated patients with recurrent malignant glioma [72]. Considering the apparent mimicry of celecoxib's anti-tumor effects by DMC, it might be worthwhile to explore the combination effects of CPT-11 and DMC in greater detail. The potential advantages of evaluating the non-coxib DMC for use in the clinic will be discussed further below.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
Our efforts to understand the mechanisms by which DMC accomplishes the down-regulation of survivin revealed that at least part of this regulation occurs at the level of transcription, i.e., our results clearly indicate that DMC is able to potently inhibit survivin expression at the gene level via the inhibition of promoter activity (Figure 4). The extent of survivin promoter inhibition is comparable to the transcriptional repression of the cyclin A and cyclin B promoters by DMC and celecoxib, which we described earlier and which represents a crucial component of the cell cycle-inhibitory function of these two drugs [20,25]. Thus, similar to the negative regulation of cell cycle components by these two drugs, transcriptional events also appear to be involved in mediating their apoptosis-inducing function (not shown for celecoxib).[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
Although the above described transcriptional events are quite prominent, additional levels of survivin regulation by celecoxib and DMC are likely. For example, it has been shown that survivin protein is stabilized and protected from degradation via its phosphorylation by the critical cell cycle regulator, cyclin-dependent kinase (CDK). In particular, phosphorylation on threonine-34 of the survivin protein, which is accomplished by the cyclinB/cdk1 complex, leads to substantial extension of survivin's half-life during mitosis [73,74]. Conversely, it has been shown that the inhibition of cyclinB/cdk1 activity by various modes of intervention leads to increased turn-over and loss of survivin protein [75-78].
In this regard, we have recently demonstrated that the transcriptional down-regulation of cyclin A and cyclin B by celecoxib or DMC, as mentioned further above, effects the complete loss of enzymatic activity of the respective CDK complexes, including cyclinB/cdk1 [20,25]. Thus, we surmise that in addition to the transcriptional down-regulation of survivin expression, DMC and celecoxib also cause its increased posttranslational degradation via the elimination of CDK enzymatic activity.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
In the past, studies investigating the COX-2 independent effects of celecoxib in vitro have been received with reservations, due to the relatively high concentrations of drugs that were required to generate such effects. While drug concentrations between 10 to 80 μM are generally needed to produce anti-proliferative and apoptosis-inducing effects in cell culture in vitro, celecoxib concentrations measured in the serum of patients or animals are in the range of 3–10 μM [79-81]. Thus, this discrepancy has led to the suggestion [17,82] that in vitro effects of celecoxib (and perhaps DMC) might be an artifact and not reflective of the mechanisms taking place in vivo. It was therefore imperative for us to demonstrate whether or not the down-regulation of survivin by celecoxib and DMC could be recapitulated in an in vivo model.
As convincingly demonstrated by our results, both celecoxib and DMC were able to potently inhibit survivin expression in tumors of a xenograft mouse tumor model (Figure 8). Even more so, similar to the events in our in vitro system, the number of apoptotic cells in tumors from drug-treated animals was substantially elevated. We therefore believe that those drug-induced events that we documented under elevated drug concentrations in vitro do not represent artifacts of the cell culture system, but rather are reflective of events that also take place in vivo in drug-treated animals.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
The experimental use of DMC alongside celecoxib encompasses an important aspect that relates to the recently revealed potentially life-threatening side effects of coxib use in the clinic. The long-term use of coxibs at high dosages – as believed to be necessary if used in anti-cancer therapy – is troubled by severe, potentially life-threatening risks, such as cardiovascular events, renal injury, and gastrointestinal toxicity [9,83-86]. Considering that these side effects are believed to be a class effect due to the inhibition of COX-2 and the resulting imbalance of prostanoids [8,87,88], it is tempting to speculate that the clinical use of a celecoxib analog such as DMC, which lacks COX-2 inhibitory function but maintains anti-tumor potency, perhaps might avoid many of these unwanted side effects – and possibly could be used at even higher dosages than celecoxib for certain anti-tumor purposes.[](https://www.ncbi.nlm.nih.gov/mesh/C506698)
## Conclusion
*In the **Conclusion** section:*
It has become clear that at least parts of celecoxib's documented anti-tumor effects are mediated via mechanisms that do not appear to involve COX-2. In this regard, our study presents the anti-apoptotic and chemoprotective protein survivin as an apparently important component that is involved in mediating the drug's COX-2-independent induction of apoptotic tumor cell death. This provides additional evidence that DMC, which does not inhibit COX-2, is able to potently mimic all known anti-tumor functions of celecoxib, and further supports our proposition [33] that it might be worthwhile to further evaluate DMC's potential anti-cancer benefit in the clinic.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Materials and methods
*In the **Materials and methods** section:*
## Materials
*In the **Materials** section:*
Celecoxib is 4- [5-(4-methylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benzenesulfonamide [89]. DMC is a close structural analog, where the 5-aryl moiety has been altered by replacing 4-methylphenyl with 2,5-dimethylphenyl, resulting in 4- [5-(2,5-dimethylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benzenesulfonamide [21,51]. Both compounds were synthesized in our laboratory according to previously published procedures; see ref. [89] for celecoxib and ref.
[51] for DMC. Each drug was dissolved in DMSO at 100 mM (stock solution). In the case of valdecoxib [90] and rofecoxib [91], commercial caplets of Bextra® (Pfizer, New York, NY) and Vioxx® (Merck, Whitehouse Station, NJ), respectively, were suspended in H2O to disintegrate the excipient, and the active ingredient was dissolved in DMSO at 25 mM. In addition, we used pure rofecoxib powder that was synthesized in our laboratory according to established procedures [92]. All traditional NSAIDs were purchased from Sigma (St. Louis, MO) in powdered form and dissolved in DMSO at 100 mM. All drugs were added to the cell culture medium in a manner that kept the final concentration of solvent (DMSO) below 0.5%.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Cell lines and culture conditions
*In the **Cell lines and culture conditions** section:*
Most cell lines were obtained from the American Tissue Culture Collection (ATCC) and were propagated in DMEM or RPMI (GIBCO BRL, Grand Island, NY) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 0.1 mg/ml streptomycin in a humidified incubator at 37°C and a 5% CO2 atmosphere. The HCT116 colon carcinoma cell line, and derivatives thereof where the p53 tumor suppressor gene or the p21Waf1 gene were disrupted by targeted homologous recombination [41,42], were kindly supplied by Bert Vogelstein, Johns Hopkins Oncology Center (Baltimore, MD). Some of the glioblastoma cell lines were provided by Frank B. Furnari and Webster K. Cavenee (Ludwig Institute of Cancer Research, La Jolla, CA).[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Immunoblots and antibodies
*In the **Immunoblots and antibodies** section:*
Total cell lysates were prepared by lysis of cells with RIPA buffer [93], and protein concentrations were determined using the bicinchoninic acid (BCA) protein assay reagent (Pierce, Rockford, IL). For Western blot analysis, 50 μg of each sample was processed as described [94]. The primary antibodies were purchased from Cell Signaling Technologies (Beverly, MA), Cayman Chemical (Ann Arbor, MI), or from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) and were used according to manufacturer's recommendations. The secondary antibodies were coupled to horseradish peroxidase, and were detected by chemiluminescence using the SuperSignal West substrate from Pierce. All immunoblots were repeated at least once to confirm the results.[](https://www.ncbi.nlm.nih.gov/mesh/C047117)
## Immunohistochemistry
*In the **Immunohistochemistry** section:*
Immunohistochemical analysis of protein expression in tumor tissues and cell lines was performed with the use of the Vectastatin ABC kit (Vector Laboratories, Burlingame, CA) according to manufacturer's instructions. This procedure employs biotinylated secondary antibodies and a preformed avidin: biotinylated enzyme complex that has been termed the ABC technique. As the primary antibody, we used anti-survivin antibody (Santa Cruz Biotech) diluted 1:100 in 2% normal goat blocking serum.
## TUNEL staining
*In the **TUNEL staining** section:*
Apoptosis was measured quantitatively with the use of the terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end-labeling (TUNEL) assay [95]. All components for this procedure were from the ApopTag In Situ Apoptosis Detection kit (Chemicon, Temecula, CA), which was used according to the manufacturer's instructions.[](https://www.ncbi.nlm.nih.gov/mesh/C027078)
## MTT assay
*In the **MTT assay** section:*
MTT assays were performed in 96-well plates as described in detail elsewhere [31] with the use of 3.0–8.0 × 103 cells per well.[](https://www.ncbi.nlm.nih.gov/mesh/C022616)
## Plasmids and stable transfections
*In the **Plasmids and stable transfections** section:*
The human LN229 glioblastoma cell line was stably co-transfected with individual luciferase reporter plasmids and the pSV2neo plasmid. The latter expresses the bacterial aminoglycoside-3'-phosphotransferase (neo) gene [96], which enables selection of transfected cells in medium containing the aminoglycoside G418 sulfate. Stable transfections were performed with the use of Lipofectamine 2000 (Invitrogen, Carlsbad, CA), and mass cultures of transfected cells were selected in G418 according to standard protocols [97].[](https://www.ncbi.nlm.nih.gov/mesh/D000617)
The following luciferase reporter plasmids were used. Cyclin B-luc harbors 555 base pairs (bp) of upstream cyclin B promoter sequences [98] and was kindly provided by William R. Taylor, Cleveland Clinic Foundation (Cleveland, OH). CMV-luc is under the control of 880 bp encompassing the promoter of cytomegalovirus (CMV) [20]. The survivin reporter plasmids -6270Surv-luc and -230Surv-luc harbor 6270 bp and 230 bp, respectively, of the upstream promoter region of the survivin gene [99] and were kindly provided by the laboratory of Dario Altieri, Yale University (New Haven, CT).
## Tumor growth in nude mice
*In the **Tumor growth in nude mice** section:*
All animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Southern California, and all applicable policies were strictly observed during the course of this study. Four- to six-week-old male athymic nu/nu mice were obtained from Harlan (Indianapolis, IN) and kept in a pathogen-free environment. To support more consistent tumor take and uniform growth [100, 101], the animals were whole-body irradiated with 300 cGy of ionizing radiation (Cesium 137) four days prior to xenotransplantation by using a low dose-rate laboratory irradiator (Gammacell 40; Atomic Energy of Canada Limited, Canada).[](https://www.ncbi.nlm.nih.gov/mesh/C000614989)
For tumor inoculation, 5 × 105 U87 glioblastoma cells were injected subcutaneously into the right flank. Once palpable tumors had developed, the animals were randomly divided into three groups: (i) treatment with celecoxib (1,000 ppm in animal chow), (ii) treatment with DMC (1,000 ppm in animal chow), and (iii) no drug treatment (regular chow without drug added). The tumor size in all animals was measured every three to four days. Tumor size was calculated by the following formula: Volume (mm3) = L û W û H û 0.5 (L: length, W: width, H: height). Student t-test was used for statistical analysis, and a P-value of <0.05 was considered significant.[](https://www.ncbi.nlm.nih.gov/mesh/D000068579)
## Authors' contributions
*In the **Authors' contributions** section:*
PP performed experiments and assembled the manuscript. NS, AK, and Y-TL performed experiments. JU and NAP were responsible for synthesizing the various drugs. C-SC supplied additional samples of drugs and provided guidance for the project. FMH and TCC participated in the design and execution of the project.
AHS conceived of the study and participated in its design, execution, and coordination. All authors read and approved of the final manuscript.
|
# Introduction
FISH mapping of Philadelphia negative BCR/ABL1 positive CML
# Abstract
*In the **Abstract** section:*
Background
Chronic myeloid leukaemia (CML) is a haematopoietic stem cell disorder, almost always characterized by the presence of the Philadelphia chromosome (Ph), usually due to t(9;22)(q34;q11) or its variants. The Ph results in the formation of the BCR/ABL1 fusion gene, which is a constitutively activated tyrosine kinase. Around 1% of CML patients appear to have a Ph negative karyotype but carry a cryptic BCR/ABL1 fusion that can be located by fluorescence in situ hybridisation (FISH) at chromosome 22q11, 9q34 or a third chromosome. Here we present FISH mapping data of BCR and ABL1 flanking regions and associated chromosomal rearrangements in 9 Ph negative BCR/ABL1 positive CML patients plus the cell line CML-T1.
Results
BCR/ABL1 was located at 9q34 in 3 patients, 22q11 in 5 patients and CML-T1 and 22p11 in 1 patient. In 3 of 6 cases with the fusion at 22q11 a distal breakpoint cluster was found within a 280 Kb region containing the RAPGEF1 gene, while in another patient and the CML-T1 the distal breakpoint fell within a single BAC clone containing the 3' RXRA gene. Two cases had a duplication of the masked Ph while genomic deletions of the flanking regions were identified in 3 cases. Even more complex rearrangements were found in 3 further cases.
Conclusion
BCR/ABL1 formation resulted from a direct insertion (one step mechanism) in 6 patients and CML-T1, while in 3 patients the fusion gene originated from a sequence of rearrangements (multiple steps). The presence of different rearrangements of both 9q34 and 22q11 regions highlights the genetic heterogeneity of this subgroup of CML. Future studies should be performed to confirm the presence of true breakpoint hot spots and assess their implications in Ph negative BCR/ABL1 positive CML.
## Background
*In the **Background** section:*
Chronic myeloid leukaemia (CML) is a pluripotent haematopoietic stem cell disorder defined by expression of the BCR/ABL1 fusion gene, a constitutively activated tyrosine kinase, harbored by the Philadelphia chromosome (Ph), which is a result of a t(9;22)(q34;q11) or a related variant translocation [1]. In ~1% of the CML patients the bone marrow cells appear to be Ph negative by G-banding, although the BCR/ABL1 fusion gene can be identified by molecular means and located by fluorescence in situ hybridisation (FISH) on chromosome 22q11, 9q34 or even a third chromosome. The biology and clinical significance of the genetic rearrangements in Ph negative BCR/ABL1 positive disease have been widely discussed following the first descriptions [2-5]. Two mechanisms for the formation of the chimeric gene in masked Ph positive cells have been postulated: either by insertion of ABL1 into the BCR region (or vice versa) or by a multiple step model where a classical t(9;22) is followed by a translocation of both products and/or another autosome, thereby restoring the normal chromosome morphology. In both instances, more than the 2 breaks associated with classical t(9;22) are implicated.
Although as early as 1990 Morris et al. [6] provided evidence that the insertion involves additional sequences distal to the 3' ABL1 site, the extent of the genomic rearrangements in this form of CML are unknown.
In view of the few studies published with a precise map of the insertions [7-10], we aimed to construct an accurate map of the insertions in the cell line CML-T1 [11] and 9 patients with Ph negative BCR/ABL1 positive CML using a range of FISH probes obtained from BAC clones. The fusion gene was identified at 9q34 (3 patients), 22q11 (5 patients and CML-T1) and 22p11 (3 patients), resulting in relocation of sequences well in excess of either 3' ABL or 5' BCR by means of a direct insertion (6 patients and CML-T1) or a sequence of events (3 patients). Recurrent distal breakpoints were found at the regions of RAPGEF1 and RXRA genes.
## Methods
*In the **Methods** section:*
Nine archival bone marrow chromosome preparation samples of CML patients (7 females and 2 males) with Ph negative BCR/ABL1 positive disease collected from Hammersmith and Royal Free Hospitals (London, UK), together with the cell line CML-T1, were investigated (Table 1). All samples tested positive for BCR/ABL1 fusion by PCR. Investigations were carried out on bone marrow samples obtained at presentation. G-banding and molecular cytogenetic analysis, including chromosome painting and FISH mapping with locus specific probes, were performed following protocols in routine use [12]. A minimum of 25 metaphase and over 100 interphase cells after short term in vitro culturing were analysed and results described following ISCN (2005).
Five of the samples in this cohort (cases 1–3, 7 & 8) were part of another study [13].
Characteristics of the samples.
In all samples, FISH with the commercially available LSI BCR/ABL1 Dual Color, Dual Fusion Translocation Probe ("D-FISH", Vysis, Downers Grove, IL, USA) was initially performed using manufacturer's protocol to identify the chromosome location of the BCR/ABL1 fusion gene. FISH mapping was carried out with Bacterial Artificial Chromosomes (BAC) clones obtained from the BACPAC Resources Center (Children's Hospital Oakland Research Institute, Oakland, CA, USA), the Sanger Centre (Cambridge, UK) and Invitrogen (Paisley, UK) (Table 2). Clones were grown in Luria-Bertani medium with Chloramphenicol (20 μg/ml), BAC DNA extracted with a QIAGEN Large-Construct Kit (Qiagen, West Sussex, UK) and directly labelled with either Spectrum Orange or Spectrum Green dUTPs with a Nick Translation Kit (Vysis). BAC DNA from chromosomes 9 and 22 belonging to the Human 32 K Clone Set from the BACPAC Resources Center was amplified with a GenomePlex Single Cell Whole Genome Amplification Kit (Sigma-Aldrich, Dorset, UK) and labelled as described before.[](https://www.ncbi.nlm.nih.gov/mesh/D002701)
Summary of the BAC clones used for FISH analysis, their genomic address and source.
BAC RP11-164N13 was used to target BCR gene. Since this BAC covers both major and minor BCR breakpoints, it is found split when BCR gene is rearranged. RP11-61N10 covers BCR non-coding sequences centromeric to the breakpoint and it was used to identify 5' BCR region. RP11-83J21, which covers the 3' end of ABL1 incorporating the whole of the coding region, was used to identify the sequences telomeric of the ABL1 breakpoint, whereas RP11-57C19 was used to identify the ABL1 sequences centromeric of the breakpoint.
The BAC clones and genes were located according to the UCSC database, hg17 (University of California Santa Cruz, CA, USA) [14]. Mapping data for the 32 K human clone set was obtained from the BACPAC Resources Center website [15] and used to assess the size of the sequences found to be rearranged. In addition, sub-telomeric probes directly labelled from the regions of 9q, 22q and 16q were used (Stretton Scientific Ltd, Stretton, UK). All tests were carried out as dual colour, dual probe FISH.
Digital imaging and karyotyping were carried out using a SmartCapture and SmartType FISH workstation (Digital Scientific Ltd, Cambridge, UK).
Array CGH analysis (aCGH) was carried out on the cell line CML-T1. The aCGH was performed using two platforms – 1 Mb BAC clone chip (SGI2600) [16] and oligo-nucleotide (105 K Agilent) [17] following manifacturer's protocol, while data processing and presentations were carried out using 'in house' software as reported [18,19].[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
## Results
*In the **Results** section:*
A summary of the molecular cytogenetic investigations carried out on BCR/ABL1 positive samples from 9 CML patients with normal bone marrow (BM) karyotype as well as the cell line CML-T1 with masked Ph chromosome is presented Figure 1.
Summary of the FISH mapping carried out on Ph negative BCR/ABL1 positive bone marrow cells from 9 CML patients and the cell line CML-T1. Header row shows the location of BCR/ABL1 fusion: 22q11.2 (cases no 1–6), 22p (case 7) and 9q34.1 (cases 8–10). (a) Map of the 9q34.1-qter region with coloured squares (key on the right) indicating the location or deletion of the BAC clones used for FISH analysis (approximate genomic distances to the breakpoint and names of the BAC clones on the left). A thick black horizontal line presents the ABL1 breakpoint, which is encompassed by the clones RP11-57C19 and RP11-83J21. When a breakpoint falls within a BAC, the probe gives a split signal in two different locations. (b) Map of the 22q11.2 region with coloured squares (key below) indicating the location or deletion of the BAC clones tested (genomic distances and names of the clones on the left). A thick black horizontal line presents the BCR breakpoint, which falls within the BAC clone RP11-164N13.
## FISH analysis using commercial BCR/ABL1 D-FISH probe (Vysis)
*In the **FISH analysis using commercial BCR/ABL1 D-FISH probe (Vysis)** section:*
The BCR/ABL1 D-FISH probe (Vysis) contains sequences covering both genes labelled in different colours, so that rearrangements affecting them can be visualised at chromosome level. The ABL1 probe (red) targets a 650 Kb region of 9q34.1 including the whole of the ABL1 gene (173.8 Kb), thus spanning the common breakpoint, and extends 5' of ABL1 to incorporate the ASS gene. The BCR probe (green) is represented by two 600 Kb regions of 22q11.2 separated by a 300 Kb gap, one of the regions covering the entire BCR gene and extending 5' of it in order to include the IGLV gene, and the other starting 300 Kb telomeric of BCR and ending at 900 Kb 3' of the gene. The application of this probe revealed the BCR/ABL1 fusion at three different chromosome sites: at 22q11.2 (samples no 1–6), at 22p11 (sample no 7) and at 9q34 (samples no 8–10). There was evidence for the formation of the reciprocal ABL1/BCR fusion in just one patient (no 4), which showed 2 fusion signals at der(22) and der(9).
In the remaining cases only 1 fusion signal was observed, irrespective of the fusion gene location.
In addition, D-FISH also revealed the presence of different clones with further rearrangements in CML-T1 and patients no 1 and 3. CML-T1 was found to harbour 3 clones: i) 7 out of 20 metaphases (35%) showed a 2 red 2 fusion signal pattern typical of diploid cells with duplication of the masked Ph and no normal 22 homologue (loss of the green signal); ii) 7 out of 20 metaphases (35%) showed a 4 red 4 fusion pattern for tetraploid cells with duplication of the masked Ph and no normal 22; iii) 6 out of 20 metaphases (30%) showed a 2 red 4 fusion pattern for tetraploid cells with duplication of the masked Ph, no normal 22 and deletion of 5' ABL1 (loss of two red signal). Patient no 3 had also developed a sub-clone with a duplication of the masked Ph and loss of the normal 22 homologue (2 red 2 fusion D-FISH pattern), as seen in 4 out of 20 metaphases (20%). On the other hand, D-FISH in patient no 1 showed a 1 red 1 green 1 fusion signal pattern, uncovering a deletion of sequences centromeric to ABL1 at the der(9)ins(22;9) in all cells screened (loss of 1 red signal).
## Patients with BCR/ABL1 fusion residing on chromosome 22q11.2
*In the **Patients with BCR/ABL1 fusion residing on chromosome 22q11.2** section:*
The BAC clone RP11-83J21 covering 3' ABL1 moved to 22q11 in 6 cases (no 1–6), whereas RP11-57C19 remained at der(9) in all but one (no 1, Figure 1). In patient no 1, we identified a cryptic genome loss at der(9) of at least 1.34 Mb in all cells. The sequences deleted, centromeric of ABL1 breakpoint, were covered by the BAC clones RP11-57C19 to RP11-17O4. However, it was not possible to assess the full extend of the deletion due to lack of material.
In all 6 cases the inserted 9q34 sequences exceeded the 3' boundaries of the ABL1 gene and estimated to be 720 Kb long in 1 case (no 1), 1 Mb in 2 cases (no 2–3), between 1.6 Mb and 2.2 Mb in 1 case (no 6) and 3.9 Mb in 2 cases (no 4–5). The estimated sizes of the insertions were calculated based on the information of the BAC clones chromosome location available at the UCSC database (genome build 35). In cases no 1–5 the insertions were found to stretch distally falling into two sub-groups: a "small" insertion (720 Kb–1 Mb) and a "large" insertion (3.9 Mb) (Figure 2). The distal boundary of the "small" insertion seen in 3 patients (no 1–3) fell within a region covered by 3 overlapping BAC clones (CTD-2107G12, RP11-40A7 and RP11-323H21), thus forming a breakpoint cluster (Figure 2) which was estimated to be 280 Kb long and found to house several genes: POMT1, UCK1 and RAPGEF1. In further two samples (patient no 4 and CML-T1) the inserted material was found to be larger (estimated size of 3.9 Mb) and the distal breakpoint fell within a single BAC clone (RP11-92B21) containing the 3' end of the RXRA gene (Figure 2).
BCR/ABL1 fusion at 22q11.2: "small" and "large" insertions with recurrent distal breakpoints. (a) Diagram showing the small size ins(22;9)(q11;q34.1q34.1) seen in 3 patients (no 1–3). The BAC clones covering the distal breakpoint region are presented with green lines. The ABL1 breakpoint marks the proximal boundary of the insertion (< 1 Mb) while the distal breakpoint (shown by red arrows) falls within a 280 Kb breakpoint cluster housing the UCK1, POMT1 and RAPGEF1 genes. (a1) A representative metaphase cell in patient no 3 with co-hybridization of FISH probes RP11-323H21 and RP11-413M3, showing a split signal from RP11-323H21 (green signals at both chromosome 9 homologues and masked Ph) and duplication of the masked Ph (green signals on 2 masked Ph). (a2) BCR/ABL1 D-FISH (Vysis) in patient no 2, showing the absence of green signal at der(9). (b) Diagram showing the large size ins(22;9)(q11.2;q34.1q34.2) seen in patient no 4 and CML-T1. The ABL1 breakpoint marks the proximal boundary of the 3.9 Mb insertion, while the distal breakpoint lies within the clone RP11-92B21 (red arrow). (b1) A representative metaphase cell in patient no 4 with co-hybridization of FISH probes RP11-92B21 and RP11-413M3, showing a split signal from RP11-92B21 (green signal at both chromosome 9 homologues and masked Ph). (b2) BCR/ABL1 D-FISH (Vysis) in CML-T1, showing absence of green signal at der(9) and duplication of the masked Ph (two fusion signals).
High-resolution aCGH analysis of the cell line CML-T1 identified a gain at 9q34.1 starting at ABL1 breakpoint at 130.6 and covering 3.9Mb in 3' direction until the distal part of the RXRA gene at 134.5, confirming the FISH mapping data (Figure 3). This gain resulted from the duplication of the der(22)ins(22;9) seen in all CML-T1 cells, which was always accompanied by loss of the normal 22 homologue. Furthermore, use of D-FISH with CML-T1 had previously revealed a deletion of 5' ABL1 in a tetraploid sub-clone. We mapped the length of the deletion, which was found to be 8.7 Mb long (from RP11-138E2 to 9q telomere) and affect not the der(9)ins(22;9) but the "normal" 9. This was demonstrated by FISH when co-hybridizing a BAC clone found within the insert (in green) and a BAC from chromosome 9 centromeric or telomeric to the inserted sequences (in red).
Tetraploid cells with a deletion displayed 2 red signals from two chromosome 9 and 4 green signals from the 4 masked Ph chromosomes. Therefore not only two red signals were deleted, but also the 2 green signals from the "normal" 9 were missing, demonstrating that the loss had occurred at the "normal" homologues.
Gains and deletions in the Ph negative BCR/ABL1 positive cell line CML-T1. (a) Array CGH reveals gains of sequences downstream of the ABL1 breakpoint. The genome profile of the 9q34.1-qter region is shown at the top and the 22q11.2-2 region is presented at the bottom, aligned at the ABL1 and BCR breakpoints (vertical dashed line). Results of the SGI2600 BAC chip are shown in red and 44 K Agilent oligonucleotide in blue. Both BAC and oligo array detect a gain of the 9q34 sequences proximally flanked by the ABL1 breakpoint and distally by the RXRA gene. (b) A representative tetraploid metaphase cell with co-hybridization of FISH probes RP11-138E2 and RP11-17O4, showing the proximal breakpoint of the 9q34 deletion arisen in this sub-clone (the arrows show the two missing green signals from RP11-138E2). (c) A representative tetraploid metaphase cell with co-hybridization of RP11-323H21 and a 9q sub-telomeric probe (Stretton), showing the duplication of the masked Ph and that the genomic loss affects not the der(9) but the "normal" homologue (the 2 green and 2 red signals from the two normal 9 are missing). The top box on the right shows 4 green, 2 red signals as seen in interphase tetraploid cells with deletion, while the bottom box on the right shows 6 green, 4 red signals as seen in the interphase tetraploid cells without deletion.[](https://www.ncbi.nlm.nih.gov/mesh/D009841)
Regarding BCR flanking regions, all probes tested from chromosome 22 remained at their original locations in cases no 1, 2, 3 and 5. However, in case no 4, sequences distal to the BCR breakpoint and housed by the BAC clones RP11-164N13, RP11-529P21, RP11-223O10 and RP11-698L6 were found embedded within the der(9) chromosome at the 9q34 region (Figure 4). Both RP11-164N13 and RP11-529P21 gave a split signal pattern being found at the masked Ph and derivative 9, apart from the normal 22. The distal breakpoint of the 22 sequences found at der(9) fell between RP11-698L6 and RP11-594L10.
BCR/ABL1 at 22q11.2 in patient no 4 results from a multiple step mechanism. (a) BCR/ABL1 D-FISH probe (Vysis) showing 1 red, 1 green, and 2 fusion signals. The presence of both BCR/ABL1 and ABL1/BCR fusion genes is an evidence of an initial t(9;22)(q34;q11.2). (b) A representative metaphase cell with co-hybridization of FISH probes RP11-61N10 and RP11-164N13, showing a split signal from RP11-164N13. Thus, the proximal boundary of the 22q11.2 sequences identified within the structure of the der(9) chromosome coincides with the BCR breakpoint. (c) A representative metaphase cell with co-hybridization of FISH probes RP11-698L6 and RP11-446O3. RP11-698L6 is identified at der(9) while RP11-446O3 is found at der(22). The distal breakpoint of the 22q11.2 fragment lies between these two BAC clones. (d) Schematic representation of the multiple step rearrangement with chromosomes 9 in red and 22 in green. Red arrowheads show the breakpoints. The presence of both 9q34 sequences inserted on der(22) and 22q11.2 sequences inserted on der(9) (black arrows) can be explained by two consecutive translocations: an initial t(9;22)(q34;q11.2) followed by a second reciprocal translocation between the two products, with breakpoints distal to both BCR/ABL1 and ABL1/BCR fusion genes.
Even more complex rearrangements involving a third chromosome were revealed in another patient (no 6) (Figure 5). FISH painting and FISH mapping identified a three way cryptic translocation t(9;22;16)(q34;q11;q?13) and found the presence of 9q34.1 sequences sandwiched in the der(22)(22pter-q22.1::9q34.1::16q?13-qter). The distal breakpoint of the 9q34.1 fragment fell between RP11-81P5 and RP11-326L24 and the insertion was therefore estimated to be 1.6 Mb–2.2 Mb long. All sequences from chromosome 22 distal to the BCR breakpoint were found at chromosome 16, while sequences proximal to the breakpoint remained at 22q11.
BCR/ABL1 resides at 22q11.2 as a result of a cryptic three-way rearrangement between chromosomes 9, 22 and 16 in patient no 6. (a) BCR/ABL1 D-FISH probe (Vysis) showing 1 fusion signal at der(22), 2 red signals at 9 and der(9), 1 green signal at 22 and 1 unexpected green signal at der(16). (b) Chromosome painting confirming the presence of a t(9;22;16)(q34;q11;q?13) in an apparently normal G banding karyotype. (c) FISH with co-hybridization of RP11-326L24 and RP11-643E14 locating RP11-643E14 at the masked Ph while RP11-326L24 is retained at the der(9). (d) A representative metaphase cell with co-hybridization of FISH probes RP11-81P5 and RP11-153P4 identifying RP11-81P5 at the masked Ph. The distal breakpoint of the 9q34 insert is therefore flanked by the BAC clones RP11-81P5 and RP11-326L24. (e) A representative metaphase cell with co-hybridization of FISH probes RP11-92B21 and RP11-223O10 showing RP11-223O10 at normal 22 and der(16), thus confirming the relocation of sequences distal to BCR breakpoint at chromosome 16. (f) Schematic representation of the three-way rearrangement with chromosomes 9 in red, 16 in yellow and 22 in green; note that the der(22) contains material from all three parties. The presence of 9q34.1 sequences embedded within the masked Ph suggests that the t(9;22;16) could be a result of a two stage event: firstly ins(22;9) followed by translocation between der(22)ins(9;22) and 16q.
## Patient with the BCR/ABL1 fusion residing at 22p11
*In the **Patient with the BCR/ABL1 fusion residing at 22p11** section:*
The BCR/ABL1 fusion was unexpectedly found at 22p11 in 1 patient (no 7) (Figure 6). FISH with RP11-164N13 showed one signal at the normal chromosome 22 and a split signal at the 22p11 and 22q11 regions from the other homologue. RP11-61N10 was found at the normal 22 and only at the 22p11 region of der(22), thus confirming the location of BCR/ABL1 fusion at the p arm of the derivative 22. RP11-529P21, RP11-223O10 and 22qter remained on 22q11, while all probes tested from RP11-83J21 to 9qter were found at 22p11.
BCR/ABL1 fusion resides at 22p11 in patient no 7. (a) BCR/ABL1 D-FISH probe (Vysis) showing a split green signal from BCR within the masked Ph. There is only 1 fusion signal located at der(22), another green signal at the same (der22), one green signal at normal 22 and 2 red signals at both chromosomes 9. (b) A representative metaphase cell with co-hybridization of FISH probes RP11-424E7 and RP11-61N10 identifying both probes at the p arm of the masked Ph. (c) A representative metaphase cell with co-hybridization of a 9q sub-telomeric probe (Stretton) and RP11-83J21, showing the presence of the two probes at 22p11 and indicating that all sequences distal to ABL1 breakpoint had moved to 22p11. (d) A representative metaphase cell with co-hybridization of a 22q sub-telomeric probe (Stretton) and RP11-83J21, showing that the 22q telomere is retained at its original location. (e) Schematic representation of the events that may have lead to the formation of BCR/ABL1 fusion gene and its unusual location at 22p11, with chromosome 9 in red and 22 in green. Sequences distal to BCR breakpoint are found at their original location while 5' BCR and 9q34 sequences distal to ABL1 breakpoint (including the telomere) are relocated at 22p11. This could be explained by an initial t(9;22)(q34;q11) followed by a three-way translocation between 9q34, 22q11 and 22p11, which would require 5 breaks (red arrowheads). (f) The unusual location of BCR/ABL1 at 22p11 could also be a result of an initial ins(9;22)(q34;q11q11) followed by a translocation between der(9) and the p arm of der(22). This sequence of events would need also the same amount of breaks (red arrowheads) and therefore cannot be rule out.
## Patients with the BCR/ABL1 fusion residing at 9q34.1
*In the **Patients with the BCR/ABL1 fusion residing at 9q34.1** section:*
The BCR/ABL1 fusion was located by FISH at band 9q34.1 in 3 patients (no 8–10) and thought to result from a direct insertion of 22q11.2 sequences. As expected, the telomeric breakpoint was found in all cases within the BAC clone RP11-164N13, which was always seen at both der(9) and der(22) (Figure 7). RP11-61N10 was always found at 9q34, however, due to lack of material it was not possible to assess the proximal breakpoint of the 22q11 insertion. All BACs tested from 9q34 remained at their respective location on chromosome 9 in all 3 patients.
BCR/ABL1 fusion resides at 9q34.1 in 3 patients. (a) A representative metaphase cell hybridized with a BCR/ABL1 D-FISH probe (Vysis) showing the presence of 1 fusion signal at der(9), 1 red signal at 9 and 2 green signals at both chromosomes 22 as seen in patients no 9 and 10. (b) A representative metaphase cell with co-hybridization of FISH probes RP11-61N10 and RP11-164N13 showing 2 fusion signals at 22 and der(9), plus 1 green signal at der(22), as seen in patients no 9 and 10. RP11-164N13 is therefore split, with the 5' BCR sequences relocated at der(9). (c) BCR/ABL1 D-FISH probe (Vysis) in patient no 8 showing 1 fusion signal at der(9), 1 red signal at 9 and 1 green signal at 22. The green signal from der(22) is deleted. (d) A representative metaphase cell with co-hybridization of FISH probes RP11-61N10 and RP11-594L10 giving a fusion signal at normal 22 and one red signal from RP11-61N10 at der(9), confirming that the BCR sequences moved to der(9) are centromeric of the breakpoint. There is again one green signal missing, because the deletion at der(22) includes not only 3' BCR but also at least 1.77 Mb distal to the breakpoint.
In patient no 8, the insertion was accompanied in all cells by loss of a region at least 1.77 Mb long from 22q11.2 and immediately distal to the BCR breakpoint, covered from the 3' end of RP11-164N13 to at least RP11-765G14 (Figure 1).
## Discussion
*In the **Discussion** section:*
Since the first description of a Ph negative BCR/ABL1 positive CML patient [2], several studies have been published reporting similar cases. Most of them are focused on the presence and location of the BCR/ABL1 fusion in CML patients with masked Ph chromosome and commonly achieved by application of commercial FISH probes, which have been proved to be very useful to identify the presence and location of the BCR/ABL1 fusion gene in CML patients with no distinguishable Ph chromosome. These studies have established the importance of FISH tests for the diagnosis and therapy monitoring of Ph negative BCR/ABL1 positive CML. However, the commercial probes don't provide enough information to understand the mechanisms involved in the formation of the masked Ph chromosome.
We used FISH mapping with BAC probes in order to study the formation of the BCR/ABL1 fusion and the underlying genomic rearrangements in 9 CML patients with Ph negative BCR/ABL1 positive CML and the cell line CML-T1. The formation of the fusion gene resulted from the relocation of not only the 3' ABL1 sequences within the BCR region at chromosome 22q11.2 or 5' BCR sequences within ABL1 region at 9q34.1, but also a considerable amount of flanking material, leaving the chromosome morphology apparently intact.
The fusion gene was located at 9q34.1 in 3 patients, at 22q11.2 in 5 patients and CML-T1, and at 22p11 in another patient. 5 out of the 6 cases with a 9q34 insertion at 22q11 displayed recurrent distal breakpoints that fell within two gene bearer regions. Thus, in 3 patients a common breakpoint cluster of 280 Kb was found. According to UCSC database, this region houses 3 genes: POMT1 (protein-O-mannosyltransferase1), UCK1 (uridine-cytidine kinase 1) and RAPGEF1 (guanine nucleotide releasing factor for RAP1; also known as C3G). In another patient and CML-T1 the breakpoint fell within a single BAC clone encompassing the 3' end of RXRA gene (retinoid × receptor alpha).
Although further investigations were not carried out in this study, both RAPGEF1 and RXRA belong to pathways the disruption of which may be relevant to the evolution of the patients. RAPGEF1 has been shown to have transformation suppressor activity towards several oncogenes [20,21] and also interact with BCR/ABL1 [22]. RXRA belongs to a family of nuclear receptors that target multiple signalling pathways [23] and its downregulation has been showed to be critical for neutrophil granulocytes differentiation [24].
Other studies found RXRA to contribute in acute promyelocytic leukaemia transformation [25,26]. Although we have no direct evidence for the immediate involvement of either RAPGEF1 or RXRA genes, their potential role merits further investigation.
Two other mapping studies that found similar insertions have been published [7,8]. In the first study, the authors used FISH mapping to identify the rearrangements involved in 2 Ph negative CML patients with variant translocations. A 3 Mb insertion from 22q11 into ABL1 was identified in 1 patient, while the other had a 9q34 insertion at the BCR region with a distal breakpoint falling within the clone RP11-353C22 (genome address: 31,278,002–131,588,248). This result matches with our findings since the latter BAC is located within the same 280 Kb common breakpoint region identified in 3 patients of our study. On the other hand, Valle et al.
[7] found a 5.6 Mb insertion of 9q34 sequences into BCR. This insertion is larger than the ones identified in our study and was not accompanied by any deletions or other rearrangements.
Deletions of 5' ABL1 and/or 3' BCR sequences at the der(9) chromosome in patients with classical and variant Ph translocations [27] have been shown to have adverse prognostic value in CML patients treated with interferon [28], although their impact in patients being treated with tyrosine kinase inhibitors is controversial [29-31]. Dual colour, dual fusion translocation FISH probes spanning the BCR and ABL1 genes are very useful for revealing these events but they have a limited value in interphase nuclei in patients with masked Ph, since often the merging of the 5' BCR and 3' ABL1 signals by simple insertion leads to an apparent loss of one fusion signal that can be falsely assessed as deletion [13]. Thus, D-FISH (Vysis) in a patient with a direct ins(22;9) gives a 2 red, 1 green, 1 fusion signal pattern, which is the same pattern obtained in case of a typical t(9;22) with deletion of 5' ABL1 at der(9). If the patient has a direct ins(9;22) the D-FISH signal pattern is 1 red, 2 green, 1 fusion, which could be mistaken for a typical t(9;22) with deletion of 3' BCR at der(9).
Furthermore, Ph negative BCR/ABL1 positive patients with either a cryptic deletion of 5' ABL1 or 3' BCR show a 1 red, 1 green, 1 fusion D-FISH pattern, also typical for a t(9;22) with deletion of the reciprocal ABL1/BCR fusion. Therefore, when a deletion signal pattern is detected by interphase FISH with a dual colour, dual fusion probe, it is essential to look also at the metaphases in order to be able to differentiate a classical t(9;22) with deletion from a simple insertion.
Batista et al [32] reported a Ph negative BCR/ABL1 positive patient with an insertion of ABL1 into BCR associated with a deletion of the ASS – 5' ABL1 region. Zagaria et al [8] also reported two cases of CML patients with masked Ph and associated deletions: one patient with a cryptic ins(9;22) and a deletion of 5' ABL1 and 3' BCR regions; the other patient with a multi-step variant translocation and a deletion of around 400 Kb telomeric of the ABL1 gene. In addition to them, De Melo et al [13] identified with commercial triple-colour FISH probes 5' ABL1 deletions in 2 patients and 3' BCR deletion in 1 patient out of 14 CML patients with masked Ph. Our study confirmed and sized such deletions in 2 patients which where also part of De Melo's cohort.
CML-T1 also had a 8.7 Mb deletion of 9q34 material in one of the sub-clones, but in this case the loss was found to affect the homologue not involved in the BCR/ABL1 formation.
We identified a duplication of the chromosome 22 harboring the BCR/ABL1 fusion accompanied by loss of the normal homologue in 1 out of 9 patients in this study plus the cell line CML-T1. Such duplications of the BCR/ABL1 bearing chromosome (either chromosome 22 or 9) seem to be a relatively common event in Ph negative BCR/ABL1 positive CML patients, being accompanied by loss of the normal homologue in most of the cases and seen both in chronic phase and blast crisis [33-35].
Regarding the formation of the fusion gene in Ph negative BCR/ABL1 positive CML patients, Morris et al [6] suggested two possible mechanisms: a one step model, where BCR/ABL1 results from a simple insertion of either 3' ABL1 into BCR or 5' BCR into ABL after three genomic breaks; and a multiple step model, with an initial classical t(9;22)(q34;q11) followed by a second translocation of both products and/or a third chromosome, requiring a minimum of 4 genomic breaks.
Our study provided evidence for the existence of both mechanisms implicated in the formation of the fusion gene in Ph negative patients. We found a simple insertion (one step event) in 6 out of 9 patients (no 1–3, 8–10) and CML-T1 (no 5), with no evidence of secondary rearrangements within the regions flanking BCR and ABL1 breakpoints apart from the accompanying deletions seen in 2 patients and CML-T1. On the other hand, traces of sequential rearrangements indicating a multiple step mechanism were found in 3 patients. Patient no 4 had a 9q34 insertion at chromosome 22 with bits from 22q11 distal to the breakpoint embedded within the der(9), suggesting an initial t(9;22)(q34;q11) followed by further translocation of both products. Patient no 7 carried the BCR/ABL1 fusion at 22p11.
Sequences downstream of the breakpoint remained at their original location, while 5' BCR and all 9q34 sequences distal to ABL1 breakpoint (including the telomere) were relocated at 22p11, which could be explained by an initial t(9;22)(q34;q11) followed by a three-way translocation between 9q34, 22q11 and 22p11. This sequence of events would have required 5 breaks (2 for the translocation and 3 for the second one). However, an initial ins(9;22) followed by a reciprocal translocation between 9q34 and 22p11 would also require 5 breaks and therefore cannot be ruled out. Finally, patient no 6 had a three way cryptic t(9;22;16)(q34;q11;q?13) with 9q34 sequences embedded within the der(22), suggesting an initial ins(22;9) followed by a translocation between chromosomes 16 and der(22)ins(22;9). These data show not only that the two mechanisms do happen, but also that they are not excluding options.
An example of the latter is patient no 6, where an initial direct ins(22;9) would be part of the spectrum of rearrangements that had restored the normal morphology of the der(22).
## Conclusion
*In the **Conclusion** section:*
In summary, we found that the BCR/ABL1 fusion resulted from relocation of not only the 3' ABL1 sequences within BCR at 22q11.2 or 5' BCR sequences within ABL but also a considerable amount of flanking material, with distal recurrent breakpoints of the excised 3' ABL1 sequences at RAPGEF1 and RXRA regions. BCR/ABL1 resulted from a direct insertion (one step mechanism) in 6 patients and CML-T1, while in 3 patients the fusion gene was a result of a sequence of events (multiple steps). Finally, the presence of different rearrangements of both 9q34 and 22q11 regions demonstrates the genetic heterogeneity of this subgroup of CML. Future studies should be performed to confirm the presence of true breakpoint hot spots and assess their implications in Ph negative BCR/ABL1 positive CML.
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interests.
## Authors' contributions
*In the **Authors' contributions** section:*
AV carried out the FISH mapping and wrote the manuscript. DB, AC and CG performed the array CGH and qPCR analysis. AR, JH, MV and VDM performed G-banding and initial FISH analysis. DM and JFA provided clinical samples. EN designed the study, supervised its execution and co-participated in the writing of the manuscript.
All authors read and approved the final manuscript.
|
# Introduction
Involvement of ER stress in retinal cell death
# Abstract
*In the **Abstract** section:*
Purpose
To clarify whether endoplasmic reticulum (ER) stress is involved in retinal cell death, using cultured retinal ganglion cells (RGC-5, a rat ganglion cell line transformed with E1A virus), and transgenic mice ER stress-activated indicator (ERAI) mice carrying a human XBP1 and venus a variant of green fluorescent protein (GFP) fusion gene.
Methods
RGC-5 damage was induced by tunicamycin, and cell viability was measured by double nuclear staining (Hoechst 33342 and either YO-PRO-1 or propidium iodide). The expressions of glucose-regulated protein 78(GRP78)/BiP, the phosphorylated form of eukaryotic initiation factor 2α (p-eIF2α), and C/EBP-homologous (CHOP) protein after tunicamycin (in vitro or in vivo) or N-methyl-D-aspartate (NMDA; in vivo) treatment were measured usin[g immunoblo](https://www.ncbi.nlm.nih.gov/mesh/D014415)t or immunostaining. ERAI mice carrying the F-XBP1-DBD-venus e[xpression gen](https://www.ncbi.nlm.nih.gov/mesh/C017807)e were used [to monit](https://www.ncbi.nlm.nih.gov/mesh/C089813)or E[R-stress in vivo](https://www.ncbi.nlm.nih.gov/mesh/D011419). Twenty-four hours after intravitreal injection of tunicamycin or NMDA, or after raising intraocular pressure (IOP), the retinal fluorescence intensity was visualized in anes[thetized an](https://www.ncbi.nlm.nih.gov/mesh/D014415)imals using an ophthalmosc[ope and in retinal f](https://www.ncbi.nlm.nih.gov/mesh/D016202)la[tmou](https://www.ncbi.nlm.nih.gov/mesh/D016202)nt or cross-section specimens using laser confocal microscopy.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
Results
Treatment with tunicamycin induced apoptotic cell death in RGC-5 and also induced production of ER stress-related proteins (BiP, the phosphorylated form of eIF2α, and CHOP protein). In vivo, tunicamycin induced retinal ganglion cell (RGC) loss and thinning of the inner plexiform layer, 7 days after intravitreal injection. In flatmounted retinas of ERAI mice, the fluorescence intensity arising from the XBP-1-venus fusion protein, indicating ER-stress activation, was increased at 24 h after tunicamycin, NMDA, or IOP elevation. In transverse cross-sections from ERAI mice, the fluorescence intensity was first increased in cells of the ganglion cell and inner plexiform layers at 12 and 24 h, respectively, after NMDA injection, and it was localized to ganglion and amacrine cells at 12 and 24 h, respectively, and to microglial cells at 72 h. BiP and CHOP were increased at 12 h after NMDA injection, and the increases persisted for the remainder of the 72 h observation period.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
Conclusions
These data indicate that ER-stress may play a pivotal role in RGC death, whether induced by NMDA or IOP elevation.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Endoplasmic reticulum (ER) stress is caused by a number of biochemical and physiological stimuli that result in the accumulation of unfolded proteins in the ER lumen, and it is closely associated with the neuronal cell injury caused by vascular and neurodegenerative diseases such as stroke, Alzheimer disease, and Parkinson disease [1-3]. However, little is known about the role, if any, of ER stress in retinal damage.
Retinal ganglion cell (RGC) death is a common feature of many ophthalmic disorders such as glaucoma, optic neuropathies, and retinovascular diseases, such as diabetic retinopathy and retinal vein occlusions. RGC death has been reported to occur via a variety of mechanisms involving, for example, oxidative stress [4], excitatory amino acids [5], nitric oxide (NO) [6], and apoptosis [7].
Glutamate, one of the excitatory amino acids, is the main neurotransmitter in the retinal signaling pathway. Excessive glutamate increases both intracellular Ca2+ and NO production through activation of the N-methyl-D-aspartate (NMDA)-type glutamate receptor, resulting in retinal cell death [8,9]. Recently, Uehara et al. [10] reported that in primary cortical culture, even mild exposure of NMDA induces apoptotic cell death. They demonstrated to be caused by an accumulation of polyubiquitinated proteins and increases in X box binding protein (XBP-1) mRNA splicing and C/EBP-homologous (CHOP) mRNA, representing activation of the unfolded-protein response (UPR) signaling pathway.
They also found that protein-disulphide isomerase (PDI), which assists in the maturation and transport of unfolded secretory proteins, prevented the neurotoxicity associated with ER stress. They suggest that neurodegenerative disorders might be mediated by S-nitrosylation of PDI, which would reduce its enzymatic activity. Their results strongly suggest that activation of ER stress may participate in the retinal cell death occurring after NMDA receptor activation and/or ischemic insult. Hence, the purpose of the present study is to examine how ER stress might induce retinal damage both in vitro using cultured retinal ganglion cells (RGC-5, a rat ganglion cell line transformed using E1A virus) and in vivo (using ER stress-activated indicator (ERAI) transgenic mice, in which effective identification of cells under ER-stress conditions is possible in vivo, as described in our previous report) [11].
Use of ERAI mice should provide valuable information regarding the dynamics of ER stress-induced retinal damage.[](https://www.ncbi.nlm.nih.gov/mesh/D018846)
## Methods
*In the **Methods** section:*
## Materials
*In the **Materials** section:*
Dulbeco's modified Eagles's medium (D-MEM) was purchased from Sigma-Aldrich (St. Louis, MO). The drugs used and their sources were as follows. Tunicamycin was obtained from Calbiochem (San Diego, CA) and Wako (Osaka, Japan). Isoflurane was acquired from Nissan Kagaku (Tokyo, Japan), and fetal bovine serum (FBS) was obtained from Valeant (Costa Mesa, CA).[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
## Retinal ganglion cell line (retinal ganglion cell-5) culture
*In the **Retinal ganglion cell line (retinal ganglion cell-5) culture** section:*
Cultures of RGC-5 were maintained in D-MEM supplemented with 10% FBS, 100 U/ml penicillin (Meiji Seika Kaisha, Ltd., Tokyo, Japan), and 100 μg/ml streptomycin (Meiji Seika Kaisha, Ltd.) in a humidified atmosphere of 95% air and 5% CO2 at 37 °C. The RGC-5 cells were passaged by trypsinization every 3 days, as in a previous report [12].[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Cell viability assay after tunicamycin
*In the **Cell viability assay after tunicamycin** section:*
RGC-5 cells were plated at a density of 1000 cells/well in 96-well culture plates (number 3072, Falcon®, Becton Dickinson and Company, Franklin Lakes, NJ). Twenty-four h later, cells were washed twice with D-MEM and then immersed in D-MEM supplemented with 1% FBS plus tunicamycin at 1 to 4 μg/ml. Twenty-four or forty-eight hours after the addition of tunicamycin, cell viability was measured using a single-cell digital imaging-based method employing fluorescent staining of nuclei. Briefly, cell death was assessed on the basis of combination staining with fluorescent dyes [namely, Hoechst 33342 (Molecular Probes, Eugene, OR) and either YO-PRO-1 (Molecular probes) or propidium iodide (PI; Molecular probes)]. Observations were made using an Olympus IX70 inverted epifluorescence microscope (Olympus, Tokyo, Japan).
At the end of the above culture period, Hoechst 33342 and YO-PRO-1 or PI dyes were added to the culture medium at 8 μM, 0.1 μM, and 1.5 μM, respectively, for 30 min. Images were collected using a digital camera (Coolpix 4500, Nikon Corp., Tokyo, Japan). In a blind manner, a total of at least 400 cells per condition were counted using image-processing software (Image-J ver. 1.33f, National Institutes of Health, Bethesda, MD). Cell mortality was quantified by expressing the number of YO-PRO-1- or PI-positive cells as a percentage of the number of Hoechst 33342-positive cells.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
## Animals
*In the **Animals** section:*
ER-stress-activated indicator (ERAI)-transgenic mice carrying the F-XBP1DDBD-venus expression gene [11] and their background wild-type mice (C57BL/6) aged 8-11 weeks or male adult ddY mice (Japan SLC, Hamamatsu, Japan) weighing 36-43 g for experiments other than the comparison with ERAI-transgenic mice were used, and were kept under controlled lighting conditions (12 h:12 h light/dark). All experiments were performed in accordance with the ARVO statement for the Use of Animals in Ophthalmic and Vision Research, and were approved and monitored by the Institutional Animal Care and Use Committee of Gifu Pharmaceutical University.
## Retinal damage induced by N-methyl-D-aspartate (NMDA)-, tunicamycin-, or intraocular pressure (IOP) elevation
*In the **Retinal damage induced by N-methyl-D-aspartate (NMDA)-, tunicamycin-, or intraocular pressure (IOP) elevation** section:*
Male mice were anesthetized with 3.0% isoflurane and maintained using 1.5% isoflurane in 70% N2O and 30% O2, delivered via an animal general anesthesia machine (Soft Lander, Sin-ei industry Co. Ltd., Saitama, Japan). The body temperature was maintained at 37.0 - 37.5 °C with the aid of a heating pad and heating lamp. Retinal damage was induced by injection (2 μl/eye) either of NMDA (Sigma-Aldrich) at 20 mM dissolved in 0.01 M phosphate-buffered saline (PBS) or of tunicamycin at 50 and 500 μg/ml, or (b) by acutely increasing the intraocular pressure (IOP). For NMDA- or tunicamycin-induced injury, the relevant agent was injected into the vitreous body of the left eye under the above anesthesia. In the IOP elevation model, the pupils were dilated with topical 2.5% phenylephrine hydrochloride and 1% tropicamide (Santen Pharmaceuticals Co. Ltd., Osaka, Japan).
After topical instillation of 0.4% oxybuprocaine hydrochloride (Santen Pharmaceuticals Co. Ltd.), the anterior chamber was cannulated with a 32-gauge needle connected to a reservoir containing 0.9% NaCl. IOP was elevated by raising the height of the reservoir, maintaining a pressure of 100 mm Hg for 45 min. Retinal ischemia was confirmed by the blanching of the iris and retinal circulation. At the end of the elevated IOP period, the needle was removed, and reperfusion of the retinal vasculature was confirmed by ophthalmoscopic examination (KOM 300; Konan Inc., Nishinomiya, Japan). One drop of levofloxacin ophthalmic solution (Santen Pharmaceuticals Co. Ltd.) was applied topically to the treated eye after each procedure (intravitreal injection or ischemia-reperfusion).[](https://www.ncbi.nlm.nih.gov/mesh/D007530)
## Monitoring endoplasmic reticulum (ER) stress using ERAI-transgenic mice
*In the **Monitoring endoplasmic reticulum (ER) stress using ERAI-transgenic mice** section:*
In anesthetized ERAI-transgenic or wild-type mice, retinal damage was induced by injection (2 μl/eye) of either NMDA at 20 mM or tunicamycin at 50 μg/ml into the vitreous body, or by elevating IOP to 100 mmHg for 45 min (see above). Twenty-four hours later, the fluorescence intensity arising from the XBP-1-venus fusion protein, which is translated from the F-XBP1DDBD-venus gene, was visualized in the retina of anesthetized animals using an ophthalmoscope (TRC-50; TOPCON, Tokyo, Japan) fitted with a fluorescence filter. In separate experiments, the distribution and time-course of changes in fluorescence intensity in the retina were measured in retinal flatmount and cross-section specimens using either laser confocal microscopy (Bio-Lad Laboratories, Inc, Hercules, CA) or epifluorescence microscopy (Power BX50; Olympus, Tokyo, Japan). At various times after the intravitreal injections (12, 24, and 72 h), eyes were enucleated, then fixed in 4% paraformaldehyde for 1 h or overnight at 4 °C as preparation for retinal flatmount and retinal cross-section, respectively. For the preparation of retinal flatmounts, detached retinas were flatmounted on slides (MAS COAT; MATSUNAMI GLASS IND., LTD., Osaka, Japan) by making radial incisions.
They were then mounted under a coverslip and observed using the epifluorescence microscope. For the preparation of retinal cross-sections, fixed eyes were immersed in 20% sucrose for 48 h at 4 °C, and embedded in optimum cutting temperature (OCT) compound (Sakura Finetechnical Co., Ltd, Tokyo, Japan). Transverse, 10 μm thick cryostat sections were cut and placed onto slides (MAS COAT) under a coverslip, and observed using the laser confocal microscope.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
## Immunoblotting
*In the **Immunoblotting** section:*
RGC-5 cells or mouse retinas were lysed using a cell-lysis buffer (RIPA buffer (R0278; Sigma) with protease (P8340; Sigma Aldrich) and phosphatase inhibitor cocktails (P2850 and P5726; Sigma), and 1 mM EDTA). Cell lysates were solubilized in SDS-sample buffer, separated on 10% SDS-polyacrylamide gels, and transferred to PVDF membrane (Immobilon-P; Millipore, Bedford, MA). Transfers were blocked for 1 h at room temperature with 5% Blocking One-P (Nakarai Tesque, Inc., Kyoto, Japan) in 10 mM Tris-buffered saline with 0.05% Tween 20 (TBS-T), then incubated overnight at 4 °C with the primary antibody. The transfers were then rinsed with TBS-T and incubated for 1 h at room temperature in horseradish peroxidase goat anti-rabbit or goat anti-mouse (Pierce, Rockford, IL) diluted 1:2000. The immunoblots were developed using chemiluminescence (Super Signal® West Femto Maximum Sensitivity Substrate; Pierce), and visualized with the aid of a digital imaging system (FAS-1000; Toyobo CO., LTD, Osaka, Japan).
The primary antibodies used were as follows: mouse anti-BiP (BD Bioscience, San Jose, CA), rabbit anti-phospho-eIF2α (Ser51; Cell Signaling, Beverly, MA), rabbit anti-eIF2α (Cell Signaling), mouse anti-CHOP (Santa Cruz, Santa Cruz, CA), and rabbit anti-actin (Santa Cruz).[](https://www.ncbi.nlm.nih.gov/mesh/D004492)
## Immunostaining
*In the **Immunostaining** section:*
To clarify the distribution and localization of the XBP1-venus fusion protein in the retina of ERAI mice (as seen in the retinal flatmounts and cross sections), double-staining immunocytochemistry was performed. At various times after intravitreal injections (12, 24, and 72 h), eyes were enucleated, fixed in 4% paraformaldehyde overnight at 4 °C, immersed in 20% sucrose for 48 h at 4 °C, and embedded in optimum cutting temperature (OCT) compound (Sakura Finetechnical Co., Ltd, Tokyo, Japan). Transverse, 10 μm thick cryostat sections were cut and placed onto slides MAS COAT (MATSUNAMI GLASS IND., LTD.). Sections were subsequently processed for immunocytochemical localization using antibodies against CHOP (1:100 dilution in PBS; Santa Cruz), glucose-regulated protein 78(GRP78)/BiP (1:100 dilution in PBS), thymus cell antigen 1 (Thy-1; 1:100 dilution in PBS; Serotec Ltd, Oxford UK), microglia (OX-42, 1:100 dilution in PBS; Serotec Ltd), and amacrine cells (HPC-1/Syntaxin, 1:100 dilution in PBS; Santa Cruz). The sections were incubated either (a) with Alexa Fluor-568-conjugated secondary antibody (1:200 dilution in PBS; Molecular Probes, Eugene, OR) for 1 h at room temperature, mounted with a coverslip, and observed under a laser confocal microscope (Bio-Lad Laboratories, Inc), or (b) with biotin-conjugated secondary antibody for 1 h at room temperature, and visualized using a VECTOR M.O.M. Immunodetection kit (Vector, Burlingame, CA).
Each image was taken using a digital camera (Coolpix 4500; Nikon, Tokyo, Japan) attached with epifluorescence microscope (Power BX50; Olympus).[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
## Histological analysis of mouse retina
*In the **Histological analysis of mouse retina** section:*
Seven days after the NMDA or tunicamycin injection, eyeballs were enucleated for histological analysis. In mice under anesthesia, produced by an intraperitoneal injection of sodium pentobarbital (80 mg/kg), each eye was enucleated, then kept immersed for at least 24 h at 4 °C in a fixative solution containing 4% paraformaldehyde. Six paraffin-embedded sections (thickness, 3 μm) cut through the optic disc of each eye were prepared in a standard manner, and stained with hematoxylin and eosin. Retinal damage was evaluated as described previously, and three sections from each eye were used for the morphometric analysis. Light-microscope photographs were taken using a digital camera (Coolpix 4500, Nikon) and the cell counts in the ganglion cell layer (GCL) and the thickness of the inner plexiform layer (IPL) at a distance between 350 and 650 μm from the optic disc were measured on the images in a masked fashion by a single observer (Y.I.).
Data from three sections (selected randomly from the six sections) were averaged for each eye, and the values obtained were used to evaluate the GCL cell count and the IPL thickness.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
## Statistical analysis
*In the **Statistical analysis** section:*
Data are presented as the means±SEM. Statistical comparisons were made using a Student's t-test or Dunnett's test, by means of STAT VIEW version 5.0 (SAS Institute Inc., Cary, NC). P<0.05 was considered to be statistically significance.
## Results
*In the **Results** section:*
## Retinal cell death and time-course of changes in endoplasmic reticulum (ER) stress-related protein induced by tunicamycin
*In the **Retinal cell death and time-course of changes in endoplasmic reticulum (ER) stress-related protein induced by tunicamycin** section:*
We examined whether tunicamycin treatment could induce cell death through ER stress in retinal ganglion cell using RGC-5. Representative fluorescence stainings of nuclei [using Hoechst 33342, YO-PRO-1, and propidium iodide (PI) dyes] are shown in Figure 1A. Vehicle-treated control cells displayed normal nuclear morphology and negative staining with both YO-PRO-1 dye (which stains early apoptotic and later-stage cells) and PI dye (which stains late-stage apoptotic cells; upper panels in Figure 1A). Treatment with tunicamycin led to shrinkage and condensation of nuclei, and to positive staining with each of these dyes (lower panels in Figure 1A). The number of cells exhibiting PI fluorescence was counted, and positive cells were expressed as the percentage of PI- to Hoechst 33342-positive cells (Figure 1B). After treatment with tunicamycin at 1, 2, or 4 μg/ml for 24 h, the percentages of PI-positive cells were 8.3±1.2% (n=6), 13.1±0.9% (n=6), and 11.3±0.6% (n=6), respectively, while in the non-treated control group the percentage was 0.5±0.2% (n=6).
After treatment with tunicamycin at 1, 2, or 4 μg/ml for a longer time period (48 h), the corresponding values were 41.5±3.5% (n=6), 43.7±2.1% (n=6), and 50.7±2.6% (n=6), respectively (1.2±0.4% (n=6) for the non-treated control group). Time-course data for the changes in the protein levels of glucose-regulated protein (GRP)78/BiP, the phosphorylated form of eukaryotic initiation factor 2α (eIF2α), total eIF2α, and C/EBP-homologous protein (CHOP) occurring after tunicamycin treatment at 2 μg/ml are shown in Figure 1C. BiP, a biomarker of ER-stress, increased time-dependently throughout the 24 h tunicamycin treatment period, while actin levels remained unchanged. Treatment with tunicamycin time-dependently induced eIF2α phosphorylation, while total eIF2α levels were not changed during the 24 h observation period. CHOP was first detected at 6 h after addition of tunicamycin and persisted thereafter. These data indicate that treatment with tunicamycin can induce expressions of ER stress-related proteins and subsequent apoptotic cell death in RGC-5 culture in vitro.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
Retinal cell death and time-course of changes in endoplasmic reticulum (ER)-stress related proteins induced by tunicamycin. A: Representative fluorescence microscopy showing nuclear stainings for Hoechst 33342 (blue), YO-PRO-1 (green), and propidium iodide (PI, red) at 48 h after addition of tunicamycin at 1 μg/ml. B: The number of cells displaying PI fluorescence was counted at two time-points, and positive cells were expressed as the percentage of PI to Hoechst 33342. Each column represents the mean±SEM (n=6). Double asterisks and double hash marks; p<0.01 versus corresponding control group (Dunnett's test). C: Representative immunoblots showing the time-course of changes in protein levels (GRP78/BiP, phosphorylated-eIF2α, total eIF2α, and CHOP) after tunicamycin treatment at 2 μg/ml.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
## Intravitreal injection of tunicamycin induces retinal cell death in mice
*In the **Intravitreal injection of tunicamycin induces retinal cell death in mice** section:*
To clarify whether tunicamycin would induce retinal cell death in vivo, we examined the histological changes in the retina at 7 days after intravitreal injection of tunicamycin. As shown in Figure 2, intravitreal injection of tunicamycin at 0.1 μg/eye (a low dose) induced a significant loss of cells in the retinal ganglion cell layer (GCL), but no thinning of the inner plexiform layer (IPL; versus vehicle-treated retinas). At a high dose of 1μg/eye, tunicamycin significantly decreased both the cell count in GCL and the IPL thickness (versus the non-treated normal retina; Figure 2). On the other hand, no retinal damage was induced by intravitreal injection of an identical volume of vehicle (versus the non-treated retina). Together, these findings suggest that tunicamycin at 0.1 μg/eye (giving an estimated concentration in the vitreous body of approximately 10 μg/ml) induces retinal ganglion cell death at a concentration similar to that inducing exhibiting the apoptotic cell death in RGC-5 in vitro.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
Intravitreal injection of tunicamycin induces retinal cell death in mice. A: Representative photographs showing non-treated normal retina, vehicle-treated retina, and low-dose (0.1 μg/eye) and high-dose (1 μg/eye) tunicamycin-treated retinas 7 days after intravitreal injection. Quantitative analysis of cell number in ganglion cell layer (B) and thickness of inner plexiform layer (IPL) C: Each column represents the mean±SEM (n=10). Double asterisks p<0.01 versus vehicle-treated control group (Dunnett's test). The horizontal scale bar represents 25 μm and the vertical bar indicates each thickness of IPL.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
## Increase in XBP-1-venus fusion protein in the retina in ER stress-activated indicator (ERAI)-transgenic mice
*In the **Increase in XBP-1-venus fusion protein in the retina in ER stress-activated indicator (ERAI)-transgenic mice** section:*
To investigate whether ER stress is induced in the mouse retina during the early stages of retinal damage in vivo, we used ERAI-transgenic mice carrying the F-XBP1DDBD-venus expression gene, which allows effective identification of cells under ER stress in vivo, as previously described by Iwawaki et al. [11]. Twenty-four h after intravitreal injection of tunicamycin at 0.1 μg or of N-methyl-D-aspartate (NMDA) at 40 nmol, the fluorescence intensity arising from the XBP-1-venus fusion protein was visualized in the retina of anesthetized animals (using an ophthalmoscope) as shown in Figure 3. Both tunicamycin and NMDA increased the fluorescence intensity of this protein, while little change in fluorescence intensity was observed in the control fellow eyes. For further elucidation of this phenomenon, the distribution and time-course of changes in the fluorescence intensity derived from the XBP-1-venus fusion protein were measured in retinal flatmount and transverse sections, as shown in Figure 4A,C. In the flatmounts, such stimulations as NMDA, an intraocular pressure (IOP) elevation, and tunicamycin all induced increases in fluorescence intensity at the time-points indicated in Figure 4A.
In the NMDA-treated retinas of ERAI mice, the background fluorescence intensity was time-dependently increased in the period from 12 to 72 h, but little change was observed in the NMDA-treated retinas of wild-type mice.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
Non-invasive imaging of XBP-1-venus fusion protein in ERAI mouse retina in vivo. Twenty-four hours after intravitreal injection of either tunicamycin at 0.1 μg/eye or N-methyl-D-aspartate (NMDA) at 40 nmol/eye, the fluorescence intensity arising from XBP-1-venus fusion protein was visualized in the retinas of anesthetized animals using an ophthalmoscope fitted with a fluorescence filter.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
Expression and localization of XBP-1-venus fusion protein in ERAI mouse retinas after various types of retinal damage. A: Representative fluorescence photographs of increased XBP-1-venus fusion protein in ERAI mouse flatmounted retina after N-methyl-D-aspartate (NMDA), intraocular pressure (IOP) elevation, or tunicamycin insult. The fluorescence (green) arising from XBP-1-venus fusion protein was observed under an epifluorescence microscope. The scale bar represents 25 μm. B: Distribution of increased XBP-1-venus fusion protein in retinal cross-sections from ERAI mice after NMDA injection at 40 nmol/eye. The distribution of fluorescence (green) arising from XBP-1-venus fusion protein was observed under a laser confocal microscope. Each large box shows an enlargement of the area within the corresponding small box. C: Localization of XBP-1-venus fusion protein in ERAI mouse retina after NMDA injection. In the retinal nerve fiber layer (upper panels), Thy-1-positive cells (red) can be seen to merge with XBP-1-venus fusion protein (green). In the middle panels, OX-42 (a microglia marker)-positive cells (red) are partly merged with XBP-1-venus fusion protein (green). In the inner plexiform layer (lower panels), HPC-1 (an amacrine marker)-positive cells (red) are merged with XBP-1-venus fusion protein (green).[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
These changes in background could reflect increases in the lower part of the ganglion cell layer, such as the inner plexiform layer and neuroepithelial layer, of the retina. In transverse sections, increases in fluorescence intensity were first observed in cells of the GCL and inner plexiform layer at 12 and 24 h, respectively, after NMDA injection, and the increases peaked in GCL cells at 24 h (Figure 4B). The increase in fluorescence had diminished at 72 h after the NMDA injection, but morphologically distinct cells (such as microglia cells) had appeared in GCL. On the other hand, the retinas of wild-type and non-treated ERAI mice showed a low fluorescence intensity (below background), while a slight fluorescence intensity was observed in the neuroepithelial layer of the retina (Figure 4B). These cells merged with Thy-1-positive cells (ganglion cells) and some OX-42-positive cells (microglia) in GCL, and with HPC-1-positive cells (amacrine cells) in IPL (Figure 4C).
Together, these results suggest that XBP-1 splicing, representing activation of the ER-stress signal pathway, may be induced in retinal ganglion and amacrine and microglia cells during the early stages of retinal cell damage.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
## Increases in GRP78/BiP and CHOP in mouse retina after NMDA injection
*In the **Increases in GRP78/BiP and CHOP in mouse retina after NMDA injection** section:*
To clarify whether ER stress-related proteins other than XBP-1 are induced in the mouse retina by NMDA stimulation, we examined the changes in BiP, a biomarker of ER stress, in the retina after intravitreal injection of NMDA. As shown in Figure 5B, cell loss in GCL and thinning of IPL were observed at 72 h after NMDA injection (versus non-treated control retinas; Figure 5A). Using immunoblots, as shown in Figure 5C, we found that BiP was significantly increased at 12 h after the NMDA injection, and that the increase persisted for the remainder of the 72 h observation period. Next, we investigated the distribution and time-course of changes in GRP78/BiP and CHOP, a proapoptosis protein, after NMDA injection. In the non-treated control retina, slight immunoreactivities for BiP and CHOP were observed in a number of cells in GCL and IPL (Figure 5D).
Increases in these immunoreactivities were observed in retinal ganglion cells at 12 h after NMDA injection, and time-dependent increases were noted in the inner retina (Figure 5D).[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
Increases in GRP78/BiP and CHOP in retinal extracts following stimulation by intravitreal injection of N-methyl-D-aspartate (NMDA) in mice. A, B: Representative photographs showing retinal cross-sections stained with hematoxylin and eosin after NMDA injection at 40 nmol/eye. C, upper panel: Representative immunoblots showing the time-course of changes in GRP78/BiP protein levels after intravitreal injection of NMDA. C, lower panel: Quantitative analysis of GRP78/BiP band densities. Data are expressed as mean±SEM (n=6) of values (in arbitrary units) obtained for single band density. Double asterisks represents p<0.01 versus vehicle-treated control group (Dunnett's test). D: Immunostainings for GRP78/BiP and CHOP in mouse retina after NMDA injection at 40 nmol. The scale bar represents 25 μm.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
## Discussion
*In the **Discussion** section:*
In the present study, we could detect pathological changes and time-dependent changes related to ER stress in retinal flatmount and transverse sections and in the retinas of living mice after retinal damage. Moreover, we demonstrated that ER stress signals were activated in the retina in vivo after tunicamycin, elevating IOP, or NMDA treatment.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
Agents or conditions that adversely affect ER protein folding lead to an accumulation of unfolded or misfolded proteins in the ER, a condition defined as ER stress. ER stress can be induced by agents or conditions that interfere with (a) protein glycosylation (e.g., glucose starvation, tunicamycin, glucosamine), (b) disulfide-bond formation (e.g., DTT, homocysteine), (c) Ca2+ balance (A23187, thapsigargin, EGTA), and/or (d) a general overloading of the ER with proteins (e.g., viral or non-viral oncogenesis) [1,13,14]. However, little is known about any involvement of ER stress in retinal damage. In the present study, we found that tunicamycin induced the ER stress-associated proteins BiP, p-eIF2α, and CHOP in cultured RGC-5 cells.
These protein levels started to increase at 2 to 6 h after the start of tunicamycin treatment, and increased time-dependently until 24 h after the start of the treatment, while apoptotic cell death with condensation and fragmentation of nuclei was observed 24 h later. BiP acts as an ER resident molecular chaperon that is induced by ER stress, and this protein refolds the unfolded proteins, thereby tending to maintain homeostasis in the ER [15,16]. Since CHOP is a member of the CCAAT/enhancer-binding protein family that is induced by ER stress and participates in ER-mediated apoptosis, CHOP may be a key molecule in retinal cell death [17]. In the present study, the phosphorylation of eIF2α was increased concomitantly with the increases in the expression of BiP and CHOP proteins, even through p-eIF2α might be expected to suppress protein synthesis.
Boyce et al. [18] reported that selective inhibition of eIF2α dephosphorylation increases both p-eIF2α and CHOP protein. These data suggest that during ER stress, p-eIF2α (inactive form) is still able to stimulate the translation of ATF4 mRNA, thereby increasing the transcription of BiP or CHOP mRNA, but that enough unphosphorylated-eIF2α (active form) may remain to translate BiP and CHOP mRNAs to proteins. On the other hand, we found that staurosporine, which mediates mitochondrial dysfunctions resulting in apoptotic cell death, did not induce any increases in BiP and CHOP proteins in RGC-5 [unpublished data]. Taken together, these findings suggest that persistent ER stress may induce apoptotic cell death through the eIF2α-CHOP signal pathway in RGC-5.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
Next, we tried to determine whether tunicamycin could induce retinal damage in vivo. Intravitreal injection of low-dose tunicamycin induced a significant loss of cells in the retinal ganglion cell layer (GCL), but no thinning of the inner plexiform layer (IPL). These findings suggest that retinal ganglion cells are more sensitive to ER stress-induced cell death than other retinal cells. High-dose tunicamycin significantly decreased both the cell count in GCL and the thickness of IPL. The concentration of tunicamycin in the vitreous body after an intravitreal injection of low-dose tunicamycin was estimated to be 10 μg/ml.
The tunicamycin concentration achieved within the retina will have been less than this. Interestingly, in the present in vitro study, tunicamycin at 1 to 4 μg/ml induced cell death with an increase in ER-stress signals, suggesting that the in vivo concentration of tunicamycin in the retina was roughly similar to that employed in vitro. Use of tunicamycin at a high dose also led to decreases in IPL, INL (inner nuclear layer), and ONL (outer nuclear layer) in the retina. In guinea pigs, a single subcutaneous injection of tunicamycin at 0.4 mg/kg has been reported to induce hepatotoxicity with dilation of the cisternae of the ER [19].
Furthermore, Zinszner et al. [20] noted that in mice, a single sublethal intraperitoneal injection of tunicamycin (1 mg/kg) induces CHOP expression and subsequent severe histological damage with an increase in TUNEL-positive cells, and a characteristic transient renal insufficiency. They also found that CHOP-deficient mice show an attenuated increase in TdT-mediated dUTP nick-end labeling (TUNEL)-positive cells during the renal damage induced by tunicamycin. These findings suggest that in vivo, tunicamycin-induced retinal cell death is due, at least in part, to an ER-stress mechanism.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
NMDA receptors may participate in the processes of excitotoxicity and neuronal death in the retina [21,22]. Previous studies have found that TUNEL-positive cells can be observed in the GCL and INL of the mouse retina at an early stage (within 24 h) after an intravitreal injection of NMDA [23,24]. The hallmark of NMDA-induced neuronal death is a sustained increase in the intracellular Ca2+ concentration accompanied by overactivation of vital Ca2+-dependent cellular enzymes [25]. Thus, the signal-transduction pathways for NMDA-mediated cell death in the retina are well studied, but not yet fully understood.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
To illuminate the role and distribution of ER stress in vivo, we focused on the retina of ERAI mice. Information about the status of ER stress during the course of a given disease might be obtained by crossing an ERAI transgenic mouse (the indicator mouse for ER stress in living cells) with a mouse model of the human disease of interest. In flatmounted retinas, fluorescence was detected following various stimulations [tunicamycin, NMDA, and intraocular pressure (IOP) elevation]. To our knowledge, this is the first report demonstrating that NMDA and ischemic insult (elevating IOP), in addition to tunicamycin, can activate the ER stress signal (measured as the splicing of the XBP-1 and venus fusion gene in ERAI transgenic mice) in the retina in vivo. Interestingly, ER stress was also induced in the retina after a transient IOP elevation, defined as an ischemia-reperfusion model.
It has been reported that this model exhibits retinal cell damage similar to that induced by NMDA, and that both of these examples of damage are protected against by MK-801, an NMDA receptor antagonist, and by NO synthetase-inhibitor treatment [8,26]. Although little is known about the precise mechanisms responsible for activation of ER stress after NMDA or IOP elevation (ischemia-reperfusion), both stimuli cause intracellular Ca2+ overload and increased NO production, resulting in apoptotic cell death. Several lines of study suggest that intracellular Ca2+ overload and excessive production of NO deplete Ca2+ in the ER, thereby resulting in ER stress [27,28]. Recently, Uehara et al. [10] reported that NO induces S-nitrosylation of protein-disulphide isomerase (PDI), an enzyme that assists in the maturation and transport of unfolded secretory proteins and thereby helps to prevent the neurotoxicity associated with ER stress.
S-nitrosylated-PDI exhibits reduced enzymatic activity and induces cell death through the ER stress pathway. These mechanisms may contribute to the activation of ER stress in the retina after NMDA stimulation or IOP elevation. Accordingly, our findings may provide important new insights into the mechanisms underlying the retinal cell damage induced by NMDA and by ischemia-reperfusion. In transverse retinal sections, we observed an increase in fluorescence intensity within the cells of the ganglion cell layer (GCL) and inner plexiform layer (IPL) at 12 and 24 h, respectively, after NMDA injection. The cells displaying increased fluorescence were ganglion cells (at 12 h after the injection), amacrine cells in IPL (at 24 h), and microglia in GCL (at 72 h).
These data indicate that ganglion cells may be more sensitive to ER stress than the other retinal cells examined.[](https://www.ncbi.nlm.nih.gov/mesh/D014415)
To further clarify the participation of ER stress, we examined the changes in BiP and CHOP in the retina after NMDA-induced injury. We found (a) that NMDA induced BiP proteins in the retina at 12 h after its injection (on the basis of immunoblots), and (b) that, NMDA induced both BiP and CHOP in the retina (especially within retinal ganglion cells and INL) at 12 h after its injection (on the basis of our immunostaining results). The expression of the CHOP gene reportedly increases in the rat retina after intravitreal injection of NMDA [29]. Furthermore, Awai et al. [30] found that treatment with MK-801, an NMDA receptor antagonist, inhibited the increases in CHOP mRNA and protein in the mouse retina that are observed after intravitreal injection of NMDA, and moreover that CHOP-deficient mice were resistant to NMDA-induced retinal damage.
However, CHOP-deficient mice partially suppressed the NMDA-induced cell death, and therefore other pathways, such as mitochondrial dysfunction, may be engaged in the retinal cell death. Collectively, the above results indicate that NMDA can cause ER stress in the retina, and that the neurotoxicity induced by NMDA is due in part to a mechanism dependent on CHOP protein induction through excessive ER stress.[](https://www.ncbi.nlm.nih.gov/mesh/D016202)
In conclusion, we have identified a close association between ER stress and retinal damage, and our results suggest that the ER stress-signal pathway might be a good target in the treatment of retinal diseases.
# References
*In the **References** section:*
|
# Introduction
7′-Phenyl-1′,3′,5′,6′,7′,7a’-hexahydrodipiro[acenaphthylene-1,5′-pyrrolo[1,2-c]thiazole-6′,2′′-indane]-2,1′′(1H)-dione
# Abstract
*In the **Abstract** section:*
In the title compound, C31H23NO2S, the pyrrolidine ring adopts an envelope conformation (with the spiro C atom as the flap), while the thiazolidine ring and the t[wo cyclope](https://www.ncbi.nlm.nih.gov/mesh/D011759)ntane rings adopt twisted conformations. The mean plane through the hexahydropyrrol[o[1,2-c]thia](https://www.ncbi.nlm.nih.gov/mesh/D053778)zole ring [r.m.s [deviation = 0](https://www.ncbi.nlm.nih.gov/mesh/D003517).400 (1) Å] forms dihedral angles of 76.83 (4), 80.70 (5) and 7[9.00 (4)° with the benzene ring and](https://www.ncbi.nlm.nih.gov/mesh/D013844) the mean planes of the dihydroacenaphthylene and the dihydroindene rings, respectively. In the crystal, [molecu](https://www.ncbi.nlm.nih.gov/mesh/D001554)les are linked by C—H⋯O hydrogen [bonds into sheets lying](https://www.ncbi.nlm.nih.gov/mesh/C042553) parallel[ to the bc pla](https://www.ncbi.nlm.nih.gov/mesh/D007192)ne. One of the ketone O atoms accepts three such bonds. Weak C—H⋯π in[teracti](https://www.ncbi.nlm.nih.gov/mesh/D006859)ons are also observed.[](https://www.ncbi.nlm.nih.gov/mesh/D007659)
## Related literature
*In the **Related literature** section:*
For related structures, see: Wei et al. (2011a ▶,b ▶, 2012 ▶). For ring conformations, see: Cremer & Pople (1975 ▶). For the stability of the temperature controller used in the data collection, see: Cosier & Glazer (1986 ▶).
## Experimental
*In the **Experimental** section:*
## Crystal data
*In the **Crystal data** section:*
C31H23NO2S
M r = 473.56
Monoclinic,
a = 8.4054 (1) Å
b = 11.3716 (1) Å
c = 23.5194 (2) Å
β = 92.259 (1)°
V = 2246.30 (4) Å3
Z = 4
Mo Kα radiation[](https://www.ncbi.nlm.nih.gov/mesh/D008982)
μ = 0.18 mm−1
T = 100 K
0.30 × 0.18 × 0.16 mm
## Data collection
*In the **Data collection** section:*
Bruker SMART APEXII CCD diffractometer
Absorption correction: multi-scan (SADABS; Bruker, 2009 ▶) T min = 0.949, T max = 0.972
39597 measured reflections
10047 independent reflections
7694 reflections with I > 2σ(I)
R int = 0.039
## Refinement
*In the **Refinement** section:*
R[F 2 > 2σ(F 2)] = 0.051
wR(F 2) = 0.124
S = 1.03
10047 reflections
316 parameters
H-atom parameters constrained
Δρmax = 0.58 e Å−3
Δρmin = −0.29 e Å−3
Data collection: APEX2 (Bruker, 2009 ▶); cell refinement: SAINT (Bruker, 2009 ▶); data reduction: SAINT; program(s) used to solve structure: SHELXTL (Sheldrick, 2008 ▶); program(s) used to refine structure: SHELXTL; molecular graphics: SHELXTL; software used to prepare material for publication: SHELXTL and PLATON (Spek, 2009 ▶).
## Supplementary Material
*In the **Supplementary Material** section:*
Supplementary data and figures for this paper are available from the IUCr electronic archives (Reference: HB6700).
# supplementary crystallographic information
*In the **supplementary crystallographic information** section:*
## Comment
*In the **Comment** section:*
As part of our ongoing search to prepare heterocyclic compounds with potential antitubercular activity (Wei et al., 2011a,b), we have synthesized the title compound as described below.
In the molecular structure (Fig 1), the pyrrolidine ring (N1/C12/C13/C22/C23) is in envelope conformation (Cremer & Pople, 1975) [puckering parameters, Q= 0.4480 (11) Å and φ= 68.75 (14)° with atom C13 at the flap]. Meanwhile, the thiazolidine ring and the two cyclopentane rings (S1/N1/C23–C25, C1/C2/C10–C12 & C13–C15/C20/C21) are twisted about C25–S1 bond [puckering parameters, Q= 0.3450 (11) Å and φ= 339.37 (19)°], C12–C1 bond [puckering parameters, Q= 0.1209 (11) Å and φ= 167.0 (5)°] and C13–S14 bond [puckering parameters, Q= 0.2875 (11) Å and φ= 190.5 (2)°], respectively, adopting half-chair conformation. In addition, the dihedral angles between the mean plane through the hexahydropyrrolo [1,2-c]thiazole ring (S1/N1/C12/C13/C22–C25)
[r.m.s deviation of 0.400 (1) Å] with the benzene ring (C26–C31) and the mean planes of the dihydroacenaphthylene and the dihydro-indene rings (C1–C10/C12 & C13–C21) are 76.83 (4), 80.70 (5) and 79.00 (4)°, respectively. The bond lengths and angles are within normal ranges and comparable to the related structure (Wei, et al., 2011b; Wei, et al., 2012).[](https://www.ncbi.nlm.nih.gov/mesh/D011759)
The crystal packing is shown in Fig. 2. The molecules are linked into sheets lying parallel to bc-plane via C7—H7A···O2, C4—H4A···O1, C23—H23A···O1 and C31—H31A···O1 (Table 1) hydrogen bonds. The crystal structure also features C18—H18A···Cg1 and C25—H25A···Cg2 (Table 1) interactions (Cg1 and Cg2 are the centroids of the C2–C6/C11 and C15–C20 rings, respectively).[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Experimental (cont.)
*In the **Experimental (cont.)** section:*
A mixture of (E)-(2-benzylidene)-2,3-dihydro-1H-indene-1-one (0.001 mol), acenaphthenequinone (0.001 mol) and thiazolidine-4-carboxylic acid (0.002 mol) (1:1:2) were dissolved in methanol (10 ml) and refluxed for 4 h. After completion of the reaction as evident from TLC, the excess solvent was evaporated slowly and the product was separated and recrystallized from methanol to reveal the title compound as yellow crystals.[](https://www.ncbi.nlm.nih.gov/mesh/D001597)
## Refinement (cont.)
*In the **Refinement (cont.)** section:*
All H atoms were positioned geometrically (C–H = 0.95 and 1.00 Å) and refined using a riding model with Uiso(H) = 1.2 Ueq(C).
## Figures
*In the **Figures** section:*
The molecular structure of the title compound, showing 30% probability displacement ellipsoids.
The crystal packing of the title compound. The H atoms not involved in the intermolecular interactions (dashed lines) have been omitted for clarity.
## Crystal data (cont.)
*In the **Crystal data (cont.)** section:*
## Data collection (cont.)
*In the **Data collection (cont.)** section:*
## Refinement (cont.)
*In the **Refinement (cont.)** section:*
## Special details
*In the **Special details** section:*
## Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2)
*In the **Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2)** section:*
## Atomic displacement parameters (Å2)
*In the **Atomic displacement parameters (Å2)** section:*
## Geometric parameters (Å, º)
*In the **Geometric parameters (Å, º)** section:*
## Hydrogen-bond geometry (Å, º)
*In the **Hydrogen-bond geometry (Å, º)** section:*
Cg1 and Cg2 are the centroids of the C2–C6/C11 and C15–C20 rings, respectively.
Symmetry codes: (i) −x+1, y−1/2, −z+1/2; (ii) −x+1, y+1/2, −z+1/2; (iii) −x+1, −y, −z; (iv) x−1, y, z; (v) x+1, y, z.
# References
*In the **References** section:*
Hydrogen-bond geometry (Å, °)[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Cg1 and Cg2 are the centroids of the C2–C6/C11 and C15–C20 rings, respectively.
|
# Introduction
[Nitric oxide](https://www.ncbi.nlm.nih.gov/mesh/D009569) activates intradomain [disulfide](https://www.ncbi.nlm.nih.gov/mesh/D004220) bond formation in the kinase loop of Akt1/PKBα after burn injury
# Abstract
*In the **Abstract** section:*
Severe burn injury is an acute inflammatory state with massive alterations in gene expression and levels of growth factors, cytokines and free radicals. During the catabolic processes, changes in insulin sensitivity and skeletal muscle wasting (unintended loss of 5–15% of lean body mass) are observed clinically. Here, we reveal a novel molecular mechanism of Akt1/protein kinase Bα (Akt1/PKBα) regulated via cross-talking between dephosphorylation of Thr308 and S-nitrosylation of Cys296 post severe burn injury, which were characterized using nano-LC interface[d w](https://www.ncbi.nlm.nih.gov/mesh/D013912)ith tandem quadrupole time-[of-](https://www.ncbi.nlm.nih.gov/mesh/D003545)fight mass spectrometry (Q-TOF)micro tandem mass spectrometry in both in vitro and in vivo studies. For the in vitro studies, Akt1/PKBα was S-nitrosylated with S-nitrosoglutathione and derivatized by three methods. The derivatives were isolated by SDS-PAGE, trypsinized [and analyzed by the ](https://www.ncbi.nlm.nih.gov/mesh/D026422)tandem MS. For the in vivo studies, Akt1/PKBα in muscle lysates from[ bu](https://www.ncbi.nlm.nih.gov/mesh/D012967)rned rats was immuno-precipitated, derivatized with HPDP-Biotin and analyzed as above. The studies demonstrated that the NO free radical reacts with the free thio[l of Cys296](https://www.ncbi.nlm.nih.gov/mesh/C501660) to produce a Cys296-SNO intermediate which accelerates in[te](https://www.ncbi.nlm.nih.gov/mesh/D009569)raction with Cys310 to form Cys296-[Cys31](https://www.ncbi.nlm.nih.gov/mesh/D013438)0 in[ th](https://www.ncbi.nlm.nih.gov/mesh/D003545)e kinase loop. MS[/MS](https://www.ncbi.nlm.nih.gov/mesh/D003545) seq[uen](https://www.ncbi.nlm.nih.gov/mesh/D026403)ce analysis indicated that the dipeptide, linked [via](https://www.ncbi.nlm.nih.gov/mesh/D003545) Cys296-Cys3[10,](https://www.ncbi.nlm.nih.gov/mesh/D003545) und[erw](https://www.ncbi.nlm.nih.gov/mesh/D003545)ent dephosphorylation at Thr308. These effects were not observed in[ lysates ](https://www.ncbi.nlm.nih.gov/mesh/D004151)from sham ani[mal](https://www.ncbi.nlm.nih.gov/mesh/D003545)s. A[s a](https://www.ncbi.nlm.nih.gov/mesh/D003545) result of this dual effect of burn [inj](https://www.ncbi.nlm.nih.gov/mesh/D013912)ury, the loose conformation that is slightly stabilized by the Lys297-Thr308 salt bridge may be replaced by a more rigid structure which may block substrate access. Together[ wi](https://www.ncbi.nlm.nih.gov/mesh/D008239)th t[he ](https://www.ncbi.nlm.nih.gov/mesh/D013912)findings of our previous report concerning mild IRS-1 integrity changes post burn, it is reasonable to conclude that the impaired Akt1/PKBα has a major impact on FOXO3 subcellular distribution and activities.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Metabolic alterations that are produced by critical illness such as burn trauma are associated with a hypermetabolic/inflammatory state, increased protein catabolism (with resulting muscle wasting) and insulin resistance. Muscle wasting can lead to muscle weakness that can result in hypoventilation, prolongation of dependence on mechanical ventilation, prolonged rehabilitation and even death. Insulin resistance is a well established state in critically ill patients and is considered to play a key role in the metabolic derangements and muscle wasting. Binding of insulin to its receptor (IR) activates IR tyrosine kinase, which then phosphorylates IR substrates (IRSs). Phosphorylation of IRS1 and IRS2 transduces the signal from IR to phosphatidylinositol-3-kinase (PI3-kinase).
Post-translational modifications (PTMs) of the insulin signaling system are considered to be major disease-dependent events that regulate glucose transport via GLUT-4 translocation and protein synthesis.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
Akt1/PKBα is a critical downstream mediator of the IR/IRS/PI3-kinase pathway of the insulin signaling system. Akt1/PKBα consists of three structural features: the N-terminal pleckstrin homology (PH) domain, a large central kinase domain and a short C-terminal hydrophobic motif. High specific binding of the PH domain with membrane lipid products of PI3-kinase recruits Akt1/PKBα to the plasma membrane where phosphorylations of Thr308 (pThr308, kinase domain) and Ser473 (pSer473, hydrophobic motif) occur. Phosphorylation of Thr308 partially stimulates kinase activity; however, additional phosphorylation of Ser473 is required for full activity. Activation is associated with a disordered to ordered transition of a specific αC helix of Akt1/PKBα via an allosteric mechanism.
A salt bridge between the side-chain of Lys297 and the phosphate group of pThr308 in this αC helix contributes to an ordered activation segment from 292DFG to APE319. Reversible dephosphorylations of Thr308 and Ser473 by protein phosphatase 2A (PP2A) and PH domain leucine-rich repeat protein phosphatase (PHLPPα) also occur in the Akt1/PKBα activation/deactivation cycle.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
In addition to the role of reversible phosphorylation/dephosphorylation in the regulation of Akt1/PKBα activity, this kinase is also reversibly inactivated by S-nitrosylation under conditions that result in persistently increased production of nitric oxide; such as after burn injury. Thiol titration and NMR data indicate that a disulfide bond (Cys60-Cys77) exists in the kinase PH domain. A second disulfide bond in the critical kinase activation loop (Cys297-Cys311) has been reported to be associated with dephosphorylation under oxidative stress in vitro. In addition, it has been shown that when Cys224 of Akt1/PKBα is mutated to a Ser residue, the kinase becomes resistant to NO donor-induced S-nitrosylation and inactivation; suggesting that this residue is a major S-nitrosylation acceptor site. In vivo S-nitrosylations of the insulin receptor β and Akt1/PKBα result in reductions in their kinase activities.
These data suggest that the redox status of Akt1/PKBα, regulated by NO, is a second factor in the PTM that modulates kinase activity (via dynamic conformational changes) and thus GLUT-4 trafficking and protein synthesis. Nevertheless, to date, published data on the reversible phosphorylation(s) and S-nitrosylation(s) relevant to Akt1/PKBα activation, conformation and regulation have not provided conclusive information concerning their interrelationships nor critical S-nitrosylation sites involved in the kinase activation/deactivation cycle.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Recent technical developments have made it feasible to study the molecular details of these important processes. These techniques include: i) sensitive and site-specific procedures for the detection of S-nitrosylation based upon nano-LC interfaced with tandem MS; ii) the Biotin-Switch method for qualitative discrimination of the thiol state between free, disulfide bonded and S-nitroylated cysteine residues under carefully defined conditions. Potential problems related to quantification with this technique have been discussed previously; and iii) highly specific anti-Akt1/ PKBα mAbs that can be used to immunoprecipitate quantities of protein that are sufficient to yield SDS-PAGE bands with Coomassie brilliant blue R-250 staining which are compatible with tandem MS analysis.[](https://www.ncbi.nlm.nih.gov/mesh/D001710)
Burn injury-associated impairments in IRS1 signaling and attenuated IR-IRS-PI3K-Akt/PKB activation have been the major focuses of our research team. Significantly reduced phosphorylations of Ser473 and Thr308, as well as decreased Akt/PKB kinase activity were observed after burn injury [55% total body surface area (TBSA), day 3] and insulin stimulation. However, the interrelationship between impaired kinase activity and the loop disulfide bond reported under oxidative stress remains unclear. In the present study we investigated the interaction between S-nitrosylation and phosphorylation at Cys296-Lys297 and Thr308-Phe309-Cys310 in the kinase loop at the proteomic level.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
Specifically, the following issues need to be studied: i) the ability of Cys296 to chemically quench elevated levels of free radicals, mainly nitric oxide; ii) loop conformational changes associated with two types of PTMs; iii) quantitative proteomics of Akt1/PKBα by stable isotope labeling in mice. In this study, we obtained MS/MS sequence data to characterize the thiol states of Cys296 in the kinase activity loop of Akt1/PKB. These measurements were possible despite the extremely low level of nitrosylated protein (at the 10−15 pmol level, the chance of positive hits is ∼25% with lysates prepared from 25 mg of soleus muscle). The biochemical role of S-nitrosylation at Cys296 was characterized as an intermediate state which reduces the kinetic barrier to form the disulfide bond with Cys310 within the activity loop.
This occurs simultaneously with dephosphorylation of pThr308 after burn injury. The facts that no other disulfide bonds associated with Cys296 were detected suggest that they may be thermodynamically forbidden; due to geometry and/or dihedral strain. The data obtained with soleus muscle from burned and sham-treated rats indicates that NO-mediated formation of the Cys296-Cys310 disulfide bond (which likely downregulates kinase activity) plays a reciprocal role with formation of a Lys297-pThr308 salt bridge (which upregulates kinase activity) during disease-associated reversible activation/deactivation processes.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## Materials and methods
*In the **Materials and methods** section:*
## Chemicals
*In the **Chemicals** section:*
Acetonitrile (ACN, LC-MS Chromasolv), formic acid (FA), glacial acetic acid, LC-MS grade water, dithiothreitol (DTT), iodoacetic acid, iodoacetamide, [Glu1]-fibrinopeptide B, methyl methanethiolsulfonate (MMTS), S-nitrosoglutathione (GSNO), sodium L-ascorbate, neocuproine, N,N-dimethylformamide (DMF), dimethyl sulfoxide (DMSO) were obtained from Sigma Chemical Co. (St. Louis, MO). SDS-PAGE Ready gels (4-15% Tris-HCl, cat. no. 161-1122), Laemmli sample buffer (cat. no. 161-0737) and Coomassie brilliant blue R-250 (no. 161-0436) were obtained from Bio-Rad. Trypsin profile IGD kits (cat.
no. PP0100) were obtained from Sigma. Anti-Akt1/PKBα monoclonal antibody (cat. no. 05-798; lot, 26860) and inactive Akt1/PKBα (cat. no. 14-279) were purchased from Upstate (Charlottesville, VA, USA). Streptavidin agarose CL-4B (cat.
no. 85881) was a product of Fluka (Milwakee, WI, USA). HPDP-Biotin (cat. no. 21341) and Iodoacetyl-LC-Biotin (cat. no. 21333) were purchased from Pierce (Rockford, IL, USA).[](https://www.ncbi.nlm.nih.gov/mesh/C032159)
## Mapping of cysteine residues in inactive Akt1/PKBα
*In the **Mapping of cysteine residues in inactive Akt1/PKBα** section:*
Inactive Akt1/PKBα (10 μg, 0.18 nmol, in 10 μl stock solution) was transferred to a siliconized Eppendorf tube (0.6 ml) containing Laemmli sample buffer (2X, 10 μl, pH was adjusted to 8.0) and DDT (2 μl, 20 nmol, PBS, pH 8.0), and the solution was kept at 95°C for 5 min. Freshly prepared Iodoacetyl-LC-Biotin (15 μl, 55 nmol, in DMF) was added to the denatured protein solution followed by stirring for an additional 15 min at room temperature. The resulting biotinylated Akt1/PKBα was purified by SDS-PAGE and stained with Coomassie brilliant blue R-250. The protein bands were excised (∼1 mm size) and digested (Akt1/PKBα: trypsin 25, overnight at 37°C) with a Trypsin Profile IGD kit according to the manufacturer's instructions. The biotinylated peptide mixture was captured by gentle stirring with streptavidin agarose CL-4B (30 μl packed) at room temperature for 1 h (final vol, 100 μl). The streptavidin beads were washed with PBS (0.5 ml ×3), followed by water/ acetonitrile (ACN 10%, 0.5 ml ×3).
Biotinylated peptides were released from the streptavidin beads with formic acid (70%, 100 μl) at room temperature for 15 min with brief vortexing. The supernatant containing biotinylated peptides was transferred to a new vial and the formic acid was evaporated with a SpeedVac. The biotinylated peptide mixture was resuspended in water/acetonitrile (ACN, 2%, with 0.1% FA, 70 μl), and the aliquots (10 μl) were injected into a Waters CapLC-tandem quadrupole time-of-fight mass spectrometry (Q-TOF) system.[](https://www.ncbi.nlm.nih.gov/mesh/C088816)
## Identification of disulfide bonds in inactive Akt1/PKBα
*In the **Identification of disulfide bonds in inactive Akt1/PKBα** section:*
Inactive Akt1/PKBα (10 μg, 0.18 nmol, in 10 μl stock solution) was transferred into a siliconized Eppendorf tube (0.6 ml) containing Laemmli sample buffer (2X, 10 μl, pH 8.0) and iodoacetamide (2 μl, 20 nmol, PBS, pH 8.0). The mixture was maintained at 95°C for 5 min and then stirred at room temperature for an additional 15 min. The Akt1/PKBα was purified by SDS-PAGE and stained with Coomassie brilliant blue R-250. The protein bands were processed as above.[](https://www.ncbi.nlm.nih.gov/mesh/C088816)
## Identification of NO acceptor sites in inactive Akt1/PKBα
*In the **Identification of NO acceptor sites in inactive Akt1/PKBα** section:*
Three samples of inactive Akt1/PKBα (10 μg, 0.18 nmol, in 10 μl stock solution) were treated with GSNO (250 nmol, 50 μl PBS, pH 8.0, 200-fold excess/thiol group) for 1 h at room temperature in the dark in siliconized Eppendorf tubes (0.6 ml). Separation of Akt1/PKBα and GSNO was achieved by two successive acetone/water precipitations (0.3 ml, 70% ACN) at −40°C for 10 min. The supernatants (containing GSNO) were removed by centrifugation at 14,000 × g for 2 min. The kinase pellets were resuspended in blocking buffer (100 μl, 20 mM Tris-HCl, pH 7.7, 2.5% SDS, 20 mM MMTS, 1 mM EDTA, 0.1 mM neocuproine) at room temperature for 1 h with gentle stirring (1 mm ID ×5 mm bar). Excess MMTS was removed by acetone (100%, 0.3 ml) precipitation (as above), and the protein pellets were resuspended in PBS (50 μl, pH 8.0).
Freshly prepared iodoacetic acid (5 μl, 2 mM in PBS, pH 8.0), HPDP-Biotin (5 μl, 2 mM in DMSO), Iodoacetyl-LC-Biotin (5 μl, 2 mM in DMF) and sodium ascorbate (20 μl, 5 mM, PBS) were added to the three vials containing nitrosylated Akt1/PKBα, respectively. The reaction mixtures were stirred at room temperature for 15 min (iodoacetic acid and Iodoacetyl-LC-Biotin) or 1 h for the thiol-disulfide exchange reaction. Aliquots of SDS sample buffer (2X, with 5% 2-mercaptoethanol, 50 μl) were added to the protein solutions, and the mixtures were incubated at 95°C for 5 min. The derivatized proteins were processed as above. Carboxymethyl cysteine (CMC)-containing peptides, were neutralized with FA (5 μl) and sequenced via parent ion discovery trigged by the CMC immonium ion (134.02±0.05 mDa) as reported previously.
Biotinylated peptides were sequenced with data-dependent acquisition after capture with streptavidin agarose beads. Ten-microliter aliquots of each final solution were injected into the CapLC-Q-TOF system.[](https://www.ncbi.nlm.nih.gov/mesh/D026422)
## Analysis of the Cys296-Cys310 disulfide bond formation in Akt1/PKBα after treatment with S-nitrosoglutathione
*In the **Analysis of the Cys296-Cys310 disulfide bond formation in Akt1/PKBα after treatment with S-nitrosoglutathione** section:*
Inactive Akt1/PKBα (10 μg, 10 μl, 0.18 nmol) and freshly prepared GSNO (5 μl, 250 nmol, PBS, pH 8.0) were stirred in an Eppendorf tube (0.6 ml) in the dark at room temperature for 1 h. Separation of Akt1/PKBα and GSNO was performed with acetone/water (70%) as above. The kinase pellet was resuspended in PBS (10 μl), and SDS sample buffer (10 μl with iodoacetamide, 20 nmol) was added. The cysteine alkylation was performed at room temperature for 15 min. The protein samples were separated with SDS-PAGE Ready gels and digested as above. Aliquots of the final solution (10 μl) were injected into the CapLC-Q-TOF system.[](https://www.ncbi.nlm.nih.gov/mesh/D026422)
## Measurement of the free and disulfide bonded Cys296 in Akt1/PKBα from soleus muscle of burned rats
*In the **Measurement of the free and disulfide bonded Cys296 in Akt1/PKBα from soleus muscle of burned rats** section:*
Soleus muscle lysates from rats with third degree burn (40% TBSA) were prepared as previously described. The lysates (∼10 mg/ml total proteins) were diluted to ∼3-5 mg protein/ml protein with PBS, and filtered through 0.22-μm membranes. Immunoprecipitation was performed as follows. Anti-Akt1/PKBα mAb (clone AW24, 5 μg; Upstate) and prewashed protein G agarose beads (50 μl, packed) were kept at 4°C (100 μl of PBS) for 1 h under gentle stirring. Without washing the beads, the soleus lysates (5 ml) were added and stirring was continued for an additional 90 min. Non-specific proteins were removed by washing with PBS (3X), Laemmli sample buffer (50 μl, pH 8) containing HPDP-Biotin (400 μM) was added and the mixtures were maintained at 95°C for 5 min.
The procedures for SDS-PAGE separation and in-gel trypsin digestion were the same as described above.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
The burn injury protocol was approved by the Committee on Research Animal Care and Use of the Massachusetts General Hospital (MGH). The MGH animal care facility is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care.
## LC-MS/MS analysis
*In the **LC-MS/MS analysis** section:*
All experiments were performed using a Waters CapLC-Q-TOFmicro system (Waters Corporation, Milford, MA, USA) as previously described. An analytical column (75 mm ID ×150 mm, C18 PepMap300, 5 mm, LC Packings) was used to connect the stream select module of the CapLC with the voltage supply adapter for ESI. Peptide mixtures were loading onto the precolumn (C18 resin) at a flow rate of 15 μl/min. Dead volume from the CapLC injector to the precolumn was measured to be ∼1.5 μl. After washing with mobile phase C (auxiliary pump, 0.1% formic acid in water/ACN, 2% ACN) for 2 min, the trapped peptides were back-washed from the precolumn onto the analytical column using the 10-position stream switching valve. Freshly prepared mobile phases A and C were sonicated under vacuum for ∼25 min, and mobile phase B was treated in this way for 5 min.
The mobile phases were degassed every week, and the CapLC pumps were wet primed for 20 cycles. A linear gradient was used to elute the peptide mixture from mobile phase A (0.1% FA in water/ACN, 2% ACN) to mobile phase B (0.1% FA in ACN). The gradient was segmented as follow: isocratic elution with 2% B for 3 min, increasing B from 2 to 70% (3-40 min), isocratic elution with 70% B (40-45 min) and decreasing B from 70 to 2% (over 2 min). The injector syringe (25 μl) was washed with degassed mobile phase A, and the injection volume was set as full loop mode (10 μl). The gradient flow rate was set at 1.5 μl/min before the 16/1 Nanotee splitter and the pressure drop from the analytical column was ∼800 psi.
The pressure drop (or the flow splitting ratio) was adjusted and maintained with 20 μm ID capillary tubing at the waste outlet position of the Nanotee splitter. The gradient flow rate was ∼95 nl/min. The electrospray voltage was set to ∼3,000 V to obtain an even ESI plume at the beginning of the gradient (high water content). As a routine sensitivity check, the PicoTip Emitter position and other parameters were adjusted to achieve ∼45 counts/sec for the capillary tubing background peak (m/z 429). Sample cone and extraction cone voltages were set at 45 and 3 V, respectively.
The instrument was operated in positive ion mode with the electrospray source maintained at 90°C. The instrument was calibrated with synthetic human [Glu1]-fibrinopeptide B (100 fmol/μl in acetonitrile/water, 10:90, 0.1% formic acid, v/v) at an infusion rate of 1 μl/min in TOF MS/MS mode. The peptide was selected at m/z 785.8 and focused into the collision cell containing argon gas at ∼3×10−5 Torr; the collision energy was set at 35 V. Instrument resolution for the [Glu1]-fibrinopeptide B parent ion, m/z 785.84, was found to be 5,250 FWHM. All data were acquired and processed using MassLynx 4.1 software. For parent ion discovery triggered by the CMC immonium ion (134.02±0.03 Da), the survey low and high collision energies were set at 5 and 30 V, respectively.
MS survey data were collected in continuum mode over the m/z 100–1,200 range. Data-dependent acquisition (DDA) was set from 450 to 1,500 m/z for the biotinylated peptides. Scan time was in the range of 1.9–3.8 sec (depending upon sample conditions), and the inter-scan delay was 0.1 sec. MS to MS/MS switch criteria were dependent upon the reporter ion intensity (5 counts/sec) and detection window (2.3 Da, charge status). The instrument was switched from MS/MS back to MS after 5 sec without intensity restriction.[](https://www.ncbi.nlm.nih.gov/mesh/D010455)
## Evaluation of the S-nitrosylated cysteine site
*In the **Evaluation of the S-nitrosylated cysteine site** section:*
Confirmations of the S-nitrosylation sites were performed by the following three step procedure. i) For parent ion discoveries by continuum MS survey, the peptide mass tolerance was 0.2 Da for the CMC immonium ion. Under these conditions, only a few false positive ions were observed and these were eliminated manually from the expected CMC parent ion list. ii) The positively discovered parent ions were analyzed with PepSeq of MassLynx V4.1 software; oxidation of methionine was searched as a variable modification.
iii) For peptides, with MS/MS scores <35, manual interpretations of candidate parent ions were performed with the following procedure: continuum MS/MS spectra were smoothed, the upper 80% was centroided and cysteine residues were confirmed with three different thiol-specifically derivatized y ions. Cysteine residue monoisotopic mass C3H5NOS = 103.01 Da was replaced with CMC residue monoisotopic mass C5H7NO3S = 161.01 Da, HPDP-Biotin derivatized adduct residue monoisotopic mass C22H37N5O4S3 = 531.20 Da and Iodoacetyl-LC-Biotin derivatized adduct residue monoisotopic mass C21H35N5O4S2 = 485.21 Da, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D010455)
## Results and Discussion
*In the **Results and Discussion** section:*
It has been reported that NO production is elevated by stressors such as burn injury and in patients with type 2 diabetes. It has also been shown that the Cys297-Cys311 disulfide bond in the critical kinase activation loop of Akt1/PKBα may be formed in association with dephosphorylation under oxidative stress in vitro. Thus, we hypothesized that reversible S-nitrosylation at either Cys296 or Cys310 in the kinase active loop may be a second PTM factor which complements reversible phosphorylation at Thr308 in the regulation of kinase activity and we sought to determine how S-nitrosylation interacts with phosphorylation during the Akt1/PKBα activation cycle. To address these issues, GSNO was used as the only NO donor in a model S-nitrosylation system to randomly target the seven cysteine residues of the kinase at pH 8. Vicinal Cys296 and Cys310 take advantage of the pKa for dissociation of the thiol to thiolate, and these electron-rich thiolate groups can lead to formation of an intradomain disulfide bond.
Under these conditions, intracellular free cysteine residues, and cysteines at the kinase surface without interactions or located in hydrophobic environments (i.e. high pKa), are unlikely to be affected by GSNO. In contrast, Cys296 and Cys310, which may have low pKa values due to weak interactions with vicinal residues inside the loop, are potential S-nitrosylation sites as predicted from the 3D structure of the kinase. NO donors, such as thioredoxin and thiol/disulfide oxidoreductases were excluded from the system to prevent possible interferences; however, a small amount of 2-mercaptoethanol (∼0.05% v/v) was necessary to prevent oxygen effects.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
The simple, but well-defined, S-nitrosylation reaction model was used to probe for particular NO acceptor sites in human Akt1/PKBα (inactive, 89% pure containing 2-mercaptoethanol and EGTA; Upstate) in three steps. i) Mapping of all cysteine residues with DTT reduction, Iodoacetyl-LC-Biotin alkylation and affinity capture provided relative MS ionization efficacies and charge states. ii) Detection of disulfide bonds with and without GSNO, provided an understanding of NO-mediated disulfide bond formation. The concentrations of the NO donor used here were similar to the levels used in reported studies. iii) MS/MS pinpointed the S-nitrosylated sites with three different thiol-specific derivatives.
As indicated above, false-negatives may occur with the Biotin-Switch method, whereas false-positives are more common with the other methods; however, thiolether derivatives can be identified with MS/MS data. The findings of these studies were used to study the biological consequences of S-nitrosylation of Akt1/PKBα in soleus muscle from burned rats. This in vivo system was used because soleus muscle is an insulin-sensitive tissue with high levels of IRS-1.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
A base peak intensity (BPI) nano-LC chromatogram of all seven affinity captured cysteine residues that were biotinylated with Iodoacetyl-LC-Biotin is shown (Fig. 1A). Cysteine residue monoisotopic mass of C3H5NOS = 103.01 Da was replaced with derivatized Cys residue monoisotopic mass of C21H35N5O4S2 = 485.21 Da. The relative simplicity of the nano-LC chromatogram indicates the high purification efficacy for removing non-biotinylated tryptic peptides from streptavidin agarose beads. Three predominate TOF MS tryptic parent ions were identified; m/z 639.79 (T41, M+2H+ = 639.83) eluting at 50.5 min, m/z 1088.49 (T9, M-CH4+2H+ = 1088.03) eluting at 51.5 min and m/z 924.67 (T44, M+3H+ = 924.43) eluting at 53 min are doubly and triply charged tryptic peptides containing Cys296, Cys310 and Cys60, respectively. Fig.
1B shows the parent ions co-eluting at ∼53 min as well as the charge state assignments. Parent ions m/z 924.67 (T44, M+3H+ = 924.43) and m/z 1386.51 (T44, M+2H+ = 1386.14) are triply and doubly charged ions from the same tryptic peptide, 308TFCGTPEYLAPEVLEDNDYGR328, which contains Cys310. Parent ion m/z 1266.09 (T58, M+3H+ = 1266.41) is triply charged and derived from the peptide, 437YFDEEFTAQMTITPPDQDDSMECVDSER465, which contains Cys460. Parent ion m/z 815.87 (T11, M+2H+ = 815.93) is doubly charged from the peptide, 77CLQWTTVIER86, which contains Cys77. Parent ion m/z 1088.49 resulted from CH4 neutral loss from m/z 1096.48.
Fig. 1C shows TOF MS parent ions that co-eluted at ∼50.8 min; chromatographic peak tailing the most intense peak at 50.5 min. Parent ions m/z 731.33 (T9, M+3H+ = 731.03) and m/z 1096.46 (T9, M+2H+ = 1096.04) are triply and doubly charged ions from the same tryptic peptide, 49ESPLNNFSVAQCQLMK64, which contains Cys60. Parent ion m/z 639.79 (T41, M+2H+ = 639.83) is a doubly charged ion from the tryptic peptide, 290ITDFGLCK297, which contains Cys296. Fig. 1D shows the TOF MS parent ions that co-eluted at ∼53.5 min. Parent ion m/z 829.00 (T45, M+3H+ = 829.05) is triply charged and derived from the tryptic peptide, 329AVDWWGLGVVMYEMMCGR346, which contains Cys344.
Parent ion m/z 872.70 (T32, M+3H+ = 872.43) is triply charged and derived from the tryptic peptide, 223LCFVMEYANGGELFFHLSR241, which contains Cys224. No doubly charged T58, T45 or T32 ions were observed. It is clear that the ionization efficacies for the peptides containing Cys296 (M+2H+), Cys310 (M+2H+ and M+3H+), Cys60 (M+2H+ and M+3H+) and Cys77 (M+2H+) are much higher than for the triply charged peptides containing Cys460 (M+3H+), Cys334 (M+3H+) and Cys224 (M+3H+) under the same conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
When Akt1/PKBα was treated with GSNO without cleavage of disulfide bonds and the free cysteine residues were alkylated with iodoacetamide, two intradomain disulfide bonds were identified: Cys60-Cys77 in the PH domain and Cys296-Cys310 in the kinase active loop. The monoisotopic parent ion with m/z 821.35, shown in Fig. 2A, represents two tryptic peptides containing the Cys296-Cys310 disulfide bond in the kinase loop. The isotopic peaks at m/z 821.61 and m/z 821.35 are attributed to the M+1 and M+0 ions. A mass difference of 0.26 Da (expected 0.25 Da) indicated four positive charges: two at N-terminals and two at side chains of the C-terminals of the dipeptides.
The expected quadruply charged disulfide bond linked Cys296 and Cys310-containing peptides (T41-SS-T44, M+4H+) were calculated to be m/z 821.38 [(894.45 + 2387.06 + 4)/4]. The monoisotopic parent ion with m/z 764.41, shown in Fig. 2B, represents the two tryptic peptides containing the Cys60-Cys77 disulfide bond in the PH domain. The quadruply charged state is calculated as m/z 764.66 (M+1) - 764.41 (M+0) = 0.25 which indicates four positive proton charges.
The quadruply charged disulfide bond linked Cys60 and Cys77 containing peptides (T9-SS-T11, M+4H+) are calculated as m/z 764.37 [(1806.86 + 1246.63 + 4)/4]. Without GSNO treatment, only the Cys60-Cys77 disulfide bond was detected. The mass accuracies for the two measurements were found to be 36 ppm (Cys296-Cys310 disulfide bond linked dipeptides) and 78 ppm (Cys60-Cys77 disulfide bond linked dipeptides). The impact of GSNO on Cys296-Cys310 disulfide bond formation is demonstrated in Fig. 2C and D. The S-nitrosylation reaction without GSNO (Fig.
2C) shows the triply charged tryptic peptide, 308TFCGTPEYLAPEVLEDNDYGR328, [carboxyamidomethyl cysteine (CAM) derivative] containing Cys310 at m/z 815.99 (expected monoisotopic parent ion, 816.03). The observed M+1 isotopic peak was at m/z 816.33. The difference between the isotopic M+1 and M+0 peak of 0.34 Da indicates three proton charges. In contrast, the triply charged ions at m/z 821.31 and 821.65 (difference = 0.31 Da) do not represent the quadruply charged Cys296-Cys310 dipeptides in Fig. 2C.
The triply charged Cys310-containing peptide was found to be totally absent with GSNO treatment as shown in Fig. 2D. The doubly charged ions at m/z 816.35 and 816.85 (difference = 0.50 Da) are not related to the triply charged tryptic peptide 308TFCGTPEYLAPEVLEDNDYGR328 (CAM derivative) containing Cys310 at m/z 815.99 as shown in Fig. 2C.
In contrast, the ions at m/z 821.33 and 821.58 (difference = 0.25 Da) are indeed from quadruply charged Cys296-Cys310-linked dipeptides. Since quadruply charged Cys296-Cys310-linked dipeptides are formed at the expense of triply charged Cys310-containing peptide after GSNO treatment, it is obvious that S-nitrosylation and disulfide bond formation occur simultaneously in the kinase loop.[](https://www.ncbi.nlm.nih.gov/mesh/D026422)
We next sought to determine which cysteine residue is the NO acceptor that initializes Cys296-Cys310 disulfide bond formation. There are three possibilities for the two cysteine residue thiol states: single S-nitrosothiol, double S-nitrosothiols and nitroxyl disulfide. The last case (nitroxyl disulfide) can be ruled out from the list, since the expected net mass increases of 28 Da (NO - 2H = 30 - 2 Da) were not observed for the corresponding dipeptides. The second case, double S-nitrosothiols of Cys296 and Cys310, may occur if both pKa values are acidic inside the kinase loop. The Biotin-Switch method was used to identify the S-nitrosothiol within the loop under gentle reaction conditions (GSNO 250 nmol, 1 h).
In addition, two other thiol-specific reagents, iodoacetic acid and Iodoacetyl-LC-Biotin (leaving molecule: HI, fast and quantitative), were evaluated.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Table I shows the expected results of Cys296 S-nitrosylation in the kinase loop with the three different chemical modifications. The resulting S-nitrosylated Cys was reduced with ascorbate and then derivatized with iodoacetic acid to afford the CMC derivative (the Cys residue with a monoisotopic mass C3H5NOS = 103.01 Da was replaced by the CMC residue with a monoisotopic mass C5H7NO3S = 161.01 Da) for sequence analysis. The CMC derivative of the y2 ion of the doubly charged tryptic peptide, 290ITDFGLCK297, was confirmed at m/z 308.17 (expected 308.13 = 161.01 + 145.10+ 2.02). The Cys HPDP-Biotin adduct (Cys residue monoisotopic mass C3H5NOS = 103.01 Da was replaced with the adduct residue monoisotopic mass C22H37N5O4S3 = 531.20 Da) was used for sequence analysis. The corresponding y2 ion of the Biotin-HPDP derivatized, 290ITDFGLCK297, was confirmed at m/z 678.29 (expected 678.32 = 531.20 + 145.10 + 2.02).
The Cys Iodoacetyl-LC-Biotin adduct (Cys residue monoisotopic mass C3H5NOS = 103.01 Da was replaced with adduct residue monoisotopic mass C21H35N5O4S2 = 485.21 Da) was used for peptide sequence analysis. The corresponding y2 ion of Iodoacetyl-LC-Biotin derivatized, 290ITDFGLCK297 was confirmed at m/z 632.38 (expected 632.33 = 485.21 + 145.10 + 2.02). Since the y2 ions of 296Cys-Lys297 produced with the three different derivatization procedures were unambiguously observed it is likely that Cys296 is a favorable S-nitrosylation site under the conditions used. Although studies with mutated Akt1/PKBα (Cys224) indicated that Cys224 is a major S-nitrosylation acceptor site in vitro, the biological role of S-nitrosylated Cys224 in kinase regulation needs to be further explored. In the current study it was determined that significant S-nitrosylation of Cys224 is improbable, since using the three alkylation approaches and trypsin digestion, the levels of positive ionization of Cys224-containing peptides were below the level of detection.
This failure in detection of S-nitrosylated Cys224 may be a false-negative under our experimental conditions and clearly warrants further investigation. Nevertheless, our findings clearly demonstrate that S-nitrosylated Cys296 is directly relevant to the kinase activation regulation cycle.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
One possible explanation for the kinetics of Cys296-Cys310 disulfide bond formation in the kinase loop may be that there is a high kinetic barrier without GSNO. Due to its highly labile nature, S-nitrosylated Cys296, which forms rapidly in the presence of GSNO, may function as an intermediate state. Since this intermediate is likely to have a lower kinetic barrier for Cys296-Cys310 disulfide bond formation, the overall speed of the reaction should increase greatly. It has been reported that trans-nitrosylation reactions between vicinal thiols can occur and accelerate disulfide bond formation. The well characterized Cys296-Cys310 disulfide bond can be used as a signature peptide for detection of S-nitrosylation of Cys296 after immunoprecipitation.
The separation of tryptic peptide mixtures with our nano-LC interfaced Q-TOFmicro is demonstrated in Fig. 3 (bottom panel). The extracted mass ion peak m/z 821.62, as shown in Fig. 3 (top panel), is the M+1 isotopic peak of the quadruply charged dipeptides (the most intense isotopic peak due to a high number of carbon atoms).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
The in vitro system allowed us to determine conditions that are favorable for evaluation of S-nitrosylation of Cys296 by MS/MS and was useful for studying the mechanism of intradomain disulfide bond formation. The reason for using inactive Akt1/PKBα (unphosphorylated) in these studies was to find possible S-nitrosylation sites in relationship with the following published data: i) Akt1/PKBα undergoes transient phosphorylation/dephosphorylation which regulates the kinase activity conformation cycle; ii) kinase disulfide bond formation, Cys297-Cys311, and dephosphorylation at pThr308 are induced simultaneously by H2O2 oxidative stress in vitro; iii) high levels of nitric oxide production occur both after burn injury and in diabetic patients. Previous results from our laboratory have indicated that there is S-nitrosylation at Cys296 in rat soleus muscle. A parent ion at m/z 690.83 containing Cys296 (T41-T42: 290ITCFGLCKEGIK301) was observed with CAM immonium trigged parent ion discovery; however, MS/MS sequencing data were not obtained. As a continuation of these studies to explore S-nitrosylation in the kinase active loop, large amounts of rat soleus muscle lysate (∼3-5 mg/ml total proteins, 3 ml for each experiment, day 4 after 40% TBSA, 3rd degree burn) were used.
In the present study, detailed MS/MS analyses of HPDP-biotinylated free Cys296 peptide and Cys296-Cys310 disulfide bound dipeptides of Akt1/PKBα were performed with lysates of rat soleus muscle after burn injury. The tryptic parent ion derivatized from free Cys296 after burn injury was observed at m/z 662.84 (M+2H+, expected 662.82) and the MS/ MS sequence data are shown in Fig. 4. A low sequence score of 18 was obtained from the parent ion with S/N = 3. However, the critical diagnostic y2, y4 and y5 ions at m/z 678.29, 849.34 and 995.51 confirmed that trace amounts of free Cys296 are indeed present after intradomain disulfide bond formation induced by burn injury.
In addition, partial sequencing data for Cys296-Cys310 disulfide-linked dipeptides are shown in Fig. 5. The C-terminal y ion series of Cys310-containing peptide, 308TFCGTPEYLAPEVLEDNDYGR328, was observed for the quadruply charged parent ion (T41-SS-T44, M+4H+). Cys296-Cys310 disulfide-linked dipeptides were not observed in muscle lysates from sham-treated animals (negative controls). The chance of obtaining the MS/MS sequence using our in vivo experimental conditions is only ∼20–25%.
This indicates that one interpretable MS/MS outcome (score >25) is expected in four or five independent experiments in which three successive injections are performed. Nevertheless, these MS/MS data for peptides containing free Cys296 and Cys296-Cys310-linked dipeptides are sufficient to verify our hypothesis that S-nitrosylation promotes intradomain disulfide bond formation and dephosphorylation at pThr308 after burn injury as illustrated in Fig. 6. Due to its high lability of Cys296-SNO, direct identification of this species in vivo was not possible.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
S-nitrosylation of Akt1/PKBα is a key factor for understanding the regulation of glucose transport and downstream protein synthesis. A recent study demonstrated that blockade of iNOS prevents the S-nitrosylations of Akt and IRS-1 and results in insulin resistance in vivo. Although it is clear that two PTMs of Akt1/PKBα, phosphorylation at Thr308 and S-nitrosylation at Cys296, are critical for the regulation of Akt1/PKBα activity under stress conditions, there are still many unanswered questions concerning how reversible phosphorylation/dephosphorylation and S-nitrosylation/denitrosylation modulate Akt1/PKBα activity. For example, it has been reported that the Cys296-Cys310 disulfide bond is present only when there is binding of substrate to the active kinase loop and phosphorylation at Thr308; indicating that both disulfide bond formation as well as phosphorylation of Thr308 are important for kinase activity. In contrast, this disulfide bond was not observed under similar conditions in two studies of the ternary structure of the kinase; even though, oxidative stress was shown to induce dephosphorylation of pThr308 and disulfide bond formation in the kinase loop in an in vitro study.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
In summary, our data establish that Cys296 is an important S-nitrosylation site in the kinase loop of Akt1/PKBα under gentle reaction conditions: i) iodoacetic acid as previously described; ii) the HPDP-Biotin switch method; and iii) the Iodoacetyl-LC-Biotin method to ensure indirect capture of Cys296-SNO which may be undetectable with HPDP-Biotin. The corresponding derivatized y2 ions (296Cys-Lys297) in the tryptic peptide (Ile-Thr-Asp-Phe-Gly-Leu-Cys-Lys) were obtained with mass sequences to eliminate false-positive discovery. Although no other S-nitrosylated cysteine residues were detected, it is possible that S-nitrosylations at Cys224, Cys344 and Cys460 were missed due to very low ionizations (i.e., false-negative discoveries). As a consequence of S-nitrosylation at Cys296, there is rapid disulfide bond formation with vicinal Cys310 in the kinase loop, which alters kinase substrate recognition as well as Akt-FOXO switch. This affords a stable disulfide bond linked quadruply charged parent ion at m/z 821.35 (M+4H+).
Partial sequencing data for Cys296-Cys310 linked dipeptides from soleus muscle lysates indicated that burn injury is associated with both dephosphorylation of pThr308 and disulfide bond formation. These two types of PTMs may provide insights for understanding negative cooperative effects on reduced Akt/PKB kinase activity after burn injury as previously reported by our laboratory. Although our results have provided important mechanistic information, quantitative measurements of Thr308/pThr308 and free Cys296/ SNO-Cys296/bound Cys296 in patients with burn injury and type 2 diabetes remain very challenging.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
# Abbreviations
*In the **Abbreviations** section:*
Akt1/PKBα
Akt1/protein kinase Bα;
CAM[](https://www.ncbi.nlm.nih.gov/mesh/C505896)
carboxyamidomethyl cysteine;[](https://www.ncbi.nlm.nih.gov/mesh/C505896)
CMC[](https://www.ncbi.nlm.nih.gov/mesh/D002233)
carboxymethyl cysteine;[](https://www.ncbi.nlm.nih.gov/mesh/D002233)
GSNO[](https://www.ncbi.nlm.nih.gov/mesh/D026422)
S-nitrosoglutathione;[](https://www.ncbi.nlm.nih.gov/mesh/D026422)
PH
pleckstrin homology;
PTM
post-translational modification;
Q-TOF
tandem quadrupole time-of-fight mass spectrometry;
TBSA
total body surface area
# References
*In the **References** section:*
Mapping of cysteine residues in inactive Akt1/PKBα. (A) Base peak intensity (BPI) nano-LC chromatogram of affinity capture of all seven cysteine residues that were biotinylated with Iodoacetyl-LC-Biotin. Sample preparation: see materials and methods section for details. Column conditions: 75 mm ID ×150 mm, C18 PepMap300, 5 mm, under linear gradient conditions at a flow rate 95 nl/min. (B) TOF MS analysis of parent ions co-eluted at retention time of ∼53 min. Parent ions m/z 924.67 and 1386.51 are triply and doubly charged ions from the same tryptic peptide 308TFCGTPEYLAPEVLEDNDYGR328 which contains Cys310. Parent ion m/z 1266.09 is a triply charged ion from the tryptic peptide, 437YFDEEFTAQMTITPPDQDDSMECVDSER465, which contains Cys460. Parent ion m/z 815.87 is a doubly charged ion derived from the tryptic peptide, 77CLQWTTVIER86, which contains Cys77. The parent ion at m/z 1088.49 results from CH4 neutral loss from m/z 1096.46 as shown in C. (C) TOF MS analysis of parent ions co-eluting at retention time of ∼50.8 min. Parent ions m/z 731.33 and 1096.46 are triply and doubly charged ions from the same tryptic peptide, 49ESPLNNFSVAQCQLMK64, which contains Cys60. Parent ion m/z 639.79 is doubly charged and is derived from tryptic peptide, 290ITDFGLCK297, which contains Cys296. (D) TOF MS analysis of parent ions co-eluting at retention time of ∼53.5 min. Parent ion m/z 829.00 is triply charged and derived from tryptic peptide, 329AVDWWGLGVVMYEMMCGR346, which contains Cys344. Parent ion m/z 872.70 is triply charged and derived from tryptic peptide, 223LCFVMEYANGGELFFHLSR241, which contains Cys224.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Detections of two intradomain disulfide bonds in Akt1/PKBα. (A) Detection of intradomain Cys296-Cys310 disulfide bond in the kinase loop. Inactive Akt1/PKBα (10 µg) was treated with GSNO and iodoacetamide (50 µM) in Laemmli sample buffer as in D. In-gel trypsin digestion was performed after SDS-PAGE separation (4–15% Tris-HCl). Monoisotopic parent ion at m/z 821.35, charge state 4. Expected quadruply charged disulfide linked Cys296 and Cys310 containing the peptide at m/z 821.38. (B) Detection of the intradomain Cys60-Cys77 disulfide bond in the PH domain. Monoisotopic parent ion at m/z 764.41, charge state 4. Expected quadruply charged disulfide bond linked Cys60 and Cys77-containing peptide at m/z 764.37. (C) Free thiol state of Cys310 in the kinase loop without NO donor. The triply charged parent ion m/z 815.99: 308TFCGTPEYLAPEVLEDNDYGR328 (expected: m/z 816.03, CAM derivative) represents the completely free thiol state of Cys310, while the triply charged m/z 821.31 is not from disulfide linked Cys296-Cys310 dipeptides (expected charge state 4). The Cys296-Cys310 disulfide bond was not detected in the absence of the NO donor. (D) Nitric oxide promotes the formation of the Cys296-Cys310 disulfide bond in the kinase loop. Inactive Akt1/PKBα (10 µg) was treated with GSNO (250 nmol, 50 μl PBS, pH 8.0, 1 h at room temperature in dark) prior to alkylation with iodoacetamide and SDS-PAGE. The doubly charged m/z 816.35 ion is not from a Cys310-containing tryptic peptide (expected charge state 3), and quadruply charged m/z 821.33 occurs at the expense of diminished triply charged Cys310 peptide. The free thiol of Cys310 is completely converted into the disulfide bond with Cys296.[](https://www.ncbi.nlm.nih.gov/mesh/D004220)
Nano-LC chromatogram of tryptic peptides of Akt1/PKBα and MS ion 821.62 chromatogram of soleus muscle. Top panel: mass ion chromatogram of the dipeptides m/z 821.62: M+1 isotopic peak of the quadruply charged dipeptides (intensity of M+0 monoisotopic peak is lower than M+1). Bottom panel: BPI chromatogram of the Akt1/PKBα tryptic peptides after immunoprecipitations and in-gel digestion from nano-LC interfaced with Q-TOF tandem mass spectrometry.[](https://www.ncbi.nlm.nih.gov/mesh/D010455)
MS/MS sequence analysis of biotinylted free Cys296 peptide of Akt1/PKBα after burn injury. Rat soleus muscle lysates (30 mg total protein) were treated with anti-Akt1/PKBα mAb and in-gel biotination was performed with HPDP-Biotin. Parent ion m/z 662.84 (M+2H+, expected 662.82) was sequenced. Cys residue monoisotopic mass C3H5NOS = 103.01 Da is replaced with the adduct residue monoisotopic mass C22H37N5O4S3 = 531.20 Da. A low sequence score 18 was obtained from the parent ion with S/N = 3; however the critical diagnostic y2, y4 and y5 ions at m/z 678.29, 849.34 and 995.51 confirmed that trace amounts of free Cys296 are present after burn injury.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
MS/MS sequence analysis of the Cys296-Cys310 disulfide-linked peptide with dephosphoryated Thr308 in soleus muscle from burned rats. Partially sequenced Cys296-Cys310 disulfide-linked dipeptides: C-terminal y ion series (y3 to y9) of Cys310-containing peptide, 308TFCGTPEYLAPEVLEDNDYGR328, were observed from the quadruply charged parent ion (T41-SS-T44, M+4H+).[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
Proposed mechanism for Akt1/PKBα kinase regulation by phosphorylation and S-nitrosylation in the muscle of burned rats. Phosphorylation of Thr308 stabilizes the disordered loop structure between 292DFG and APE319 via a salt bridge with Lys297 as illustrated in the loop Peptide 1, which upregulates Akt1/PKBα kinase activity. NO free radical production is increased after burn injury. A large portion of Cys296 undergoes S-nitrosylation at Cys296 (Peptide 2); however, some free Cys296 remains (Peptide 3). S-nitrosylation activates Cys296-Cys310 intradomain disulfide bond formation (Peptide 4). S-nitrosylation at Cys296 is associated with dephosphorylation of Thr308 and inaccessibility to the kinase site; which downregulates kinase activity.[](https://www.ncbi.nlm.nih.gov/mesh/D013912)
Characterization of the thiol-specifically modified Akt1/PKBα peptide 290ITDFGLCK297.[](https://www.ncbi.nlm.nih.gov/mesh/D013438)
|
# Introduction
Diabetes changes expression of genes related to [glutamate](https://www.ncbi.nlm.nih.gov/mesh/D018698) neurotransmission and transport in the Long-Evans rat retina
# Abstract
*In the **Abstract** section:*
Purpose
This study investigated changes in the transcript levels of genes related to glutamate neurotransmission and transport as diabetes progresses in the Long-Evans rat retina. Transcript levels of vascular en[dothelial](https://www.ncbi.nlm.nih.gov/mesh/D018698) growth factor (VEGF), erythropoietin, and insulin-like growth factor binding protein 3 (IGFBP3) were also measured due to their protective effects on the retinal vasculature and neurons.
Methods
Diabetes was induced in Long-Evans rats with a single intraperitoneal (IP) injection of streptozotocin (STZ; 65 mg/kg) in sodium citrate buffer. Rats with blood glucose >300 mg/dl were deemed diabetic. Age-matched controls received a single IP injection of sodium citrate buffer only. The retinas were dissected at 4 and 12 weeks after induction of diabetes, and mRNA and protein were extracted from the left and right retinas of each rat, respectively. Gene expression was analyzed using quantitative real-time reverse-transcription PCR. Enzyme-linked immunosorbent assay was used to quantify the concentration of VEGF prot[ein in each re](https://www.ncbi.nlm.nih.gov/mesh/D013311)ti[na.](https://www.ncbi.nlm.nih.gov/mesh/D013311) Statistical si[gnificance was](https://www.ncbi.nlm.nih.gov/mesh/D000077559) determined using 2[×2 analysis o](https://www.ncbi.nlm.nih.gov/mesh/D001786)f variance followed by post-hoc analysis using Fisher’s protected least squares differenc[e.](https://www.ncbi.nlm.nih.gov/mesh/D000077559)
Results
Transcript levels of two ionotropic glutamate receptor subunits and one glutamate transporter increased after 4 weeks of diabetes. In contrast, 12 weeks of diabetes decreased the transcript levels of several genes, including two glutamate transporters, four out of five N-methyl-D-aspartate (NMDA) receptor subunits, and all five kainate receptor subunits. Diabetes had a greater effect on gene expression of NMDA and kainate receptor subunits than on the α-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA) receptor subunits, for which only GRIA4 significantly decreased after 12 weeks. VEGF protein levels were significantly increased in 4-week diabetic rats compared to age-matched control rats whereas the increase was not significant after 12 weeks. Transcript levels of VEGF and VEGF receptors were unchanged with diabetes. Erythropoietin and IGFBP3 mRNA levels significantly increased at both time points, and IGFBP2 mRNA levels increased after 12 weeks.
Conclusions
Diabetes caused significant changes in the transcriptional expression of genes related to ionotropic glutamate neurotransmission, especially after 12 weeks. Most genes with decreased transcript levels after 12 weeks were expressed by retinal ganglion cells, which include glutamate transporters and ionotropic glutamate receptors. Two genes expressed by retinal ganglion cells but unrelated to glutamate neurotransmission, γ-synuclein (SNCG) and adenosine A1 receptor (ADORA1), also had decreased mRNA expression after 12 weeks. These findings may indicate ganglion cells were lost as diabetes progressed in the retina. Decreased expression of the glutamate transporter SLC1A3 would lead to decreased removal of glutamate from the extracellular space, suggesting that diabetes impairs this function of Müller cells. These findings suggest that ganglion cells were lost due to glutamate excitotoxicity. The changes at 12 weeks occurred without significant changes in retinal VEGF protein or mRNA, although higher VEGF protein levels at 4 weeks may be an early protective response. Increased transcript levels of erythropoietin and IGFBP3 may also be a protective response.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Diabetic retinopathy, a major complication of type 1 and 2 diabetes, is characterized by damage to the retinal microvasculature, which can eventually lead to impaired vision and blindness. In addition to producing vascular dysfunction in the retina, diabetes also damages the neurons. The purpose of investigating transcriptional gene expression was to concurrently measure changes to the neurons, glia, and vasculature as diabetes progresses in the rat retina. These genes included those related to glutamate neurotransmission and transport. The expression of vascular endothelial growth factor (VEGF), erythropoietin (EPO), and insulin-like growth factor-1 (IGF-1) were also measured since they have neuroprotective properties in addition to their effects on the retinal vasculature.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Glutamate is the predominant excitatory neurotransmitter in the retina. This study focused on ionotropic glutamate receptors, which are divided into three classes based on their affinity for the glutamatergic agonists N-methyl-D-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA), and kainate. In previous work, diabetes was found to alter the expression of selected glutamate receptor subunits in the retina of Wistar rats with up to 4 months of diabetes and in diabetic patients without signs of retinopathy, but an overall pattern of changes was not apparent. The NMDA receptor antagonist memantine was shown to reduce retinal vascular and neuronal changes in a rat model of diabetes; therefore, understanding ionotropic glutamate receptor dysfunction in diabetes may have therapeutic importance. Glutamate transporters are also key constituents in glutamatergic neurotransmission because they regulate the extracellular concentration of glutamate.
The glutamate transporter SLC1A3 (also known as GLAST) takes up extracellular glutamate into Müller cells. Another type, the vesicular glutamate transporters (VGLUTs), mediates glutamate uptake into the synaptic vesicles of excitatory neurons. Previous work showed that diabetes impairs glutamate metabolism and transport in the retina.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
To provide further information on retinal and glial cells, the transcript levels of the neural- and glial-related genes γ-synuclein (SNCG), glial fibrillary acidic protein (GFAP), and adenosine A1 receptor (ADORA1) were measured. SNCG is expressed in retinal ganglion cells. It is implicated in the pathogenesis of breast tumors and Alzheimer disease, but the normal physiologic function of SNCG is unknown. GFAP is an intermediate filament protein expressed in glial cells. Increased protein levels of GFAP, which indicates glial reactivity, were found in one study of early diabetic retinopathy.
ADORA1 is expressed on retinal ganglion cells and blood vessels.
VEGFA and EPO have neuroprotective and angiogenic properties. In addition to being found in retinal blood vessels, VEGF expression has been found in several retinal layers, including the inner nuclear layer (INL), outer nuclear layer (ONL), and ganglion cell layer (GCL), and within the cytoplasm of retinal ganglion cells and glial cells. Similar to VEGFA, EPO has angiogenic properties. It is mainly expressed in the kidney, but the retina also expresses EPO along with the EPO receptor (EPOR). In addition to effects on the vasculature, VEGFA and EPO have neuroprotective effects in the brain and retina.
Similar to VEGF and EPO, the IGF-1 system is involved in angiogenesis and neuroprotection. IGF-1 is normally bound to one of six binding proteins (IGFBP1–6), which prolongs its half-life in the circulation. The IGFBPs also function independently of IGF-1, although their roles have not been fully elucidated. A study of patients with proliferative diabetic retinopathy showed that they had significantly increased IGF-1 and IGFBP2 protein levels in the vitreous.
This study measured transcriptional gene expression in the pigmented Long-Evans rat retina at 4 and 12 weeks of diabetes. The results showed concurrent changes in the expression of genes related to glutamate neurotransmission, glutamate transport, VEGF, EPO, and IGFBPs.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
## Methods
*In the **Methods** section:*
## Induction of diabetes
*In the **Induction of diabetes** section:*
These experiments were approved by the Northwestern University IACUC and conformed to the NIH Guide for the Care and Use of Laboratory Animals. Pigmented Long-Evans rats between 50 and 57 days old (Harlan Laboratories, Madison, WI) were maintained on a 12 h:12 h light-dark cycle, and had access to standard rat chow and water ad libitum. The rats were assigned to four groups: 4-week control rats, 4-week diabetic rats, 12-week control rats, and 12-week diabetic rats. Each group comprised of six rats. Diabetes was induced with a single intraperitoneal (IP) injection of streptozotocin (STZ; Axxora, San Diego, CA; 65 mg STZ/kg rat, 6.5 mg/ml) in 0.05 M sodium citrate buffer (pH 5).
Rats with blood glucose levels greater than 300 mg/dl 2 days after induction were deemed diabetic. Three rats each from the 4-week diabetic and 12-week diabetic groups had blood glucose levels below 300 mg/dl and were reinjected with STZ. None of the rats were treated with insulin. Age-matched control rats received a single IP injection of an equivalent volume of sodium citrate buffer (0.01 ml/g rat). Blood glucose levels were measured from the tail vein 2 days after the injections and weekly thereafter using a Bayer CONTOUR Meter (Bayer HealthCare, Mishawaka, IN).
The meter read “HI” if blood glucose exceeded 600 mg/dl. Those readings were set to 600 mg/dl for averaging. Readings were usually taken in the morning under non-fasting conditions. Rats were also weighed every week.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## Sample collection and preparation
*In the **Sample collection and preparation** section:*
After 4 and 12 weeks of diabetes, the rats were anesthetized with 5% isoflurane and decapitated. Each retina was immediately dissected from the eye as previously described by Winkler. The retina was then frozen on dry ice and stored at −80 °C.[](https://www.ncbi.nlm.nih.gov/mesh/D007530)
## Quantitative real-time reverse-transcription polymerase chain reaction
*In the **Quantitative real-time reverse-transcription polymerase chain reaction** section:*
Total RNA was extracted from the right retinas using RNeasy Lipid Tissue Mini Kit (Qiagen, Valencia, CA). The cDNA was synthesized by reverse transcription of 1 µg RNA primed with oligo(dT) and random 9-mers. The primers were designed using PerlPrimer as previously described. The forward and reverse primers were limited to 18–20 base pairs (bp) in length. The generated amplicon varied from 69 to 110 bp.
The primer sequences and PCR conditions for each gene are given in Appendix 1. The cDNA synthesized from the samples was used as a substrate for quantifying messenger RNA (mRNA) expression levels by quantitative RT–PCR in the presence of SYBR Green (Stratagene, La Jolla, CA). The amount of mRNA of each gene was normalized to acidic ribosomal phosphoprotein (P0) mRNA for each rat. Then, data from the six rats in each group were averaged. Graphs in the Results section show the normalized mRNA levels and relative mRNA scaled to 4-week control rats.[](https://www.ncbi.nlm.nih.gov/mesh/C098022)
## VEGF protein measurements
*In the **VEGF protein measurements** section:*
Protein was extracted from the left retinas by homogenization in lysis buffer (10 mM Tris pH 7.4, 1.0 mM Na3VO4, and 1% sodium dodecyl sulfate) at 95 °C. The lysates were incubated at 95 °C for 5 min. The samples were then centrifuged and the supernatant collected. Protein samples were stored at −80 °C until analyzed. Sodium dodecyl sulfate was removed using Pierce Detergent Removal Spin Columns (Pierce Biotechnology, Rockford, IL).
Total protein concentration was quantified using the Pierce bicinchoninic acid (BCA) Protein Assay Kit (Pierce Biotechnology). The Quantikine Rat VEGF Immunoassay (R&D Systems, Minneapolis, MN) was used to quantify the concentration of VEGF protein in each retina. The antibody in the immunoassay recognized the VEGFA 120 and 164 isoforms. VEGF protein concentration was then normalized to the total protein concentration for each rat.[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
## Statistics
*In the **Statistics** section:*
All values are reported as mean ± standard error of the mean (SEM) unless otherwise stated. A data point was considered an outlier if it was greater than two standard deviations from the mean of the group. Data sets with outliers were GRIN2D, GRIA2, VGLUT2, VGLUT3, insulin-like growth factor binding protein 2 (IGFBP2), and IGFBP3. These data sets were Winsorized at the fifth percentile to minimize the effects of the outlier. The first step in the Winsorization process was to first sort all 24 measurements in a data set from lowest to highest.
Then, the lowest and highest values were replaced with the next value in the data set. Thus, the mRNA levels for the five genes listed above are reported as the Winsorized mean and SEM. The data for IGFBP2 were averaged from two separate qRT-PCR runs. Statistical significance was determined using a two-factorial analysis of variance (ANOVA) with two levels in each factor (2×2 ANOVA) and was defined as p<0.05. The factors for the ANOVA were time point (levels: 4 weeks and 12 weeks) and treatment (levels: control and diabetic).
Fisher’s protected least significant difference was used for post-hoc analysis. StatView (SAS Institute, Cary, NC) was used to perform the statistical analyses.
## Results
*In the **Results** section:*
## Streptozotocin-induced diabetes
*In the **Streptozotocin-induced diabetes** section:*
All the STZ-treated rats exhibited characteristics of diabetes. The rats’ blood glucose levels were over 300 mg/dl and remained consistently hyperglycemic until the animals were euthanized (Figure 1A). Several rats lost weight after STZ treatment, and all the diabetic rats gained weight slower than the age-matched control rats (Figure 1B). They also showed symptoms of polyuria. The age-matched control rats had normal glucose levels, consistently gained weight until euthanized, and showed no signs of polyuria.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
Blood glucose and weight. Weekly measurements of (A) blood glucose and (B) weight for control and diabetic rats combined for 4- and 12-week time points (mean±SD).[](https://www.ncbi.nlm.nih.gov/mesh/D001786)
## Transcriptomic analyses
*In the **Transcriptomic analyses** section:*
The significant changes in mRNA expression found from post-hoc tests following ANOVA are discussed below. The complete results of the ANOVA are summarized in Appendix 2.
## NMDA receptor subunits
*In the **NMDA receptor subunits** section:*
All the ionotropic glutamate receptors are tetrameric proteins that form cation channels. The NMDA receptor is a heterotetramer formed by two conserved NR1 subunits encoded by the gene GRIN1 and two NR2 subunits encoded by the genes GRIN2A–D. GRIN1 is more abundantly expressed in the retina than the other subunits (Figure 2A). Its expression levels in the 12-week diabetic rats were significantly lower than in the 12-week control rats and the 4-week diabetic rats (p<0.05). The 12-week diabetic rats also had lower GRIN1 mRNA levels than the 4-week control rats, but the difference did not reach significance (p=0.0610). For the genes GRIN2A, GRIN2B, and GRIN2D, the transcript levels were significantly decreased in the 12-week diabetic rats compared to each of the three other groups (p<0.002, p<0.01, and p<0.03, respectively).
The mRNA expression pattern for GRIN2C differed from that of the other NMDA receptor subunits (Figure 2B). GRIN2C was significantly increased in the 4-week diabetic rats compared to the age-matched control rats (p<0.02).
Effect of diabetes on expression of NMDA receptor subunits. qRT-PCR analysis was performed on cDNA isolated from control and STZ-induced diabetic rat retina after 4 and 12 weeks. Expression of each gene was normalized to acidic ribosomal phosphoprotein (P0) for each rat (A), and then scaled to the 4-week control rats for each gene (B; mean ± SEM). Compared to the age-matched control rats, the 12-week diabetic rats had significantly reduced transcript levels of GRIN1, GRIN2A, GRIN2B, and GRIN2D (*, p<0.05; **, p<0.01; *** p<0.005). Transcript levels of GRIN2C were significantly increased in the 4-week diabetic rats compared to the age-matched control rats (*, p<0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## AMPA receptors
*In the **AMPA receptors** section:*
Similar to the NMDA receptors, the AMPA receptors are heterotetramers. Each AMPA receptor is composed of two conserved subunits of GluR2, which is encoded by GRIA2 and is the most abundantly expressed subunit in the retina (Figure 3A). The other two subunits are GluR1, GluR3, or GluR4, which are encoded by GRIA1, GRIA3, and GRIA4, respectively. The AMPA receptor subunits each have two isoforms, flip and flop, which result from alternative splicing of the mRNA transcript. The retina predominantly expresses the flop isoforms of GRIA1, GRIA2, and GRIA4, and the flip and flop isoforms of GRIA3.
The primers for GRIA1, GRIA2, and GRIA4 were not specific for a particular isoform, while the flip isoform of GRIA3 was analyzed. Figure 3B shows the expression patterns of the AMPA receptors scaled to the 4-week control rats for each gene. The 12-week diabetic rats had significantly lower GRIA1 transcript levels than the 4-week diabetic rats (p<0.05). The 12-week diabetic rats also had significantly lower expression of GRIA4 mRNA than each of the three other groups (p<0.01). In contrast, diabetes did not alter the expression of GRIA2 and GRIA3 flip at either 4 or 12 weeks (p>0.02).
Effect of diabetes on mRNA expression of AMPA receptor subunits. qRT-PCR analysis was performed on cDNA isolated from control and STZ-induced diabetic rat retinas after 4 and 12 weeks. Expression of each gene was normalized to acidic ribosomal phosphoprotein (P0) for each rat (A), and then scaled to the 4-week control rats for each gene (B; mean ± SEM). The 12-week diabetic rats had significantly lower GRIA1 mRNA than the 4-week diabetic rats (#; p<0.05). The 12-week diabetic rats also had significantly reduced mRNA levels of GRIA4 compared to the age-matched control rats (**, p<0.01). Diabetes did not affect the mRNA levels of GRIA2 and GRIA3 flip at 4 or 12 weeks.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## Kainate receptors
*In the **Kainate receptors** section:*
The genes GRIK1, GRIK2, GRIK3, GRIK4, and GRIK5 encode for the protein subunits GluR5, GluR6, GluR7, KA1, and KA2, respectively. GluR5, GluR6, and GluR7 can form homomers and heteromers, whereas KA1 and KA2 must complex with GluR5, GluR6, or GluR7 to form a functional receptor. Figure 4A shows the abundance of transcript for each kainate receptor subunit. All five kainate receptor subunits showed similar mRNA expression patterns (Figure 4B) where the 12-week diabetic rats had significantly lower mRNA levels than each of the three other groups (p<0.03).
Effect of diabetes on expression of the kainate receptor subunits. qRT-PCR analysis was performed on cDNA isolated from control and STZ-induced diabetic rat retinas after 4 and 12 weeks. Expression of each gene was normalized to acidic ribosomal phosphoprotein (P0) for each rat (A), and then scaled to the 4-week control rats for each gene (B; mean ± SEM). Compared to the age-matched control rats, the 12-week diabetic rats had significantly reduced mRNA levels of all the kainate receptor subunits (*, p<0.05; **, p<0.01; ***, p<0.005).[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## Glutamate transporters
*In the **Glutamate transporters** section:*
The mRNA expression levels of four glutamate transporters were measured: SLC1A3, VGLUT1, VGLUT2, and VGLUT3 (Figure 5). SLC1A3 is expressed on Müller cells and is responsible for uptake of glutamate for reprocessing. SLC1A3 mRNA levels in 12-week diabetic rats were significantly lower than those of the 4-week control rats and the 4-week diabetic rats (p<0.005 and p<0.001, respectively). The 12-week diabetic rats also had lower SLC1A3 mRNA levels than the 12-week control rats, with the difference approaching significance (p=0.0522). The VGLUTs mediate glutamate uptake into synaptic vesicles.
VGLUT1 mRNA was more abundantly expressed than the two other vesicular transporters (Figure 5A), and its expression was significantly lower in 12-week diabetic rats than each of the three other groups (p<0.05). VGLUT2 expression was significantly decreased in the 12-week diabetic rats compared to the 4-week diabetic rats (p<0.02). VGLUT2 mRNA levels in 4-week diabetic rats trended toward a 1.25-fold increase over the 4-week control rats (p=0.0503). In contrast to its effect on the other glutamate transporters, diabetes had no effect on the mRNA expression of VGLUT3 (p>0.15).[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Effect of diabetes on expression of the glutamate transporters SLC1A3, VGLUT1, VGLUT2, and VGLUT3. qRT-PCR analysis was performed on cDNA isolated from control and STZ-induced diabetic rat retinas after 4 and 12 weeks. Expression of each gene was normalized to acidic ribosomal phosphoprotein (P0) for each rat (A), and then scaled to the 4-week control rats for each gene (B; mean ± SEM). The SLC1A3 mRNA levels in the 12-week diabetic rats were significantly lower than those of the 4-week control rats and 4-week diabetic rats (###, p<0.005; ####, p<0.001). The 12-week diabetic rats also had lower SLC1A3 mRNA levels than the 12-week control rats, but the difference was not quite significant (p=0.0522). The 12-week diabetic rats had significantly lower VGLUT1 mRNA levels than the age-matched control rats (***, p<0.005). VGLUT2 mRNA was significantly higher in the 4-week diabetic rats compared to the 12-week diabetic rats (#, p<0.05). VGLUT2 mRNA levels in the 4-week diabetic rats increased 1.25 fold over the 4-week control rats, but the difference was not quite significant (p=0.0503).[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## SNCG, GFAP, and ADORA1
*In the **SNCG, GFAP, and ADORA1** section:*
The mRNA expression patterns of other neural- and glial-related genes not directly connected to glutamate signaling were also studied. Figure 6A shows the relative transcript levels of each gene in the retina, and Figure 6B shows their expression patterns. The 12-week diabetic rats had significantly lower levels of SNCG and ADORA1 mRNA than each of the three other groups (p<0.01). In contrast, diabetes did not affect GFAP expression at either 4 or 12 weeks of diabetes (p>0.05).[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Effect of diabetes on expression of SNCG, GFAP, and ADORA1. qRT-PCR analysis was performed on cDNA isolated from control and STZ-induced diabetic rat retinas after 4 and 12 weeks. Expression of each gene was normalized to acidic ribosomal phosphoprotein (P0) for each rat (A), and then scaled to the 4-week control rats for each gene (B; mean ± SEM). The 12-week diabetic rats had significantly lower SNCG and ADORA1 mRNA levels than the age-matched control rats (**, p<0.01; ***, p<0.005). Diabetes did not affect the mRNA levels of GFAP at 4 or 12 weeks.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## VEGF, EPO, and their receptors
*In the **VEGF, EPO, and their receptors** section:*
The relative transcript levels for VEGF, EPO, and their respective receptors are shown in Figure 7A. Figure 7B shows the expression patterns. Diabetes did not affect VEGFA mRNA levels (p>0.05). The VEGF receptors FLT1 and KDR significantly decreased after 12 weeks in the control and diabetic rats (p<0.02), which is most likely due to age effects, not diabetes. Diabetes significantly increased EPO mRNA levels 1.49-fold after 4 weeks and 1.51-fold after 12 weeks (p<0.05). Unlike its ligand, EPO receptor (EPOR) expression did not change with diabetes (p>0.15).
Effect of diabetes on expression of VEGF and VEGF–associated genes. qRT-PCR analysis was performed on cDNA isolated from control and STZ-induced diabetic rat retinas after 4 and 12 weeks. Expression of each gene was normalized to acidic ribosomal phosphoprotein (P0) for each rat (A), and then scaled to the 4-week control rats for each gene (B; mean ± SEM). The retina expresses erythropoietin (EPO) at low levels, yet diabetes significantly increased EPO transcript levels (significant main effect for treatment factor; +, p<0.05). The transcriptional expression of VEGFA, the vascular endothelial growth factor (VEGF) receptors FLT1 and KDR, and erythropoietin receptor (EPOR) did not change with diabetes.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## IGF-1 receptor and binding proteins
*In the **IGF-1 receptor and binding proteins** section:*
IGFBP2 transcript levels were much greater than the levels of IGFBP1 and IGFBP3 (Figure 8A). IGFBP2 mRNA levels were significantly higher in the 12-week diabetic rats than in the age-matched control rats and the 4-week diabetic rats (p<0.02). The 4-week diabetic rats trended toward lower IGFBP2 mRNA levels than age-matched control rats (p=0.0532). Although its expression was low in the retina, IGFBP3 mRNA levels in the 4-week and 12-week diabetic rats were significantly increased over their age-matched control rats (p<0.05). The effect of diabetes on mRNA expression at each time point varied among IGF1R, IGFBP1, IGFBP2, and IGFBP3 (Figure 8B).
Diabetes did not alter the mRNA expression of IGF1R or IGFBP1 (p>0.15).
Effect of diabetes on mRNA expression of IGF-1 associated genes. qRT-PCR analysis was performed on cDNA isolated from control and STZ-induced diabetic rat retinas after 4 and 12 weeks. Expression of each gene was normalized to acidic ribosomal phosphoprotein (P0) for each rat (A), and then scaled to the 4-week control rats for each gene (B; mean ± SEM). The 12-week diabetic rats had significantly higher IGFBP2 transcript levels than the age-matched control rats (**, p<0.01). The 4-week diabetic rats had lower IGFBP2 mRNA levels than the 4-week control rats, but the difference did not reach significance (p=0.0532). Although expression of IGFBP3 in the retina was low, the IGFBP3 transcript levels for the 4-week and 12-week diabetic rats were significantly increased over those of the age-matched control rats (*, p<0.05). Diabetes did not affect the mRNA expression levels of IGF1R or IGFBP1 at 4 or 12 weeks.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## VEGF protein levels
*In the **VEGF protein levels** section:*
The diabetic rats had elevated VEGF protein levels compared to the age-matched control rats. Figure 9 shows VEGF protein levels normalized to total protein for each group. The 4-week diabetic rats had 1.5 times the VEGF protein of the age-matched control rats (53.5±4.6 pg VEGF/mg total protein versus 36.1±7.6 pg VEGF/mg total protein, p<0.04). The 12-week diabetic rats also had higher VEGF protein levels at 1.2 times the levels of the age-matched control rats (72.5±4.9 pg VEGF/mg total protein versus 59.6±4.1 pg VEGF/mg total protein), but this difference was not significant (p=0.1025). Total protein levels were not significantly different between the groups (p>0.1).
Effect of diabetes on VEGF protein levels. Total protein was extracted from one retina of each rat. VEGF protein levels were measured with enzyme-linked immunosorbent assay (ELISA) and normalized to total protein for diabetic (black bars) and age-matched control (gray bars) rats at 4 and 12 weeks (mean ± SEM). The asterisk (*) indicates significantly different from age-matched control rats, p<0.05).
## Discussion
*In the **Discussion** section:*
Diabetic retinopathy is clinically defined as injury to the retinal microvasculature. In addition to vascular changes, patients with diabetes demonstrate retinal functional changes, which can appear early in non-proliferative diabetic retinopathy (NPDR) before signs of microvascular injury. Thus, dysfunction in the diabetic retina encompasses vascular and neural changes. This study evaluated genes related to glutamate neurotransmission and transport, and genes that have protective effects on the retinal vasculature and neurons (Table 1). Several studies have shown that diabetes increased apoptosis and ganglion cell loss in the rat retina.
STZ-induced diabetes significantly increased TUNEL–positive cells in Sprague-Dawley rat retinas after 1, 3, 6, and 12 months of diabetes. Another group found significantly lower retinal ganglion cell counts after 4 weeks of diabetes in Brown Norway rats. Likewise, diabetic patients exhibit structural changes to the inner retina early in diabetes. Patients with non-proliferative diabetic retinopathy had a significantly thinner nerve fiber layer (NFL) than non-diabetic control rats. In patients with minimal NPDR, the GCL and the NFL were significantly thinner than in non-diabetic control rats, while the outer retina was unaffected.
Thinning of the GCL and the NFL suggests that ganglion cell injury or death occurs early in diabetes.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Summary of gene expression changes.
The results of this study combined with previous work suggest that the loss of ganglion cells in diabetes may be caused by glutamate excitotoxicity. Most glutamate receptor subtypes have been implicated in excitotoxicity by allowing excessive influx of Ca2+ into neurons. Elevated intracellular calcium levels can trigger various downstream effects, including cell death. Although the exact mechanisms leading from excess glutamate to cell death are not fully understood, Ca2+ influx through NMDA receptors is a key contributor. NMDA receptors are the primary mediators because they are directly coupled to Ca2+ signaling pathways that lead to cell death.
Thus, the pathway of Ca2+ influx through NMDA receptors is pathologically more detrimental than the concentration of intracellular Ca2+.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
## Glutamate transporters (cont.)
*In the **Glutamate transporters (cont.)** section:*
Diabetes was previously found to impair glutamate transport and glutamate recycling in Müller cells. Müller cells maintain a low extracellular concentration of glutamate by taking it up via the transporter SLC1A3, also known as GLAST. The activity of SLC1A3 in Müller cells isolated from Long-Evans rats was reduced after 4 weeks and decreased further after 13 weeks. Consistent with those results, this study found that SLC1A3 mRNA levels were significantly reduced after 12 weeks of diabetes. The changes in SLC1A3 expression are most likely specific to that gene and do not reflect a general loss of Müller cells since the GFAP mRNA levels were not significantly altered.
Within the Müller cells, glutamine synthetase converts glutamate to the less neuroactive glutamine, which is then taken up by neurons and converted to glutamate. The content and activity of glutamine synthetase in the retina decreased after 2, 3, and 6 months of diabetes in Sprague-Dawley rats. These Müller cell dysfunctions in diabetes may lead to accumulation of glutamate in the extracellular space of the retina and contribute to glutamate excitotoxicity.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
In addition to SLC1A3, the expression of VGLUT1 and VGLUT2 transcripts was also altered by STZ-induced diabetes. The main function of the VGLUTs is to load glutamate from the cytoplasm into synaptic vesicles. In the rat retina, VGLUT1 is expressed in photoreceptor and bipolar cell terminals. In this study, VGLUT1 expression was significantly decreased after 12 weeks of diabetes. In contrast, VGLUT2 mRNA was upregulated after 4 weeks of diabetes, but the increase was not sustained after 12 weeks.
VGLUT2 is expressed on horizontal and ganglion cells in the rat retina. VGLUT3 is expressed in non-glutamatergic amacrine cells in the rat retina, and its mRNA expression was not affected by diabetes in this study. Diabetes decreased VGLUT1 and VGLUT2 protein levels in retinal synaptosomes after 2 weeks but not after 8 weeks in Wistar rats. Diabetes affects the expression of VGLUT1 and 2, but more studies are needed to determine the pathological consequences.[](https://www.ncbi.nlm.nih.gov/mesh/D013311)
## Ionotropic glutamate receptors
*In the **Ionotropic glutamate receptors** section:*
Diabetes also altered the expression of the NMDA receptor subunit transcripts. With the exception of GRIN2C, the NMDA receptor subunits showed significantly reduced mRNA expression after 12 weeks of diabetes. However, in contrast to the results of this study, another study using Wistar rats found that GRIN1 mRNA expression was increased after 1 and 4 weeks of diabetes and did not change after 12 weeks, and GRIN2C mRNA levels did not change at any time point. Retinal ganglion cells express the NMDA receptor subunits GRIN1 and GRIN2A-D, but the combination of NMDA receptor subunit expression can depend on the individual cell. GRIN1, GRIN2A, GRIN2B, and GRIN2D are much more likely to be expressed on ganglion cells than GRIN2C.
Conversely, amacrine cells also express GRIN1 and GRIN2A–D but NMDA receptors are not found in all amacrine cells. It is unclear why GRIN2C had a different expression pattern than the other NMDA receptor subunits. GRIN2C mRNA increased after 4 weeks of diabetes, but qRT-PCR of the entire retina cannot distinguish whether ganglion cells or amacrine cells were responsible. The depressed expression of GRIN1, GRIN2A, GRIN2B, and GRIN2D mRNA could indicate downregulation of NMDA receptors but, taken with other evidence, strongly indicates ganglion cell and possibly amacrine cell loss after 12 weeks. Ganglion cells may be more susceptible to glutamate excitotoxicity than other neurons because they are the primary cell type expressing NMDA receptors.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Ganglion cells also express AMPA and kainate receptors, as do other retinal neurons. In situ hybridization studies showed strong labeling for GRIK1, GRIK2, GRIK3, and GRIK5 mRNA in ganglion cells. In the present study, all the kainate receptor subunits were downregulated after 12 weeks of diabetes.
Most of the AMPA receptor subunits exhibited different mRNA expression patterns than the kainate and NMDA receptor subunits, which may be due to differences on which retinal cell types the subunits are expressed. In situ hybridization showed that GRIA2 and GRIA3 were expressed in the cells of the INL, the ONL, and some ganglion cells. In this study, diabetes did not change the expression of GRIA2 and GRIA3. In agreement with these results, the GRIA2 and GRIA3 mRNA levels did not change in the Long-Evans rat retina assessed with in situ hybridization after 2 and 6 weeks of diabetes, nor did they change in the Wistar rat retina after 1, 4, or 12 weeks of diabetes as measured with qRT-PCR. In addition to ganglion cells, photoreceptors, bipolar cells, and amacrine cells express GRIA1 and GRIA3.
Diabetes did not change GRIA2 and GRIA3 mRNA expression possibly because photoreceptors, bipolar cells, and amacrine cells were less likely to be affected by glutamate excitotoxicity than ganglion cells. GRIA1 is expressed predominantly by amacrine cells and bipolar cells, and to a lesser extent by ganglion cells. The mRNA expression of GRIA1was biphasic with an increase after 4 weeks and a decrease after 12 weeks. However, Wistar rats showed no change in GRIA1 transcript levels after 1, 4, and 12 weeks of diabetes. GRIA4 is almost exclusively expressed by ganglion cells.
Its mRNA expression levels were significantly decreased after 12 weeks of diabetes, supporting the conclusion that ganglion cells were lost at that time point in Long-Evans rats.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Although there could be selective downregulation of GRIA4 and specific NMDA and kainate receptor subunits with no loss of ganglion cells, it is more likely that the ganglion cells expressing these receptors were lost by 12 weeks of diabetes. The evidence that glutamate uptake is reduced implicates glutamate excitotoxicity in this process. With continued elevation of glutamate, ganglion cell loss would be expected to continue past the time points investigated here. Other evidence obtained in this study supports the conclusion that ganglion cells were lost in the 12-week diabetic rat retina. SNCG was used as a marker for ganglion cells, and its mRNA levels decreased significantly in the 12-week diabetic rats compared to each of the three other groups.
ADORA1 mRNA levels also decreased significantly in the 12-week diabetic rats. It is expressed in the ganglion cell layer, and interaction with adenosine reduces glutamate-induced calcium influx into the ganglion cells. As noted in the introduction, other work also supports the conclusion that ganglion cells are lost as diabetes progresses in rodent models and humans. The loss of ganglion cells could partially account for the decreased visual function of diabetic Long-Evans rats. At 8 weeks of diabetes, rats with or without cataracts exhibited similar losses in contrast sensitivity and acuity.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
## The VEGF, EPO, and IGF-1 system
*In the **The VEGF, EPO, and IGF-1 system** section:*
In this study, VEGF protein levels significantly increased after 4 weeks but not 12 weeks of diabetes, which is consistent with other studies. The VEGF protein levels were higher in the diabetic rats than the control Sprague-Dawley rats after 2, 4, and 6 weeks of diabetes but not after 12 weeks. Another study found that VEGF protein levels were increased after 4 weeks of diabetes in Sprague-Dawley and Long-Evans rats and significantly in Brown Norway rats, yet no change was found in any strain after 12 weeks. The early increase in VEGF protein levels was not accompanied by an increase in mRNA expression in this study or the study by Schrufer et al., while Brucklacher et al. found VEGF mRNA levels decreased in diabetic rats at 4 and 12 weeks. These results suggest that post-transcriptional mechanisms or translational regulation acts to alter VEGF protein levels, or that qRT-PCR of samples from the whole retina is not sensitive enough to detect VEGF mRNA changes occurring in specific cell types.
In addition, the results show that elevated VEGF protein levels are not persistent in the retina early in diabetes and can vary as the disease progresses. Diabetes did not alter the mRNA expression of the VEGF receptors FLT1 and KDR.
EPO also has angiogenic properties. VEGF and EPO are reported to have neuroprotective properties as well. In a post-mortem analysis, retinas from diabetic patients without diabetic retinopathy had higher EPO mRNA levels than age-matched controls. In this study, the EPO mRNA levels were elevated in the rat retina at 4 and 12 weeks of diabetes, possibly to protect the neural and vascular cells in the retina.
Diabetes increased the expression of IGFBP2 after 12 weeks and IGFBP3 after 4 and 12 weeks, but did not change the expression of IGF1R or IGFBP1. The interaction between IGF-1 and its receptors regulates VEGF expression and can induce blood–retinal barrier breakdown and retinal neovascularization. The effects of the IGF-1 system can be seen in the vasculature and the central nervous system. The IGF binding proteins modulate the activity of IGF-1 but also have effects independent of IGF-1 and IGF1R. IGFBP3 has been shown to have anti- and proapoptotic characteristics and to promote and inhibit proliferation in various cell and tissue types (see for reviews). These activities are most likely dependent on tissue type and pathological condition.
In agreement with this study, Kirwin et al. found IGFBP3 transcript levels significantly increased after 4 weeks and 3 months of diabetes in the Long-Evans rat retina. In the mouse model of retinopathy of prematurity, exogenous IGFBP3 promoted vessel survival during the vaso-obliterative hyperoxic phase and increased vessel regrowth during the relative hypoxic phase independent of IGF-1, and reduced apoptosis in retinal neurons. Even less is known about the functions of IGFBP1 and IGFBP2 in the retina. How IGF binding proteins impact the progression of diabetic retinopathy has yet to be fully evaluated.
## Conclusion
*In the **Conclusion** section:*
Diabetes caused significant changes in the expression of genes related to glutamate neurotransmission and transport. Evidence suggests diabetes causes dysfunction in glutamate processing resulting in ganglion cell loss. The effect of diabetes on the expression of ionotropic glutamate receptor subunits varies between humans and rats and between rat strains. Nonetheless, diabetes alters the expression of the various ionotropic glutamate receptors, and the changes vary over the duration of diabetes. Mounting evidence indicates that diabetes disrupts glutamate signaling in the retina and affects retinal neurons as well as the retinal vasculature.
Alterations in gene expression varied with the duration of diabetes. Most of the genes with elevated mRNA levels after 4 weeks did not have sustained increases after 12 weeks. In addition, more genes had altered expression after 12 weeks, indicating that diabetes leads to more changes in the retina over time. Increased expression of EPO and IGFBP3 and increased VEGF protein levels may be protective responses to damage caused by diabetes, but these responses may not provide sufficient protection. This study shows that diabetes not only injures the retinal vasculature but also affects the neurons in the retina.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
# Appendix 1. Primer sequences.
*In the **Appendix 1. Primer sequences.** section:*
To access the data, click or select the words “Appendix 1.”
# Appendix 2: Results of two-factorial ANOVA (2×2 ANOVA) for each gene.
*In the **Appendix 2: Results of two-factorial ANOVA (2×2 ANOVA) for each gene.** section:*
To access the data, click or select the words “Appendix 2.”
# References
*In the **References** section:*
|
# Introduction
Crystal structure of c5321: a protective antigen present in uropathogenic Escherichia coli strains displaying an SLR fold
# Abstract
*In the **Abstract** section:*
Background
Increasing rates of antimicrobial resistance among uropathogens led, among other efforts, to the application of subtractive reverse vaccinology for the identification of antigens present in extraintestinal pathogenic E. coli (ExPEC) strains but absent or variable in non-pathogenic strains, in a quest for a broadly protective Escherichia coli vaccine. The protein coded by locus c5321 from CFT073 E. coli was identified as one of nine potential vaccine candidates against ExPEC and was able to confer protection with an efficacy of 33% in a mouse model of sepsis. c5321 (known also as EsiB) lacks functional annotation and structurally belongs to the Sel1-like repeat (SLR) family. Herein, as part of the general characterization of this potential antigen, we have focused on its structural properties.
Results
We report the 1.74 Å-resolution crystal structure of c5321 from CFT073 E. coli determined by Se-Met SAD phasing. The structure is composed of 11 SLR units in a topological organisation that highly resembles that found in HcpC from Helicobacter pylori, with the main difference residing in how the super-helical fold is stabilised. The stabilising effect of disulfide bridges in HcpC is replaced in c5321 by a strengthening of the inter-repeat hydrophobic core. A metal-ion binding site, uncharacteristic of SLR proteins, is detected between SLR units 3 and 4 in the region of the inter-repeat hydrophobic core. Crystal contacts are observed between the C-terminal tail of one molecule and the C-terminal amphipathic groove of a neighbouring one, resembling interactions between ligand and proteins containing tetratricopeptide-like repeats.[](https://www.ncbi.nlm.nih.gov/mesh/D012645)
Conclusions
The structure of antigen c5321 presents a mode of stabilization of the SLR fold different from that observed in close homologs of known structure. The location of the metal-ion binding site and the observed crystal contacts suggest a potential role in regulation of conformational flexibility and interaction with yet unidentified target proteins, respectively. These findings open new perspectives in both antigen design and for the identification of a functional role for this protective antigen.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
## Background
*In the **Background** section:*
CFT073 Escherichia coli is an uropathogenic strain responsible for conditions like cystitis and pyelonephritis (an ascending form reaching pelvis and kidneys), severe cases of which may lead to sepsis [1]. Uropathogenic E. coli (UPEC) bacteria are a subclass of ExPEC (Extraintestinal Pathogenic E. coli), a group of pathogens responsible for neonatal meningitis and septicaemia [2]. Increasing rates of antimicrobial resistance among uropathogens, complicating the future treatment of such infections, led to the development of vaccine preparations based on specific virulence factors, which unfortunately did not demonstrate long-term protection [3]. Hence, a broader approach to vaccine design, including the identification of non-virulence factors through methods such as immunoproteomics and reverse vaccinology (targeting of possible vaccine candidates starting from genomic information) is necessary. Recently, subtractive reverse vaccinology was used to identify a number of antigens present in ExPEC but absent or variable in non-pathogenic strains, suggesting that a broadly protective E. coli vaccine may be possible [2].
The 52 kD protein coded by locus c5321 from CFT073 E. coli was identified as one of nine potential vaccine candidates against ExPEC and was able to confer protection with an efficacy of 33% in a sepsis mouse model [2]. Although an antibody-mediated response is likely to be responsible for the capacity of c5321 to induce protection in mice, the actual mechanism of action of anti-c5321 antibodies is still unknown. Recent data from our laboratories have suggested a role for c5321 in impairing the effector functions of human immunoglobulins indicating that antibodies directed against c5321 may affect the ability of E. coli to evade the immune system [4].
Sequence-based analysis performed with SMART [5] and PFAM [6] indicates that the protein is composed of Sel1-like repeats (SLRs, PFAM: PF08238). These repeats share a consensus sequence that is responsible for their helix-turn-helix (α/α) motif and are named after Caenorhabditis elegans sel-1 gene product [7]. Such motifs are flexible in length, usually comprising 36–38 amino-acid residues, with few key positions of small and large hydrophobic residues. The crystal structure of the Helicobacter pylori cysteine-rich protein B (HcpB)
[8], considered as a prototype of the structural fold consisting of SLR units, reveals the modular architecture with the α/α motifs arrayed in tandems and resulting in a super-helical fold. Structural domains composed of several such motifs are thought to act as interaction scaffolds to mediate protein-protein interactions. SLR units can be present in tandem arrays of up to 30 motifs or in groups dispersed throughout the protein sequence. SLR-containing proteins are found in both prokaryotic (more prevalent) and eukaryotic organisms, and are thought to have been acquired by horizontal gene transfer. Unfortunately, only few functional annotations are available for SLR proteins.
There is accumulating evidence that C. elegans Sel1 is involved in degradation of proteins from the endoplasmic reticulum, while the yeast Hrd3 protein is thought to act as an adaptor protein for membrane-bound complexes and HcpA/B from H. pylori is speculated to be responsible for the adaptation of this bacterium to different hosts. It could be said that these molecular functions of SLR proteins are related, in that they are associated with signal transduction pathways [9].[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
SLR proteins share similar consensus sequence with the much more abundant TPR (tetratricopeptide repeat) protein family, in which TPR units are composed of 34 amino-acid residues [9,10]. The structural topology of TPR-containing proteins was revealed by the structure of the TPR domain of the protein phosphatase 5 (PP5) [11]. It displays a super-helical fold similar to the one characteristic of the SLR family. However, the superposition of this TPR domain with HcpB highlights different super-helix parameters, consequence of different packing angles within and between the repeats.
The region of specific ligand binding, as observed in different TPR domains, is located in the amphipathic groove of the super-helix, with three tandem repeats likely being the optimal minimal length for binding [10,12]. Similar interactions most likely facilitate self-assembly into higher order structures [10,13]. TPR domains, as mediators of protein-protein interactions, have been implicated in a wide variety of cellular functions, such as transcription, cell cycle, protein translocation, protein degradation and host defence.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Due to the non-globular, rather elongated architecture of the repeating TPR and SLR units, where stabilization of the fold is achieved mostly through short-range interactions (along the primary sequence), the energy landscape of these proteins is distinct from that of globular proteins. Inter- and intra-element interactions of such quasi-one-dimensional structures are balanced in such a way that small local perturbations yield large effects, readily facilitating structural transitions that may be related to their biological function [14].
Unlike the case of TPR proteins, limited knowledge is available for SLR proteins, including fewer available crystal structures as well as functional annotations. Here, we report the 1.74 Å-resolution crystal structure of c5321 from CFT073 E. coli determined by Se-Met SAD phasing. The structure is composed of 11 SLR units, which to our knowledge represents the bacterial protein with the highest number of Sel1-like repeats solved up to date. It displays similar packing angles to those found in HcpB/C proteins from H. pylori, however with a distinct mode of overall fold stabilisation. Furthermore, we report the presence of a metal-ion binding site, generally uncharacteristic of TPR and SLR proteins.
Crystal contacts between the C-terminal tail of c5321 and the C-terminal section of the amphipathic groove of a molecule belonging to the adjacent asymmetric unit are analysed and their possible biological relevance discussed. As part of a study of the antigenic properties of c5321, the regions of the protein that are recognised by antibodies have been mapped using murine monoclonal IgGs.[](https://www.ncbi.nlm.nih.gov/mesh/D012645)
## Results and discussion
*In the **Results and discussion** section:*
## Overall structure
*In the **Overall structure** section:*
The crystallographic structure of the functional unit (aa 24–490) of c5321 has been solved by SAD phasing and refined to a resolution of 1.74 Å. According to predictions by SignalP [15], the first 23 amino acids constitute a signal sequence. The final model has R and Rfree values of 15.5% and 19.2%, respectively. The model has acceptable root-mean-square differences for bond lengths and angles (0.006 Å and 0.884 degrees, respectively) and none of the residues lie in disallowed regions of the Ramachandran plot (Table 1). One molecule is present in the asymmetric unit of the crystal.
Amino-acid numbering in the model sequence reflects that of the functional unit (1–467 aa).[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Data collection and refinement statistics
c5321 displays a super-helical fold (Figure 1), containing eleven Sel1-like repeats, an N-terminal and two C-terminal helices with probable capping function and a (partly helical) C-terminal tail. Each repeat consists of two helices, helix-1 and helix-2, formed predominantly by thirteen residues, and connected predominantly by a 7-residue loop (Table 2). Individual repeats, tethered by three-residue loops, stack on top of each other creating an extended super-helical molecule with a continuous hydrophobic core. This structure can also be viewed as an overlapping array of three-helix bundles. The right-handed super-helix is approximately 115 Å in length, with a diameter of ~50 Å and a pitch (length of one complete helical turn measured parallel to the helix axes) of 60–65 Å. A complete helical turn comprises about seven to eight SLR units.
The N- and C-terminal helices do not have a true SLR consensus sequence, but they share structural homology. A closer inspection of the amino-acid sequence suggests a role in ‘neutralizing’ hydrophobic surfaces on solvent exposed parts of the first and last repeats, hence facilitating the molecule’s solubility. To date, this is the known structure with the highest number of SLR repeats for a bacterial protein.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Crystal structure of c5321 solved at 1.74 Å resolution. A. Side view of the super-helix, with a Sel1-like-repeat (SLR) unit circled and the Mg2+-binding site between repeats 3 and 4 indicated in pale green. Helix-1 of the repeats is coloured in violet and helix-2 in green. N-terminal and C-terminal (likely capping) helices, as well as the C-terminal tail are represented in grey. B. Top-down view (along the vertical axis) from the C-terminus of the super-helix, depicting its concave (formed by the first helices of the repeats) and convex (formed by the second helices) surfaces.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
Sel1-like repeat units in c5321
Uncharacteristic of SLR and TPR proteins, a metal-ion binding site, occupied by magnesium, was found between repeats 3 and 4 (Figure 1A). It resides in the negatively charged patch of the amphiphilic concave surface of the super-helix. Finally, the C-terminal region of the molecule contacts the C-terminal tail of the molecule in the adjacent asymmetric unit.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
## Similarities with other SLR and TPR proteins
*In the **Similarities with other SLR and TPR proteins** section:*
The packing angles of repeats are similar to those observed in the H. pylori cysteine-rich protein C (HcpC, PDB id 1OUV) [16], comprised of 267 residues (following signal-peptide cleavage) and sharing 47-53% sequence similarity (25-33% sequence identity) with c5321 (the highest structural homology with proteins in the PDB). An important difference between the two proteins resides in the inter-repeat disulfide bonds stabilizing the super-helical packing in HcpC, not present in c5321. Likewise, a shorter homologue (138 residues), H. pylori cysteine rich protein B (HcpB) (PDB id 1KLX) [8], has a 39-46% sequence similarity with c5321 and its repeats are also cross-linked by disulfide bonds.
Finally, the putative Sel1-repeat protein kpn_04481 (228 residues) from Klebsiella pneumoniae ssp. pneumoniae (PDB id 3RJV, not published) shares 45-47% sequence similarity with c5321. While HcpC displays a pattern of repeat interactions similar to that found in c5321, with slight variation in packing angles within and between the repeats, kpn_04481 presents an uncommon packing angle of its fifth repeat that allows contacts with the intra-repeat loops belonging to the second and third repeats. Structure resolution and functional annotation of more SLR family proteins shall lead to understanding the necessity for this observed variety.[](https://www.ncbi.nlm.nih.gov/mesh/D004220)
The first structure of a TPR-motif-containing protein was solved in 1998 [11] and, to date, the structure with the largest number of repeats (11.5) is that of the TPR domain of O-linked GlcNAc transferase (PDB id 1W3B) [17]. A distinct packing angle of the TPR units, as well as within repeats, allows a narrower super-helix in which, unlike SLR proteins, convex-face helical contacts are absent and those between concave helices are less extensive. The significance of these differences in packing between SLR and TPR assemblies is yet to be understood, but is likely related to distinguishing target proteins.
## Structural analysis of inter- and intra-SLR interactions
*In the **Structural analysis of inter- and intra-SLR interactions** section:*
Even though the SLR family of proteins is known for a low conservation of the consensus sequence of the repeat, c5321 reproduces this sequence particularly well (Figure 2). Here, the definition of SLR derived from that of TPR, as annotated in the SMART database has been used. A repeat is composed of the more tightly packed helices, named helix-1 and helix-2 (Figures 1, 3 and 4). Alternatively, a repeat can be defined as constituted by the two helices packed at a wider angle (helix-2 and helix-1′), as suggested by Lüthy and collaborators based on the structures of HcpB and HcpC [8,16].
The rational for this alternative definition lies in the greater conservation of the latter repeat, reflected mainly in the constant length of the shorter loop between these helices. Herein we retain the SMART definition, as one could argue that the repeat should correspond to the entity containing the longer loop, which is an important component in defining the super-helical geometry.
Sequences of the eleven SLR repeats of c5321, aligned with the SLR consensus sequence. First two rows provide a counter for intra-SLR position (first digit of the counter given in first row and second digit in second row). Second column corresponds to sequence number in full-length protein of first residue in the row. Specific amino-acid residues in the consensus sequence (from the SMART database [5]) are indicated with capital letters, while lower case letters stand for: p-polar; h-hydrophobic; t-turn like; s-small; c-charged; a-aromatic; l-leucine, valine or isoleucine. Secondary-structure elements are outlined in the last row, with loop regions being represented by hyphens (except SLR9, where helix2 is longer at the expense of the standard intra-repeat loop length). Residues matching specific conserved amino acids in the SLR consensus sequence are coloured in orange, while consensus amino-acid types with a dominant representative in c5321 SLR units are shown in blue (with the exception of W28r, which is coloured green for easy identification).[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Close-up view of the interactions within and between SLR repeats – convex face. Residue numbers reflect intra-SLR positions as given in Figure 2. The amino-acid code (e.g. Y11r) is given when the specific amino acid is conserved in the position (as indicated in the consensus in Figure 2). Otherwise, only the position is specified (e.g. 24r). The colouring scheme is the same as in Figure 2. Residues in grey, with their side chains represented as sticks, reflect consensus areas with lack of dominant amino-acid representative. It does not necessarily reflect a reduced importance in packing.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
Close-up view of the interactions within and between SLR repeats – concave face. Residue numbers reflect intra-SLR positions as given in Figure 2. The colouring scheme is the same as in Figure 2. Concave face is of less-conserved character than convex face (Figure 3).
The structure-based sequence alignment in Figure 2 shows that the majority of the 11 repeats retain particularly conserved amino-acid residues at positions 3, 7, 8, 11, 14, 17, 21, 25, 32, 33 and 36, indicated in orange, and conserved amino-acid types with dominant representative at positions 4, 18, 24, 28, 29, 30, 31, 34 and 35, coloured in blue. Position 28, occupied by aromatic residues in SLRs, is interestingly almost exclusively represented by tryptophan throughout the c5321 repeats.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
The SLR consensus sequence highlights conserved glycine residues at positions 8, 14, 17 and 36, of which the last three facilitate turns in both intra- and inter-repeat loops while the first one, along with conserved alanine residues at positions 25 and 32, allows a close packing of the helices within the repeat. In turn, alanine residues at positions 3 and 33 favour the specific packing angle between repeats (Figure 3).[](https://www.ncbi.nlm.nih.gov/mesh/D005998)
SLR proteins, unlike their TPR analogues, have few contacts across helix-2 units (convex-face helices). In c5321, they are mostly facilitated by the conserved tryptophan at position 28 (we shall use W28r to indicate repeat position), characteristic for this protein (Figure 2). Interestingly, c5321 helix-2 units have a slight kink, not found in HcpC, which effectively reduces the distance between these units at the side of the inter-repeat loop hence allowing the key positioning of W28r (Figure 3). On the other hand, existing interactions between helix-1 units (concave-face helices) tend to be less conserved (Figure 4). The angular geometry between repeats is mainly dictated by an inter-repeat hydrophobic core centred at conserved residue L7r in helix-1′ (Figure 3).
L7r is in almost all cases in close contact with W28r in helix-2′, with residues 26r (hydrophobic or R), 29r (F or Y), 30r (primarily R) and A33r in helix-2 and with Y11r in the same helix-1′. Further contacts between helix-2 and helix-1′ include hydrophobic interactions between residues 10r and 26r, the interaction between residues 6r (polar) and 29r and those between the conserved Y11r and 30r, with the frequent presence of a hydrogen bond between the respective hydroxyl and guanidinium groups. Also, 30r (primarily R) is often hydrogen bonded via structural water to 24r (primarily Q) (not shown) (Figure 3). In the convex face, the inter-repeat hydrophobic core is protected from solvent by residues 26r and 30r from helix-2, Y11r from helix-1′, V18r from the intra-repeat loop and W28r from helix-2′.[](https://www.ncbi.nlm.nih.gov/mesh/D014364)
Other conserved residue types (Figure 2) play an important role in intra-repeat interactions, such as hydrogen-bonded glutamine residues at positions 4 of helix-1 and 35 of helix-2, the stacking of Y11r and 24r (primarily Q), along with the already mentioned L7r and W28r contact (Figure 3).[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Intra-repeat loops mainly consist of seven residues, with SLR8 containing one additional residue and SLR9 four residues less. Such relatively long loops are important for sufficient inter-repeat packing at wide angles, i.e. the length of these loops governs to a certain extent the stability of the inter-repeat geometry by capping the inter-repeat hydrophobic core. Loop tethering in the conformation observed in the structure is achieved mainly by interactions of conserved V18r with the inter-repeat hydrophobic core (26r and Y11r). The inter-repeat loop is shorter, three-residue long, and allows anti-parallel helical packing as well to limit the angle of inter-repeat helices to a certain degree (Figure 3).[](https://www.ncbi.nlm.nih.gov/mesh/D014633)
Compared to its closest homologue HcpC, c5321 possesses greater sequence conservation and a different means for SLR-fold stabilisation. The Hcp family is unique among SLR proteins in that, in addition to the contributions from the constituent hydrophobic inter-repeat patch and the intra-repeat loop, fold stabilisation is achieved by disulfide bond tethering the C-terminal end of one repeat and the N-terminal end of the next repeat [8,16]. In the case of c5321, which displays a very similar inter-repeat angle, the disulfide-bond effect is most likely substituted by W28r-mediated interactions.[](https://www.ncbi.nlm.nih.gov/mesh/D004220)
## Metal-ion binding site between SLRs 3 and 4
*In the **Metal-ion binding site between SLRs 3 and 4** section:*
Electron density corresponding to a magnesium ion is identified in the c5321 structure, with the metal-ion-binding site located between repeats SLR3 and SLR4. H99 in SLR3, E136 in SLR4 and its intra-repeat-loop residue D138, along with three water molecules, constitute the octahedral coordination ligands of the magnesium ion (Figure 5). Sequence alignment of SLR units does not reveal any other potential metal-ion-binding sites, of similar composition, in the protein.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
Magnesium-ion-binding site in c5321. A. Mg2+ coordination by residues H99 of SLR3, E136 of SLR4 and D138 of the intra-repeat loop. Alanine at repeat position 10 (A133) is indicated to highlight the lack of VdW contacts in the corresponding area. B. Electron density map (σA-weighted 2Fo - Fc contoured at 2σ) at the metal-ion-binding site, including the three water molecules that complete the coordination.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
Interestingly, the hydrophobic inter-repeat core is not as efficiently packed between repeats 3 and 4. This appears mainly due to alanine in position 10 of SLR4 (A133), occupied by larger residues in other repeats, and the corresponding loss of the interaction with positions 12 and 26 of SLR3 (Figure 5A). CASTp, an application for the detection of cavities in proteins [18], identifies a pocket between SLR3 and 4 (volume of 80.8 Å3, area of 95.6 Å2) as one of the top two in the SLR regions, where the other identified pocket is located between SLRs 8 and 9 that lacks the intra-repeat loop. In the case of SLRs 3–4, fewer important interactions between helix-1′ and helices 1 and 2 along with somewhat weaker intra-repeat loop tethering, in the absence of other compensatory mechanisms, would decrease the stability of the inter-repeat packing. Mg2+ coordination by residues of the intra-repeat loop, helix-1 and helix-1′ likely represents a means of stability/flexibility regulation in this region, for a currently unknown purpose.
The significance of the intra-repeat-loop absence between SLRs 8 and 9, which could lead to a certain degree of instability in this region (due to greater solvent exposure of the hydrophobic core), is also unknown.[](https://www.ncbi.nlm.nih.gov/mesh/D000409)
As the likely origin of the magnesium ion in the c5321 structure is MgCl2 used in the crystallization conditions, the natural metal ion for this system and its binding specificities remain to be determined. Real-time quantitative PCR revealed that c5321 mRNA levels in the uropathogenic strain CFT073 grown in Luria-Bertani (LB) medium were higher in the presence of specific ion chelators (including desferal and EDTA), suggesting a scenario where ions that are ligands to the protein are also regulators of its expression (Pastorello et al., unpublished results). Unconventional iron binding, like the triad His-Glu-Asp involved in magnesium-ion coordination in the crystal structure of c5321, is found in X-ray structures of proteins closely related to ferritin and DNA-binding proteins from starved cells (Dps), e.g. Dps from Mycobacterium smegmatis (PDB id 1VEQ) or antigen TpF1 from Treponema pallidum (PDB id 2FJC). Interestingly, all these species are dodecameric entities presenting a spherical shell with a large inner cavity. The iron ion binds inside the cavity at the interface between two adjacent monomers.
The histidine on one side, and glutamate and aspartate on the other, are provided by different subunits and configure the metal-binding site (Figure 6). In all Dps-like structures, the iron ion presents a possible tetrahedral or trigonal bipyramid geometry, given that water molecules involved in coordination might not be observed due to low resolution (e.g. 3.98-2.5 Å for the indicated structures). It cannot be excluded that the coordination of the metal-ion in c5321 could be of such lower order in a potentially native iron-bound structure.[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
Similarities between the Mg-ion binding site in c5321 and Fe-ion binding site in Dps proteins. A. The Fe-ion binding site involving the triad His-Glu-Asp found in X-ray structures of Dps proteins (PDB ID 2FJC). B. Octahedral Mg-ion coordination observed in c5321 (water molecules not indicated).[](https://www.ncbi.nlm.nih.gov/mesh/D008274)
Metal-ion binding, as evident from the PDB repository, is not common in SLR and TPR-containing proteins. However, an example where metal-ion binding might be partly responsible for dynamics and ligand-binding regulation is that of human Pex5p receptor. Sr2+ binding (physiological equivalent unknown) in the protein's TPR domain hinge region, even though in a coordination not resembling the case of c5321, leads to near rigid-body movement of its two halves (lobes) and less overall conformational flexibility of the domain [19].[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
Metal ions play a role in many important functions in proteins, including stability, conformational changes, folding and assembly. One can speculate that for c5321 the stabilisation of the SLR 3–4 region could represent a means of regulation of overall conformational flexibility, and in turn affect ligand (protein/peptide) binding (suspected to be in one of the major grooves, as discussed in the next section). Clearly, further investigation of metal-ion binding, its specificity and functional role will be required in order to assess these hypotheses.[](https://www.ncbi.nlm.nih.gov/mesh/D008670)
## Crystal contacts between the C-terminal super-helical groove of one molecule and the C-terminal tail of the molecule belonging to the adjacent asymmetric unit
*In the **Crystal contacts between the C-terminal super-helical groove of one molecule and the C-terminal tail of the molecule belonging to the adjacent asymmetric unit** section:*
In the crystal packing the concave surface of the super-helix in the region of repeats 8, 9 and 10 of one molecule interacts with the C-terminal tail of a symmetry-related molecule (Figure 7A). Interestingly, similar C-terminal tail interactions have already been observed in other SLR proteins, such as HcpC (Figure 8).
Crystal contacts: binding of the C-terminal tail of c5321 to the C-terminal concave groove of a neighbour molecule. A. Relative orientation of the two molecules, with SLRs 7–11 alternating in purple and light blue (the rest of the molecule is represented in grey). B. Close-up view of the interactions (see main text).
Similarities between crystal contacts observed in c5321 and in HcpC and interactions of Hop with the C-terminus of its Hsp70 target (PDB ID 1ELW). A. Structural superposition highlights similar mode of interaction involving concave region of the super-helix (molecule 1) and the extended peptide conformation of the C-terminus (molecule 2). B. Close-up of the hydrogen bonding interactions between the conserved asparagine residue (molecule 1, N310 in c5321) and the main chain of the C-terminus peptide (molecule 2). Note that the additional asparagine residue in c5321 (N345) is also involved in interactions with the main chain.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
Interactions are predominantly of polar character accompanied by van der Waals contacts. N345, belonging to repeat 10, participates in bidentate hydrogen-bonding interactions with the backbone carbonyl oxygen and nitrogen atoms of the C-terminal tail's Q459. N310 of repeat 9 is near the C-terminal tail's A461 and could be engaged in similar interactions with its backbone atoms, upon rotameric change (Figure 7B). These interactions can be categorised as anchoring, non-specific for peptide binding in such type of groove. Specificity might be governed by non-polar interactions of the tail peptide with the inter-repeat core and polar interactions with the solvent exposed repeat area.
Thus, residues responsible for shape complementarity comprise A458 and A461 of the C-terminal tail in van der Waals contact with A313 and A341, A309, T306 of the second molecule, respectively. Polar interactions involve S460 hydrogen bonding to R317 in SLR9 and K280 in SLR8, while D304 and T306 in SLR9 take part in charge-charge and hydrogen-bond interactions with K463 in the C-terminal tail, stacking against Y272 of SLR8. Additional stacking interactions are observed between F414 belonging to one of the C-terminal helices and K455 at the C-terminal end (Figure 7B). The last three residues of the C-terminus, for which lower electron density is observed, do not really contribute to the interactions within this concave pocket.[](https://www.ncbi.nlm.nih.gov/mesh/D001216)
Some similarity is shared with the mode of target-protein/peptide binding of TPR proteins, such as receptor for peroxisomal uptake-Pex5, Hsp70/Hsp90 organizing protein-Hop (Figure 8), FKBP52 and PP5. In these, TPR tandems recognise the C-terminal EEVD signal sequence of the target protein/peptide by a “carboxylate clamp” (group of conserved positively charged residues) [10,20-22] and likely also utilizing bidentate hydrogen-bonding interactions between conserved asparagine residues lining the super-helical groove and target peptide backbone atoms [17]. Figure 8 highlights this type of hydrogen bonding, involving a conserved asparagine residue, in c5321, HcpC and Hop/Hsp70(C-terminus). In most cases, the binding pocket establishing primary interactions through the target’s C-terminal tail is composed of three TPR repeats, in line with three tandem TPR domains being the most populated, suggesting that these represent the minimal functional binding unit.
Secondary interactions are often important for establishing specificity and are thought to lie outside of this primary region, as shown for Hop [21,23]. Cortajarena et al. also emphasised the importance of both short-range interactions and long-range electrostatics as determinants of specificity [24].[](https://www.ncbi.nlm.nih.gov/mesh/D002264)
On the other hand, there are examples of self association for TPR proteins, such as dimerisation of the Sgt1 plant protein [25] or oligomerisation of the MamA bacterial protein (demonstrated in vivo), that involve their terminal helices binding in the super-helical groove regions [26]. For a number of TPR-containing proteins it has been shown that self-association can serve to regulate their biological function. SE-HPLC analysis of c5321 shows that aggregates/oligomers are present in very little amount (less than 5%, data not shown), although an exhaustive study of c5321 oligomerisation has not been performed.
The potential biological significance of the observed crystal contact in c5321 is revealed by comparison with the known TPR ligand-binding examples. Indeed, they share common features like binding in the concave area of the super-helix with similar peptide-backbone anchoring mode and three-tandem SLR domains as binding pocket. Correlation between the crystal contacts observed in the TPR protein Cyp40 and its interactions with the natural ligand (Hsp90) in solution [27] further suggests the possibility of a similar scenario in the c5321 case, with yet unidentified target protein/peptide (or self-association).
## Epitope mapping
*In the **Epitope mapping** section:*
Mapping of c5321 epitopes for murine monoclonal antibodies was performed as part of the general characterization of this protein as a potential vaccine candidate, alongside presenting an opportunity to further investigate binding regions in c5321. Proteolytic digestion of the antigen following its incubation with monoclonal antibodies (see Methods) did not result in identification of any epitope-containing peptides. However, epitope-containing peptides were captured from partial digestion of c5321 (performed prior to incubation with monoclonal antibodies) with GluC (IgGs 17A7-C2 and 14E7/D10) or LysC (IgG 16H8/G6) (Figure 9).
Epitope-mapping results. A. MS spectra of the proteolytic-digestion products of c5321 with GluC (upper spectrum) and the peptide immunocaptured with mAb14E7/D10 (111-SVKWFRLAAEQGRDSGQQSMGDAYFE-136, lower spectrum). B. MS spectra of the proteolytic-digestion products of c5321 with LysC (upper spectrum) and the peptide immunocaptured with mAb16H8/G6 (18-AQLELGYRYFQGNETTK-34, lower spectrum). C. MS spectra of the proteolytic-digestion products of c5321 with GluC (upper spectrum) and the peptide immunocaptured with mAb17A7-C2 (340-KAAQFNLGNALLQG-KGVKKDE-360, lower panel). D. Identified epitope-containing regions mapped onto the c5321 structure.
The sequences of the epitope-containing peptides map to helix-1 of SLR1 for mAb16H8/G6, helix-1 of SLR10 for mAb17A7-C2 and helix-2 of SLR3 and helix-1 of SLR4 for mAb14E7/D10 (Figure 9D). Failure to immunocapture the products of LysC or trypsin cleavage (at the C-terminal side of arginine or lysine residues) by mAb17A7-C2 and mAb14E7/D10 further narrows down the important epitope components to the intra-repeat loop of SLR10 (containing three lysines) for mAb17A7-C2 and to the outer helix-2 of SLR3 (containing one lysine and two arginines) for mAb14E7/D10, respectively. These are in agreement with no steric hindrance to the access and binding of the antibody to these convex areas of the super-helix, and partly overlap with regions that are involved in Mg2+ binding (between SLRs 3 and 4) or belong to the C-terminal-tail binding groove (SLRs 8, 9 and 10). How relevant this observation is with respect to the previously discussed roles of these regions remains to be further investigated.[](https://www.ncbi.nlm.nih.gov/mesh/D001120)
## Conclusions
*In the **Conclusions** section:*
We have solved the structure of c5321 from uropathogenic Escherichia coli to 1.74 Å resolution. This antigen displays a super-helical Sel1-like repeat fold with eleven SLR units and a remarkably preserved consensus repeat sequence. It shares high structural similarity with its closest homologue of known three-dimensional structure, HcpC from Helicobacter pylori, albeit with differences in how the SLR-fold is stabilized. While disulfide bridges in HcpC lock the characteristic inter-repeat geometry, in c5321 a conserved tryptophan residue at repeat position 28 appears to contribute fundamentally in maintaining the same geometry by strengthening the inter-repeat hydrophobic core. Metal ion binding, generally uncharacteristic of SLR proteins, is observed between SLR units 3 and 4, suggesting a regulatory role in conformational flexibility.
Furthermore, crystal contacts observed between molecules belonging to neighbour asymmetric units share similarity to contacts characteristic for TPR-protein interactions with their physiological targets, suggesting a potential physiological interaction mode of c5321 with yet unidentified targets. The structure of c5321 is a first step for its functional characterisation and opens the door to the possibility of redesigning this antigen for vaccine-development purposes.[](https://www.ncbi.nlm.nih.gov/mesh/D004220)
## Methods
*In the **Methods** section:*
## Cloning, expression and purification of c5321
*In the **Cloning, expression and purification of c5321** section:*
c5321 gene, without the predicted signal sequence, was amplified by PCR from the CFT073 genomic DNA template, cloned in pET-21b vector (Novagen) and transformed in DH5α-T1R chemically competent cells for propagation. BL21(DE3) chemically competent cells were used for His-tagged protein expression (6xHis at the C-terminus). Purification of the recombinant protein was performed from the bacterial soluble fraction using nickel-affinity chromatography as already described [28,29]. Cleavage of the His-tag was not performed.[](https://www.ncbi.nlm.nih.gov/mesh/D009532)
## Expression and purification of selenomethionine-labelled c5321
*In the **Expression and purification of selenomethionine-labelled c5321** section:*
The plasmid DNA containing the c5321 gene downstream of the T7 promoter was transformed into B834 DE3 cells and the protein was expressed during 8 h at 25°C using the Overnight Express Autoinduction System 2 medium (Novagen) supplemented with 100 nM vitamin B12, 0.125 mg/ml selenomethionine and 50 μg/ml ampicillin. The cells were harvested by centrifugation at 5000 g for 30 min.[](https://www.ncbi.nlm.nih.gov/mesh/D014805)
The cell pellet was suspended in lysis buffer (50 mM Na phosphate pH 7.5, 150 mM NaCl, 10 mM imidazole) and the cells were lysed by sonication. The insoluble fraction was removed by centrifugation (14000 g, 15 min) and the cleared lysate was applied onto a Ni-NTA sepharose column (Qiagen), equilibrated with the lysis buffer. The column was washed with 10 volumes of wash buffer (50 mM Na phosphate pH 7.5, 150 mM NaCl, 20 mM imidazole) and the protein was eluted with a gradient of increasing imidazole concentration (50 mM Na phosphate pH 7.5, 150 mM NaCl, 100 mM-300 mM imidazole). The protein-rich fractions were pooled and dialyzed 3 times against 100 volumes of 50 mM Na phosphate pH 7.5, 150 mM NaCl buffer. The protein sample was concentrated to 10 mg/ml using a 3 kD cut-off Amicon Ultra concentrator (Millipore).
His-tag cleavage was not performed. The yield of the labelled c5321 was estimated to be about 20 mg of protein per litre of bacterial cell culture.[](https://www.ncbi.nlm.nih.gov/mesh/C018279)
## Crystallisation and data collection
*In the **Crystallisation and data collection** section:*
The protein was buffer exchanged with 5 mM Tris–HCl pH 8.0, 1 mM β-mercaptoethanol. Crystals were grown at 18°C using the hanging drop vapour diffusion method. The protein solution, at ~10 mg/ml concentration, was combined in a 1:1 ratio (v/v) with a well solution consisting of 20% PEG3350, 100 mM Tris–HCl pH 8.5, and 200 mM MgCl2. Prior to X-ray diffraction analysis, crystals were transferred to a cryo-protectant solution (20% ethylene glycol, 20% PEG3350, 100 mM Tris–HCl pH 8.5, and 200 mM MgCl2) and flash cooled in liquid nitrogen. Diffraction data were collected at 100 K at the beam-line ID14-4, ESRF, Grenoble.
These data were indexed, integrated and scaled using MOSFLM and SCALA.[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
## Structure solution and refinement
*In the **Structure solution and refinement** section:*
The structure was solved by the Se-Met SAD method, using crystals that contained one molecule in the asymmetric unit. Se-Met SAD data at 2.28 Å resolution were collected to a satisfactory redundancy of 7.1 and relatively good signal/noise ratio of 13.9. Se atoms were located using Autosol, PHENIX [30] and initial phases were calculated from their positions. Model building was performed with the Autobuild module of PHENIX as well as Buccaneer [31], combined with manual reconstruction.
Subsequent refinement was carried out with Phenix/Coot using a Se-Met SAD 1.74-Å resolution dataset and yielded an R factor of 15.5% and an Rfree of 19.2%. Mg2+ was identified as the most likely representative for the electron density peak in proximity of repeats 3 and 4. 2 Cl-, 29 ethylene glycol and 466 water molecules were included in the final model. The quality of the final model was assessed with PROCHECK [32]. None of the residues lie in the disallowed region of the Ramachandran plot. The model was analysed using Pymol, which was used for figure preparation as well.
Crystallographic statistics are shown in Table 1.[](https://www.ncbi.nlm.nih.gov/mesh/D012645)
## Production of monoclonal antibodies against c5321
*In the **Production of monoclonal antibodies against c5321** section:*
The purified recombinant c5321 was used to immunize CD1 mice. The first dose was a 50-μg injection, whereas the second and third doses at days 14 and 21 were of 25 μg. At day 28, anti-c5321 titres were measured in mice sera using ELISA plates coated with the recombinant c5321. A fourth dose was then administered, and 3 days later mice spleen cells were fused with myeloma cells (NS0). After 2 weeks of incubation in hypoxanthine-aminopterinthymidine selective medium, the hybridoma supernatants were screened for antibody-binding activity by ELISA.
Hybridomas secreting anti-c5321 antibodies, selected by Western Blot to determine their capacity to recognize the antigen in bacterial extracts, were cloned by limiting dilution and then expanded and frozen for subsequent purification of mAbs. The mAb subclasses were determined using a mouse mAb isotyping kit (Roche). The mAbs were purified from culture supernatant by Protein G affinity columns (GE Healthcare), and after exhaustive dialysis in PBS buffer, the concentration of the purified mAb was determined by spectrophotometric reading at 280 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D019271)
## Epitope mapping of monoclonal antibodies
*In the **Epitope mapping of monoclonal antibodies** section:*
The epitope-mapping protocols are based on the approach described by [33], which we adapted to the two different protocols used here [34]:
1) Immunocapturing of peptides from antigen partial digestion. Peptide mixtures were obtained by digestion of c5321 with trypsin, LysC and GluC (separately) in 50 mM ammonium bicarbonate buffer in a ratio of 10:1 at 37°C for 3 h. To capture the epitope-containing peptide, a 25-μl suspension of Dyanbeads Pan Mouse IgG (uniform, super-paramagnetic polystyrene beads of 4.5 μm diameter coated with monoclonal human anti-mouse IgG antibodies) was used. The beads were washed twice with PBS using a magnet and re-suspended in the initial volume. 1 μg of the probe (murine) mAb was added and incubated for 30 min at room temperature, the beads were then washed twice with PBS to remove mAb excess.
0.5 μl of Protease Inhibitor Mix (GE Healthcare) was added before the peptide mixture to avoid potential degradation of the antibodies. The sample was incubated for 30 min at room temperature with gentle mixing. After incubation, the beads were washed three times with 1 ml PBS, and the bound peptide was then eluted with 50 μl of 0.2% TFA. The elute fraction was concentrated and washed with C18 ZipTips (Millipore) and eluted in 3 μl of 50% ACN and 0.1% TFA. For MALDI-MS analysis, 1 μl of sample was mixed with the same volume of a solution of alpha-cyano-4-hydroxy-transcinnamic acid matrix (0.3 mg/ml in H2O:ACN:TFA at 6:3:1), spotted onto the MALDI target plate and air-dried at room temperature.
MALDI-mass spectra were recorded in the positive ion mode on an UltrafleXtreme MALDI TOF/TOF instrument (Bruker Daltonics). Ion acceleration was set to 25 kV. All mass spectra were externally calibrated using a standard peptide mixture. For MS/MS analysis, the MASCOT search engine (Matrix Science. London, UK) was used with the following parameters: one missed cleavage permission, 20-ppm measurement for MS and 0.3 Da for MS/MS tolerance. Positive identification were accepted with p < 0.05.
In the searches, modification of methionine to methionine sulfoxide was allowed.[](https://www.ncbi.nlm.nih.gov/mesh/C027043)
2) Partial digestion of immunocaptured antigens. To capture conformational epitopes, the order of the steps in the previous protocol was inverted. The intact protein (20 μg) was added to the beads, allowing it to bind to the immobilised mAb. The protease was then added to the sample in a ratio 50:1, and incubated at 37°C for 3 h. After proteolysis, the beads were washed ten times with 1 ml PBS, and the bound peptide was then eluted as previously described.
To avoid the analysis of proteolysed antibody fragments within the elute fraction, c5321 was substituted by PBS in negative controls.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## Availability of supporting data
*In the **Availability of supporting data** section:*
The coordinates and merged structure factors for c5321 have been deposited in the Protein Data Bank repository under accession code 4BWR [DOI:10.2210/pdb4bwr/pdb].
## Competing interests
*In the **Competing interests** section:*
The authors declare that they have no competing interests.
## Authors’ contributions
*In the **Authors’ contributions** section:*
Designed the project: LS, MS, XD; designed and performed the experiments: DU, MFN, IP, EC, DZ; analysed the data: DU, MFN, LC, JDM, AL, DR, LS, MS, XD; wrote the manuscript: DU, MFN, LC, JDM, AL, LS, MS, XD. All authors read and approved the final manuscript.
## Acknowledgements
*In the **Acknowledgements** section:*
The authors thank M. Bolognesi for valuable discussion on the manuscript. This work has been supported by funding under the Sixth Research Framework Programme of the European Union (ref. LSHB-CT-2006-037325, BacAbs). DU is recipient of a postdoctoral grant of Universitat Autònoma de Barcelona (UAB). Crystallographic-data collection received the support of the European Synchrotron Radiation Facility (ESRF) (ref. MX-1104).
|
# Introduction
[Ropivacaine](https://www.ncbi.nlm.nih.gov/mesh/D000077212)-Induced Contraction Is Attenuated by Both Endothelial [Nitric Oxide](https://www.ncbi.nlm.nih.gov/mesh/D009569) and Voltage-Dependent Potassium Channels in Isolated Rat Aortae
# Abstract
*In the **Abstract** section:*
This study investigated endothelium-derived vasodilators and potassium channels involved in the modulation of ropivacaine-induced contraction. In endothelium-intact rat aortae, ropivacaine concentration-response curves were generated in the presence or [absence of ](https://www.ncbi.nlm.nih.gov/mesh/D000077212)the following inhibitors: the nonspecific nitric oxide s[ynthase (NO](https://www.ncbi.nlm.nih.gov/mesh/D000077212)S) inhibitor N ω-nitro-L-arginine methyl ester (L-NAME), the neuronal NOS inhibitor N ω-propyl-L-arginine hydrochloride, the inducible NOS inhibitor 1400W d[ihydrochloride, the nitric oxide-](https://www.ncbi.nlm.nih.gov/mesh/D019331)se[nsitiv](https://www.ncbi.nlm.nih.gov/mesh/D019331)e guanylyl cyclase (GC) inhibi[tor ODQ, the NOS and GC inhibitor m](https://www.ncbi.nlm.nih.gov/mesh/C109599)ethylene blue, the phosphoinos[itide-3 kinase inhibi](https://www.ncbi.nlm.nih.gov/mesh/C496401)tor wortmannin, the cytochrome p450 epoxygenase inhibitor flu[con](https://www.ncbi.nlm.nih.gov/mesh/C095284)azole, the voltage-dependen[t potassium ch](https://www.ncbi.nlm.nih.gov/mesh/D008751)annel inhibitor 4-aminopyridine (4-AP), th[e calcium-](https://www.ncbi.nlm.nih.gov/mesh/D000077191)activated potassium channel inhibitor tetrae[thylammoniu](https://www.ncbi.nlm.nih.gov/mesh/D015725)m (TEA), the inward-rectifying potassium channel inh[ibitor barium c](https://www.ncbi.nlm.nih.gov/mesh/D015761)hl[orid](https://www.ncbi.nlm.nih.gov/mesh/D015761)e, and the ATP-sensitive potassium channel inhibitor [glibenclamide. The](https://www.ncbi.nlm.nih.gov/mesh/D019789) e[ffe](https://www.ncbi.nlm.nih.gov/mesh/D019789)ct of ropivacaine on endothelial nitric oxide synthas[e (eNOS) phosph](https://www.ncbi.nlm.nih.gov/mesh/C024986)orylation in human umbilical vein endothelial cells [was examined ](https://www.ncbi.nlm.nih.gov/mesh/D005905)by western blott[ing. Ropiva](https://www.ncbi.nlm.nih.gov/mesh/D000077212)caine-induced contraction was weaker in endothelium-intact aortae than in endothelium-denuded aortae. L-NAME, ODQ, and methylene blue enh[anced ropiv](https://www.ncbi.nlm.nih.gov/mesh/D000077212)acaine-induced contraction, whereas wortmannin, N ω-propyl-L-arginine hydrochloride, 1400W dihydr[ochlor](https://www.ncbi.nlm.nih.gov/mesh/D019331)id[e, ](https://www.ncbi.nlm.nih.gov/mesh/C095284)and fl[uconazole had ](https://www.ncbi.nlm.nih.gov/mesh/D008751)no effect.[ 4-AP and T](https://www.ncbi.nlm.nih.gov/mesh/D000077212)EA enhanced ropivacaine-induce[d contract](https://www.ncbi.nlm.nih.gov/mesh/D000077191)io[n; however, barium chloride and gli](https://www.ncbi.nlm.nih.gov/mesh/C109599)be[nclamide had no effec](https://www.ncbi.nlm.nih.gov/mesh/C496401)t. eNO[S phosphory](https://www.ncbi.nlm.nih.gov/mesh/D015725)lation was induc[ed b](https://www.ncbi.nlm.nih.gov/mesh/D015761)y rop[iva](https://www.ncbi.nlm.nih.gov/mesh/D019789)caine. The[se results ](https://www.ncbi.nlm.nih.gov/mesh/D000077212)suggest that ropivacaine-induce[d contraction i](https://www.ncbi.nlm.nih.gov/mesh/C024986)s att[enuated prima](https://www.ncbi.nlm.nih.gov/mesh/D005905)rily by both endothelial nitric oxide and voltage-de[pendent pot](https://www.ncbi.nlm.nih.gov/mesh/D000077212)assium channels.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
## 1. Introduction
*In the **1. Introduction** section:*
Ropivacaine is an aminoamide local anesthetic with a long duration that produces vasoconstriction both in vivo and in vitro, suggesting that intrinsic vasoconstriction induced by ropivacaine contributes to the drug's long-lasting analgesic effect [1–4]. Ropivacaine produces vasoconstriction at low concentrations, followed by vasodilation at 1 × 10−3 M [4]. The clinical profile of ropivacaine is similar to that of racemic bupivacaine, but its toxicity is relatively low compared with that of bupivacaine [5]. Ropivacaine is an aminoamide local anesthetic of the n-alkyl-substituted pipecolyl xylidine family, which includes levobupivacaine and mepivacaine [5].
Vasoconstriction induced by levobupivacaine and mepivacaine is attenuated by endothelial nitric oxide (NO) [6–8]. In endothelium-denuded aortae, ropivacaine-induced contraction is mediated mainly by the lipoxygenase pathway and partly by the cyclooxygenase pathway [4]. However, in endothelium-intact aortae, endothelium-derived vasodilators, including NO, endothelium-derived hyperpolarizing factor (EDHF), and prostacyclin, are involved in the modulation of vascular tone via vasodilation [9]. Ropivacaine induces endothelial NO-dependent relaxation in isolated vessels precontracted with phenylephrine and attenuates phenylephrine-induced contraction [10, 11].
In addition, the change of the membrane potential of vascular smooth muscle induced by the activation or inhibition of various potassium channels, including voltage-dependent, calcium-activated, inward-rectifying, and adenosine triphosphate-sensitive potassium channels, modulates vascular tone via vasodilation and vasoconstriction [12]. However, the endothelium-derived vasodilators and various potassium channels involved in the modulation of ropivacaine-induced contraction remain unknown. Therefore, the goal of this in vitro study was to investigate both endothelium-derived vasodilators and potas[sium channe](https://www.ncbi.nlm.nih.gov/mesh/D000077212)ls prim[arily invo](https://www.ncbi.nlm.nih.gov/mesh/D000577)lved in modulating ropivacaine-induced contraction in isolated endothelium-intact aortae.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
## 2. Materials and Methods
*In the **2. Materials and Methods** section:*
All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee (Jinju, Gyeongnam, Republic of Korea) at Gyeongsang National University and were performed in accordance with the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences.
## 2.1. Preparation of Aortic Rings for Tension Measurement
*In the **2.1. Preparation of Aortic Rings for Tension Measurement** section:*
Experimental preparation was performed as previously described [13]. Male Sprague-Dawley rats weighing 250–300 g were anesthetized via intramuscular injections of Zoletil 50 (15 mg/kg). The descending thoracic aorta was dissected free, and surrounding connective tissues and fat were removed under microscopic guidance in a Krebs solution bath (118 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO4, 1.2 mM KH2PO4, 2.4 mM CaCl2, 25 mM NaHCO3, and 11 mM glucose). The aorta was cut into 2.5 mm rings, suspended on Grass isometric transducers (FT-03, Grass Instrument, Quincy, MA, USA) under a 3.0 g resting tension in 10 mL of Krebs bath at 37°C, and aerated continuously with 95% O2 and 5% CO2 to maintain the pH within the range of 7.35–7.45. The rings were equilibrated for 120 min, changing the bathing solution every 30 min. Endothelium was removed from some aortic rings by inserting a 25-gauge needle tip into the lumen of the rings and gently rubbing for a few seconds.
Once phenylephrine (1 × 10−7 M)-induced contraction had stabilized, acetylcholine (1 × 10−5 M) was added to assess the endothelial integrity. Endothelial integrity was confirmed by the observation of more than 70% acetylcholine-induced relaxation. Contraction in response to isotonic 60 mM KCl was measured for all aortic rings and defined as the reference value (100%). After washing out the KCl from the organ bath and allowing a return to the baseline resting tension, a cumulative concentration-response curve induced by ropivacaine was obtained as described in subsequent sections.[](https://www.ncbi.nlm.nih.gov/mesh/C006131)
## 2.2. Experimental Protocols
*In the **2.2. Experimental Protocols** section:*
The first series of experiment assessed the effect of endothelial denudation and nonspecific nitric oxide synthase (NOS) inhibitor N ω-nitro-l-arginine methyl ester (l-NAME, 1 × 10−4 M) on the cumulative concentration (1 × 10−5 to 1 × 10−3 M)-response curves induced by ropivacaine in isolated aortae. l-NAME was directly added to the organ bath containing endothelium-intact aorta 20 min before the addition of ropivacaine. Subsequent concentrations of ropivacaine were directly added to the organ bath after the previous concentration had produced a sustained and stable response.[](https://www.ncbi.nlm.nih.gov/mesh/D019331)
The second series of experiments assessed the cumulative concentration-response curves induced by ropivacaine in isolated endothelium-intact aortae in the presence or absence of the following inhibitors: the neuronal NOS inhibitor N ω-propyl-l-arginine hydrochloride (5 × 10−8 M), the inducible NOS inhibitor 1400W dihydrochloride (1 × 10−6 M), the NO-sensitive guanylyl cyclase (GC) inhibitor 1H-[1,2, 4] oxadiazolo[4,3-a]quinoxalin-1-one (ODQ, 1 × 10−6 and 1 × 10−5 M), the NOS and GC inhibitor methylene blue (1 × 10−6 M), the cytochrome P450 epoxygenase inhibitor fluconazole (1 × 10−5 M), and the cyclooxygenase inhibitor indomethacin (1 × 10−5 and 3 × 10−5 M). The aforementioned inhibitors were directly added to the organ bath 20 min before the addition of ropivacaine. Inhibitor concentrations were chosen on the basis of the concentrations used in previous experiments similar to this experiment [6, 10, 13–18].[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
The third series of experiments assessed which specific potassium channels are primarily involved in the attenuation of ropivacaine-induced contraction in endothelium-intact aortae. In endothelium-intact aortae, ropivacaine concentration-response curves were generated in the presence or absence of the following potassium channel inhibitors: the voltage-dependent potassium channel inhibitor 4-aminopyridine (4-AP, 2 × 10−3 M), the calcium-activated potassium channel inhibitor tetraethylammonium (TEA, 2 × 10−3 M), the adenosine triphosphate-sensitive potassium channel inhibitor glibenclamide (1 × 10−5 M), and the inward-rectifying potassium channel inhibitor barium chloride (3 × 10−5 M) [19–22]. In addition, in the endothelium-intact aortae pretreated with 1 × 10−4 M l-NAME, cumulative ropivacaine concentration-response curves were generated in the presence or absence of either 4-AP (2 × 10−3 M) or TEA (2 × 10−3 M). In endothelium-intact aortae, cumulative phenylephrine concentration (1 × 10−8 to 1 × 10−5 M)-response curves were generated in the presence or absence of either 4-AP (2 × 10−3 M) or TEA (2 × 10−3 M).
We also investigated whether ropivacaine-induced contraction involves endothelium-independent activation of voltage-dependent and calcium-activated potassium channels of vascular smooth muscle. After the ropivacaine (10−4 M)-induced contraction in endothelium-denuded aortae reached a plateau, TEA (2 × 10−3, 5 × 10−3, 1 × 10−2 M) or 4-AP (2 × 10−3, 5 × 10−3, 1 × 10−2 M) was cumulatively added to the organ bath to generate cumulative concentration-response curves for TEA or 4-AP.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
Finally, we assessed the ropivacaine concentration-response curves in endothelium-intact aortae in the presence or absence of the phosphoinositide-3 kinase (PI3K) inhibitor wortmannin (1 × 10−7 M) to determine whether the NO-mediated attenuation of ropivacaine-induced contraction is associated with the pathway involving PI3K-Akt-endothelial nitric oxide synthase (eNOS) [23, 24].[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
## 2.3. Cell Culture
*In the **2.3. Cell Culture** section:*
Human umbilical vein endothelial cells (HUVECs; EA.hy 926 cells, American Type Culture Collection, Manassas, VA, USA) were grown in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% fetal bovine serum (FBS), 2 mmol/L l-glutamine, 100 IU/mL penicillin, and 10 μg/mL streptomycin as previously described [6]. Cells were cultured in 100 mm dishes and grown in a humidified 5% CO2 incubator. HUVECs were plated at a density of 1 × 107 cells per 100 mm dish. Cells were used between passage numbers 6 and 12.[](https://www.ncbi.nlm.nih.gov/mesh/D005973)
## 2.4. Cell Stimulation
*In the **2.4. Cell Stimulation** section:*
Cells were plated at a density of 1 × 107 cells per 100 mm dish. The cells were stimulated with ropivacaine (1 × 10−4 M). To detect phosphorylated eNOS (p-eNOS), cells were treated with ropivacaine (1 × 10−4 M) for 5, 10, 30, and 60 min, harvested, and subjected to western blot analysis.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
## 2.5. Western Blot Analysis
*In the **2.5. Western Blot Analysis** section:*
Western blot analysis was performed as previously described [6]. Briefly, cells were lysed in PRO-PREP protein extract solution to isolate total cell extracts. After centrifugation at 16,000 ×g for 20 min at 4°C, the protein concentration was determined by the Bradford method. Thirty micrograms of protein was subjected to 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis. The separated proteins were transferred to a polyvinylidene difluoride membrane using the SD semidry transfer cell system (Bio-Rad, Hercules, CA, USA).
The membranes were incubated with primary antibodies (anti-eNOS antibodies: rabbit polyclonal, Cell Signaling Technology, Beverly, MA, USA; anti-phospho-eNOS antibodies: Ser1777 rabbit polyclonal, Cell Signaling Technology) at a 1 : 500 concentration (4 μg/mL) in 5% skim milk in Tris-buffered saline with Tween (TBST) overnight at 4°C, and the bound antibody was detected by horseradish peroxidase-conjugated anti-rabbit IgG. The membranes were washed and then developed using the Luminol Reagent system (Animal Genetics, Suwon, Republic of Korea).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## 2.6. Materials
*In the **2.6. Materials** section:*
All drugs were of the highest purity available commercially. Phenylephrine, l-NAME, 1400W dihydrochloride, ODQ, indomethacin, wortmannin, 4-AP, TEA, barium chloride, and glibenclamide were obtained from Sigma-Aldrich (Saint Louis, MO, USA). N ω-propyl-l-arginine hydrochloride was obtained from Tocris Bioscience (Bristol, UK). Methylene blue and fluconazole were purchased from SALF Laboratorio Farmacologico (Bergamo, Italy) and Pfizer Global Manufacturing (France), respectively. Ropivacaine was donated by AstraZeneca Korea (Seoul, Republic of Korea).
Zoletil 50 was purchased from Virbac (Virbac Laboratories, Carros, France). DMEM, FBS, penicillin, streptomycin, and glutamine were supplied by Gibco BRL (Rockville, MD, USA). All concentrations are expressed as the final molar concentration in the organ bath. ODQ, N ω-propyl-l-arginine hydrochloride, 1400W dihydrochloride, wortmannin, indomethacin, and glibenclamide were dissolved in dimethyl sulfoxide (DMSO) (final organ bath concentration: 0.1% DMSO). Unless stated otherwise, all other drugs were dissolved in distilled water.[](https://www.ncbi.nlm.nih.gov/mesh/D010656)
## 2.7. Data Analysis
*In the **2.7. Data Analysis** section:*
Data are expressed as the mean ± SD. Contractile responses induced by ropivacaine are expressed as the percentage of the maximum contraction in response to isotonic 60 mM KCl. Vascular responses induced by TEA or 4-AP in endothelium-denuded aortae precontracted with 1 × 10−4 M ropivacaine are expressed as the percent change from baseline precontraction induced by 1 × 10−4 M ropivacaine. N indicates the number of rats from which descending thoracic aortic rings were derived. The effects of endothelial denudation and various inhibitors on the concentration-response curves induced by ropivacaine or phenylephrine were analyzed by two-way analysis of variance (ANOVA) followed by Bonferroni's post-hoc test using GraphPad Prism version 5.0 for Windows (GraphPad Software, San Diego, CA, USA).
The band intensities from western blotting analysis were analyzed by Student's t-test. Reponses to each concentration of ropivacaine, 4-AP, and TEA were analyzed by repeated-measures ANOVA followed by Bonferroni's post-hoc test. P values less than 0.05 were considered significant.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
## 3. Results
*In the **3. Results** section:*
Ropivacaine produced vasoconstriction at 3 × 10−4 M in endothelium-intact aortae, followed by vasodilation at 1 × 10−3 M (3 × 10−4 M: P < 0.001 versus 1 × 10−5 M; 1 × 10−3 M: P < 0.05 versus 3 × 10−4 M; Figures 1 and 2(a)).[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
Ropivacaine-induced contraction was weaker in endothelium-intact aortae than in endothelium-denuded aortae (P < 0.05 versus endothelium-denuded aortae at 1 × 10−4 to 1 × 10−3 M ropivacaine; Figures 1 and 2(a)), suggesting that attenuation of ropivacaine-induced contraction is endothelium dependent. Pretreatment of endothelium-intact aortae with inhibitors including l-NAME (1 × 10−4 M), N ω-propyl-l-arginine hydrochloride (5 × 10−8 M), 1400W dihydrochloride (1 × 10−6 M), ODQ (1 × 10−5 M), methylene blue (1 × 10−6 M), fluconazole (1 × 10−5 M), indomethacin (3 × 10−5 M), wortmannin (1 × 10−7 M), 4-AP (2 × 10−3 M), TEA (2 × 10−3 M), barium chloride (3 × 10−5 M), and glibenclamide (1 × 10−5 M) did not significantly alter the baseline resting tension (supplementary Figure 1 in Supplementary Material available online at http://dx.doi.org/10.1155/2013/565271). Pretreatment with the nonspecific NOS inhibitor l-NAME (1 × 10−4 M) significantly increased ropivacaine-induced contraction in endothelium-intact aortae (P < 0.001 versus endothelium-intact aortae at 1 × 10−4 to 1 × 10−3 M; Figure 2(a)), whereas the neuronal NOS inhibitor N ω-propyl-l-arginine hydrochloride (5 × 10−8 M) and the inducible NOS inhibitor 1400W dihydrochloride (1 × 10−6 M) had no effect (Figure 2(b)), suggesting that endothelium-dependent attenuation of ropivacaine-induced contraction involves endothelial NO. Pretreatment with the NO-sensitive GC inhibitor ODQ (1 × 10−6 and 1 × 10−5 M) and the NOS and GC inhibitor methylene blue (1 × 10−6 M) significantly increased ropivacaine-induced contraction in endothelium-intact aortae (P < 0.001 versus control at 1 × 10−4 to 1 × 10−3 M; Figures 3(a) and 3(b)), suggesting that endothelium-dependent attenuation of ropivacaine-induced contraction involves the NO-GC pathway. The cytochrome P450 epoxygenase inhibitor fluconazole had no effect on ropivacaine-induced contraction in endothelium-intact aortae (Figure 3(b)), but the cyclooxygenase inhibitor indomethacin (1 × 10−5 and 3 × 10−5 M) attenuated ropivacaine-induced contraction (P < 0.05 versus control at 1 × 10−4 to 1 × 10−3 M; Figure 3(c)).[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
Pretreatment with the voltage-dependent potassium channel inhibitor 4-AP (2 × 10−3 M) greatly enhanced ropivacaine-induced contraction in endothelium-intact aortae (P < 0.001 versus control at 1 × 10−4 to 1 × 10−3 M), and pretreatment with the calcium-activated potassium channel inhibitor TEA (2 × 10−3 M) slightly increased ropivacaine-induced maximal contraction (P < 0.001 versus control at 3 × 10−4 M) (Figure 4(a)), suggesting that ropivacaine-induced contraction is attenuated by voltage-dependent and calcium-activated potassium channels. However, pretreatment with the inward-rectifying potassium channel inhibitor barium chloride (3 × 10−5 M) and the adenosine triphosphate-sensitive potassium channel inhibitor glibenclamide (1 × 10−5 M) had no effect on ropivacaine-induced contraction in endothelium-intact aortae (Figure 4(a)). Ropivacaine-induced contraction was stronger in endothelium-intact aortae pretreated with l-NAME (1 × 10−4 M) plus 4-AP (2 × 10−3 M) or l-NAME (1 × 10−4 M) plus TEA (2 × 10−3 M) than in endothelium-intact aortae pretreated with l-NAME (1 × 10−4 M) alone (P < 0.001 versus 1 × 10−4 M l-NAME alone at 3 × 10−5 and 1 × 10−4 M; Figure 4(b)). Pretreatment with 4-AP (2 × 10−3 M) or TEA (2 × 10−3 M) enhanced phenylephrine-induced contraction in endothelium-intact aortae (P < 0.01 versus control at 3 × 10−7 to 10−5 M; Figure 4(c)), suggesting that phenylephrine-induced contraction is attenuated by voltage-dependent and calcium-activated potassium channels.[](https://www.ncbi.nlm.nih.gov/mesh/D015761)
4-AP (2 × 10−3 to 10−2 M) and TEA (2 × 10−3 to 10−2 M) induced contraction in endothelium-denuded aortae that were precontracted with ropivacaine (1 × 10−4 M) (Figure 5, P < 0.001), suggesting that ropivacaine-induced contraction involves endothelium-independent activation of voltage-dependent and calcium-activated potassium channels of vascular smooth muscle.[](https://www.ncbi.nlm.nih.gov/mesh/D015761)
The PI3K inhibitor wortmannin (1 × 10−7 M) had no effect on ropivacaine-induced contraction in endothelium-intact aortae (Figure 6), suggesting that endothelium-dependent attenuation of ropivacaine-induced contraction does not involve the PI3K-Akt-eNOS pathway.[](https://www.ncbi.nlm.nih.gov/mesh/D000077191)
eNOS phosphorylation was induced in HUVECs at 30 and 60 min after treatment with 1 × 10−4 M ropivacaine (P < 0.05; Figure 7).[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
## 4. Discussion
*In the **4. Discussion** section:*
This study presents novel information suggesting that ropivacaine-induced contraction is attenuated primarily by endothelial NO and voltage-dependent potassium channels in endothelium-intact aortae. The major findings of this in vitro study were as follows: (1) ropivacaine-induced contraction was attenuated in endothelium-intact aortae; (2) l-NAME, ODQ, and methylene blue enhanced ropivacaine-induced contraction in endothelium-intact aortae; (3) 4-AP and TEA enhanced ropivacaine-induced contraction in endothelium-intact aortae with or without l-NAME; (4) eNOS phosphorylation was induced by ropivacaine in HUVECs.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
NO is produced from l-arginine in the endothelium by eNOS [9, 25]. Endothelial NO stimulates GC in the vascular smooth muscle and subsequently induces the formation of cyclic guanosine monophosphate (cGMP) and stimulation of cGMP-dependent protein kinase, which promote vascular smooth muscle relaxation [9, 25]. The attenuation of ropivacaine-induced contraction is endothelium dependent.
In endothelium-intact aortae, the nonspecific NOS inhibitor l-NAME enhanced ropivacaine-induced contraction, whereas the highly selective neuronal NOS inhibitor N ω-propyl-l-arginine hydrochloride and the inducible NOS inhibitor 1400W dihydrochloride did not affect contraction. Taken together, these results suggest that endothelium-dependent attenuation of ropivacaine-induced contraction is associated with eNOS. Ropivacaine produces endothelium-dependent vasodilation in isolated guinea pig aortae precontracted with phenylephrine via a pathway involving NO-GC [10]. In addition, ropivacaine attenuates phenylephrine-induced contraction of endothelium-intact aortae in an endothelial NO-dependent manner [11].
Similar to the results of previous studies that used different methods from those used here, our findings that the NOS inhibitor l-NAME, the NO-sensitive GC inhibitor ODQ, and the combined NOS and GC inhibitor methylene blue enhanced ropivacaine-induced contraction in endothelium-intact aortae suggest that endothelium-dependent attenuation of ropivacaine-induced contraction is associated with activation of the NO-GC-cGMP pathway [10, 11]. Ropivacaine-induced contraction is dependent on calcium influx via voltage-operated calcium channels [4, 26]. Ropivacaine-induced contraction appears to be mediated by cytosolic phospholipase A2 activated by calcium influx [4].
This calcium influx may contribute to activation of eNOS because the eNOS that produces NO binds calmodulin in a calcium-dependent manner [9]. PI3K stimulates Akt (protein kinase B) as a downstream signal molecule and subsequently induces eNOS phosphorylation and vasodilatation, which is a calcium-independent novel mechanism for eNOS activation [23]. The PI3K inhibitor wortmannin had no effect on ropivacaine-induced contraction (Figure 6), suggesting that endothelial NO-mediated attenuation of ropivacaine-induced contraction is not associated with the pathway involving PI3K-Akt-eNOS. Further research on the effect of ropivacaine on the endothelial intracellular concentration of free calcium, which is required for the classic signal pathway of eNOS activation, is needed to elucidate the detailed cellular mechanism of ropivacaine-induced NO release.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Reinforced by the results obtained from isometric tension measurements in the current study, ropivacaine induced eNOS phosphorylation in HUVECs. Because we used HUVECs instead of rat aortic endothelial cells, and considering the heterogeneity of endothelial cells, we should be very cautious about interpreting data obtained from western blotting using HUVECs [27]. In this in vitro study, the time (30 min) required for ropivacaine-induced eNOS phosphorylation in HUVECs appears to be slightly longer than that required for ropivacaine-induced contraction inhibited by endothelial NO release. This difference may be ascribed to differences in vessel location and species. In addition, both levobupivacaine and mepivacaine induce endothelium-dependent NO-mediated attenuation of vasoconstriction and eNOS phosphorylation
[6–8, 16]. As ropivacaine belongs to the family of n-alkyl-substituted pipecolyl xylidine aminoamide local anesthetics that includes levobupivacaine and mepivacaine, endothelium-dependent NO-mediated attenuation of ropivacaine-induced contraction may be a common characteristic of this family of local anesthetics.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
The activation of various potassium channels results in potassium efflux via the opening of potassium channels and subsequently induces membrane hyperpolarization, which leads to the relaxation of vascular smooth muscle through the inhibition of voltage-operated calcium channels [12]. In endothelium-intact aortae, ropivacaine-induced contraction was greatly enhanced by 4-AP and slightly enhanced by TEA, suggesting that ropivacaine-induced contraction involves both the primary activation of voltage-dependent potassium channels and the partial activation of calcium-activated potassium channels. Ropivacaine increases the intracellular free calcium concentration in vascular smooth muscle, which may contribute to the stimulation of calcium-activated potassium channels observed in this study [26]. Glibenclamide and barium chloride had no effect on ropivacaine-induced contraction in endothelium-intact aortae, suggesting that ropivacaine-induced contraction does not involve the activation of adenosine triphosphate-sensitive and inward-rectifying potassium channels. Procaine, an aminoamide local anesthetic, produces vasodilation in aortae precontracted with phenylephrine via both endothelial NO and endothelium-independent calcium-activated potassium channels [28].
Conversely, endothelial NO produced by endothelium-dependent vasodilators stimulates the opening of various potassium channels including voltage-dependent, calcium-activated, and adenosine triphosphate-sensitive potassium channels via the stimulation of cGMP-dependent protein kinase and subsequently produces vasodilation [20, 29, 30]. In the current study, the endothelium-dependent attenuation of ropivacaine-induced contraction appeared to involve endothelial NO release. If l-NAME-mediated enhancement of ropivacaine-induced contraction involves inactivation of the opening of potassium channels induced by a NO-mediated pathway, there would be no significant difference in ropivacaine-induced contraction between endothelium-intact aortae pretreated with l-NAME alone and endothelium-intact aortae pretreated with l-NAME plus potassium channel inhibitor (2 × 10−3 M TEA or 4-AP). However, as 4-AP- and TEA-mediated enhancement of ropivacaine-induced contraction was observed in l-NAME-pretreated endothelium-intact aortae (Figure 4(b)), these results suggest that ropivacaine-induced, voltage-dependent, and calcium-activated potassium channel activation may be mediated by an endothelial NO-independent mechanism.
In addition, 4-AP or TEA produced vasoconstriction in endothelium-denuded aortae precontracted with ropivacaine (Figure 5). Taken together, these results suggest that ropivacaine-induced contraction is attenuated by two independent mechanisms including endothelial NO and endothelium-independent activation of voltage-dependent and calcium-activated potassium channels in vascular smooth muscle. Further research into the effect of potassium channel inhibitors on the voltage-dependent and calcium-activated potassium channel current induced by ropivacaine in vascular smooth muscle cells is needed to elucidate the detailed cellular mechanism. Local anesthetics including bupivacaine and ropivacaine inhibit voltage-dependent and tandem pore domain potassium channels, which may contribute to local anesthetic toxicity, whereas the activation of voltage-dependent and calcium-activated potassium channels accompanied by ropivacaine-induced vasoconstriction observed in this study may be associated with a negative feedback mechanism in which voltage-operated calcium channel-mediated vasoconstriction induced by a contractile agonist (e.g., phenylephrine) limits muscle contraction via the opening of voltage-dependent and calcium-activated potassium channels [12, 31, 32].
Furthermore, 4-AP and TEA increased phenylephrine-induced contraction in the present study (Figure 4(c)), suggesting that phenylephrine-induced contraction also induces the activation of voltage-dependent and calcium-activated potassium channels. Thus, the opening of voltage-dependent and calcium-activated potassium channels accompanied by ropivacaine-induced vasoconstriction appears to be associated with a nonspecific negative feedback mechanism that limits ropivacaine-induced vasoconstriction in vascular smooth muscle cells.[](https://www.ncbi.nlm.nih.gov/mesh/D011188)
One of the major proposed mechanisms responsible for EDHF-induced vasodilation is potassium channel activation induced by epoxyeicosatrienoic acid, which is produced from arachidonic acid via cytochrome P450 epoxygenase [33]. The cytochrome P450 epoxygenase inhibitor fluconazole had no effect on ropivacaine-induced contraction, suggesting that cytochrome P450 epoxygenase-mediated EDHF-induced vasodilation does not contribute to the endothelium-dependent attenuation of ropivacaine-induced contraction. Indomethacin attenuated ropivacaine-induced contraction, suggesting that the endothelium-dependent attenuation of ropivacaine-induced contraction does not involve endothelial prostacyclin. Further investigation into the effect of ropivacaine on the production of arachidonic acid metabolite in endothelial cells is needed.[](https://www.ncbi.nlm.nih.gov/mesh/C067410)
Ropivacaine at lower concentrations induces both vasoconstriction and decreased skin blood flow [1–4]. The combined topical application of ropivacaine and epinephrine does not further reduce sciatic nerve blood flow compared with the topical application of ropivacaine alone, suggesting that adding epinephrine to ropivacaine does not synergistically induce vasoconstriction, which may be due to the strong intrinsic vasoconstriction induced by ropivacaine alone [34]. The clinical relevance of ropivacaine-induced vasoconstriction revealed in this study must be tempered by the fact that the aorta is a conduit vessel, whereas blood flow is controlled by small-resistance arterioles such as rat mesenteric arteries with diameters of less than 100–300 μm [35]. Even with this limitation, vasoconstriction induced by 3 × 10−4 M ropivacaine, which corresponds to 0.093% ropivacaine and is within the clinically relevant concentration (0.2%) of ropivacaine used for local infiltration, may contribute to the vasoconstriction and decreased blood flow observed in previous studies [1–4, 34].
As ropivacaine-induced contraction is attenuated by both endothelial NO and voltage-dependent and calcium-activated potassium channels, the magnitude of ropivacaine-induced contraction may be enhanced in patients with decreased endothelial function and impaired potassium channel function associated with hypertension and diabetes, leading to a longer duration of ropivacaine-induced analgesia [20].[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
In conclusion, these results suggest that ropivacaine-induced contraction is attenuated primarily by both endothelial NO release and the activation of voltage-dependent potassium channels. The activation of voltage-dependent and calcium-activated potassium channels that is induced by ropivacaine-induced contraction seems to be associated with a negative feedback mechanism. In addition, the endothelial NO-mediated attenuation of ropivacaine-induced contraction does not appear to involve the activation of the pathway associated with PI3K-Akt-eNOS.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
## Supplementary Material
*In the **Supplementary Material** section:*
Traces showing the change in tension in endothelium-intact (a) and endothelium-denuded (b) aortae in response to 60 mM KCl and ropivacaine.[](https://www.ncbi.nlm.nih.gov/mesh/D011189)
(a) The effect of endothelial denudation and N ω-nitro-l-arginine methyl ester (l-NAME) on ropivacaine concentration-response curves in isolated aortae. Data are shown as the mean ± SD and expressed as a percentage of the maximal contraction induced by isotonic 60 mM KCl (100% = 2.29 ± 0.19 g [n = 7], 100% = 2.78 ± 0.39 g [n = 6], and 100% = 2.34 ± 0.33 g [n = 7] for untreated endothelium-intact aortae, untreated endothelium-denuded aortae, and endothelium-intact aortae treated with 1 × 10−4 M l-NAME, resp.). N indicates the number of rats from which descending thoracic aortic rings were derived. *P < 0.001 and † P < 0.05 versus endothelium-intact aortae. # P < 0.001 versus 1 × 10−5 M ropivacaine and § P < 0.05 versus 3 × 10−4 M in endothelium-intact aortae. (b) The effect of N ω-propyl-l-arginine hydrochloride and 1400W dihydrochloride on ropivacaine concentration-response curves in endothelium-intact aortae. Data are shown as the mean ± SD and expressed as a percentage of the maximal contraction induced by isotonic 60 mM KCl (100% = 2.44 ± 0.46 g [n = 6], 100% = 2.28 ± 0.27 g [n = 6], and 100% = 2.33 ± 0.33 g [n = 6] for untreated endothelium-intact aortae, endothelium-intact aortae treated with 5 × 10−8 M N ω-propyl-l-arginine hydrochloride, and endothelium-intact aortae treated with 1 × 10−6 M 1400W dihydrochloride, resp.). N indicates the number of rats from which descending thoracic aortic rings were derived.[](https://www.ncbi.nlm.nih.gov/mesh/D019331)
The effect of 1H-[1,2, 4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) (a), methylene blue (b), fluconazole (b), and indomethacin (c) on ropivacaine concentration-response curves in endothelium-intact aortae. Data are shown as the mean ± SD and expressed as a percentage of the maximal contraction induced by isotonic 60 mM KCl. N indicates the number of rats from which descending thoracic aortic rings were derived. (a) 100% = 2.40 ± 0.48 g (n = 6), 100% = 2.55 ± 0.55 g (n = 6), and 100% = 2.70 ± 0.61 g (n = 6) for untreated endothelium-intact aortae, endothelium-intact aortae treated with 1 × 10−6 M ODQ, and endothelium-intact aortae treated with 1 × 10−5 M ODQ, respectively. (b) 100% = 2.00 ± 0.32 g (n = 9), 100% = 2.12 ± 0.40 g (n = 7), and 100% = 2.10 ± 0.42 g (n = 6) for untreated endothelium-intact aortae, endothelium-intact aortae treated with 1 × 10−6 methylene blue, and endothelium-intact aortae treated with 1 × 10−5 M fluconazole, respectively. (c) 100% = 2.08 ± 0.27 g (n = 6), 100% = 2.09 ± 0.40 g (n = 6), and 100% = 2.47 ± 0.67 g (n = 6) for untreated endothelium-intact aortae, endothelium-intact aortae treated with 1 × 10−5 M indomethacin, and endothelium-intact aortae treated with 3 × 10−5 M indomethacin, respectively. (a) and (b): *P < 0.001 versus control. (c): *P < 0.001, † P < 0.05, and # P < 0.01 versus control.[](https://www.ncbi.nlm.nih.gov/mesh/C095284)
(a) and (b) The effect of 4-aminopyridine (4-AP), tetraethylammonium (TEA), barium chloride, and glibenclamide on ropivacaine concentration-response curves in endothelium-intact aortae without or with 1 × 10−4 M N ω-nitro-l-arginine methyl ester (l-NAME). (c) The effect of 4-AP and TEA on phenylephrine concentration-response curves in endothelium-intact aortae. (a) and (b) Data are shown as the mean ± SD and expressed as a percentage of the maximal contraction induced by isotonic 60 mM KCl. N indicates the number of rats from which descending thoracic aortic rings were derived. (a) 100% = 2.38 ± 0.27 g (n = 10), 100% = 2.48 ± 0.47 g (n = 6), 100% = 2.20 ± 0.22 g (n = 6), 100% = 2.32 ± 0.31 g (n = 5), and 100% = 2.39 ± 0.31 g (n = 5) for untreated endothelium-intact aortae and endothelium-intact aortae treated with 2 × 10−3 M 4-AP, 2 × 10−3 M TEA, 3 × 10−5 M barium chloride, and 1 × 10−5 M glibenclamide, respectively. (b) 100% = 2.37 ± 0.22 g (n = 6), 100% = 2.81 ± 0.37 g (n = 6), 100% = 2.71 ± 0.30 g (n = 6), and 100% = 2.77 ± 0.14 g (n = 6) for untreated endothelium-intact aortae, endothelium-intact aortae pretreated with 1 × 10−4 M l-NAME alone, endothelium-intact aortae pretreated with 1 × 10−4 M l-NAME plus 2 × 10−3 M 4-AP, and endothelium-intact aortae pretreated with 1 × 10−4 M l-NAME plus 2 × 10−3 M TEA, respectively. (c) Data are shown as the mean ± SD and expressed as a percentage of the maximal contraction induced by isotonic 60 mM KCl (100% = 2.74 ± 0.32 g (n = 5), 100% = 2.77 ± 0.35 g (n = 5), and 100% = 2.66 ± 0.38 g (n = 5) for untreated endothelium-intact aortae and endothelium-intact aortae treated with 2 × 10−3 M 4-AP and 2 × 10−3 M TEA, resp.). N indicates the number of descending thoracic aortic rings. (a): *P < 0.001 versus control. (b) *P < 0.001, † P < 0.01 and # P < 0.05 versus control. § P < 0.001 versus 1 × 10−4 M l-NAME alone. (c) *P < 0.001 and † P < 0.01 versus control.[](https://www.ncbi.nlm.nih.gov/mesh/D015761)
Cumulative concentration-response curves induced by 4-aminopyridine (4-AP) and tetraethylammonium (TEA) in endothelium-denuded aortae precontracted with 1 × 10−4 M ropivacaine. Data are shown as the mean ± SD and expressed as a percentage of the maximal contraction induced by ropivacaine (1 × 10−4 M) (100% = 0.94 ± 0.28 g (n = 10) and 100% = 0.97 ± 0.25 g (n = 10) for endothelium-denuded aortae with 4-AP and TEA, resp.). N indicates the number of descending thoracic aortic rings. *P < 0.001 versus ropivacaine (1 × 10−4 M).[](https://www.ncbi.nlm.nih.gov/mesh/D015761)
The effect of wortmannin (1 × 10−7 M) on ropivacaine concentration-response curves in endothelium-intact aortae. Data are shown as the mean ± SD and expressed as a percentage of the maximal contraction induced by isotonic 60 mM KCl (100% = 2.61 ± 0.14 g [n = 5] and 100% = 2.46 ± 0.20 g [n = 5] for untreated endothelium-intact aortae and endothelium-intact aortae treated with 1 × 10−7 M wortmannin, resp.). N indicates the number of rats from which descending thoracic aortic rings were derived.[](https://www.ncbi.nlm.nih.gov/mesh/D000077191)
Effect of ropivacaine on the activation of endothelial nitric oxide synthase (eNOS; n = 4) by phosphorylation at Ser1777 in human umbilical vein endothelial cells (HUVECs). HUVECs were treated with ropivacaine (1 × 10−4 M) for 5, 10, 30, and 60 min. (a) Phosphorylation of eNOS was examined by western blotting, as described in Methods. (b) Band intensities at 5, 10, 30, and 60 min were assessed by scanning densitometry. Data are shown as the mean ± SD. N indicates the number of independent experiments. *P < 0.05 versus control. t-eNOS: total eNOS; P-eNOS: phosphorylated eNOS.[](https://www.ncbi.nlm.nih.gov/mesh/D000077212)
|
# Introduction
Changes in Oxidative Damage, Inflammation and [[NAD(H)](https://www.ncbi.nlm.nih.gov/mesh/D009243)] with Age in Cerebrospinal Fluid
# Abstract
*In the **Abstract** section:*
An extensive body of evidence indicates that oxidative stress and inflammation play a central role in the degenerative changes of systemic tissues in aging. However a comparatively limited amount of data is available to verify whether these processes also contribute to normal aging within the brain. High levels of oxidative damage results in key cellular changes including a reduction in available nicotinamide adenine dinucleotide (NAD+), an essential molecule required for a number o[f vital cellular processes includ](https://www.ncbi.nlm.nih.gov/mesh/D009243)in[g DN](https://www.ncbi.nlm.nih.gov/mesh/D009243)A repair, immune signaling and epigenetic processing. In this study we quantified changes in [NAD(H)] and markers of inflammation and oxidative damage (F2-isoprostanes, 8-OHdG, tota[l anti](https://www.ncbi.nlm.nih.gov/mesh/D009243)oxidant capacity) in the cerebrospinal fluid (CSF) o[f healthy human](https://www.ncbi.nlm.nih.gov/mesh/D028441)s [across](https://www.ncbi.nlm.nih.gov/mesh/C067134) a wide age range (24–91 years). CSF was collected from consenting patients who required a spinal tap for the administration of anesthetic. CSF of participants aged >45 years was found to contain increased levels of lipid peroxidation (F2-isoprostanes) (p = 0.04) and inflammation (IL-6) (p = 0.00) and [decre](https://www.ncbi.nlm.nih.gov/mesh/D008055)ased levels of [both total anti](https://www.ncbi.nlm.nih.gov/mesh/D028441)oxidant capacity (p = 0.00) and NAD(H) (p = 0.05), compared to their younger counterparts. A positive association was a[lso ob](https://www.ncbi.nlm.nih.gov/mesh/D009243)served between plasma [NAD(H)] and CSF NAD(H) levels (p = 0.03). Further analysis of the data identified a rel[ations](https://www.ncbi.nlm.nih.gov/mesh/D009243)hip betwee[n alco](https://www.ncbi.nlm.nih.gov/mesh/D009243)hol intake and CSF [NAD(H)] and markers of inflammation. The CSF of participants wh[o consu](https://www.ncbi.nlm.nih.gov/mesh/D000438)med >1 standard d[rink o](https://www.ncbi.nlm.nih.gov/mesh/D009243)f alcohol per day contained lower levels of NAD(H) compared to those who consumed no alco[hol (p<](https://www.ncbi.nlm.nih.gov/mesh/D000438)0.05). An increase in CSF IL-6 was [observ](https://www.ncbi.nlm.nih.gov/mesh/D009243)ed in participants who reported dri[nking >](https://www.ncbi.nlm.nih.gov/mesh/D000438)0–1 (p<0.05) and >1 (p<0.05) standard alcoholic drinks per day compared to those who did not drink alcohol. Taken together these data suggest a progressive age associated increase in oxi[dative ](https://www.ncbi.nlm.nih.gov/mesh/D000438)damage, inflammation and reduced [NAD(H)] in the brain which may be exacerbated by alcohol intake.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Aging is an unavoidable biological process characterized by a progressive decline in physiological and biochemical function resulting in an increased predisposition to disease. In 1956 Harman proposed the oxidative stress theory of aging suggesting that the accumulation of unrepaired oxidative damage results in the typical aging phenotype . The term ‘oxidative stress’ describes a significant imbalance between antioxidant defenses and the bodies’ formation of reactive nitrogen and/or oxygen species (ROS). While there are several sources of ROS within the body, the primary source is generally agreed to be the leakage of electrons to ground state oxygen from early components of the mitochondrial electron transport chain, resulting in the production of the superoxide radical (O2 •–) , . Importantly, at modest concentrations, ROS are used in a variety of normal physiological functions.
Although there is the potential for damage, this is kept in check by an intricately connected antioxidant defense and repair system . However, under conditions of reduced antioxidant capacity or excess production, ROS can cause indiscriminant damage to cellular constituents (DNA, proteins and lipids) that, if unrepaired, may lead to cell death and tissue dysfunction.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
The brain is particularly vulnerable to oxidative damage as a consequence of its high oxygen demand, high level of both polyunsaturated fatty acids and transition metals, and poor antioxidant defenses –. As we age, the vulnerability of the brain to oxidative damage increases due to reduced integrity of the blood brain barrier and amplified mitochondrial dysfunction , . Indeed animal and tissue studies have shown the aging brain to be accompanied by an accumulation of markers of lipid, protein and DNA oxidative damage –. Failure to repair this damage has been demonstrated to cause genomic instability and neuronal apoptosis and is associated with the development of neuropathologies such as Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis –.[](https://www.ncbi.nlm.nih.gov/mesh/D010100)
Both normal brain aging and neurodegenerative disease are characterized by increased inflammation associated with microglial over activation and a subsequent rise in pro-inflammatory cytokines –. Excessive release of pro-inflammatory cytokines further promotes a pro-oxidative state and neuronal degradation. Elevated levels of the inflammatory cytokine IL-6 have been associated with cognitive impairment and the induction of Alzheimer’s-type hyperphosphorylation of tau protein , .
As inflammation and oxidative damage rise with age a decrease in available nicotinamide adenine dinucleotide (NAD+) has been observed in multiple organs of the rat , including the brain (data unpublished). NAD+ is a ubiquitous molecule that is required for a number of vital cellular processes. In addition to its role in cellular energy and metabolism there are several enzymes, including poly(ADP-ribose) polymerase 1 (PARP) and silent information regulators (e.g. SIRT1), that use NAD+ as their substrate –. Importantly PARP activation in response to DNA damage catalyzes the successive cleavage of the ADP-ribose moiety from NAD+ resulting in the formation of poly(ADP-ribose) subunits. Under conditions of mild-to-moderate DNA damage this process facilitates DNA repair .
However over-activation of PARP, due to excessive DNA damage, can result in neuronal death as a consequence of decreased ATP production due to NAD+ depletion –. In order to preserve cellular energy and concomitantly SIRT1 (associated with maintaining cellular longevity) and PARP activity, adequate levels of NAD+ must be sustained.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
Inflammation and oxidative activity, when adequately regulated, form part of the normal physiology in all age groups. However during the course of life, each individual will experience (to varying degrees) a gradual increase in oxidative damage burden in many of the body’s tissues. If experienced in the brain, neurodegenerative changes are likely. While age is the major risk factor for the development of most neurodegenerative disorders , , a number of lifestyle choices can also promote pathogenesis by increasing oxidative stress. Previous studies have indicated that chronic alcohol exposure in humans’ results in neurodegeneration, ranging from minor synaptic and dendritic changes to neuronal cell death .
More recently administration of alcohol to rats was shown to significantly increase brain mitochondrial lipid and protein oxidation and decrease superoxide dismutase mRNA expression and ATP-ase activity. The authors postulated that the alcohol-induced production of ROS alters mitochondrial membrane properties leading to mitochondrial dysfunction and subsequently further ROS production .[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
Collectively these reports indicate that certain lifestyle choices may accelerate the development of an age associated oxidative-inflammatory state leading to increased tissue damage and reduced DNA repair capacity (through reduced NAD+ availability) within the central nervous system. While evidence from cell culture, animal and limited postmortem brain tissue studies support this hypothesis, to date no study has investigated this in a healthy human cohort.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
In this study we investigated whether markers of oxidative and inflammatory activity increase with age in the central nervous system of relatively healthy humans and whether this was associated with a decrease in available NAD(H). We further correlated changes in CSF oxidative/inflammatory markers and [NAD(H)] with specific modifiable lifestyle factors.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## Materials and Methods
*In the **Materials and Methods** section:*
## Ethics Statement
*In the **Ethics Statement** section:*
This study was conducted in accordance with the Helsinki declaration. Ethical approval was obtained from the Human Research Ethics Committee, Sydney Adventist Hospital (HREC# 2011-005). Written informed consent was obtained from all participants prior to commencement.
## Participants
*In the **Participants** section:*
Male (n = 20) and female (n = 50) participants, who required a spinal tap for the administration of anesthetic as part of routine care, were recruited at Sydney Adventist Hospital, Australia. The average age of participants was 53 years (SD = 19.9, interquartile range = 38). Participants were excluded from the cohort if they were smokers or had a confirmed diagnosis of a neurological/neurodegenerative disorder or central nervous system infection. In total 70 CSF and 38 matched blood samples were collected from consenting participants considered in general good health.
## Sample Collection
*In the **Sample Collection** section:*
Fasting (≥10 h) blood and CSF samples were collected by an accredited anesthetist no longer than 30 minutes apart. CSF samples were collected, prior to injection of spinal anesthetics, via standard lumber puncture. Blood samples were collected into heparinized tubes from an intravenous cannula inserted into a superficial vein on an upper limb, prior to the administration of fluids or anesthetics. Samples were prepared by centrifuging at 1800 rpm for 10 minutes and stored within 1 hour of collection, at −194 degrees Celsius until analysis. Samples intended for F2-isoprostane analysis where stored in the presence of a glutathione/butylated hydroxytoluene preservative.[](https://www.ncbi.nlm.nih.gov/mesh/D028441)
## Assessment of Alcohol Consumption
*In the **Assessment of Alcohol Consumption** section:*
Alcohol consumption was assessed via questionnaire upon hospital admission. Specifically participants were asked ‘Do you drink alcohol.’ If this was affirmative participants were asked to provide the number of standard drinks consumed per day.[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
## Total NAD(H)
*In the **Total NAD(H)** section:*
Total NAD(H) concentrations in plasma and CSF samples were measured spectrophotometrically using a thiazolyl blue microcycling assay established by Bernofsky and Swan (1973) , and adapted to a 96 well plate format by Grant and Kapoor (1998) . In brief, 125 µL of the reaction mixture containing 120 mM bicine (pH 7.8), 0.5 mM MTT, 2 mM PMS, 0.6 M ethanol and alcohol dehydrogenase (300 units/mL) was added to either 6 µL of plasma or 20 µL of CSF. Following a 10 minute incubation at 37 degrees Celsius, the concentration of total NAD(H) was measured, using a Model 680XR microplate reader (BioRad, Hercules), as the change in absorbance at 570 nm.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## Total Antioxidant Capacity (TAC)
*In the **Total Antioxidant Capacity (TAC)** section:*
CSF total antioxidant capacity was measured using a standardized commercial kit (Antioxidant Assay Kit, Cayman Chemical Company, Ann Arbor, MI USA). This assay relies on the ability of antioxidants to inhibit oxidation of 2,2′-Azino-di-[3-ethylbenzthiazoline sulphonate] (ABTS) by metmyoglobin. Briefly, 10 µL of metmyoglobin was added to 10 µL of diluted sample. 150 µL of chromagen, containing ABTS, was subsequently added and the reaction initiated by adding 40 µL of 441 µM hydrogen peroxide. The plate was incubated on a shaker for 5 minutes and the amount of oxidized ABTS was measured spectrophotometrically, at an absorbance of 750 nm, using a Model 680XR microplate reader (BioRad, Hercules).[](https://www.ncbi.nlm.nih.gov/mesh/C002502)
## Interleukin-6 (IL-6)
*In the **Interleukin-6 (IL-6)** section:*
IL-6 was measured using a standardized commercial solid phase sandwich enzyme linked-immuno-sorbent assay (ELISA) (Human IL6 High Sensitivity ELISA Kit, Abcam, Cambridge, MA USA). Briefly, 100 µL of CSF was added to a plate pre-coated with a monoclonal antibody specific for IL-6. 50 µL of biotinylated anti-IL6 was then added and the plate incubated for 3 hours. After incubation, the plate was washed before the addition of the enzyme, horseradish peroxidase. The plate was incubated for 30 minutes and then washed again to remove any unbound enzymes.
The 3,3′,5,5′-Tetramethylbenzidine substrate was added and the plate was incubated in the dark for 12–15 minutes, after which H2SO4 was added to the wells to stop the enzyme-substrate reaction. The intensity of this colored product is directly proportional to IL-6 concentration. Absorbance was measured at 450 nm using a Model 680XR microplate reader (BioRad, Hercules).[](https://www.ncbi.nlm.nih.gov/mesh/C021758)
## 8-hydroxy-2′-deoxyguanosine (8-OHdG)
*In the **8-hydroxy-2′-deoxyguanosine (8-OHdG)** section:*
8-OHdG was measured in CSF samples using a standardized commercial competitive ELISA (Highly Sensitive 8-OHdG Check, Japan Institute for the Control of Aging, Shizuoka Japan). Briefly 50 µL of sample or standard and 50 µL monoclonal antibody was adsorbed onto a 96-well plate precoated with 8 OHdG. Following an overnight incubation at 4 degrees Celsius the plate was washed and incubated with a secondary antibody for 1 hour. The plate was washed again before the addition of a chromatic solution for 15 minutes, after which the reaction was terminated and absorbance was measured at 450 nm using a Model 680XR microplate reader (BioRad, Hercules).[](https://www.ncbi.nlm.nih.gov/mesh/C067134)
## F2-Isoprostanes
*In the **F2-Isoprostanes** section:*
Total F2-Isoprostanes were measured in CSF by gas chromatography–mass spectrometry (GC-MS) using electron capture negative ionization according to a modification in the method of Mori et al (1999) . Briefly, after the addition of an internal standard (15-F2t-IsoP-d4, 5 ng), plasma and CSF samples (200 µL) were hydrolyzed with KOH in methanol, acidified, and applied to prewashed Certify II cartridges (Varian). Following washing with methanol:water (1∶1) and hexane:ethyl acetate (75∶25) the F2-Isoprostanes were eluted with ethyl acetate:methanol (90∶10), dried, and derivatized. The F2-Isoprostanes were quantitated by monitoring ions at m/z 569 and 573 for F2-Isoprostanes and 15-F2t-IsoP-d4 respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D028441)
## Statistical Analysis
*In the **Statistical Analysis** section:*
Statistical analyses were performed using SPSS version 16.0 and GraphPad Prism version 5 for Windows. Data is presented as means ± standard deviation unless otherwise stated. Multiple linear regression, controlling for age, was used to identify significant relationships between CSF [NAD(H)], plasma [NAD(H)], and markers of oxidative damage and inflammation. The Independent T Test or Mann-Whitney U Test was employed to analyze the effect of age on markers of oxidative damage, inflammation and metabolism. The Wilcoxon Signed Ranks Test was used to identify the association between mean plasma and CSF NAD(H) levels.
Kruskal-Wallis with Dunn’s post-hoc test was used to assess the effect of alcohol consumption on markers of oxidative damage, inflammation and metabolism. Both the Kolmogorov-Smirnov, Shapiro-Wilk and histogram analysis was used to check normality of the variables. When required the Levene’s Test of Equality was used to check homogeneity of variances between groups. If the variances of the groups were found to be either not homogenous and/or normality tests for the variables were not significant then further investigation with graphical displays was performed to assess the distributions of the variables. Both adjusted and non-adjusted P-values are provided throughout with test significance set at P value ≤0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## Results
*In the **Results** section:*
## Age Associated Differences in CSF Markers of Oxidative and Inflammation
*In the **Age Associated Differences in CSF Markers of Oxidative and Inflammation** section:*
A number of studies have shown that lifestyle behaviors in midlife (i.e. around 45–50 years) are associated with reduced cognitive function in later life –. Thus to assess age related differences in CSF markers of oxidative damage and inflammation, for this analysis participants were divided into two groups, aged ≤45 years and >45 years.
In this cohort older age was associated with an increase in a number of CSF oxidative and inflammatory markers (Table 1). Specifically CSF lipid peroxidation (F2-isoprostane) was significantly increased in those aged >45 years; 417.49±34.39 pmol/L compared to those ≤45 years; 395.09±34.04 pmol/L (p = 0.04). Those over 45 years also showed significantly increased levels of the inflammatory cytokine IL-6 (p = 0.00, 2.37±1.93 vs. 0.71±0.43 pg/mL for those ≤45 years) and reduced CSF TAC levels (p = 0.00, 0.90±0.28 vs. 1.49±0.51 nmol/mg protein for those <45 years). Those aged >45 years also tended to have raised levels of CSF 8-OHdG, a marker of oxidative DNA damage, although this did not quite reach statistical significance (p = 0.06) (Table 1).[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
Due to small sample volume, some tests have one or more missing data. Comparisons made using the Independent T Test or Mann-Whitney U Test.
p≤0.05 compared to ≤45 years,
p≤0.001 compared to ≤45 years.
Differences in selected CSF markers according to age and gender.
Assessing these trends in each gender revealed that the CSF of females >45 years contained significantly higher levels of IL-6 (p = 0.00, 1.71±1.23 vs. 0.69±0.43 pg/mL) and lower TAC levels (p = 0.00, 0.98±0.30 vs. 1.49±0.51 nmol/mg protein) than their younger counterparts. Due to low number of male participants ≤45 years, valid comparisons were not possible.
## Age Associated Decrease in CSF NAD(H)
*In the **Age Associated Decrease in CSF NAD(H)** section:*
CSF NAD(H) levels were significantly lower in participants aged >45 years compared to those aged ≤45 years; 75.88±30.14 vs. 88.59±21.07 µg/mL respectively (p = 0.05) (Table 1). Assessing these trends in each gender revealed that, the CSF of females >45 years contained significantly lower levels of NAD(H) compared to their younger counterparts (p = 0.01, 73.13±25.99 vs. 89.98±19.75 µg/mL respectively). Due to low number of male participants ≤45 years, valid comparisons were not possible.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## Inter-correlation between CSF NAD(H), Oxidative Damage and Inflammatory Markers
*In the **Inter-correlation between CSF NAD(H), Oxidative Damage and Inflammatory Markers** section:*
After controlling for age, a significant inverse association was observed between CSF log TAC and CSF 8 OHdG (p = 0.05, n = 37) (Figure 1A). A significant positive association was observed between CSF DNA (8 OHdG) and lipid oxidation (F2-isoprostane) markers (p = 0.01, n = 34) (Figure 1B). An inverse association between CSF [NAD(H)] and F2-isoprostane levels was also found (p = 0.02, n = 48) although this did not remain statistically significant after controlling for age (p = 0.06) (Figure 1C).
No further associations were apparent between CSF [NAD(H)], oxidative damage and inflammatory markers.[](https://www.ncbi.nlm.nih.gov/mesh/C067134)
Inter-correlation between CSF NAD(H), oxidative damage and inflammatory markers. (A) Positive association between CSF 8 OHdG and CSF total antioxidant capacity.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
A significant positive association was observed between CSF 8(TAC) (p = 0.05, n = 37). Comparisons were made using multiple linear regression controlling for age. (B) Positive association between CSF 8 OHdG and CSF F2 Isoprostane levels. A significant positive association was observed between CSF 8 OHdG and F2 Isoprostane levels (p = 0.01, n = 34). Comparisons were made using multiple linear regression controlling for age. (C) Inverse association between CSF [NAD(H)] and CSF F2 Isoprostane levels. A significant inverse association was observed between CSF [NAD(H)] and F2 Isoprostane levels (p = 0.02, n = 48). Comparisons were made using the Pearson correlation coefficient and multiple linear regression controlling for age.[](https://www.ncbi.nlm.nih.gov/mesh/C067134)
## CSF [NAD(H)] Correlates with Peripheral [NAD(H)]
*In the **CSF [NAD(H)] Correlates with Peripheral [NAD(H)]** section:*
After controlling for age, a significant positive relationship was observed between plasma and CSF NAD(H) concentrations (p = 0.03, n = 38) (Figure 2). An increase of one µg/mL in plasma [NAD(H)] was associated with a 0.11 µg/mL increase in CSF [NAD(H)]. The mean level of NAD(H) was significantly lower in CSF (82.24±26.59 µg/mL, n = 38) compared to plasma (358.81±98.56 µg/mL, n = 38) (p = 0.00).[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
Positive association between plasma and CSF NAD(H) levels.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
A significant positive relationship was observed between plasma and CSF NAD(H) concentrations (p = 0.03, n = 38). An increase of one µg/mL in plasma [NAD(H)] was associated with a 0.11 µg/mL increase in CSF [NAD(H)]. Comparisons were made using multiple linear regression controlling for age.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## Influence of Alcohol Intake on CSF [NAD(H)], Oxidative Damage and Inflammation
*In the **Influence of Alcohol Intake on CSF [NAD(H)], Oxidative Damage and Inflammation** section:*
The data was further analyzed to identify possible relationships between alcohol intake and CSF [NAD(H)] and markers of inflammation and oxidative damage. For this analysis, participants were divided into three groups, those who consumed zero (n = 32), >0–1 (n = 14) and >1 (n = 8) standard alcoholic drinks per day. CSF NAD(H) levels were significantly different between the groups (p = 0.02). Specifically the CSF of participants who consumed >1 standard drink of alcohol per day contained significantly lower levels of NAD(H) compared to those who reported consuming zero drinks per day; 62.39±19.93 vs. 86.93±25.32 µg/mL respectively (p<0.05) (Figure 3A). A significant increase in CSF IL-6 was also observed in participants who drank >1 (p<0.05) and 0–1 (p<0.05) standard alcoholic drinks per day compared to those who did not consume alcohol; 2.11±1.28, 2.25±1.66 vs. 1.16±1.67 pg/mL respectively (Figure 3B).
No associations were found between alcohol consumption and CSF markers of DNA (8-OHdG) and lipid (F2-isoprostanes) oxidative damage or total antioxidant capacity. When the data was stratified according to both age and gender CSF IL6 levels remained significantly higher in female participants who drank >0–1 or ≥1 standard alcoholic drinks per day. No other observations remained statistically significant after stratifying for age and gender.[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
Association between alcohol consumption and CSF (A) [NAD(H)] (B) IL-6 levels. (A) Alcohol consumption is associated with decreased levels of CSF NAD(H).[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
Participants who consumed zero (n = 32), >0–1 (n = 14) and >1 (n = 8) standard alcoholic drinks per day were found to contain significantly different levels of CSF NAD(H) (p = 0.02). Specifically the CSF of participants who consumed >1 standard drink of alcohol per day contained significantly lower levels of NAD(H) compared to those who reported consuming zero drinks per day; 62.39±19.93 vs. 86.93±25.32 µg/mL respectively (p<0.05). Comparisons were made using the Kruskal-Wallis with Dunn’s post-hoc test. Error bars represent 95% confidence intervals. (B) Alcohol consumption is associated with increased levels of CSF IL-6. A significant increase in CSF IL-6 was observed in participants who drank >1 (p<0.05) and >0–1 (p<0.05) standard alcoholic drinks per day compared to those who drank zero; 2.12±1.28, 2.25±1.66 vs. 1.16±1.67 pg/mL respectively. Comparisons were made using the Kruskal-Wallis with Dunn’s post-hoc test. Error bars represent 95% confidence intervals.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## Gender Associated Differences in CSF Markers of Oxidative Damage and Inflammation
*In the **Gender Associated Differences in CSF Markers of Oxidative Damage and Inflammation** section:*
In this cohort, the CSF of male participants (n = 18) contained significantly higher levels of IL-6 compared to females (n = 42) (p = 0.00); 2.82±2.23 vs. 1.03±0.91 pg/mL respectively. However when the data was stratified into age groups, this observation did not remain statistically significant (Table 1). No significant differences were observed between males and females for any other markers of inflammation, oxidative damage or [NAD(H)].[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
## Discussion
*In the **Discussion** section:*
Considerable evidence now indicates that both inflammation and oxidative stress contribute to the development of various neuropathologies including Alzheimer’s and Parkinson’s disease , , . While age is the major risk factor for the development of most neurodegenerative disorders it has yet to be confirmed if oxidative stress and inflammation increase during normal brain aging in humans. To our knowledge this is the first study to show that both oxidative damage and inflammation increase after the age of 45 in the central nervous system (CNS) of relatively healthy humans. In this study we report that the CSF of participants aged over 45 years contained statistically higher amounts of the oxidative damage marker F2-isoprostane and the inflammatory cytokine IL-6. Those aged over 45 years also tended to have increased CSF levels of the DNA damage marker 8-OHdG.
These data are consistent with previous results from both our laboratory and others showing that DNA and lipid oxidation increase with age in multiple organs, including the brain in animals , , , . While limited research has been conducted within the CNS of living humans, an age related accumulation in markers of both oxidative damage (8-OHdG) and inflammation (IL-6) has been previously reported in postmortem brain tissue , .[](https://www.ncbi.nlm.nih.gov/mesh/D028441)
It is well established, that oxidative DNA damage activates the NAD-dependent DNA repair enzyme, PARP, which is involved in base excision repair . Utilizing unexposed human skin, our laboratory has previously shown that that PARP activity increases with age and correlates with NAD+ depletion . In the present study we investigated whether levels of CSF NAD(H) were also associated with age and report for the first time that [NAD(H)] does decline with age in the CNS of healthy humans. As expected an inverse trend between CSF [NAD(H)] and markers of central DNA (8-OHdG) and lipid (F2-isoprostanes) oxidative damage was also observed. In addition, as would be predicted, after controlling for age, increased CSF total antioxidant capacity was significantly correlated with higher CSF levels of NAD(H).[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
Adequate levels of NAD(H) are required to maintain normal cellular functions. In addition to its role in cellular energy metabolism there are several enzymes, in addition to PARP, that use the oxidized form, NAD+ as the sole substrate for their activities –. Notably SIRT1, a member of a highly conserved family of histone deacetylases modulates key transcription factors such as FOXO and pro-apoptotic p53, and is thought to play a central role in cell longevity and aging , . As both PARP and SIRT1 compete for the same intracellular pool of NAD+ it has been suggested that depletion of NAD+ results in reduced SIRT1 deacetylase activity , . Further PARP over-activation has been shown to reduce ATP production due to NAD+ depletion resulting in neuronal death –.
The theory that excessive NAD+ depletion facilitates cell death is supported by observations in rodent models of brain ischemia and Alzheimer’s disease where significantly reduced levels of total cellular NAD+ occur prior to neuronal death –. Adequate NAD+ levels are therefore required to maintain cellular energy and robust SIRT1 activity. However further functional studies are required to determine the level of biochemical impact the relatively modest (∼14%) decrease in NAD(H) levels would have on cell metabolism.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
While the physiologic and pathologic importance of NAD+/NADH dependant mechanisms’ in both the central nervous system and periphery is apparent, it was not previously known whether peripheral stores of NAD(H) influence central NAD(H) levels. This study is the first to show a positive correlation between matched CSF and plasma NAD(H) levels in a healthy human cohort. This is consistent with a previous study by Rex and colleagues (2002) who observed that NAD(H) in both its oxidized and reduced forms is capable of crossing the blood brain barrier in rats . While evidence indicates that the brain is capable of independently synthesizing NAD+ , , results from the present study suggest that NAD(H) levels in the brain are potentially influenced by peripheral NAD(H) levels and consequently lifestyle choices that affect the peripheral NAD(H) pool.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
Although age is the major risk factor for the development of most neurodegenerative disorders , , a number of lifestyle choices can increase central oxidative damage and inflammation and thereby promote disease. In this cohort we observed that consumption of less than half a glass or more of alcohol per day was associated with a statistically significant increase in CSF levels of the inflammatory cytokine IL-6. While excessive alcohol consumption is generally agreed to cause alteration in brain structure, function and loss of brain mass , , the effect of low to moderate alcohol consumption on brain health is still debated within the literature. Some authors suggest that low to moderate alcohol consumption may improve cognitive functioning and even reduce the risk of Alzheimer’s disease , . In contrast to these largely epidemiological studies, a meta-analysis conducted by Verbaten (2009) assessing the effects of alcohol consumption on brain integrity, determined by both MRI and cognitive performance, concluded that the consumption of even light to moderate doses of alcohol lead to shrinkage of the brain and to decreases in grey matter volume .[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
The suggested pathways by which alcohol may damage the brain are numerous and include, but are not limited to, disruption of neural cell adhesion molecules –, promotion of endoplasmic reticulum protein misfolding , neuronal hypersensitivity to excitotoxic insults , reduction of endogenous antioxidants , and increased free radical damage to both blood brain barrier endothelial cells as well as neurons , . Consistent with our findings, others have also shown that alcohol, even at low/moderate concentrations, can act as a ligand for toll like receptor 4 (TLR4), stimulating the mitogen-activated protein kinases (MAPKs) and the transcription factor NF-κB pathways, leading to the production of nitric oxide and inflammatory cytokines , .[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
We also report for the first time an inverse relationship between alcohol consumption and CSF [NAD(H)]. Specifically the CSF of participants who consumed greater than one glass of alcohol per day had significantly lower levels of NAD(H) compared to those who did not drink alcohol. While research investigating the effect of alcohol on NAD(H) is scarce our results are consistent with a very early report by McElfresh and McDonald (1983) who also observed in Drosophila that NAD+ levels decrease by at least 20% in response to ethanol stress . Additionally recent data from our laboratory (unpublished) indicates that acute ethanol exposure (10 mM, equivalent to a blood alcohol of 0.05%) decreases intracellular [NAD(H)] in cultured human primary astrocytes (brain metabolic support cells) by as much as 64% within 30 minutes in conjunction with an increase in oxidative damage and PARP activity. By increasing CNS inflammation and oxidative damage, alcohol consumption may stimulate PARP over-activation and subsequently decrease central NAD(H) levels promoting senescence and neurodegeneration.[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
While the observations reported in this study are statistically valid it is recognized that these associations have been obtained from a modest number of participants (38 plasma/CSF matched and 70 CSF). Though both genders were well represented in the older age groups, due to the difficulty in obtaining CSF samples from essentially healthy individuals greater numbers of females were represented in the younger age range. The disproportionately low number of younger male participants, as well as the small number of participants for which alcohol consumption data was available, prevented a comprehensive analysis on how gender may influence our findings. Self-reported alcohol consumption was also relied upon in this study and may have introduced some error into our analysis. However while such error may cause the levels of alcohol reported to slightly differ from the number of glasses actually consumed, it is unlikely to significantly affect the rank order of participants.
Finally the restricted volume of sample collected as part of this study limited our analytical profile negating comparisons with other important molecular species such as the range of anti-inflammatory cytokines. Future studies overcoming these limitations are required to verify the consistency of our observations.[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
## Conclusion
*In the **Conclusion** section:*
An extensive body of evidence indicates that oxidative stress and inflammation play a central role in the physiology of aging. However, comparatively limited data are available to verify whether these processes also contribute to normal aging within the brain. This study reports for the first time a potential link between aging, increased oxidative stress, inflammation and alcohol consumption and a decline in the essential pyridine nucleotide [NAD(H)], in the CSF of a healthy human population. We also provide evidence of a relationship between peripheral and central NAD(H) stores.[](https://www.ncbi.nlm.nih.gov/mesh/D000438)
Taken together these data suggest a progressive age associated increase in oxidative damage, inflammation and reduced NAD(H) in the brain which may be exacerbated by certain lifestyle choices such as regular alcohol consumption.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
As reduced NAD(H) levels impact at least PARP and SIRT1 activity the observed decrease in NAD(H) availability within the aging brain may facilitate cell metabolic and genomic instability increasing an individuals’ susceptibility to degenerative disease. However further follow-up of the participants characterized in this study is required to confirm this hypothesis.[](https://www.ncbi.nlm.nih.gov/mesh/D009243)
# References
*In the **References** section:*
|
# Introduction
Low-dose angiostatic tyrosine kinase inhibitors improve photodynamic therapy for cancer: lack of vascular normalization
# Abstract
*In the **Abstract** section:*
Photodynamic therapy (PDT) is an effective clinical treatment for a number of different cancers. PDT can induce hypoxia and inflammation, pro-angiogenic side effects, which may counteract its angio-occlusive mechanism. The combination of PDT with anti-angiogenic drugs offers a possibility for improved anti-tumour outcome. We used two tumour models to test the effects of the clinically approved angiostatic tyrosine kinase inhibitors sunitinib, sorafenib and axitinib in combination with PDT, and compared these results with the effects of bevacizumab, t[he anti-V](https://www.ncbi.nlm.nih.gov/mesh/D000077210)EG[F antibod](https://www.ncbi.nlm.nih.gov/mesh/D000077157)y, fo[r the im](https://www.ncbi.nlm.nih.gov/mesh/D000077784)provement of PDT. Best results were obtained from the combination of PDT [and low-dos](https://www.ncbi.nlm.nih.gov/mesh/D000068258)e axitinib or sorafenib. Molecular analysis by PCR revealed that PDT in combination with axitinib suppressed VEGFR-2 expre[ssion in](https://www.ncbi.nlm.nih.gov/mesh/D000077784) tum[our vascu](https://www.ncbi.nlm.nih.gov/mesh/D000077157)lature. Treatment with bevacizumab, although effective as monother[apy, did](https://www.ncbi.nlm.nih.gov/mesh/D000077784) not improve PDT outcome. In order to test for tumour vessel normaliz[ation effec](https://www.ncbi.nlm.nih.gov/mesh/D000068258)ts, axitinib was also applied prior to PDT. The absence of improved PDT outcome in these experiments, as well as the lack of[ increas](https://www.ncbi.nlm.nih.gov/mesh/D000077784)ed oxygenation in axitinib-treated tumours, suggests that vascular normalization did not occur. The current data imply that there is a fut[ure for ](https://www.ncbi.nlm.nih.gov/mesh/D000077784)certain anti-angiogenic agents to further improve the efficacy of photodynamic anti-cancer therapy.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Photodynamic therapy (PDT) is a minimally invasive therapy in which visible or near infrared light irradiation is combined with light sensitive molecules (photosensitizers) to produce reactive oxygen species (ROS). These ROS can damage blood vessels in such a way that vascular occlusion occurs. Several photosensitizers have been approved by the FDA to treat a number of oncological applications by PDT (see Table S1). Photodynamic therapy is also used in ophthalmology and for many years PDT was the mainstay for treating exudative age-related macular degeneration, the main cause of blindness in the aged western population. Angio-occlusive PDT can cause tissue responses, such as hypoxia and inflammation, both of which play a role in inducing angiogenesis.
This angiogenic tissue response following PDT can in principle counteract the angio-occlusive effect of PDT, thus leading to a reduced tumoricidal outcome. Therefore, PDT results may be improved by co-treatment with an angiogenesis inhibitor. We previously showed in a tumour-free model that vessel regrowth after angio-occlusive PDT can effectively be inhibited by anti-angiogenic agents. In the present study, we tested the effect of combining PDT with an anti-angiogenic drug by monitoring tumour vasculature and tumour growth. This was done in two different tumour models on the chorioallantoic membrane (CAM) of the chicken embryo.[](https://www.ncbi.nlm.nih.gov/mesh/D017382)
Therapeutic anti-angiogenesis strategies have been established in the clinical management of cancer, both as monotherapies and in combination with other anti-tumour modalities. Among these are bevacizumab (Avastin®, an antibody-based drug that neutralizes VEGF), and the broad-spectrum (tyrosine) kinase inhibitors (TKIs) that inhibit the signalling of growth factor receptors. Examples of the latter are sunitinib (Sutent®), clinically approved for the treatment of metastatic renal cell carcinoma, imatinib-resistant gastrointestinal stromal-and pancreatic neuroendocrine tumours. We also tested sorafenib (Nexavar®), approved for metastatic renal cell cancer and unresectable hepatocellular carcinoma. While sunitinib inhibits VEGF receptors 1, 2 and 3 (VEGFR-1,-2 and-3), platelet-derived growth factor receptor beta (PDGFR-β) and mast/stem cell growth factor receptor (c-KIT) with medium affinity, and FGFR-1 with low affinity, sorafenib inhibits the RAF/MEK/ERK pathways, as well as VEGFR-1,-2 and-3, c-KIT and PDGFR-β with relatively low affinity.
A second-generation TKI with improved affinity to VEGFR-2 and a better toxicity profile, is axitinib (Inlyta®). Axitinib has fewer targets and has a higher affinity for the VEGF receptors. It should be noted that the combination of PDT with the antibody-based agents bevacizumab and ranibizumab has been tested clinically for the treatment of wet age-related macular degeneration. For cancer, only pre-clinical studies are available. Very limited research has been focused so far on the combination of PDT with TKIs.[](https://www.ncbi.nlm.nih.gov/mesh/D000068258)
It has previously been shown that angiogenesis inhibition can normalize cancer vessels. As the efficacy of PDT depends on tissue oxygenation, we tested sequencing of the combination therapy. We found that PDT treatment can be significantly improved by angiostatic compounds. The tested TKIs were more effective enhancers of PDT effects than bevacizumab. In addition, anti-angiogenic drugs were found to be best applied after PDT.
These results, as well as tissue oxygenation measurements, suggested that the observed improvements were not dependent on vascular normalization.[](https://www.ncbi.nlm.nih.gov/mesh/D000068258)
## Materials and methods
*In the **Materials and methods** section:*
## Cell culture, preparation and implantation on the CAM model
*In the **Cell culture, preparation and implantation on the CAM model** section:*
A2780 human ovarian carcinoma cells (ECACC, Salisbury, UK) were maintained in RPMI-1640 cell culture medium supplemented with GlutaMAX™ (Gibco, Carlsbad, CA, USA), 10% bovine calf serum (Sigma-Aldrich, St. Louis, MO, USA) and 1% antibiotics (Sigma-Aldrich). Human colorectal carcinoma (HCT-116; ECACC) cells were maintained in DMEM medium (Gibco) supplemented as above. Fertilized chicken eggs were incubated in a hatching incubator (relative humidity 65%, 37°C), as previously described. On EDD 7, 106 HCT-116 cells were mixed with ice-cold Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) and transplanted on the surface of the CAM as a 30 μl drop. 106 A2780 cells were prepared as a spheroid in a 25 μl hanging drop and 3 hrs later were transplanted on the surface of the CAM.
## Image acquisition and quantification
*In the **Image acquisition and quantification** section:*
Visualization of the CAM vasculature and irradiation with light during PDT was performed under an epi-fluorescence microscope (Eclipse E 600 FN; Nikon AG, Tokyo, Japan) with objectives (Plan Apo 4×/0.2, working distance: 20 mm or Plan Fluor 10×/0.3, working distance: 16 mm; Nikon AG), as previously described. Shortly, PDT was performed (λex = 420 ± 20 nm, λem ≥ 470 nm; Nikon) using Visudyne® (Novartis Pharma Inc., Hettlingen, Switzerland). Visualization of blood vessels was achieved through fluorescence angiography after intravenous (i.v.) injection of fluorescein isothiocyanate dextran (FITC-dextran, 20 kD, 20 μl, 25 mg/ml, Sigma-Aldrich). A volume of 20 μl of India ink from Pelikan (Witzikon, Switzerland) was administered to enhance vascular contrast.
Fluorescence images were taken using an F-view II 12-bit monochrome Peltier-cooled digital CCD camera run by ‘analySIS DOCU’ software (Soft Imaging System GmbH, Munster, Germany). Image processing and quantification of the fluorescence angiographies was achieved by using a macro written in ImageJ (version 1.40 a; National Institutes of Health, Bethesda, MD, USA), as previously described. The four concentric circles with ‘1’ being the central area, and ‘4’ being the most peripheral area, create four zones of revascularization, each of which is analysed separately by the software.[](https://www.ncbi.nlm.nih.gov/mesh/D000077362)
## Combination therapy on the CAM
*In the **Combination therapy on the CAM** section:*
Bevacizumab was purchased from Genentech (San Francisco, CA, USA), sunitinib from Pfizer Inc. (New York, NY, USA), axitinib and sorafenib from LC Laboratories (Woburn, MA, USA). Drugs were administered intravenously (20 μl) on EDD 10 and 11 at two concentrations: axitinib (6.5 or 13 μg/kg), sorafenib (21 or 85 μg/kg), sunitinib (35.5 or 71 μg/kg) and bevacizumab (99 or 497 μg/kg). Concentrations were calculated for an estimated embryo weight of 3 g. Angiograms of the CAM were taken on EDD 12. Visudyne®-PDT (subsequently referred to as PDT) was performed at a low-fluence rate (5 J/cm2, with irradiance of 35 mW/cm2 at 420 ± 20 nm). The irradiation area was limited to a circular spot of 0.02 cm2 using an optical diaphragm.
Directly after PDT, 20 μl of the angiogenesis inhibitors was administered intravenously in the CAM at the following effective doses: axitinib (13 μg/kg), sorafenib (85 μg/kg), sunitinib (71 μg/kg) and bevacizumab (497 μg/kg). Treatment was repeated 24 hrs after PDT.[](https://www.ncbi.nlm.nih.gov/mesh/D000068258)
## Tumour treatment
*In the **Tumour treatment** section:*
Vascularized tumours appeared ˜3 days after inoculation beneath the surface of the CAM and the average tumour volume was 1.66 ± 0.09 mm3. Visudyne®-PDT, as described above, was performed at this moment, while adjusting the diaphragm to the tumour size. Angiostatic therapy was performed by administering 20 μl of axitinib (13 μg/kg), sorafenib (85 μg/kg), sunitinib (71 μg/kg) and bevacizumab (497 μg/kg) intravenously at EDD 10 and 11.[](https://www.ncbi.nlm.nih.gov/mesh/D000077362)
## Combination therapy
*In the **Combination therapy** section:*
Tumours receiving combination treatment were injected twice intravenously with 20 μl of each angiogenesis inhibitor (at doses as above) according to two different schedules: (i) right after PDT and 24 hrs after PDT or (ii) 24 hrs before PDT and right after PDT (Fig. 6A). Photodynamic therapy with 5 J/cm2 and 35 mW/cm2 at 420 ± 20 nm was applied. Tumours were measured daily, volume = (the largest diameter)2 × (perpendicular diameter) × 0.5.
## Immunohistochemistry
*In the **Immunohistochemistry** section:*
Tumours were resected at treatment day 8, fixed overnight in zinc fixative solution and stained as previously described. In short, 4 μm sections were treated with 0.3% H2O2 in methanol for 30 min., a citrate buffer (20 min. at 95°C) antigen retrieval step was applied, blocking with 10% goat serum and 1% BSA was performed. Primary antibody (DIA-310; Dianova, Hamburg, Germany) incubations were performed overnight.[](https://www.ncbi.nlm.nih.gov/mesh/D006861)
## RNA isolation, cDNA synthesis and quantitative real-time RT-PCR
*In the **RNA isolation, cDNA synthesis and quantitative real-time RT-PCR** section:*
Total RNA isolation, cDNA synthesis and quantitative real-time RT-PCR (qRT-PCR) were performed as previously described. Each target gene was quantified relative to the expression of the reference genes (β-Actin and Cyclophilin-A). Chicken (gg) and human (hs) primers were synthesized by Eurogentec (Liege, Belgium).
## pO2 measurements
*In the **pO2 measurements** section:*
Intra-tumoral oxygenation was measured 24 hrs after the first treatment intervention (corresponding to treatment day 2). Measurements of the partial pressure of oxygen (pO2) within the treated tumours were obtained using an OxyLab pO2 meter (Oxford Optronix Ltd., Oxford, UK) coupled to a calibrated fibre optic probe (NP/O/E) placed in a 23G surgical steel needle. Each measurement was taken over 60 sec. after the intra-tumoral probe insertion.[](https://www.ncbi.nlm.nih.gov/mesh/D010100)
## Statistical analysis
*In the **Statistical analysis** section:*
Values are given as mean values ± SEM. Data are represented as averages of independent experiments. Statistical analysis was done using the anova test and t-test. *P indicating P-values lower than 0.05 and **P indicating P-values lower than 0.01 were considered statistically significant. Synergy was calculated using the CompuSyn application.
## Results
*In the **Results** section:*
## Clinically used angiostatic TKIs prolong the vaso-occlusive effect of PDT
*In the **Clinically used angiostatic TKIs prolong the vaso-occlusive effect of PDT** section:*
Visudyne®-PDT (PDT) was performed on the CAM at embryo development day (EDD) 10 (Fig. 1A), leading to blood flow stasis in the smaller blood vessels and in the capillary bed. Vessels with a diameter >70 μm stayed perfused (Fig. 1B). New capillaries were first seen in the most peripheral zone of the treated area (Fig. 1B) and a completely regrown vasculature was observed after 48 hrs (Fig. 1C). Quantification of the data was performed by digital image analysis in four concentric areas (Fig. 1C, most right image).[](https://www.ncbi.nlm.nih.gov/mesh/D000077362)
Clinically used angiostatic tyrosine kinase inhibitors prolong the vaso-occlusive effect of PDT. (A) Fluorescence angiograms of the CAM before PDT. The circle represents the diaphragm, which limits CAM exposed with light. (B) 24 hrs and (C) 48 hrs after PDT showing the start of micro-vascular regrowth and complete revascularization of the treated area, respectively. (C) Right panel shows the skeletonization and area numbers used for the image processing. (D and E) Natural growth of CAM vasculature and inhibition of angiogenesis by axitinib and skeleton images of EDD 12. White arrows indicate the avascular zones induced by axitinib. (F) Quantification of the number of branching points per mm2 after treatment with an ineffective and an effective dose of each compound. Effective doses: axitinib (13 μg/kg; N = 7), sorafenib (85 μg/kg; N = 7), sunitinib (71 μg/kg; N = 5) and bevacizumab (497 μg/kg; N = 5). (G) Fluorescence angiogram of the CAM treated with PDT+axitinib at its effective dose taken 48 hrs post PDT. (H) Quantification of the results for all four tested compounds. Data are shown as means (±SEM, **P < 0.01 as compared to the control in each respective area of vascular regrowth (1–4), N = 3–6 per condition). The scale bars in (A, D and G) represent 200 μm.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
To prolong the effect of PDT, treatment with anti-angiogenic compounds, axitinib, sorafenib, sunitinib or bevacizumab, was performed. Angiostatic compounds were first tested alone by administering i.v. injection on EDD 10 and 11, followed by imaging and quantification performed on EDD 12. Representative fluorescence images of the CAM treated with 0.9% NaCl (control) or axitinib (13 μg/kg) are presented in Fig. 1D and 1E, respectively. Low concentrations of all four drugs were identified where a statistically significant inhibitory effect was observed (**P < 0.01, Fig. 1F).
These doses were tested in combination with PDT. All drugs were administered twice, immediately after PDT and 24 hrs later. Interestingly, all three tested TKIs markedly suppressed the regrowth of blood vessels, as determined by a significant reduction in the number of branching points. This activity was not observed for bevacizumab. Axitinib and sunitinib were the most effective drugs (Fig.
1G and H). An ˜90% reduction in the number of branching points per mm2 was observed in treatment area 1, while bevacizumab was completely ineffective.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
## Angiostatic TKIs, but not bevacizumab, improve the anti-tumour effect of PDT
*In the **Angiostatic TKIs, but not bevacizumab, improve the anti-tumour effect of PDT** section:*
A2780 ovarian carcinoma cells were inoculated at EDD 7 and monitored for 11 days. Established and vascularized tumours were detected 3 days post implantation (EDD 10). Tumours grew to an average size of ˜140 mm3 by EDD 17 when left untreated (Fig. S1A). The chicken vasculature in these tumours was efficiently perfused, as demonstrated by the prompt distribution of India ink throughout the tumour vasculature within 5 sec. after intravenous injection (Fig. S1B).
As expected, the tumour vessels were leaky, as the ink was present in the extracellular space of the tumour already after 20 sec. (Fig. S1C).[](https://www.ncbi.nlm.nih.gov/mesh/C028433)
Sub-optimal treatment strategies were defined, both for PDT (Fig. 2A) and angiostatic compounds (Fig. 2C) in A2780. The PDT conditions were selected such that tumour growth was inhibited by ˜60% (Fig. 2B). Dose selection for axitinib is shown in Figure 2D. For sorafenib, sunitinib and bevacizumab, the sub-optimal doses in A2780 model were defined at 85, 71 and 497 μg/kg, respectively. The same doses were applied in the HCT-116 model.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
Defining sub-optimal drug concentrations and PDT conditions for tumour treatment on the CAM. Treatment regimens for CAM tumours tested for PDT alone (A) or drug alone (C). Tumour growth curves for PDT (B) and angiostatic drug (D) are shown. Arrows indicate treatment days. Data are shown as means (±SEM). N = 3–10 per condition; **P < 0.01.
Combination of PDT and i.v. drug administration immediately after and 24 hrs later was performed in the A2780 xenographs (Fig. 3A). The representative images of tumours resected on treatment day 8 from different treatment groups are presented in Figure 3B. Photodynamic therapy in combination with axitinib and sorafenib significantly improved PDT outcome (**P = 0.0033 and **P = 0.0025, respectively, Fig. 3C, N = 6–10). Surprisingly, sunitinib and bevacizumab did not or only marginally improve the effect of PDT. Synergy, as defined by the Chou-Talalay equation as combinations with a ‘combination index’ (CI) less than 1, was calculated for the combination of PDT with axitinib (CI = 0.36) and PDT with sorafenib (CI = 0.59).
Neither sunitinib nor bevacizumab gave a statistically significant difference in tumour size together with PDT as compared to PDT alone. Similar experiments with axitinib and sorafenib were performed on human HCT-116 colorectal carcinoma tumours. In this model, comparable results were observed for PDT+axitinib (N = 6, **P = 0.0008 as compared to the control) and PDT+sorafenib (N = 9, *P = 0.02), as shown in Figure 6D (schedule 1) and G, respectively, as a percentage of the control at the last day of the experiment.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
PDT and anti-angiogenesis combination therapy. (A) Treatment regimens for CAM tumours treated with schedule 1. (B) Representative images A2780 human ovarian tumours for control and various treated groups resected on treatment day 8. (C) Tumour growth curves of tumours treated by each anti-angiogenic drug, PDT and the combination of both therapies (**P = 0.0033 for PDT+axitinib and **P = 0.0025 for PDT+sorafenib as compared to PDT alone, (C) N = 6–10 per condition). S indicates synergy (CI<1).[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
## Combination therapy reduces vessel density and modulates vascular morphology and angiogenesis-related gene expression
*In the **Combination therapy reduces vessel density and modulates vascular morphology and angiogenesis-related gene expression** section:*
Immunohistochemical staining for CD31 was performed 3 and 8 days after treatment (Fig. 4) in both tumour models. It was found that the combination of PDT and TKIs (both axitinib and sorafenib) suppressed microvessel density as shown at the last (8th) experiment day (Fig. 4A, **P = 0.0009, *P = 0.022, respectively and N = 6–14). Angiogenesis inhibitors alone did not significantly suppress microvessel density (Fig. S2). Microvessel density in the bevacizumab combination group was not different from the PDT monotherapy group, while sunitinib combination group was increased as compared to the control. Another interesting difference was observed in the morphology of the tumour vessels.
While control tumours had a large numbers of small vessels with compressed lumens, the combination of PDT with axitinib and sorafenib resulted in larger vessels with an open lumen (**P < 0.001, *P = 0.051, respectively, N = 6–22, Fig. 4A). Photodynamic therapy initially (treatment day 3, Fig. 4B) suppressed microvessel density significantly, but after a longer period (day 8) this effect had largely disappeared, presumably because of the PDT-induced angiogenesis. Combination treatment of PDT + axitinib of HCT-116 tumours revealed a statistically significant decrease in vessel density (**P = 0.0006, N = 10) as compared to control tumours resected at the last (8th) experiment day (Fig. 4C).[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
Histology of resected tumours showing microvessel density. (A) CD31-stained sections of the A2780 tumours excised at day 8 for control, PDT and combination treatment groups. Graphs of microvessel density and the percentage of vessels with open lumens showing a statistically significantly decrease in microvessel density and increase in the number of vessels with an open lumen for PDT+axitinib and PDT+sorafenib treated tumours as compared to the control tumours. (B) CD31-stained sections of the A2780 tumours excised at day 3 for the most effective treatment group (PDT+axitinib, 13 μg/kg) and quantification of microvessel density (right). (C) CD31-stained sections of HCT-116 tumours excised on day 8 and quantification of microvessel density (right) showing significant inhibition of vessel density in the combination PDT+axitinib treatment group. **P < 0.01; N = 5–22 per condition.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
Based on the above-described results, we performed real-time quantitative PCR studies for tumours exposed to axitinib and its combination with PDT (Fig. 5A–C). We investigated the expression of angiogenic growth factor receptors in vasculature (chicken specific primers, 5A and B) and growth factors secreted by tumour cells (human specific primers, Fig. 5C). It was observed that early after treatment (day 3), i.v. administered axitinib, but not PDT, suppressed VEGFR-2 in the vasculature. VEGFR-2 was still down-regulated 8 days after treatment, at which time this effect was also seen for the expression of PDGFR-β. Assessment of growth factor expression in the tumour cells (Fig. 5C) did not reveal a strong angiogenic response.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
Real-time RT-PCR molecular profiling of the tumours treated with PDT, axitinib (13 μg/kg), or their combination. The expression of some of the angiogenesis-related genes determined by quantitative real-time PCR performed at day 3 (A) and day 8 (B) post PDT using chicken (gg)-specific primers for: VEGFA, VEGFR2, PDGFR-β. (C) Quantification of human genes in tumours excised on day 3 using human (hs)-specific primers for VEGFA, bFGF and PLGF. Mean relative expressions are shown with the SEM. N = 5–7 per condition.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
The change in the Ct values of the control and treatment group tumours was examined between tumours excised on treatment days 3 and 8. A detectable, but not significant, change in gene levels was noticed between days 3 and 8 in control tumours (data not shown). The only significant increase in gene expression levels was noted for VEGFA, whose expression was up-regulated by 10.7% in the host cells on Day 3 versus Day 8 (*P = 0.044, N = 5–7) and by 11.7% in the tumour cells between days 3 and 8 (*P = 0.052, N = 2–3).
## Scheduling of PDT and angiostasis: lack of vascular normalization
*In the **Scheduling of PDT and angiostasis: lack of vascular normalization** section:*
Next to the above used schedule (Fig. 6A, now called schedule 1), a treatment schedule starting with angiostatic compounds axitinib (Fig. 6B–D) or sorafenib (Fig. 6E–G) 24 hrs prior to PDT (schedule 2) was also tested in the two tumour models. Interestingly, none of the angiostatic compounds applied prior to PDT (schedule 2) resulted in significantly better anti-tumour photodynamic activity than for schedule 1 at the conditions applied. While for sorafenib similar results for schedule 1 and 2 were observed (Fig.
6E and G), for axitinib treatment schedule 2 resulted in a worse outcome (Fig. 6B and D), as compared to schedule 1 in both tumour models. In the HCT-116 model, all tumours treated with combination therapy using either schedule were inhibited significantly as compared to the control tumours (control: N = 6–12; axitinib schedule 1: **P = 0.0008; axitinib schedule 2: *P = 0.01; sorafenib schedule 1: *P = 0.022; and sorafenib schedule 2: *P = 0.024).[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
Tumour growth rate depends on the treatment schedule. Treatments were performed at day 1 and 2 (black arrows) as indicated by the two different schedules, shown in (A). Graphs show the effects of combination therapies with two different treatment schedules for axitinib in A2780 (B) and in HCT-116 (D) tumours. Also for sorafenib (E) and bevacizumab (H) in A2780 and sorafenib in HTC-116 (G) tumours. In all cases, the most effective treatment was combination therapy with treatment schedule 1. Measurements of intra-tumoral oxygenation in control, PDT, axitinib (C), sorafenib (F) and bevacizumab (I) treated A2780 tumours. Each group represents the mean ± SEM (N = 3–8 per condition; **P < 0.01). Human HCT-116 colon carcinoma growth rate inhibited by PDT+axitinib (**P = 0.0008) or PDT+sorafenib (**P = 0.022) applied at schedule 1, was similar to that obtained in the A2780 model. Data are shown as means (±SEM); N = 6–12 per condition.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
The most unexpected result was that bevacizumab pre-treatment even resulted in a loss of the anti-tumour activity resulting from the PDT treatment (Fig. 6H). To further investigate the origin of differences in tumour growth after treatment with the two schedules, intra-tumoral oxygenation was measured at 24 hrs (when PDT was performed in schedule 2) after the first bolus injection of axitinib (13 μg/kg), sorafenib (85 μg/kg) and bevacizumab (497 μg/kg), see Figure 6C, F, and I, respectively. The pO2 measurements performed 24 hrs after the first injection with the inhibitors showed a small and not significant increase in oxygenation (e.g. 6.7% for bevacizumab, as compared to control tumours, P = 0.27, N = 10). Moreover, there was no difference between the latter groups and the PDT group.[](https://www.ncbi.nlm.nih.gov/mesh/D000068258)
## Discussion
*In the **Discussion** section:*
A major limitation in the use of PDT against cancer is the PDT-induced angiogenic tissue response. As there are now many clinically approved effective angiogenesis inhibitors, it is proposed that these compounds can significantly prolong the beneficial angio-occlusive effect of PDT. The results of the present study show that angiostatic small molecule TKI can synergistically improve the anti-tumour effect of PDT, in both an ovarian and a colorectal tumour model. A major observation of this study is that this improvement of PDT outcome was because of the inhibition of PDT-induced angiogenesis, and not to the vascular normalization processes, as TKI-induced enhancement of tumour oxygenation was not observed. Synergy between PDT and anti-angiogenic TKIs for tumour growth suppression was best observed for axitinib when applied at a sub-optimal dose and combined with a sub-optimal PDT regimen.
Sorafenib also showed a synergistic activity, but these results were not observed for sunitinib and bevacizumab. The results suggest that a combination of PDT and axitinib might be a promising strategy for translation into the clinic.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
Photodynamic therapy has been most successfully used in the treatment of ophthalmological neovascularization-based disorders. These were in the past mainly wet age-related macular degeneration patients and at present mainly patients with polypoidal choroidal vasculopathy. The treatment of solid tumours with PDT is currently receiving renewed interest because it is being realized that its combination with anti-angiogenesis therapy has promising applications. Several studies have been reported on such combinations for the treatment cancer. These include pre-clinical studies assessing the activity of cetuximab and/or bevacizumab with hypericin-PDT in a human bladder carcinoma model, SU5416 and SU6668 with hypericin-PDT in a human nasopharyngeal carcinoma model and PD166285 and PD173074 with hexylether pyropheophorbide-a-PDT in a murine mammary carcinoma model.
In all these studies, the anti-angiogenic drugs were applied after PDT, and the combination treatment was shown to be more potent than the monotherapies. A comparative study in which the PDT was combined in varying treatment schedules, with clinically approved TKIs has not yet been performed. In our study, the best results, i.e. a synergistic improvement of PDT, were observed in combination with axitinib, making a clinical translation of this treatment a promising option. This would most likely be best developed for tumour types that have been shown to be successfully treated with PDT, such as basal cell carcinoma (BCC) or non-metastatic base of the tongue squamous cell carcinoma.[](https://www.ncbi.nlm.nih.gov/mesh/D000068818)
In BCC-diagnosed patients, the average recurrence was shown to be 10% at 12 months after topical Metvix® (methyl aminolevulinate)-mediated PDT. Unfortunately, the follow-up of these patients is not continued longer than 1 year post PDT, whereas it was shown in many studies that the recurrence peak post Metvix®-PDT is at 36 months. Moreover, patients treated with such PDT strategies had a better cosmetic outcome and the treatment outcome was typically superior to that achieved with existing standard therapies. Recurrent base of the tongue malignancies develop usually loco-regionally at previously irradiated fields. Also, interstitial PDT (with metatetra(hydroxyphenyl)chlorin, mTHPC) of recurrent non-metastatic carcinoma of the tongue base showed promising results.[](https://www.ncbi.nlm.nih.gov/mesh/C475457)
It is most interesting to see that when PDT was followed by angiogenesis inhibition at the applied conditions, synergisms were observed for axitinib and sorafenib, but not for sunitinib and bevacizumab. For the latter, there even seemed to be a lack of additive effect (Fig. 3). It should be noted that part of the VEGF signalling in this model may be derived from chicken VEGF, and bevacizumab probably binds chicken VEGF with a lower affinity than human VEGF. However, a number of studies have shown the efficacy of anti-VEGF antibodies (bevacizumab or ranibizumab) against chicken VEGF, so this argument may not be very significant. The question remains why neutralizing vascular endothelial growth factor (VEGF) does not work so well, while inhibition of VEGFR signalling does.
This could mean that neutralization of VEGF by a large molecule—an antibody—is much less efficient inside the microenvironment of a tumour in situ than inhibiting the VEGFR by a small molecule. Another explanation could be the broader activity spectrum of axitinib. This then raises the question why axitinib works better than sunitinib. However, the most likely explanation for this is that the affinity of axitinib for VEGFR-2 is some 40 times higher.[](https://www.ncbi.nlm.nih.gov/mesh/D000077784)
A similar discussion is valid for the situation of angiogenesis inhibition prior to PDT. Here, bevacizumab not only lacks improvement of PDT but also seems to counteract the efficacy of PDT. Apparently, the presence of VEGF is necessary for an effective PDT outcome. It can be assumed that VEGF-induced active cell metabolism is necessary for effective PDT. This also suggests that the major effect of PDT, at the applied conditions, is through its effect on the vasculature.
The fact that the results from the axitinib treatment groups do not seem to support this option may be explained by the broad activity spectrum of TKIs. Relatedly, this may also explain the overt difference between axitinib and sunitinib, being the two drugs mainly inhibiting the VEGFRs. Although VEGFRs and other growth factor receptors are considered the primary targets of these compounds, it has been shown before that more than one hundred kinases are affected by sunitinib, and it would thus be quite difficult to pinpoint the exact mechanism of action of these drugs. Moreover, it cannot be ruled out that part of the success of axitinib is through a direct activity on the tumour cells.[](https://www.ncbi.nlm.nih.gov/mesh/D000068258)
Another aim of this study was to study the consequences of the treatment sequence. Previous studies on such combination therapies for cancer were all performed by timing the angiostatic therapy starting either at the same time as PDT, or after PDT. As suggested by Jain, angiogenesis inhibition can normalize the tumour vasculature, as well as the blood flow, interstitial pressure, vessel wall permeability and oxygenation. We and others have shown that this effect of angiogenesis inhibitors can improve the combination with e.g. chemo-and radiotherapy. For example, Dings et al. found a time-window of increased tumour oxygenation over the first 4 days of treatment with either bevacizumab (10 mg/kg i.v. in a single injection) or anginex (10 or 20 mg/kg/d i.p.).
Elevated oxygenation was also accompanied by reduced vessel density and increased pericyte coverage. When radiotherapy was initiated within this window, tumour growth delay was significantly enhanced in relation to alternative treatment schedules. Huber et al. showed that SU11657 (a multi-target small molecule inhibitor of VEGFRs and PDGFR) was more effective when administered 1 day prior to radiotherapy as compared to 1 day after radiotherapy. As PDT, like radiotherapy, is dependent on oxygenation of the tissue, we put forward the hypothesis that anti-angiogenesis, at least in some cases, could effectively be given prior to PDT. In the present study, we observed that the latter treatment schedule does not improve the anti-tumour activity, or even, it can make the overall outcome worse.
This suggests that vascular normalization does not take place to a significant extent at the applied conditions. Indeed, in our experimental conditions, we did not observe significantly increased oxygenation after treatment with axitinib (13 μg/kg), sorafenib (85 μg/kg) or bevacizumab (497 μg/kg) over a period of 24 hrs. It should, however, be noted that in these studies we used very low drug doses, i.e. 0.497 mg/kg of bevacizumab, as compared to a dose of 10 mg/kg reported to induce vascular normalization by Dings et al..[](https://www.ncbi.nlm.nih.gov/mesh/D000068258)
To summarize the data from the current study, it can be concluded that PDT and anti-angiogenic therapy can synergistically inhibit tumour growth. Through the indirect neutralization of VEGF, and the direct inhibition of growth factor receptors, the anti-tumour effect of PDT can be improved.
## Conflicts of interest
*In the **Conflicts of interest** section:*
The authors confirm that there are no conflicts of interest.
# References
*In the **References** section:*
## Supporting information
*In the **Supporting information** section:*
Additional Supporting Information may be found in the online version of this article:
|
# Introduction
Anti-amyloid Compounds Inhibit α-Synuclein Aggregation Induced by Protein Misfolding Cyclic Amplification (PMCA)*
# Abstract
*In the **Abstract** section:*
Background: α-Synuclein filaments can be formed in vitro, but the process is slow and unreliable, unsuitable for high throughput screening.
Results: Protein misfolding cyclic amplification (PMCA) rapidly assembles α-synuclein into filaments that form aggregates in cell culture and whose formation is inhibited by anti-amyloid compounds.
Conclusion: Anti-amyloid compounds affect PMCA cell-transmissible α-synuclein fibril formation.
Significance: α-Synuclein PMCA is useful for the screening of anti-aggregation compounds as treatments for α-synucleinopathies.
Filaments made of α-synuclein form the characteristic Lewy pathology in Parkinson and other diseases. The formation of α-synuclein filaments can be reproduced in vitro by incubation of recombinant protein, but the filament growth is very slow and highly variable and so unsuitable for fast high throughput anti-aggregation drug screening. To overcome this obstacle we have investigated whether the protein misfolding cyclic amplification (PMCA) technique, used for fast amplification of prion protein aggregates, could be adapted for growing α-synuclein aggregates and thus suitable for screening of drugs to affect α-synuclein aggregation for the treatment of the yet incurable α-synucleinopathies. Circular dichroism, electron microscopy, and native and SDS-polyacrylamide gels were used to demonstrate α-synuclein aggregate formation by PMCA, and the strain imprint of the α-synuclein fibrils was studied by proteinase K digestion. We also demonstrated that α-synuclein fibrils are able to seed new α-synuclein PMCA reactions and to enter and aggregate in cells in culture. In particular, we have generated a line of “chronically infected” cells, which transmit α-synuclein aggregates even after multiple passages. To evaluate the sensitivity of the PMCA system as an α-synuclein anti-aggregating drug screening assay a panel of 10 drugs was tested. Anti-amyloid compounds proved efficient in inhibiting α-synuclein fibril format[ion](https://www.ncbi.nlm.nih.gov/mesh/D012967) [induced by PMC](https://www.ncbi.nlm.nih.gov/mesh/C016679)A. Our results show that α-synuclein PMCA is a fast and reproducible system that could be used as a high throughput screening method for finding new α-synuclein anti-aggregating compounds.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
α-Synucleinopathies are characterized by the misfolding and aggregation of the abundant CNS protein α-synuclein that is expressed predominantly in nerve cells and is concentrated at presynaptic nerve terminals, where it plays a role in synaptic vesicle transport (1). Aggregated α-synuclein forms the neuronal inclusions of Parkinson disease and dementia with Lewy bodies (Lewy bodies and neurites) and the oligodendroglial inclusions of multiple system atrophy (glial cytoplasmic inclusions) (1). Evidence that α-synuclein aggregation causes these diseases has accumulated through a variety of neuropathological, biochemical, and genetic studies (2, 4). In particular, dominantly inherited mutations and duplications and triplications of SNCA, the α-synuclein gene, cause familial forms of Parkinson disease, dementia with Lewy bodies, and multiple system atrophy (5), and α-synuclein has been identified as a risk factor for Parkinson disease in all genome-wide association studies done to date (6). Despite being a point of intense research, the critical question of how to prevent, stop, or revert the aggregation of α-synuclein remains unresolved.
α-Synuclein is a 140-amino acid protein with little ordered structure that binds to lipid membranes. It comprises seven repeats, in the positively charged amino-terminal region and the hydrophobic middle part, with the carboxyl-terminal region being negatively charged. Monomeric α-synuclein adopts an α-helical structure upon binding to lipid membranes containing acidic phospholipids. This conformation involves amino acids 1–98, with residues 99–140 being considered unstructured (7).[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
In recent years, the mechanism of α-synuclein aggregation has been compared with that of the prion protein, whose misfolding causes transmissible spongiform encephalopathies (8). Common biochemical hallmarks are the propensity to aggregate, insolubility in mild detergents, and partial resistance to protease digestion (9–11). The development and use of protein misfolding cyclic amplification (PMCA)3 of the prion protein (12) has helped to understand the underlying prion replication, infectivity, and strain formation.
## EXPERIMENTAL PROCEDURES
*In the **EXPERIMENTAL PROCEDURES** section:*
## Expression and Purification of Recombinant Wild-type α-Synuclein
*In the **Expression and Purification of Recombinant Wild-type α-Synuclein** section:*
BL21(DE3) Escherichia coli was transformed with human full-length α-synuclein in pRK172, and the protein was then purified as described (13). Briefly, bacterial cells were harvested and resuspended in Tris/EDTA buffer, lysed 4 °C (with 25 kg/square inch using a cell disruptor (Constant Systems Ltd.) and centrifuged). α-Synuclein protein was purified from the lysate supernatant by anion exchange using HiTrap Capto adhere (GE Healthcare), (NH4)2SO4 precipitation, gel filtration, and anion exchange using Mono Q GL (GE Healthcare). The pooled protein fractions collected from the purification steps were concentrated and solvent-exchanged using Amicon Ultra-15 centrifugal filters with 10-kDa molecular mass cutoff (Millipore). Aliquots of protein were stored at −20 °C prior to use.
A 10-μl aliquot was hydrolyzed in 6 m HCl for amino acid analysis. Protein concentrations were determined by quantitative amino acid analysis, performed in-house (LMB-MRC, UK), and confirmed at the Protein and Nucleic Acid Chemistry Facility, University of Cambridge, UK.[](https://www.ncbi.nlm.nih.gov/mesh/D014325)
## PMCA
*In the **PMCA** section:*
PMCA was carried out by subjecting recombinant wild-type full-length human α-synuclein to repeated cycles of sonication and incubation. α-Synuclein was prepared as indicated (13) and diluted to a final 90 μm concentration in conversion buffer (1% Triton X-100, 150 mm NaCl, Complete Protease Inhibitor Mixture (Roche Applied Science; in 1×PBS). For PMCA, 60-μl aliquots from 200 μl of the 90 μm reaction mixtures were transferred into 200-μl PCR tubes (Axygen) containing 37 ± 3 mg of 1.0-mm zirconia/silica beads (Biospec Products), and samples were subjected to cycles of 20-s sonication and 30-min incubation at 37 °C, for different times depending on the experiment, using a Misonix 4000 sonicator at 70 power setting. All reactions were performed in triplicate. When drugs or seeds were used, 2 μl of concentrated drugs were added into 200 μl of the PMCA reaction mixture.
Seeded reactions (for the study of substrate concentrations and the serial PMCA) were done by diluting 1:100 of 90 μm α-synuclein fibrils, previously generated by PMCA, into fresh soluble α-synuclein recombinant substrate.[](https://www.ncbi.nlm.nih.gov/mesh/D017830)
## Thioflavin T Assay
*In the **Thioflavin T Assay** section:*
From each sample, 5 μl was added to 495 μl of ThT solution (20 μm ThT, 50 mm glycine in H2O, pH 8.5, with KOH). Fluorescence was measured with a PerkinElmer Life Sciences luminescence spectrophotometer LSS5 with 450-nm excitation and 480-nm emission settings.[](https://www.ncbi.nlm.nih.gov/mesh/C009462)
## Far-UV Circular Dichroism Spectroscopy (CD)
*In the **Far-UV Circular Dichroism Spectroscopy (CD)** section:*
Conformational changes in α-synuclein PMCA samples were monitored using a CD spectrometer (Jasco J-810), taking an average of five scans at 100 nm/min over the spectral range of 190–260 nm. The samples, first tested for ThT fluorescence, were loaded into a 0.5-mm path length quartz cuvette (Hellma) and scanned in Peltier temperature-controlled unit (Jasco), at 20 °C. The CD spectrum of the buffer alone was also evaluated and found to produce negligible spectra. The relative increase in secondary structure, corresponding to α-synuclein aggregation, was determined based on the decrease in negative absorbance, with a peak ∼200 nm and subsequent simultaneous increases in negative absorbance with a peak ∼ 218 nm, consistent with a change of structure from disordered monomers to β-sheet-rich amyloid fibrils.[](https://www.ncbi.nlm.nih.gov/mesh/C009462)
## Transmission Electron Microscopy
*In the **Transmission Electron Microscopy** section:*
The morphology of α-synuclein aggregates in PMCA samples was examined by transmission electron microscopy using a Phillips model EM208S microscope operated at 80 keV. Three-μl aliquots of 24 h PMCA or 8-day incubated samples were placed directly on carbon-coated 400-mesh grids, briefly washed with ddH2O, and negatively stained with 1–2% (w/v) phosphotungstic acid. Observations were made over a wide range of magnifications up to ×110,000 using a built-in CCD camera.[](https://www.ncbi.nlm.nih.gov/mesh/D002244)
## Native and SDS Gels
*In the **Native and SDS Gels** section:*
Three-μl aliquots of α-synuclein PMCA or non-PMCA control samples were either mixed with 1 μl of 4× loading buffer (NuPAGE LDS®; Invitrogen) and incubated at 100 °C for 10 min (for SDS gels) or mixed with 1 μl of 4× native loading buffer (NativePAGE®; Invitrogen), and 3.5 μl of the mixture was loaded into 4–12% SDS (Bis-Tris)- or native gels (Invitrogen). Either low molecular mass standard (Bio-Rad) or SeeBlue Plus2 (Invitrogen) protein ladders were used as molecular mass markers for Bis-Tris gels, whereas NativeMark® unstained protein standards were used for native gels. In some cases gels were stained using only Coomassie Blue, whereas in other experiments proteins were transferred onto PVDF Immobilon membranes (Millipore), and α-synuclein was visualized by incubation with monoclonal or polyclonal anti-α-synuclein antibodies. Chemiluminescence was induced by ECL-Plus (Pierce) and recorded with the Alliance software (Uvitec Cambridge).[](https://www.ncbi.nlm.nih.gov/mesh/D012967)
## Proteinase K (PK) Digestion
*In the **Proteinase K (PK) Digestion** section:*
Aliquots of 20 μl of α-synuclein PMCA samples or controls (non-PMCA) were incubated for 30 min at 37 °C with 4 μl of 0, 15, 60, or 600 μg/ml PK (Roche Applied Science) in conversion buffer with a final concentration of 0, 2.5, 10, or 100 μg/ml PK. Enzymatic reactions were terminated by adding 6 μl of 4× loading buffer (NuPAGE LDS®) and heating for 10 min at 95 °C. Fifteen-μl samples were loaded onto 4–12% or 12% Bis-Tris gels, and SeeBlue Plus2 was used as molecular mass standard. Gels were either stained with Coomassie Blue or electrotransferred for immunoblotting.[](https://www.ncbi.nlm.nih.gov/mesh/C026272)
## α-Synuclein PMCA “Defibrillation”
*In the **α-Synuclein PMCA “Defibrillation”** section:*
Drugs were added to 90 μm α-synuclein fibrils, previously generated by PMCA in a final 50 μm concentration and incubated at 37 °C with agitation at 750 rpm for 30 min . Samples were cooled, and 5 μl was taken to determine the presence of α-synuclein fibrils using the ThT assay as described previously.[](https://www.ncbi.nlm.nih.gov/mesh/C009462)
## Drugs and Antibodies
*In the **Drugs and Antibodies** section:*
Congo red, curcumin, quinacrine, resveratrol, lacmoid acid, tannic acid, ibuprofen, acetaminophen, and aspirin (all Sigma) were diluted in DMSO at various concentrations and then diluted also in DMSO at 1 and 5 mm concentrations. Of these aliquots, 2 μl were added to 198 μl of PMCA reactions for final 10 and 50 μm concentrations. The anti-α-synuclein antibodies Syn1 (BD Biosciences), 5C2 (Novus Biologicals), Per7 (14) and Per4 (3) were used for immunoblotting. The antibody Syn1 and Hoechst 33342 dye were used for immunofluorescence. The epitopes of the anti-α-synuclein antibody are: Per7, 1–120; 5C2, 61–95; LB509, 115–122; Per4, carboxyl-terminal; Syn1, 91–99.[](https://www.ncbi.nlm.nih.gov/mesh/D003224)
## Cell Infection with α-Synuclein PMCA Fibrils
*In the **Cell Infection with α-Synuclein PMCA Fibrils** section:*
SH-SY5Y (5 × 105) cells stably overexpressing human full-length α-synuclein (15) were seeded with either sonicated α-synuclein PMCA fibrils or monomeric recombinant α-synuclein (used as control) at a 3 μg/ml concentration in the cell media. Confluent cells were split 4 days after infection, and in every following passage a cell aliquot was plated on glass coverslips and immunostained fluorescently to detect α-synuclein aggregates. Ten fields per sample were counted in three different experiments.
## RESULTS
*In the **RESULTS** section:*
## Establishment of a Reproducible and Sensitive Method to Produce α-Synuclein Aggregates
*In the **Establishment of a Reproducible and Sensitive Method to Produce α-Synuclein Aggregates** section:*
In view of the propensity of α-synuclein to aggregate in vitro we set out to establish an α-synuclein PMCA to generate recombinant wild-type α-synuclein fibril assembly. The PMCA technique combines cycles of incubation at 37 °C (to grow fibrils) and sonication (to break fibrils into smaller growing fractions) of samples containing Triton X-100 for solubility, avoiding precipitation of the aggregates. We compared the kinetics of full-length α-synuclein fibril growth by PMCA with the traditional incubation/shaking method, over 8 days and at nine different time points. ThT 480-nm emission was used as the readout for fibril assembly (16). The results (Fig. 1A) show that PMCA induces a faster kinetic of filamentous aggregate formation compared with incubation and shaking.
Fibril formation was detected following 6 h of PMCA with the maximal signal reached between 12 and 24 h following the beginning of the reaction. By the time α-synuclein fibrils were obtained by PMCA, and using the same concentration of recombinant protein, no fibrils were seen with the incubation/shaking method. With the latter method some ThT signal was observed after 4 days but with high variability. Furthermore, no ThT signal was present when PMCA was performed using β- instead of α-synuclein (Fig. 1B). After 24/48 h there was a decline in the PMCA-induced ThT signal.
To clarify the reasons for this decrease the biochemical characteristics of the samples at 1 and 8 days following PMCA were studied by SDS-PAGE with Coomassie Blue staining and Western blotting. The results showed that following 8 days of PMCA no monomeric α-synuclein was detectable, and all of the protein was concentrated in a high molecular mass smear (data not shown). It is likely that ThT does not have easy access to big aggregates, as those present in our system at this time point, and hence the reduced signal, although more work is needed to confirm this hypothesis.[](https://www.ncbi.nlm.nih.gov/mesh/D017830)
α-Synuclein PMCA. A, growth kinetics of full-length recombinant α-synuclein fibril assembly by PMCA and incubation methods. B, α- and β-synuclein PMCA fibril formation compared with non-PMCA samples in a 24-h reaction. C and D, kinetics of α-synuclein fibril formation with PMCA or incubation (Inc) using 30 μm (C) or 90 μm (D) recombinant α-synuclein substrate, with or without seeding with recombinant α-synuclein PMCA fibrils. Assembly was monitored by the enhancement over time of ThT fluorescence intensity at 480 nm. Each point represents mean ± S.D. (error bars) of three replicates and is representative of two experiments. Fibril formation is faster with PMCA compared with the incubation method both in the presence and absence of α-synuclein fibril seeds.[](https://www.ncbi.nlm.nih.gov/mesh/C009462)
To explore the effect of seeding on the initial substrate concentration and time required for fibril formation, we investigated the kinetics of fibril formation of 30 and 90 μm (Fig. 1, C and D) recombinant α-synuclein substrates in the absence or presence of 0.9 μm α-synuclein fibrils in a 24-h PMCA or in a regular incubation reaction. When recombinant α-synuclein fibrils were added to the reaction at both concentrations, a small increase in the ThT signal was observed with the incubation method, whereas using PMCA fibrils formed as fast as 2 h, with the maximum level reached between 4 and 8 h (Fig. 1, C and D). The results of the 24-h 90 μm PMCA samples shown (Fig. 1, A, B, and D), have a ThT signal average of 210.3 ± 20.4 with a 9.7% standard deviation, exhibiting high reproducibility between experiments.[](https://www.ncbi.nlm.nih.gov/mesh/C009462)
## Biochemical Characterization of α-Synuclein Fibrils Generated by PMCA
*In the **Biochemical Characterization of α-Synuclein Fibrils Generated by PMCA** section:*
The biochemical characteristics of the α-synuclein material generated by PMCA were investigated by several techniques to confirm aggregate formation. Circular dichroism (CD) was performed to compare the product of PMCA α-synuclein and non-PMCA control samples (Fig. 2A). Comparison of the spectra showed an increase in β-sheet content in α-synuclein PMCA samples compared with the non-PMCA-treated α-synuclein that remained mainly unfolded. Negative staining electron microscopy was performed on the samples. Fibrils of heterogeneous sizes were present in high amounts in the PMCA α-synuclein sample reaction whereas in the incubated sample filaments were less abundant and longer (Fig.
2B). Native gel electrophoresis (Fig. 2C) was also used to compare samples subjected or not to 24-h PMCA. Both Coomassie Blue staining and immunoblotting with specific α-synuclein antibodies showed that only after PMCA were large aggregates of α-synuclein present. The proteinase K resistance of recombinant α-synuclein PMCA samples compared with non-PMCA samples was then determined (Fig. 2D).
Non-PMCA-treated protein was easily digested by 10 μg/ml PK whereas the α-synuclein PMCA sample was resistant to digestion up to 100 μg/ml PK. Immunoblotting of 2.5 μg/ml PK-digested samples with antibodies against five α-synuclein epitopes located in the amino-terminal, central, and carboxyl-terminal part of the protein was performed. The result suggested that the amino-terminal fragment of the protein was resistant to PK digestion (Fig. 2E). Finally the existence of “strain-like” modifications in the fibril formation by performing serial PMCAs was investigated, but the pattern of bands after PK digestion remained constant after eight passages (Fig. 2F).[](https://www.ncbi.nlm.nih.gov/mesh/C048139)
Characterization of α-synuclein PMCA-derived fibrils. A, circular dichroism of recombinant α-synuclein before and after 24-h PMCA. B, transmission electron microscopy of 24-h PMCA (upper panels) and incubation (lower panels) α-synuclein fibrils in carbon coated grids at two different magnifications. C, Coomassie Blue-stained (left) and anti-α-synuclein fluorescence-immunostained (right) native gels of recombinant α-synuclein subjected (+) or not (−) to 24-h PMCA. An increase in high molecular mass species is present following PMCA. D, Coomassie Blue-stained (left) or anti-α-synuclein immunostained (right) Bis-Tris gel before and after 24-h PMCA samples following digestion with different concentrations of PK. Whereas non-PMCA samples contain mainly monomeric protein that is completely degraded by PK, the PMCA-derived samples show specific bands following PK digestion. Asterisk indicates the position of the PK band in the Coomassie Blue-stained gel. E, α-synuclein 24-h PMCA, before (−) and after (+) PK digestion, epitope mapping using several anti-α-synuclein-specific antibodies. F, Coomassie Blue-stained SDS (Bis-Tris) gel of serial α-synuclein 24-h PMCA samples after 2.5 μg/ml PK digestion. No clear significant difference is observed in band pattern after PK digestion in samples from different PMCA passages.[](https://www.ncbi.nlm.nih.gov/mesh/D002244)
## α-Synuclein PMCA for Anti-amyloid Drug Testing
*In the **α-Synuclein PMCA for Anti-amyloid Drug Testing** section:*
Our aim was to set up a rapid system for screening compounds affecting α-synuclein aggregation. We therefore investigated the effects on PMCA aggregation of α-synuclein of compounds previously described to affect amyloid aggregation differently.
Congo red and curcumin were selected because their effect has been widely studied in prions and they have been also reported to interact with α-synuclein filamentous aggregates (17–19). As a negative control quinacrine was selected because it is known not to inhibit prion aggregation in vitro although it is effective in vivo (20). Other previously studied drugs, non-steroidal anti-inflammatory drugs such as Ibuprofen (21), acetaminophen (22), and aspirin (23) were used to evaluate the specificity of the assay. The remaining drugs tested included lacmoid and resveratrol, with reported binding to α-synuclein (24, 25); tannic acid and (−)epigallocatechin gallate (EGCG), previously described as potential inhibitors of α-synuclein aggregation (26–29).[](https://www.ncbi.nlm.nih.gov/mesh/D003224)
We initially established that DMSO, used to dilute the drugs, did not affect α-synuclein PMCA, then we tested two different concentrations of each drug, 10 and 50 μm, to determine presence and potency of their inhibition in a 16-h PMCA reaction. The drug screening results (Fig. 3A) showed a great percentage of inhibition of ThT signal (70–90%) by Congo red and curcumin at both concentrations tested, lower inhibition (35–40%) with EGCG, tannic acid, and lacmoid, and no inhibition with the remaining drugs. To avoid artifacts, the results were confirmed by SDS-PAGE before and after PK digestion followed by Coomassie Blue staining. This revealed that also resveratrol (which has fluorescence emission in the presence of β-sheet structures that overlaps with the emission of ThT (25)) was inhibiting α-synuclein aggregation during PMCA (Fig. 3B).
Furthermore, we studied the disaggregating properties of the same battery of drugs in preformed PMCA α-synuclein fibrils. The same drugs that inhibited the aggregation of α-synuclein during PMCA (Fig. 3C) were also able to disaggregate preformed α-synuclein aggregates.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
Compound screening using α-synuclein PMCA. A, 16-h α-synuclein (alpha-syn) PMCA reaction, alone or in the presence of solvent (DMSO) and two concentrations of 10 different drugs added at the beginning of the reaction. ThT signal is measured as readout for α-synuclein fibril assembly. Data are normalized for α-synuclein signal (α-synuclein) without drugs and presented as mean ± S.D. (error bars) of triplicate samples in two independent experiments. B and C, Coomassie Blue staining of SDS (Bis-Tris) gel of the drug-PMCA samples before (B) and after (C) digestion with 2.5 μg/ml PK. The order of the drugs (as indicated by name abbreviation) corresponds to that in A. The lanes indicated as syn show α-synuclein without drugs. D, ThT signal of defibrillated α-synuclein (alpha-syn) PMCA samples following 30-min incubation with 50 μm concentration of 10 different drugs added to preformed fibrils.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
## α-Synuclein “Chronically Infected” Cells
*In the **α-Synuclein “Chronically Infected” Cells** section:*
To confirm the efficacy of the compounds selected by PMCA, we set up a cellular system from which to obtain α-synuclein aggregates. SH-SY5Y neuroblastoma cells stably transfected with human full-length α-synuclein were exposed to α-synuclein PMCA material and then split on confluence (Fig. 4A). The presence of α-synuclein aggregates was investigated by immunofluorescence with anti-α-synuclein antibodies, and the result showed accumulation of α-synuclein for up to 10 divisions following the initial exposure to the aggregates (Fig. 4C). The percentage of cells with accumulated α-synuclein remained fairly constant at approximately 25% during the passages, and without further exposure, indicating that the cells were chronically infected (Fig. 4B).
α-Synuclein aggregation cell model. A, diagram representing the cell “infection” experiment using SH-SY5Y neuroblastoma cells overexpressing human α-synuclein. B, cell count expressed as percentage of infected cells (showing accumulated α-synuclein) per passage. These results are from three independent experiments. C, immunofluorescence staining using Syn1 anti-α-synuclein antibody (red) and Hoechst dye staining (blue) in α-synuclein-transfected SH-SY5Y control cells and cells collected at several passages after incubation with PMCA-derived aggregates.[](https://www.ncbi.nlm.nih.gov/mesh/D001562)
## DISCUSSION
*In the **DISCUSSION** section:*
α-Synuclein is a critical protein in Parkinson disease and other neurodegenerative diseases called α-synucleinopathies. Several missense mutations, duplications and triplications of the α-synuclein gene (SNCA) are associated with hereditary forms of Parkinson disease. Additionally, all sporadic Parkinson cases as well as those associated with α-synuclein mutations have aggregated α-synuclein in Lewy bodies. These aggregates are believed to be involved in toxicity and contribute to the loss of neuronal function (for review, see Ref. 30). Moreover, the ability of α-synuclein aggregates to travel from cell to cell, spreading as seeds to form newly misfolded α-synuclein aggregates in host cells (31–33), makes the protein a target for therapy.
Indeed, inhibiting the aggregation of α-synuclein would impede cell-to-cell transmission of seeds and stop progression of the disease.
Recombinant α-synuclein can form in vitro filamentous aggregates similar to those found in human brain (32–35). Although recombinant α-synuclein aggregates have been formed in vitro for 15 years, the methods used have some limitations mainly for two reasons: unless mutant α-synuclein is used, the generation of fibrils takes days and the concentration of recombinant protein needed is very high, between 300 and 500 μm. Recently, an α-synuclein-adapted PMCA was published (36), but it still required high concentration of recombinant protein, seeding of the reaction with preformed fibrils and long times.
In this study we have set up a highly efficient PMCA for α-synuclein based on modifications of prion PMCA (12). In a concentration-dependent reaction we can generate α-synuclein fibrils in 6 h, or 2 h when seeds of preformed fibrils are added. The concentration of recombinant α-synuclein is much lower than in most methods described previously (37, 38) for aggregation; moreover, the method is highly reproducible.
When α-synuclein fibril formation rate was compared between PMCA and the traditional incubation method a significant difference was found. Fibrils were present already after 6 h of PMCA whereas using the incubation method and the same low substrate concentration, filaments only started to appear after 4 days. Furthermore, the PMCA method showed great reproducibility between the triplicates or between different experiments as indicated by the low S.D. values, averaging 10% S.D. both within the same experiment and between different experiments, whereas this was not the case for the incubation method that had greater variability.
To study whether the α-synuclein aggregation was specific and not an artifact of a system that would generate aggregates out of any protein, we performed PMCA using as substrate β-synuclein. This protein has a 63% homology with α-synuclein, and is not present in Lewy body filaments. Furthermore, it does not aggregate in vitro unless in the presence of metals, glycosaminoglycans, or molecular crowding (39). The result demonstrated that the PMCA was specific for α-synuclein, as an aggregation prone protein, because no fibrils were obtained when β-synuclein was used.
In vitro generated α-synuclein aggregates have been shown to present biophysical and biochemical characteristics similar to in vivo α-synuclein aggregates, and therefore we wanted to determine whether our PMCA-generated α-synuclein aggregates had the same hallmarks. The folding pattern of α-synuclein after PMCA revealed a high content in β-sheet structure by circular dichroism compared with the predominantly unfolded non-PMCA control. Electron microscopy and negative staining showed in the PMCA samples a heterogeneous population of fibrils with different lengths, in contrast to the longer filaments obtained with the incubation method. The difference in the fibril length is probably a reflection of the breakup of the PMCA α-synuclein fibrils during sonication. When the PMCA samples were run on native gels they showed high molecular mass aggregates compared with the low molecular mass forms of the soluble native protein in the non-PMCA control.
As for the resistance of newly generated α-synuclein aggregates to digestion with proteinase K, there were fragments still resistant to high concentrations of the enzyme in the PMCA sample compared with the non-PMCA control. Those fragments consisted mainly of amino-terminal regions, as anti-α-synuclein antibodies recognizing epitopes toward the carboxyl terminus, such as LB509 and Per4, failed to recognize some of the PK-resistant fragments. The unchanged pattern of PK-resistant bands in a serial PMCA indicated an absence of conformational or “strain” differences (40) in the samples.
Together, these results show that α-synuclein PMCA promotes the formation of α-synuclein aggregates with all biochemical features characteristic of α-synuclein aggregates in vivo. Therefore, the α-synuclein PMCA is a fast and low protein consuming method to mimic α-synuclein fibril growth. Thereafter, we explored the potential of the α-synuclein PMCA to screen for compounds that by interfering with α-synuclein aggregation would be candidates for therapy in α-synucleinopathies.
A panel of 10 compounds was chosen to prove our concept. The anti-amyloid properties of some compounds (Congo red, curcumin, resveratrol) previously established for aggregation-prone proteins, such as prions, β-amyloid, or α-synuclein. Non-steroidal anti-inflammatory drugs, initially described as α-synuclein aggregate modulators (23), but without real effect in patients, as demonstrated by epidemiological studies (41, 42), were included to test PMCA specificity. Other compounds were reported to bind α-synuclein or to alter its aggregation in other in vitro assays. Finally, quinacrine did not have any relationship with amyloids or protein aggregation and therefore was chosen as a negative control.
In our drug screening α-synuclein aggregation was strongly inhibited by Congo red, curcumin, and resveratrol (as shown by PK digestion but not ThT assay for the latter, because resveratrol has its own fluorescence), and to a lower extent by EGCG, tannic acid, and lacmoid. Very recently it was reported that lacmoid does not prevent α-synuclein aggregation as measured by “amyloid intrinsic fluorescence” (43); however, here we showed by Coomassie Blue and Western blotting that the content of high molecular α-synuclein bands and PK resistance were decreased following lacmoid. This confirmed that the decrease in ThT labeling in the presence of lacmoid corresponded in fact to a reduction of aggregated α-synuclein. None of the non-steroidal anti-inflammatory drugs influenced the formation of α-synuclein aggregates. These results, where only the drugs with proven anti-amyloid activity and/or interaction with α-synuclein have an effect inhibiting the reaction, revealed a high specificity of the system in detecting drugs with a high probability to interfere with α-synuclein pathology.
Moreover, PMCA was done in a high-throughput format that will allow screening of a large number of compounds using low amounts of recombinant proteins and in short time. Therefore we believe that the α-synuclein PMCA model that we present here is a good tool for anti-aggregation drug screening.[](https://www.ncbi.nlm.nih.gov/mesh/D003224)
To further extend the α-synuclein fibril characterization, looking at spreading, and to generate an ex vivo assay for further drug screening, we inoculated SH-SY5Y neuroblastoma cells overexpressing human full-length α-synuclein with α-synuclein PMCA fibrils. Although the cells were in contact with the recombinant fibrils for just 4 days at the beginning of the experiment, a constant percentage of infected cells (cells with accumulated α-synuclein) was present at all analyzed cell passages. We hypothesize that probably there are two mechanisms involved in the maintenance of the persistent infection: first, the turnover of cells dying and other naïve cells taking up the released aggregates, and secondly the existence of cell-to-cell transmission. Further studies to verify our hypothesis are currently ongoing. Nevertheless, these chronically infected cultures provide an expandable and reproducible cellular system for α-synuclein aggregation that can be used for drug testing, as well as for investigating the pathways involved in the spread and cell response to α-synuclein aggregation.
In summary, our results show that α-synuclein PMCA is a fast and reproducible system that can be used for high-throughput screening for α-synuclein anti-aggregating compounds. This system, complemented with chronically infected cells, is relevant for identifying therapeutic compounds for Parkinson disease and other α-synucleinopathies.
This work was supported by a grant from the Parkinson's UK (to R. A. B., M. G. S., and M. E. H.) and by funding from the UK Medical Research Council (to M. G., S. Z., and G. F.).
PMCA
protein misfolding cyclic amplification
Bis-Tris[](https://www.ncbi.nlm.nih.gov/mesh/C026272)
bis(2-hydroxyethyl)iminotris(hydroxymethyl)methane[](https://www.ncbi.nlm.nih.gov/mesh/C026272)
DMSO[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
dimethyl sulfoxide[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
EGCG[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
(−)epigallocatechin gallate[](https://www.ncbi.nlm.nih.gov/mesh/C045651)
PK
proteinase K
ThT[](https://www.ncbi.nlm.nih.gov/mesh/C009462)
thioflavin T.[](https://www.ncbi.nlm.nih.gov/mesh/C009462)
The abbreviations used are:
# REFERENCES
*In the **REFERENCES** section:*
|
# Introduction
A Combined Experimental and Computational Study of [Vam3](https://www.ncbi.nlm.nih.gov/mesh/C000598179), a Derivative of [Resveratrol](https://www.ncbi.nlm.nih.gov/mesh/D000077185), and Syk Interaction
# Abstract
*In the **Abstract** section:*
Spleen tyrosine kinase (Syk) plays an indispensable role through preliminary extracellular antigen-induced crosslinking of Fc receptor (FcR) in the pathogenesis of autoimmune disorders, such as rheumatoid arthritis. In this study, we identify Vam3, a dimeric derivative of resveratrol isolated from grapes, as an ATP-competitive inhibitor of Syk wit[h an](https://www.ncbi.nlm.nih.gov/mesh/C000598179) IC50 of 62.95 nM in an in[ vitro kina](https://www.ncbi.nlm.nih.gov/mesh/D000077185)se assay. Moreover, docking a[nd ](https://www.ncbi.nlm.nih.gov/mesh/D000255)molecular dynamics simulation approaches were performed to get more detailed information about the binding mode of Vam3 and Syk. The results show that 11b-OH on ring-C and 4b-OH on ring-D could form two hydrogen bonds wit[h Gl](https://www.ncbi.nlm.nih.gov/mesh/C000598179)u449 and Phe382 of Syk, respectively. In addition, arene-cation interaction between [ring-D o](https://www.ncbi.nlm.nih.gov/mesh/D006859)f Vam3 and L[ys4](https://www.ncbi.nlm.nih.gov/mesh/D018698)02 of Sy[k w](https://www.ncbi.nlm.nih.gov/mesh/D010649)as also observed. These results indicat[e tha](https://www.ncbi.nlm.nih.gov/mesh/D006841)t ring-C and D play an essential role [in V](https://www.ncbi.nlm.nih.gov/mesh/C000598179)am3–S[yk ](https://www.ncbi.nlm.nih.gov/mesh/D008239)interaction. Our studies may be helpful in the structural optimization of Vam3, and also aid the [desi](https://www.ncbi.nlm.nih.gov/mesh/C000598179)gn of novel Syk inhibitors in the future.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
## 1. Introduction
*In the **1. Introduction** section:*
Allergic and autoimmune disorders share significant functional overlap in the biologic pathways responsible for the activation of signal transduction events leading to production of numerous proinflammatory factors involved in disease initiation and progression. Given the reciprocal connections in these mechanistic pathways, it would be advantageous to target strategic master regulators with novel therapeutics to treat allergic and autoimmune diseases. One such crucial regulator is spleen tyrosine kinase (Syk). Syk is a cytosolic non-receptor tyrosine kinase. It serves as a key mediator of B-cell receptor and Fc receptor mediated signaling in inflammatory cells such as B cells, mast cells, macrophages, dendritic cells, and neutrophils and is involved in bone resorption by osteoclasts.
Activation of Syk occurs through preliminary extracellular antigen-induced crosslinking of FcεR1 and FcγRs I, IIA, and IIIA. Upon abnormal activation, Syk, a master upstream regulator of signal transduction, propagates downstream signaling molecules, resulting in initiation of disease. Furthermore, in specific contexts, uncontrolled activation of B cell receptor (BCR) signaling via Syk would lead to development of lymphomas and leukemia. Murine studies have shown that Syk expression is required for the survival of Non-Hodgkin Lymphomas-like (NHL-like) tumors in vitro. Pharmacologic inhibition of Syk induced apoptosis in murine B-cell lymphomas in vitro and resulted in regression of NHL-like B-cell lymphomas.
Full-length Syk is composed of two N-terminal Src homology 2 (SH2) domains followed by an interdomain linker and a C-terminal kinase domain. The tandem SH2 (tSH2) module is also separated by an inter-SH2 linker and serves as a docking platform for immune receptor tyrosine-based activating motifs (ITAMs) which are displayed at the cytosolic side of the plasma membrane. The truncated kinase domain of Syk (Syk-KD), which contains an ATP-binding pocket shows significant catalytic activity and has been extensively used for inhibitor design.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
In recent years, a number of Syk inhibitors have been discovered for treatment of autoimmune, allergic and autoinflammatory diseases and most of them are ATP-competitive Syk inhibitors. Several research groups have independently reported the crystal structures of the Syk catalytic domain, co-crystallized and ligand soaked with small molecule inhibitors. For example, as shown in Figure 1, OSB and 1B6 are two ATP-competitive inhibitors of Syk with an IC50 of 60 and 26 nM, respectively. The crystal structures of OSB and 1B6 with the catalytic domain of Syk were reported by Marcos Castillo et al. and Fernando Padilla et al., respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D000255)
Structures of 1B6 and OSB.[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
Vam3 (amurensis H), a resveratrol dimer, was first isolated from ethanol extracts of Vitis amurensis Rupr as a secondary natural product. Previous studies indicated that Vam3 has anti-inflammatory effects, including alleviate the asthmatic inflammation in asthmatic mice and decrease cigarette smoke-induced autophagy in human bronchial epithelial cells. However, the molecular basis by which Vam3 inhibits inflammation is not clear. In this study, we identified Vam3 as a potent ATP-competitive inhibitor of Syk kinase and it might exert its anti-inflammatories through the Syk pathway. As depicted in Figure 2c, Vam3 is a polyphenol hydroxyl natural product.
Compared with other Syk inhibitors which contain different amounts of N atoms, Vam3 owns a polyphenol hydroxyl scaffold with no N atoms. This might provide a new strategy to design novel Syk inhibitors. However, the solubility of Vam3 in water is poor. Structural changes on Vam3 to improve its solubility should not decrease the binding affinity of Vam3. Therefore, interaction between Vam3 and Syk interaction should be understood first.
Indeed, characterizing the 3D-structure of Syk–Vam3 complex using crystallization or nuclear magnetic resonance (NMR) techniques is the best way, but it is time and resource consuming.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
(a) IC50 determination of Vam3 with recombination Syk protein; (b) Ki determination of Vam3 with recombination Syk protein; (c) Chemical structure of Vam3.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
Fortunately, the comparably fast and inexpensive docking protocols can be combined with accurate but more expensive molecular dynamics (MD) simulation techniques to predict more reliable protein–ligand complex structures. In our work, molecular docking and dynamics simulation were carried out to investigate the binding mode of the Vam3 with Syk. To investigate the reliability of our stimulation methods, OSB and 1B6 were employed as controls during the docking studies and dynamics simulations. Resveratrol, the monomer of Vam3, was used as a negative control to validate the binding mode of Vam3–Syk complex. We hope that we can reveal the mechanism of the Vam3–Syk interaction and give some useful information to structure optimization of Vam3 as Syk selective inhibitor with good properties.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
## 2. Results and Discussion
*In the **2. Results and Discussion** section:*
## 2.1. Vam3 Inhibited Syk Kinase Activity in Vitro
*In the **2.1. Vam3 Inhibited Syk Kinase Activity in Vitro** section:*
Resveratrol is a polyphenolic compound found in grapes. Previous studies reported that resveratrol was a major Syk inhibitor and inhibits activation of Syk kinase in mast cell. Vam3 is a derivative of resveratrol. Ring-C and D of Vam3 share the same structure with Resveratrol. This suggests that Vam3 may also have the capacity for Syk inhibition.[](https://www.ncbi.nlm.nih.gov/mesh/D000077185)
To confirm that Syk was the cellular target of Vam3, in vitro kinase assays were performed by using purified Syk protein. As shown in Figure 2, Vam3 inhibited Syk kinase activity with an IC50 of 62.95 nM and Vam3 was shown to be an ATP-competitive inhibitor of Syk kinase with a Ki of 61.09 nM.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
## 2.2. Extra Precision Docking Studies
*In the **2.2. Extra Precision Docking Studies** section:*
Extra precision docking of Glide was carried out to investigate the binding mode of Vam3 with Syk. As for 1B6 and OSB, as revealed in Figure 3, two binding conformations of docking were performed respectively and there was no large difference between them. Therefore the conformations which achieved the highest GlideScore (G-score) were used as the initial structures for future binding mode analysis including a 15 ns MD simulation. As for Vam3, however, only one binding conformation was performed. This mainly came from the large rigidity of Vam3 and special shape of the ATP-binding pocket of Syk.
Therefore the only credible docking result of Vam3 was used in future binding mode analysis. As shown in Figure 4, the three molecules (1B6, OSB and Vam3), as all of them are ATP-competitive inhibitor of Syk, were docked into the APT-binding pocket of Syk and all of them were positioned in the same location of Syk. 1B6 and OSB possessed a “U”-shaped conformations in the pocket while Vam3 shown the “ψ”-shape conformation. The binding modes of 1B6, OSB and Vam3 are shown in panels of Figure 4b–d, respectively. The detailed interactions will be discussed further in the following molecular dynamics simulations.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
Superposition of conformations of docking results of 1B6 (a) and OSB (b).[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
(a) Docked structures of 1B6 (green), OSB (yellow) and Vam3(pink) with Syk; (b) The binding site positioned around 1B6; (c) The binding site positioned around OSB; (d) The binding site positioned around Vam3.[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
## 2.3. Molecular Dynamics Simulation Studies
*In the **2.3. Molecular Dynamics Simulation Studies** section:*
In the docking studies, flexibility of the protein was not taken into consideration. In order to find the key residues and position of Vam3–Syk interaction, we performed 15 ns MD simulations with the Desmond program in which flexibility of proteins were taken into consideration. Three different systems were studied, including 1B6-bound system, OSB-bound system and Vam3-bound system. 1B6-bound system and OSB-bound system were taken as controls. The root mean square deviation (RMSD) values of the backbone atoms relative to the initial structure were calculated to measure the convergence of the systems and ensure the rationality of the sampling method.
As depicted in Figure 5, the RMSD of the three were about 3.5 Å after 10 ns and all of them almost remained at this level in the following simulation processes. This indicated that the three systems were stable after 10 ns of simulation.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
The results of molecular dynamics (MD) simulation. The MD simulation time vs. root mean-square deviation (RMSD) of the backbone atoms for 1B6-bound system (black), OSB-bound system (red) and Vam3-bound system (green).[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
To estimate the difference between the MD average structures and crystal structures of controls, the average structures of the MD-simulated complexes from the last 5 ns of MD simulations were superimposed with the crystal structure of 1B6–Syk and OSB–Syk complexes. As shown in Figure 6 and Figure 7, the MD average structures of the two control complexes are overall very similar to their crystal structures. As for 1B6-bound system, 1B6 formed the same H-bonds with Glu449 and Ala451 of Syk just as in crystal structure. The same hydrophobic interactions with Leu377, Pro455 and Leu501 were also observed. Similarly in OSB-bound system, the same H-bonds and hydrophobic interactions in crystal structure of OSB with Syk were also found in MD average structure of OSB–Syk complex.
However, there were also little difference between MD average structures and crystal structures of controls. For example, H-bond with Lys402 in the crystal structure of 1B6 with Syk did not exist in crystal structure of 1B6 with Syk. OSB formed an H-bond with ASN499 but it was not found in MD average structure of OSB–Syk complex. The little difference between the MD average structures and crystal structures come from the little location drift and the later might due to the reason that we used whole Syk in our stimulations while crystal structures only contains catalytic domain. These results suggested that our methods, which using docking and molecular dynamics stimulation to investigate the interaction between Vam3 and Syk was reasonable and the MD average structures were very similar with the crystal structures.[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
Moreover, the analysis of root-mean-square fluctuation (RMSF) vs. the residue number for these three systems is illustrated in panel of Figure 8. In the figure, the ligand was not included. The residues (amino acids 377–628) of the catalytic domain was evaluated. RMSF can reflect the mobility of the residue around its mean position and is helpful to find the residues and regions of Syk with major conformational changes. Obviously, the RMSF distribution of Vam3-bound system was different with 1B6- and OSB-bound systems, which indicating that Vam3 could have a distinct interaction mode with Syk.
This may come from the difference between the structures of Vam3, 1B6 and OSB.[](https://www.ncbi.nlm.nih.gov/mesh/D000596)
(a) Superposition of MD average structures of 1B6 with Syk (purple) and crystal structures of 1B6–Syk complex (green); (b) Superposition of MD average structures of OSB with Syk (purple) and crystal structures of OSB–Syk complex (green).[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
The key interactions between 1B6 and OSB with Syk in the crystal structures are shown in (a,c); The key interactions between 1B6 and OSB with Syk in the MD average structures are preformed in (b,d), respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
The root-mean-square fluctuation (RMSF) of the catalytic domain of Syk for each residue for the 1B6-bound (black), OSB-bound (red) and Vam3-bound (green) system.[](https://www.ncbi.nlm.nih.gov/mesh/D000577)
## 2.4. Binding Mode Analysis of Vam3–Syk Complex
*In the **2.4. Binding Mode Analysis of Vam3–Syk Complex** section:*
The binding mode for the MD average structures of Syk with vam3 is displayed in Figure 9a,b. It can be observed that Vam3 extend deeply into the binding site of Syk. As is revealed previously, compared to 1B6 and OSB of which both could form three H-bonds with Syk, Vam3 could form two H-bonds with Syk. The H atom on 11b-OH of Vam3 could form hydrogen bond with the backbone atom of Glu449 and this kind of H-bond was also observed in 1B6–Syk interaction. The O atom on 4b-OH of Vam3 also had hydrogen bond interaction with the backbone atom of Phe382 and it is not observed in 1B6–Syk and OSB–Syk interactions or other inhibitors–Syk interactions.
However, Vam3 could not form H-bond with Ala451 and Asp512 which many Syk inhibitors favors to form H-bonds with. In addition, arene–cation action between ring-D of Vam3 and Lys402 was also observed. These results indicated that ring-C and ring-D and the two –OH groups on them were necessary for the activity of Vam3.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
Hydrophobic interactions are also presented between Syk and Vam3. The phenyl ring-B of Vam3 formed hydrophobic interaction with the side chain of Val385 and ring-A formed hydrophobic interactions with the side chain of Leu377 and Leu501. Fernando Padilla et al. had demonstrated that optimizing interactions of ATP-competitive Syk inhibitors with Pro455 and Asn457, present in only nine aligned kinases of a total of 433, was an attractive way of introducing high levels of Syk specificity. Unfortunately, interaction between Pro455 and Vam3 or Asn457 and Vam3 was not observed. This suggested that Vam3 may not have good Syk selectivity profile.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
Binding mode of Vam3 with Syk. (a) The key Vam3–Syk interactions in the Syk ATP binding pocket in 3-dimensional structure; (b) The two-dimensional projection of Vam3-Syk interaction.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
Furthermore, to validate the binding mode of Vam3 and Syk, the same molecular docking and 15 ns MD simulation studies of resveratrol and Syk was carried out. Resveratrol is the monomer of Vam3. It also possesses the same polyphenol hydroxyl structure like Vam3 and shows weak potency for Syk inhibition. Therefore, molecular docking and MD simulation studies on resveratrol may help us better understand the binding mode of Vam3 and Syk.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
Resveratrol was docked into the ATP-binding pocket of Syk and the conformation with the higher G-score was used in the MD simulation study. As depicted in Figure 10, the RMSD of the resveratrol–Syk system had reached a plateau after 8 ns. There were RMSD fluctuations from 10 to 15 ns as expected, and the fluctuations may come from the small size and weak binding ability of resveratrol. This indicated that the simulation was not necessarily converged when it reached a plateau in the RMSD.[](https://www.ncbi.nlm.nih.gov/mesh/D000077185)
The average structure of resveratrol–Syk complex from the last 5 ns of the MD simulation was carried out, as shown in Figure 11a. As for resveratrol, it formed H-bond with Glu 449 of Syk and hydrophobic interactions with Leu 377 and Leu 501. As revealed in Figure 11b, resveratrol positioned the same location with ring-A and C of Vam3. Similarly, Vam3 could formed H-bond with Glu 449 of Syk and hydrophobic interactions with Leu 377 and Leu 501. Besides, ring-D of Vam3 formed another H-bond with Phe 382 of Syk and arene–cation action with Lys 402 of Syk.
This could explain the potency for Syk inhibition of Vam3 is much better than resveratrol. In addition, to compare the binding modes of Vam3 and resveratrol with Syk, the RMSF analysis was carried out. As depicted in Figure 12, the RMSF distribution of resveratrol-bound system was similar with Vam3-bound system. This may come for the similar structures between resveratrol and Vam3. In summary, resveratrol and Vam3 which have the same skeleton share the similar inhibition mechanism for Syk.
More interactions were observed between Vam3 and Syk than resveratrol, which could explain the better inhibiting capacity for Syk of Vam3. These evidence suggest that the binding mode of Vam3–Syk complex is reasonable.[](https://www.ncbi.nlm.nih.gov/mesh/D000077185)
The MD simulation time vs. root mean-square deviation (RMSD) of the backbone atoms for resveratrol–Syk system.[](https://www.ncbi.nlm.nih.gov/mesh/D000077185)
(a) The two-dimensional projection of resveratrol–Syk interaction. (b) Superposition of MD average structures of resveratrol with Syk (cyan) and Vam3 with Syk (pink).[](https://www.ncbi.nlm.nih.gov/mesh/D000077185)
The RMSF of the catalytic domain of Syk for resveratrol-bound system (purple) and Vam3-bound system (green).[](https://www.ncbi.nlm.nih.gov/mesh/D000077185)
Therefore, as the essential roles of ring-C and D in Vam3–Syk interaction, structural optimization of Vam3 could focus on ring-A and B of Vam3 using chemical approach, such as bioisosteres. In addition, optimization of ring-A, the hydrophobic ring, with hydrophilic groups may improve its solubility.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
## 3. Materials and Methods
*In the **3. Materials and Methods** section:*
## 3.1. Experimental Studies
*In the **3.1. Experimental Studies** section:*
## 3.1.1. Plant Material
*In the **3.1.1. Plant Material** section:*
Vam3 was isolated from the ethanol extracts of a methanol extracts of Vitis amurensis Rupr, as described previously. This compound was prepared by dissolving in dimethyl sulfoxide (DMSO) and the final concentration of DMSO was adjusted to 0.1% (v/v) in culture media.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
## 3.1.2. In Vitro Fluorescence Polarization Kinase Assay
*In the **3.1.2. In Vitro Fluorescence Polarization Kinase Assay** section:*
The reactions were carried out in a total volume of 25 μL in 96-well microtiter plates. The Syk tyrosine kinase activity at single dose concentration of 12.5 ng/μL, 10 μL of volume, was carried out served as the enzyme source. The total volume of 10 μL mixture containing 0.2 μg/μL Poly (Glu, Tyr) sodium salt (4:1, Glu:Tyr, Sigma–Aldrich, St. Louis, MO, USA) and 10 μM ATP (Promega, Madison, WI, USA) served as the standardized substrate. The concentration range of the tested inhibitors employed in reactions was 0.0032, 0.016, 0.08, 0.4, 2, 10 μM or DMSO with 5 μL volume. All of the enzymatic reactions were conducted at 37 °C for 60 min.
The assay was terminated by adding 25 μL of ADP-GloTM Reagent (Promega, Madison, WI, USA). The 96-well plate was shaken and then incubated for 40 min at ambient temperature. Fifty microliter of Kinase detection reagent was added and the 96-well reaction plate was then read using the ADP-Glo Luminescences Protocol on a GloMax plate reader (Promega: Catalog #E7031). For each concentration of Vam3, the rate of reaction at each concentration of ATP was determined and plotted against the ATP concentration to determine the apparent Km and Vmax (maximal rate). Finally, the apparent Km (or apparent Km/Vmax) was plotted against the inhibitor concentration to determine the Ki.
All data analysis was performed using Prism and Prism enzyme kinetics programs.[](https://www.ncbi.nlm.nih.gov/mesh/C027606)
## 3.2. Computational Studies
*In the **3.2. Computational Studies** section:*
## 3.2.1. Preparation of Protein Target Structure
*In the **3.2.1. Preparation of Protein Target Structure** section:*
The crystal structure of full-length Syk in complex with ANP (PDB code: 4FL2, with the resolution of 2.19 Ǻ) was retrieved from the RCSB Brookhaven Protein Data Bank (PDB). The structure is the only well-defined full-length Syk with a resolution of 2.19 Ǻ except for the first part of the N-terminus (amino acids 1–8) and for the interdomain linker region (amino acids 265–336). Indeed, using the catalytic domain alone in the simulation would save time. However, the simulation of the whole protein with inhibitor would provide more information about the SH2 domains, which is helpful to further study.
Therefore 4FL2 was used as the receptor. Furthermore, 4FL2 is a full-length Syk in complex with AMP–PNP revealing an autoinhibited conformation which is close to the conformation of Syk–ATP complex. Therefore, using 4FL2 as the receptor is more suitable than the other structures and can make docking and MD simulation results persuasive and convincing. As for the missing residues being far away from the ATP-binding pocket, lacking of these 73 residues would not largely affect the results of our simulations. Therefore, we do not model this part.
Then the structure was prepared using the following procedures by the Protein Preparation Wizard in the Schrödinger software suite, including adding hydrogen atoms, assigning partial charges using the OPLS_2005 force field and assigning protonation states, and structure minimizing in vacuum. Finally, the cocrystal ANP was removed, and the resulting structure was used as the receptor model in the following studies.[](https://www.ncbi.nlm.nih.gov/mesh/D000227)
## 3.2.2. Ligand Preparation
*In the **3.2.2. Ligand Preparation** section:*
The structure of Vam3 and resveratrol was constructed using Mastro, while the ligands OSB and 1B6 were retrieved from the Protein Data Bank (PDB code: 4F4P and 4Y0T, respectively). All the ligands were prepared by using the LigPrep and then to proceed with stereoisomer generation, neutralization of charged structures and determination of the most probable ionization state at pH 7.2 ± 0.2. The OPLS-2005 forcefield was used for optimization to produce the low-energy conformer of the ligand.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
## 3.2.3. Molecular Docking
*In the **3.2.3. Molecular Docking** section:*
The ligands Vam3, resveratrol, OSB and 1B6 were docked into the receptor using Glide software. Glide approximated a complete systematic search of the conformational, orientational and positional space of the docked ligand, and a series of hierarchical filters was used to search for possible locations of the ligand in the active-site region. In this work, grid box was centered on the ATP centroid in the X-ray crystal structure of Syk and the ligands were docked into the box using the “extra precision” glide docking (Glide XP) which docks ligands flexibly and the protein rigidly. The quality of the geometric matches of the docked binding structures with the lowest GlideScore was visually checked and the best one was selected as the initial complex for further studies. GlideScore is based on ChemScore, but includes a steric-clash term and adds buried polar terms devised by Schrödinger to penalize electrostatic mismatches:where vdW, Coul, Lipo, H-bond, Metal, BuryP, RotB and Site are the van der Waals energy, Coulomb energy, Lipophilic contact term, Hydrogen-bonding term, Metal-binding term, Penalty for buried polar groups, Penalty for freezing rotatable bonds and Polar interactions in the active site, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
## 3.2.4. Molecular Dynamics Stimulation
*In the **3.2.4. Molecular Dynamics Stimulation** section:*
The initial coordinates for the MD calculations were taken from the docking results. For each system, MD studies were performed using OPLS_2005 force field in an explicit solvent with the TIP3P model of water within the Desmond software. The dimensions of each orthorhombic water box were 100 × 100 × 100 Å, which ensured that the entire surface of each complex was covered by the solvent model and the systems were neutralized by adding Cl− counter ions to balance the net charges of the systems. Before equilibration and long production MD simulations, the systems were minimized and pre-equilibrated using the default relaxation routine implemented in Desmond.
The solvated system was minimized first with solute restrained and then again minimized without solute restraints by using hybrid method of steepest descent and the LBFGS (limited-memory Broyden–Fletcher–Goldfarb–Shanno) algorithm with a maximum of 2000 steps including initial 10 steps of steepest descent. The minimized system was passed through a short 12 ps simulation in the NVT ensemble using a temperature of 10 K with nonhydrogen solute atoms restrained. Subsequently, the system was simulated for 12 ps in the NPT ensemble using temperature 10 K with restraints on nonhydrogen solute atoms. In the next step, the system was simulated for 24 ps in NPT ensemble using a temperature of 300 K restraining the nonhydrogen solute atoms. In the last step of equilibration process, the system was further simulated for 24 ps in the NPT ensemble with no restraints at temperature 300 K.
The temperatures and pressures in the short initial simulations were controlled using Berendsen thermostats and barostats, respectively. Then, each system was performed for a 15 ns long production MD simulation. The OPLS_2005 force field was used along with the MacroModel module to provide and check the necessary force field parameters for the ligands. When MacroModel performs an energy calculation, the program checks the quality of each parameter in use. The use of low quality parameters, especially torsional ones, may result in inaccurate conformational energy differences and geometries.
Bond, angle, torsional angle and improper angle checked parameters were listed as high- and medium-quality force field parameters for all ligands studied. During the MD simulations, the equations of motion were integrated with a 2 fs time step in the NPTensemble. The Shake algorithm was applied to all hydrogen atoms; the van der Waals (VDW) cutoff was set to 9 Å. The temperature was maintained at 300 K, employing the Nose-Hoover thermostat method with a relaxation time of 1 ps. Long-range electrostatic forces were taken into account by means of the particle-mesh Ewald (PME) approach.
Data were collected every 12 ps during the MD runs. Visualization of protein–ligand complexes and MD trajectory analyses were carried out with the VMD software package. The equilibration was monitored by examining the stability of the temperature, energy, and the density of the system as well as the RMSD of the backbone atoms.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## 4. Conclusions
*In the **4. Conclusions** section:*
In this study, we first demonstrated that Vam3 is an ATP-competitive inhibitor of Syk with IC50 of 62.95 nM and Ki of 61.09 nM by using in vitro fluorescence polarization kinase assay. Moreover, to investigate the mechanism of Vam3–Syk interaction, docking studies and molecular dynamics stimulations were performed. Through our stimulations, we have predicted optimal binding conformation of Vam3 with Syk. 11b-OH and 4b-OH of Vam3 formed two H-bonds with Glu449 and Phe382 in the active site of Syk, respectively. Arene-cation action was also found in Vam3–Syk interaction.
Together with hydrophobic interactions, these actions form the basis of the well inhibitory activity of Vam3. These results may not only useful for the structural optimization of Vam3 but also for the rational design of novel Syk inhibitors with new scaffold.[](https://www.ncbi.nlm.nih.gov/mesh/C000598179)
# Author Contributions
*In the **Author Contributions** section:*
Renpin Liu and Ying Chen designed and performed the in vitro fluorescence polarization kinase assay, and Ming Jiang and Qisheng Zheng performed all the computational studies. Ming Jiang, Renpin Liu, Saijun Fan, and Peixun Liu reviewed the data and wrote the paper. Ming Jiang, Qisheng Zheng and Peixun Liu checked, revised and finalized the paper.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
|
# Introduction
[Hypochlorous acid](https://www.ncbi.nlm.nih.gov/mesh/D006997) via [peroxynitrite](https://www.ncbi.nlm.nih.gov/mesh/D030421) activates protein kinase Cθ and insulin resistance in adipocytes
# Abstract
*In the **Abstract** section:*
We recently reported that genetic deletion of myeloperoxidase (MPO) alleviates obesity-related insulin resistance in mice in vivo. How MPO impairs insulin sensitivity in adipocytes is poorly characterized. As hypochlorous acid (HOCl) is a principal oxidant product generated by MPO, we evaluated the effects of[ HOCl on insulin ](https://www.ncbi.nlm.nih.gov/mesh/D006997)si[gnal](https://www.ncbi.nlm.nih.gov/mesh/D006997)ing in adipocytes differentiated from 3T3-L1 cells. Exposure of 3T3-L1 adipocyt[es t](https://www.ncbi.nlm.nih.gov/mesh/D006997)o exogenous HOCl (200 μmol/l) attenuated insulin-stimulated 2-deoxyglucose uptake, GLUT4 translocation, and insul[in s](https://www.ncbi.nlm.nih.gov/mesh/D006997)ignals, including tyrosine phosphorylation o[f insulin rece](https://www.ncbi.nlm.nih.gov/mesh/D003847)ptor substrate 1 (IRS1) and phosphorylation of Akt. Furthermo[re, trea](https://www.ncbi.nlm.nih.gov/mesh/D014443)tment with HOCl induced phosphorylation of IRS1 at serine 307, inhibitor κB kinase (IKK), c-Jun NH2-terminal kin[ase ](https://www.ncbi.nlm.nih.gov/mesh/D006997)(JNK), and phosphorylation of PKCθ ([PKCθ).](https://www.ncbi.nlm.nih.gov/mesh/D012694) In addition, genetic and pharmacological inhibition of IKK and JNK abolished serine phosphorylation of IRS1 and impairment of insulin signaling by HOCl. Furthermore, knockdown of[ PKCθ ](https://www.ncbi.nlm.nih.gov/mesh/D012694)using siRNA transfection suppressed phosphorylation of IKK and J[NK a](https://www.ncbi.nlm.nih.gov/mesh/D006997)nd consequently attenuated the HOCl-impaired insulin signaling pathway. Moreover, activation of PKCθ by peroxynitrite was accompanie[d by](https://www.ncbi.nlm.nih.gov/mesh/D006997) increased phosphorylation of IKK, JNK, and IRS1-serine 307. In contr[ast, ONOO− in](https://www.ncbi.nlm.nih.gov/mesh/D030421)hibitors abolished HOCl-induced phosphorylation of PKCθ, IKK, JNK, a[nd IRS](https://www.ncbi.nlm.nih.gov/mesh/D012694)1-serine 307, as we[ll as](https://www.ncbi.nlm.nih.gov/mesh/D030421) insulin resistance. F[inal](https://www.ncbi.nlm.nih.gov/mesh/D006997)ly, high-fat diet (HFD)-induced insulin resistance wa[s asso](https://www.ncbi.nlm.nih.gov/mesh/D012694)ciated with enhanced phosphorylation of PKCθ, IKK, [JNK](https://www.ncbi.nlm.nih.gov/mesh/D004041), and IRS1 at serine 307 in white adipose tissues from WT mice, all of which were not found in Mpo knockout mice fe[d HFDs](https://www.ncbi.nlm.nih.gov/mesh/D012694). We conclude that HOCl impairs insulin signaling pathway by increasing ONOO− mediated phosphorylation of PKCθ, resultin[g in](https://www.ncbi.nlm.nih.gov/mesh/D006997) phosphorylation of IKK/JNK and consequent serine[ phos](https://www.ncbi.nlm.nih.gov/mesh/D030421)phorylation of IRS1 in adipocytes.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Insulin resistance, a hallmark of obesity and fundamental cause of type 2 diabetes, is characterized by a diminished ability of insulin to regulate glucose homeostasis in insulin-sensitive organs including liver, skeletal muscle, and adipose tissue. Insulin resistance is partly caused by chronic low-level inflammation and oxidative stress in adipose tissue (Guo 2014). Infiltration of inflammatory cells, which produce cytokines and oxidants, leads to a local inflammatory environment in adipose tissue (Olefsky & Glass 2010). Accumulated evidence indicates that increased infiltration of neutrophils in adipose tissue is strongly related to the development of insulin resistance in human obesity (Elgazar-Carmon et al. 2008, Talukdar et al. 2012). Recently, we reported that neutrophil-derived myeloperoxidase (MPO) activation plays an important role in obesity-induced insulin resistance (Wang et al. 2014).
Hypochlorous acid (HOCl) is a potent oxidant formed from hydrogen peroxide and chloride ions in a reaction catalyzed by MPO (Harrison & Schultz 1976). Recently, a high correlation between the production of HOCl and metabolic disorder was identified, i.e. the concentration of HOCl in plasma was elevated in obese subjects and hypertensive patients (Yang et al. 2013). In addition, HOCl-modified proteins were present in liver and adipose tissue of obese patients (Rensen et al. 2009). However, the molecular mechanism and role of HOCl in the pathogenesis of insulin resistance remain to be determined.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
The insulin signaling transduction cascade initiated by insulin binding to its receptor causing receptor autophosphorylation and tyrosine phosphorylation of insulin receptor substrate 1 (IRS1) subsequently activating phosphoinositide 3-kinase and Akt and finally inducing translocation of intracellular GLUT4 vesicles to the cell membrane in order to enhance the uptake of glucose (Bevan 2001). Insulin resistance is caused by impaired insulin signal transduction accompanied by decreased activation of downstream obligate molecular intermediates (Saltiel & Kahn 2001). Accumulating lines of evidence have indicated that serine phosphorylation of IRS1 led to inhibition of insulin signals via interference with tyrosine phosphorylation of IRS1 and acceleration of its degradation (Gual et al. 2005). Moreover, activation of several inflammatory kinases, including inhibitor κB kinase (IKK), c-Jun NH2-terminal kinase (JNK), and protein kinase C (PKC), induces serine phosphorylation of IRS1 by inflammatory cytokines and pro-oxidants (Aguirre et al. 2000, Morino et al. 2005, Weigert et al. 2008). HOCl is known as a potent oxidant and a major inflammatory mediator inducing tissue injury in a number of inflammatory diseases (Souza et al. 2011).
Therefore, we propose the hypothesis that HOCl mediates insulin resistance through activation of PKC, IKK, and JNK, resulting in subsequent serine phosphorylation of IRS1 in adipocytes. Here, we report that exogenous HOCl impaired the insulin signaling pathway and induced phosphorylation of IRS1 at Ser307, IKK, JNK, and PKCθ in 3T3-L1 adipocytes. In contrast, HOCl-impaired insulin signals were abolished after knockdown of IKKβ and JNK using siRNA or a pharmaceutical inhibitor. Moreover, PKCθ knockdown attenuated phosphorylation of IKK and JNK, resulting in restoration of insulin sensitivity. Strikingly, exogenous HOCl-induced insulin resistance and phosphorylation of PKCθ were prevented by an ONOO− scavenger.
Overall, our findings provide a novel mechanistic basis for understanding how MPO-derived HOCl mediates insulin resistance in adipocytes.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
## Materials and methods
*In the **Materials and methods** section:*
## Materials
*In the **Materials** section:*
Mouse 3T3-L1 preadipocytes were obtained from American Type Culture Collection (Manassas, VA, USA). 3T3-L1 preadipocyte medium, 3T3-L1 adipocyte differentiation medium, and 3T3-L1 adipocyte maintenance medium were obtained from Zen-Bio, Inc. (Research Triangle Park, NC, USA). [1,2-3H] 2-deoxy-d-glucose was purchased from Perkin Elmer Life Sciences (Waltham, MA, USA). Hypochlorous sodium solution, human insulin, Nω-nitro-l-arginine methyl ester hydrochloride (l-NAME), SP600125, and antibody against phospho-IRS1 (Tyr612) were purchased from Sigma–Aldrich. Protein A/G-agarose, RIPA lysis buffer, PS-1145, IKKβ siRNA, JNK2 siRNA, PKCθ siRNA, control siRNA, and antibodies against β-actin, GAPDH, and Na+/K+ ATPase were obtained from Santa Cruz Biotechnology, Inc.
Antibodies against phospho-IRS1 (Ser307), IRS1, phospho-Akt (Ser473), phospho-Akt (Thr308), Akt, phospho-GSK3β (Ser9), GSK3β, phospho-IKKα/β (Ser176/180), IKKα, IKKβ, phospho-SAPK/JNK (Thr183/Tyr185), JNK, phospho-PKCθ (Thr538), PKCθ, GLUT4, IκBα, and HRP-linked secondary antibodies were purchased from Cell Signaling Technology (Beverly, MA, USA). ONOO− was obtained from Calbiochem (Billerica, MA, USA). All other chemicals were of the highest commercial grade available.[](https://www.ncbi.nlm.nih.gov/mesh/D003903)
## Cell treatment
*In the **Cell treatment** section:*
3T3-L1 preadipocytes were cultured and differentiated into adipocytes as described previously (Worrall & Olefsky 2002). By days 8–12 after induction of differentiation, more than 90% of the cells exhibited the characteristic lipid-vesicle-filled adipocyte phenotype. After overnight incubation in DMEM supplemented with 0.1% BSA, 3T3-L1 adipocytes were exposed to various concentrations of HOCl in the presence of 0.1% BSA for 1 h. HOCl-treated 3T3-L1 adipocytes were further stimulated with insulin for 15 min. HOCl was prepared by treating hypochlorous sodium with hydrochloric acid to adjust the pH to 4.0 according to a protocol published previously (Wang et al. 2007). The concentrations of HOCl were determined spectrophotometrically at 292 nm in 0.1 mol/l NaOH (ϵ=350/(mol/l) per cm).[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
## 2-Deoxyglucose (2-DG) uptake
*In the **2-Deoxyglucose (2-DG) uptake** section:*
The glucose uptake was assayed as described previously (Worrall & Olefsky 2002). Following overnight serum starvation, adipocytes were starved of glucose for 1 h in HEPES buffer containing 0.1% BSA and exposed to different concentrations of HOCl for 1 h. Then, cells were stimulated with 100 nM insulin for 15 min following addition of 0.2 μCi [1,2-3H] 2-DG for 5 min. Finally, cells were washed three times with ice-cold PBS buffer and solubilized with 1% Triton X-100.
3H-2-DG uptake was determined using a liquid scintillation counter. The intracellular concentration of 2-DG was normalized to total protein content.[](https://www.ncbi.nlm.nih.gov/mesh/D005947)
## Western blotting analysis
*In the **Western blotting analysis** section:*
Proteins were extracted from 3T3-L1 cells with RIPA lysis buffer (Santa Cruz Biotechnologies) containing 1 mM Na3VO4, 1 μg/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride. Protein concentration was measured using the BCA method (Pierce, Rockford, IL, USA). Cell lysates were resolved by SDS–PAGE and transferred to PVDF membrane (Millipore Corp., Bedford, MA, USA). Membranes were blocked with 5% milk and probed with specific antibodies and subsequently incubated with HRP-linked secondary antibodies. Proteins were visualized using an ECL detection system (Pierce).[](https://www.ncbi.nlm.nih.gov/mesh/D014638)
## siRNA transfection
*In the **siRNA transfection** section:*
IKKβ, JNK1, PKCθ, and control siRNA (10 μmol/l) were added to OPTI-MEM-reduced serum media (Life Technologies) with Lipofectamine RNAiMAX (Invitrogen Corp.). Adipocytes in six-well plates were transfected with siRNA in transfection medium for 6 h. The transfection medium was then replaced with a culture medium containing 10% FBS and incubated for 48 h.[](https://www.ncbi.nlm.nih.gov/mesh/C086724)
## Assay of in vitro Akt kinase activity
*In the **Assay of in vitro Akt kinase activity** section:*
A total of 500 μg protein was incubated with anti-Akt antibody and Sepharose beads overnight at 4 °C. After binding, the beads were washed four times with lysis buffer. Akt activity was measured using a non-radioactive Akt kinase assay kit (Cell Signaling Technology, Inc., Beverly, MA, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D012685)
## Experimental animals
*In the **Experimental animals** section:*
MPO knockout (Mpo − / −) mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). C57BL/6J mice were used as WT controls. Mice were housed in temperature-controlled cages under a 12 h light:12 h darkness cycle. Starting at 6 weeks of age, male mice were fed a high-fat diet (HFD, D12492; Research Diets, New Brunswick, NJ, USA) consisting of 60% fat, 20% protein, and 20% carbohydrate or a normal chow diet (ND) consisting of 13% fat, 29% protein, and 58% carbohydrates (LabDiet, St Louis, MO, USA) for up to 16 weeks. At the end of the experiments, mice were fasted for 6 h. Body weight and food intake were measured.
Blood samples were collected for detecting fasting glucose and plasma insulin (ALPCO, Salem, NH, USA). Epididymal white adipose tissue (WAT) was collected and stored at −80 °C. The animal protocol was reviewed and approved by the University of Oklahoma Institutional Animal Care and Use Committee.[](https://www.ncbi.nlm.nih.gov/mesh/D004041)
## Immunohistochemical analysis
*In the **Immunohistochemical analysis** section:*
Epididymal WAT from HFD-fed WT and Mpo − / − mice were fixed in 4% paraformaldehyde for 16 h and embedded in paraffin. Sections were deparaffinized, rehydrated, and microwaved in citrate buffer for antigen retrieval. Sections were successively incubated in 3% hydrogen peroxide, protein block buffer, and primary antibody against HOCl-oxidized LDL (Millipore Corp.) antibody against 3-chlorotyrosine (Hycult Biotech. PA, USA) overnight at 4 °C. Then, sections were rinsed in PBS buffer and incubated with labeled polymer-HRP anti-rabbit antibody and DAB chromogen.[](https://www.ncbi.nlm.nih.gov/mesh/C003043)
## Statistical analysis
*In the **Statistical analysis** section:*
Values are expressed as mean±s.e.m. One-way ANOVA was used to compare the differences among the three groups followed by Bonferroni's multiple comparison tests as applicable, and P<0.05 was considered significant.
## Results
*In the **Results** section:*
## HOCl impairs insulin-stimulated glucose uptake, insulin signals, and GLUT4 translocation
*In the **HOCl impairs insulin-stimulated glucose uptake, insulin signals, and GLUT4 translocation** section:*
We first investigated the effect of HOCl on insulin-stimulated glucose uptake measurement with 2-deoxyglucose (2-DG) labeled with tritium in 3T3-L1 adipocytes. Results of previous studies by our group and other laboratories (Ginion et al. 2011, Liu et al. 2013) on time–action and dose–response curves for insulin-stimulated glucose uptake revealed that stimulation with 100 nmol/l insulin for 15 min could induce maximal insulin action, this dose was used for the entire study to induce glucose uptake and transduction of transduction of insulin signaling in 3T3-L1 adipocytes. HOCl dose-dependently decreased insulin-stimulated glucose uptake but not the basal rate of glucose uptake, indicating an insulin-resistant state. Treatment with HOCl at a pathologically relevant oxidant concentration (200 μmol/l; Hawkins et al. 2001) resulted in approximately 70% suppression of uptake of glucose (Fig. 1A). Therefore, 200 μmol/l HOCl was used to treat adipocytes in the subsequent experiments.
Insulin could induce translocation of GLUT4 to the cellular membrane, which increases the uptake of glucose into the cells. Treatment with HOCl inhibited insulin-stimulated translocation of GLUT4 to the plasma membrane, while not affecting the expression of GLUT4 in the whole-cell lysates (Fig. 1B).[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
Next, we measured the molecular targets of HOCl-induced changes in glucose uptake. Insulin treatment stimulated phosphorylation of IRS1 at Tyr612 (Fig. 2A), Akt, and its downstream substrate GSK3β. Pretreatment of adipocytes with HOCl inhibited insulin-induced phosphorylation of Akt and GSK3β (Fig. 1C). In parallel, treatment with HOCl suppressed the activity of Akt kinase as evaluated by phosphorylation of GSK3β fusion protein (Fig. 1D).
These results indicated that treatment with HOCl induced insulin resistance in adipocytes.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
## HOCl promotes phosphorylation of IRS1 at Ser307, IKK, and JNK
*In the **HOCl promotes phosphorylation of IRS1 at Ser307, IKK, and JNK** section:*
Results from recent studies indicated that serine phosphorylation of IRS1, mediated by JNK and IKK, was associated with inhibition of the insulin signaling pathway by inflammatory cytokines (Aguirre et al. 2000, Morino et al. 2005, Weigert et al. 2008). Thus, we examined the effect of HOCl on serine phosphorylation of IRS1 at residue 307. As shown in Fig. 2A, HOCl treatment dramatically increased phosphorylation of IRS1 at Ser307 in the presence or absence of insulin, but not the expression of IRS1. Meanwhile, insulin-stimulated tyrosine phosphorylation of IRS1 was suppressed after pretreatment with HOCl.
In parallel, treatment with HOCl dramatically increased phosphorylation of IKKα/β at Ser176/180 and JNK at Thr183/Tyr185, but not protein expression (Fig. 2B and C).[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
## Phosphorylation of IKKβ is involved in HOCl-induced insulin resistance
*In the **Phosphorylation of IKKβ is involved in HOCl-induced insulin resistance** section:*
Next, we investigated whether IKKα/β was required for HOCl-triggered phosphorylation of IRS1-Ser307 and insulin resistance. To this end, the IKK-selective inhibitor PS-1145 (10 μmol/l) was used to pretreat adipocytes before the addition of HOCl. PS-1145 suppressed HOCl-induced phosphorylation of IKKα/β and reduced serine phosphorylation of IRS1 at 307. Furthermore, PS-1145 significantly abrogated HOCl-impaired insulin signals, including tyrosine phosphorylation of IRS1 and phosphorylation of Akt and GSK3β (Fig. 3A). To confirm this result, IKKβ-specific siRNA was used to suppress the expression of IKK and then HOCl-impaired insulin signals were evaluated.
IKKβ siRNA significantly decreased the expression of IKKβ and HOCl-induced phosphorylation of IKKβ. Furthermore, IKKβ siRNA attenuated serine phosphorylation of IRS1 at position 307 and restored phosphorylation of IRS1 at Tyr612, phosphorylation of Akt and GSK3β by insulin (Fig. 3B). These data provide strong evidence that IKKβ is required for serine phosphorylation of IRS1 at position 307 and impairment of insulin signaling after treatment with HOCl.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
## Phosphorylation of JNK is required for HOCl-induced insulin resistance
*In the **Phosphorylation of JNK is required for HOCl-induced insulin resistance** section:*
We also determined whether phosphorylation of JNK mediated phosphorylation of IRS1-Ser307 and insulin resistance after stimulation with HOCl. As expected, the JNK-specific inhibitor SP-600125 (30 μmol/l) markedly suppressed HOCl-induced phosphorylation of JNK phosphorylation accompanied by serine phosphorylation of IRS1 at 307 residues. In parallel, SP-600125 significantly abrogated the HOCl-induced impairment of insulin signals, as demonstrated by the restoration of phosphorylation of IRS1 at Tyr612, Akt, and GSK3β by insulin (Fig. 4A).[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
Next, we assayed whether genetic suppression of JNK altered the HOCl-impaired insulin signaling pathway. As shown in Fig. 4B, transfection with JNK siRNA significantly decreased the expression and HOCl-induced phosphorylation of JNK. Also HOCl-induced phosphorylation of IRS1 at Ser307 was dramatically decreased in cells transfected with JNK, compared with control. Moreover, JNK siRNA abrogated HOCl-impaired insulin signaling, as demonstrated by restoration of insulin-stimulated tyrosine phosphorylation of IRS1 and phosphorylation of Akt and GSK3β.
These results indicate that phosphorylation of JNK is required for insulin resistance and phosphorylation of IRS1 at Ser307 in 3T3-L1 adipocytes treatment with HOCl.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
## PKCθ mediates HOCl-induced insulin resistance via activation of IKK and JNK
*In the **PKCθ mediates HOCl-induced insulin resistance via activation of IKK and JNK** section:*
PKCθ could activate IKKβ and JNK, leading to phosphorylation of IRS1 at Ser307 and Ser302 (Werner et al. 2004). Thus, we verified whether activation of PKCθ contributes to HOCl-induced phosphorylation of IKK and JNK and insulin resistance. Exposure of 3T3-L1 adipocytes to HOCl induces phosphorylation of PKCθ at Thr538. Also knockdown of PKCθ by siRNA transfection partly attenuated HOCl-induced phosphorylation of IKK and JNK, indicating that HOCl activates IKK and JNK in a PKCθ-dependent manner (Fig. 5A).
In parallel, transfection with PKCθ siRNA dramatically decreased HOCl-induced serine phosphorylation of IRS1 at residue 307 and restored insulin-stimulated tyrosine phosphorylation of IRS1 (Fig. 5B). Moreover, PKCθ siRNA partially recovered insulin-enhanced Akt kinase activity (Fig. 5C) and translocation of GLUT4 to the plasma membrane (Fig. 5D). Taken together, these results indicate that PKCθ phosphorylation is involved in HOCl-induced insulin resistance via activation of IKK and JNK in 3T3-L1 adipocytes.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
## ONOO− mediates phosphorylation of PKCθ and IRS1-Ser307 after HOCl treatment
*In the **ONOO− mediates phosphorylation of PKCθ and IRS1-Ser307 after HOCl treatment** section:*
Our group has demonstrated that HOCl enhanced ONOO− production, a potent oxidative molecular formed by superoxide (O2·−) and nitric oxide (NO) (Wang et al. 2014). Also ONOO− plays a causal role in the pathogenesis of insulin resistance in obesity and type 2 diabetes (Randriamboavonjy & Fleming 2009). To determine whether ONOO− was involved in HOCl-induced activation of inflammatory kinases and insulin resistance, 3T3-L1 adipocytes were preincubated with l-NAME (1 mmol/l) to inhibit the production of NO, Cu/Zn SOD (SOD1, 150 U/ml) to remove O2·−, or uric acid (50 μmol/l) to scavenge ONOO− before stimulation with HOCl, and then phosphorylation of PKCθ, IKK, JNK, and IRS1 was evaluated. As shown in Fig.
6A, HOCl-induced phosphorylation of PKCθ was blocked by SOD1, l-NAME, and uric acid treatment. In addition, ONOO− dose-dependently induced phosphorylation of PKCθ, but not expression of the protein (Fig. 6B), indicating that ONOO− contributes to the phosphorylation of PKCθ by HOCl.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
We further confirmed the role of ONOO− in PKCθ-mediated downstream signals, including serine phosphorylation of IKK, JNK, and IRS1. As shown in Fig. 6C, SOD1, l-NAME, and uric acid suppressed phosphorylation of IKK, JNK, and IRS1-Ser307 in adipocytes treated with HOCl. In addition, ONOO− dose-dependently induced phosphorylation of IKK and JNK. Meanwhile, ONOO− increased serine phosphorylation of IRS1 and reduced tyrosine phosphorylation of IRS1 by insulin (Fig.
6D). Collectively, these results indicate that ONOO− mediates HOCl-induced Ser307 phosphorylation of IRS1 via activation of PKCθ, IKK, and JNK.[](https://www.ncbi.nlm.nih.gov/mesh/D030421)
## HOCl-induced insulin resistance is ONOO− dependent
*In the **HOCl-induced insulin resistance is ONOO− dependent** section:*
Pretreatment with SOD1, l-NAME, and uric acid prevented HOCl-induced impairment of insulin signals, as demonstrated by restoration of tyrosine phosphorylation of IRS1, phosphorylation of Akt, and Akt kinase activity by insulin (Fig. 7A, B and C). Moreover, SOD1, l-NAME, and uric acid abolished the deleterious effect of HOCl on translocation of GLUT4 to the plasma membrane (Fig. 7D). In contrast, the hydrogen peroxidase scavenger catalase offered no protection against HOCl-induced impairment of insulin signaling. Taken together, these results indicate that ONOO− is involved in HOCl-induced insulin resistance.[](https://www.ncbi.nlm.nih.gov/mesh/D019331)
## Deletion of MPO attenuates HFD-induced phosphorylation of IKK, JNK, and IRS1-Ser307
*In the **Deletion of MPO attenuates HFD-induced phosphorylation of IKK, JNK, and IRS1-Ser307** section:*
MPO is the sole mammalian oxidant enzyme to catalyze production of HOCl in the presence of hydrogen peroxide and chloride ions (Olza et al. 2012). To extend our in vitro findings, we measured the insulin signals and molecular pathway involved in insulin resistance in WT and MPO knockout (Mpo − / −) mice. The initial body weights of WT and Mpo − / − male mice at 6 weeks of age were similar. After 16 weeks of HFD feeding, WT mice displayed higher body weights, fasting blood glucose, and plasma insulin levels than Mpo − / − mice. In parallel, high level of homeostasis model assessment of insulin resistance (HOMA-IR) indicated that WT mice developed severe insulin resistance than Mpo − / − mice (Table 1).
In epididymal WAT of WT mice fed with HFD, infiltration of neutrophils (Wang et al. 2014), and expression of MPO were increased compared with ND-fed WT mice (Fig. 8A). In addition, 3-chlorotyrosine, a biomarker for HOCl, was present in the WAT from WT mice, mainly in the crown-like structure that contains neutrophils and macrophages. This 3-chlorotyrosine stain was absent in Mpo − / − mice (Fig. 8B). In parallel, HFD increased phosphorylation of IRS1 at Ser307, PKCθ, JNK, and IKK in WAT, while not affecting protein expression. However, the effect of HFD on phosphorylation of IRS1-Ser307, PKCθ, JNK, and IKK was absent in Mpo − / − mice (Fig.
8C and D).[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
## Discussion
*In the **Discussion** section:*
This study has unveiled the mechanism whereby HOCl induces insulin resistance in adipocytes. We showed that a clinically relevant concentration of HOCl impairs insulin-stimulated glucose uptake, reduces the amplitude of the insulin signal, and enhances phosphorylation of IRS1 at serine 307. Most importantly, HOCl induces PKCθ-dependent activation of IKK/JNK by ONOO−, causing serine phosphorylation of IRS1 and insulin resistance (Fig. 8E).[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
The key lines of evidence can be summarized as follows: first, HOCl induces phosphorylation of IKK and JNK, whereas inhibition of IKK or JNK blocks serine phosphorylation of IRS1 and impairment of insulin signals. These results indicate that activation of both IKK and JNK by HOCl is required for insulin resistance; secondly, HOCl induces phosphorylation of PKCθ, and suppression of PKCθ attenuates phosphorylation of IKK/JNK and restores insulin-stimulated glucose uptake, implying that PKCθ serves as an upstream kinase of IKK/JNK; thirdly, ONOO− treatment dose dependently induces phosphorylation of PKCθ, whereas an ONOO− scavenger reduces HOCl-stimulated phosphorylation of PKCθ, IKK, JNK, and IRS1-Ser307. These results indicate that ONOO− is the initial trigger for the development of insulin resistance in response to HOCl. Finally, deletion of MPO protects against HFD-induced phosphorylation of PKCθ, IKK, JNK, and IRS1-Ser307 accompanied by insulin resistance in WAT. Overall, our results indicate that ONOO− mediated PKCθ-dependent serine phosphorylation of IRS1 might be responsible for HOCl-induced insulin resistance.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
IRS1 is an essential insulin signal transducer for activating downstream signals after tyrosine phosphorylation. But serine/threonine phosphorylation of IRS1 exerts the opposite regulatory effect on insulin signaling via accelerated degradation and interference with tyrosine phosphorylation. Several inflammatory cytokines, such as tumor necrosis factor alpha (TNFα) and interleukin 1 beta (IL1β), induce serine phosphorylation of IRS1 at position 307 and insulin resistance. Interestingly, mutation of serine 307 to alanine (IRS1 S307A) causing dephosphorylation at this site eliminates TNFα-decreased tyrosine phosphorylation of IRS1 in Chinese hamster ovary cells, underlining the important inhibitory function of this site in transduction of insulin signaling (Aguirre et al. 2000). Weigert et al. (2008) confirmed that IRS1 S307A mutation enhanced phosphorylation of Akt in skeletal muscle cells.
Consistently, muscle-specific mutations at IRS1-Ser302, 307, and 612 increased uptake of glucose and insulin signaling (Morino et al. 2008). Thus, we proposed the hypothesis that inhibition of tyrosine phosphorylation of IRS1 and impairment of the insulin signaling pathway might be due to serine phosphorylation of IRS1 by HOCl. Indeed, we found that treatment with HOCl increased serine phosphorylation of IRS1 at position 307 and reduced insulin-stimulated tyrosine phosphorylation of IRS1 in 3T3-L1 adipocytes. Interestingly, these opposite effects of HOCl on serine/tyrosine phosphorylation of IRS1 were simultaneously reversed after blockage of activation of serine/threonine protein kinases, including IKK, JNK, and PKCθ. These results are indicative of a significant negative association between serine phosphorylation at residue 307 and tyrosine phosphorylation of IRS1 in adipocytes.
Consistent with the conclusion, serine phosphorylation of IRS1 also strongly correlates with impairment of the insulin signaling pathway in liver and muscle of diabetic mice and humans (Morino et al. 2005, Dong et al. 2008). Recently, this finding has been challenged by Copps and colleagues who global IRS1 S307A knock-in mice and observed insulin resistance in 5-month-old mice fed with a chow diet, indicating that IRS1-Ser307 makes a positive contribution to maintaining good insulin sensitivity (Copps et al. 2010). However, direct evidence is lacking to support the hypothesis that phosphorylation of IRS1-Ser307 causes inhibition of tyrosine phosphorylation and insulin resistance in adipocytes treated with HOCl. Further investigation is warranted.[](https://www.ncbi.nlm.nih.gov/mesh/D014443)
The inflammatory kinases IKKβ and JNK serve as critical molecular links between obesity, metabolic inflammation, and insulin resistance. Activation of IKK and JNK could phosphorylate IRS1 on a number of serine residues, resulting in inhibition of insulin signaling. IL1β activates JNK1 and subsequently phosphorylates IRS1 at Ser307 in adipocyte (He et al. 2006). Another cytokine, TNFα, also IKK-dependently increases IRS1-Ser307 phosphorylation in 3T3-L1 adipocytes (Nakamori et al. 2006). Treatment with HOCl induces phosphorylation of JNK in primary biliary cells (Salunga et al. 2007) and activates NF-κB in endothelial cells (Pullar et al. 2002).
Therefore, we propose the hypothesis that HOCl induces serine phosphorylation of IRS1 via activation of IKK and JNK. Indeed, exposure of 3T3-L1 adipocytes to HOCl significantly increases phosphorylation of JNK and IKK. In contrast, pharmaceutical inhibitors and siRNA-mediated knockdown of JNK and IKKβ prevent impairment of insulin signaling and phosphorylation of IRS1-Ser307, indicating that HOCl-mediated insulin resistance is dependent on activation of JNK and IKK. This conclusion is further supported by results from our in vivo studies indicating that knockout of MPO reduces phosphorylation of IKK, JNK, and IRS1-Ser307 in WAT, and in parallel protects against insulin resistance in HFD-fed obese mice.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
PKCθ has been reported to induce phosphorylate of IRS1 on Ser307 and Ser302 via activation of IKKβ and JNK (Gao et al. 2004). Free fatty acid is an inducer of PKCθ phosphorylation, causing to cause development of insulin resistance in adipocytes (Gao et al. 2004) and skeletal muscle cells (Kadotani et al. 2009). In this study, we have demonstrated HOCl to be a novel mediator of activation of PKCθ in adipocytes, which might contribute to adipose inflammation and insulin resistance. Treatment with HOCl induced phosphorylation of PKCθ in 3T3-L1 adipocytes. Moreover, knockdown of PKCθ using siRNA transfection attenuated phosphorylation of IKK, JNK, and IRS1-Ser307 and restored impairment of the insulin signaling pathway by HOCl.
These results indicate that PKCθ functions upstream of IKK and JNK to induce insulin resistance. It is noteworthy that knockdown of PKCθ could not fully inhibit HOCl-induced activation of IKK and JNK, indicating that treatment with HOCl may activate IKK and JNK in other ways independent of PKCθ. Besides PKCθ, PKCζ is also involved in the development of insulin resistance (Lee et al. 2010). It has also been reported that HOCl could induce phosphorylation of PKCζ, causing activation of NADPH oxidase in endothelial cells (Xu et al. 2006). Whether other PKC isoforms are involved in these processes requires additional investigation.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
We have recently reported that exogenous HOCl treatment increased ONOO− production in 3T3-L1 adipocytes and endothelial cells (Xu et al. 2006, Wang et al. 2014). ONOO− plays a critical role in the pathogenesis of insulin resistance through multiple pathways. For instance, ONOO− induces tyrosine nitration of insulin signaling proteins, including insulin receptor β and IRS1, leading to inactivation and degradation in adipocytes (Nomiyama et al. 2004). ONOO− induces S-glutathionylation of p21ras and serine phosphorylation of IRS1 in endothelial cells as well (Clavreul et al. 2006). In this study, we describe a novel signal transduction mechanism by which HOCl-mediated insulin resistance is ONOO− dependent.
The ONOO− scavenger uric acid offers considerable protection against HOCl-induced phosphorylation of PKCθ and IRS1-Ser307. Because ONOO− is formed by the rapid reaction of NO with O2·−, O2·− and NO inhibitors show similar protective effects on phosphorylation of inflammatory kinases. On the other hand, treatment with ONOO− directly induces phosphorylation of PKCθ, leading to a reduction of tyrosine phosphorylation of IRS1 by insulin. These observations indicate that treatment with ONOO− is essential for activation of inflammatory kinases, which triggers insulin resistance after stimulation with HOCl.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
In conclusion, the current findings strongly indicate that HOCl is a novel contributor to the development of insulin resistance in adipocytes, and a clinically relevant concentration of HOCl induces production of ONOO− and activation of inflammatory kinases, resulting in impairment of the insulin signaling pathway. HOCl-induced insulin resistance might represent a common pathological pathway in the development of the metabolic syndrome and type 2 diabetes.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
## Author contribution statement
*In the **Author contribution statement** section:*
J Zhou and Q Wang contributed equally to the study design, performed experiments, and wrote the manuscript. Y Ding performed some experiments. M-H Zou contributed to the study design and interpretation and wrote the manuscript. M-H Zou is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
(J Zhou and Q Wang contributed equally to this work)
## Declaration of interest
*In the **Declaration of interest** section:*
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
## Funding
*In the **Funding** section:*
This work was supported by National Institutes of Health grants (HL079584, HL080499, HL074399, HL089920, HL096032, HL10488, HL105157, and AG047776), a research award from the American Diabetes Association, and funds from the Warren Chair in Diabetic Research from the University of Oklahoma Health Sciences Center. Dr M-H Zou is a recipient of the National Established Investigator Award of the American Heart Association. Q Wang is a recipient of an American Heart Association Postdoctoral fellowship.
# References
*In the **References** section:*
HOCl induces insulin resistance in 3T3-L1 adipocytes. (A) HOCl dose dependently inhibits insulin-stimulated glucose uptake. Serum-starved 3T3-L1 adipocytes were pretreated with various concentrations of HOCl (0, 50, 100, 200, 500, and 1000 μmol/l) for 1 h and stimulated with 100 nmol/l insulin or left unstimulated for 15 min and then incubated with [3H]-2-deoxyglucose (2DG) for 5 min. Glucose uptake was evaluated from [3H]-2-DG counts in each cell lysate. Results are expressed as mean±s.e.m. (n=5, *P<0.05 and **P<0.01 vs no HOCl treatment). (B, C and D) HOCl impairs insulin signaling pathway. 3T3-L1 adipocytes were pretreated with 200 μmol/l HOCl for 1 h and stimulated with or without 100 nmol/l insulin for 15 min. Membrane proteins were isolated using the Mammalian Membrane Protein Extraction Kit (Abcam, Cambridge, MA, USA). Cell lysis was subjected to SDS–PAGE and immunoblotted with indicated antibodies (B and C). Cell lysates were immunoprecipitated using Akt antibody and immunoblotted with phospho-GSK3β for in vitro Akt kinase assay using non-radioactive Akt kinase assay kit (Cell Signaling Technology, Inc., Beverly, MA, USA) (D). The blot is a representative of results obtained from five independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
HOCl promotes phosphorylation of IRS1 at Ser307 in association with JNK and IKKα/β. 3T3-L1 adipocytes were pretreated with 200 μmol/l HOCl for 1 h before treatment with 100 nmol/l insulin for 15 min or being left untreated, and levels of expression of phospho-IRS1-Ser307 and phospho-IRS-Tyr612 (A), IKK (B), and JNK (C) were determined by western blot analysis. The blot is representative of results obtained from five independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
Inhibition of IKK blocks HOCl-induced serine phosphorylation of IRS1 and insulin resistance. (A) 3T3-L1 adipocytes were pretreated with 10 μmol/l PS1145 for 90 min and incubated with 200 μmol/l HOCl for 1 h and then stimulated with 100 nmol/l insulin or left unstimulated for 15 min. (B) 3T3-L1 adipocytes were transfected with IKKβ siRNA or control siRNA for 48 h and treated with 200 μmol/l HOCl for 1 h and then stimulated with or without 100 nmol/l insulin for 15 min. Western analysis of protein expression and phosphorylation of IKK, IRS1, Akt, and GSK3β was performed. Blots are representative of the results from five independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
Inhibition of JNK prevents HOCl-induced phosphorylation of IRS1 at Ser307 and insulin resistance. (A) 3T3-L1 adipocytes were pretreated with 30 μmol/l SP600125 for 90 min and incubated with 200 μmol/l HOCl for 1 h and then stimulated with 100 nmol/l insulin or left unstimulated for 15 min. (B) 3T3-L1 adipocytes were transfected with JNK2 siRNA or control siRNA for 48 h and treated with 200 μmol/l HOCl for 1 h, and then stimulated with 100 nmol/l insulin or left unstimulated for 15 min. Western analysis of protein expression and phosphorylation of JNK, IRS1, Akt, and GSK3β was performed. Blots are representative of the results from five independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
PKCθ mediates phosphorylation of IKK, JNK, and IRS1 and insulin resistance. 3T3-L1 adipocytes were transfected with PKCθ siRNA or control siRNA for 48 h and treated with 200 μmol/l HOCl for 1 h and subsequently stimulated with 100 nmol/l insulin for 15 min or left unstimulated. Western blot analysis of protein expression and phosphorylation of PKCθ, IKK, JNK, and IRS1 (A and B). Akt kinase activity in cell lysates was measured (C). The membrane fraction was isolated and proteins were subjected to SDS–PAGE and immunoblotted with GTLU4 antibody (D). The blots are representative of results obtained from five independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
HOCl mediates inflammatory kinase phosphorylation through ONOO−. (A and C) 3T3-L1 adipocytes were serum starved overnight, then treated with 200 μmol/l HOCl for 1 h in the presence of SOD1 (150 U/ml), l-NAME (1 mmol/l), or uric acid (UA, 50 μmol/l), followed by stimulation with 100 nmol/l insulin for 15 min. Cell extracts were immunoblotted with indicated antibodies. (B and D) 3T3-L1 adipocytes were serum starved overnight, then treated with ONOO− (50, 100, and 200 μmol/l) for 1 h. Western blot analysis of protein expression and phosphorylation of PKCθ, IKK, JNK, and IRS1 was performed. The blot is a representative of blots obtained from five independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
HOCl-induced insulin resistance is ONOO− dependent. 3T3-L1 adipocytes were serum starved overnight and treated with 200 μmol/l HOCl for 1 h in the presence of SOD1 (150 U/ml), catalase (100 U/ml), l-NAME (1 mmol/l), or uric acid (50 μmol/l), then stimulated with 100 nmol/l insulin for 15 min or left unstimulated. Cell extracts were immunoblotted with antibodies against IRS1 and Akt (A and B). Akt kinase activity in cell lysis was measured (C). Membrane proteins were isolated and then subjected to SDS–PAGE and immunoblotted with GTLU4 antibody (D). Blots are representative of results obtained from five independent experiments.[](https://www.ncbi.nlm.nih.gov/mesh/D006997)
MPO deficiency prevents phosphorylation of IKK, JNK, and IRS1-Ser307 in WAT from HFD-fed mice. (A, C and D) WT and Mpo − / − mice were fed with ND or HFD for 16 weeks. Homogenates of epididymal WAT were prepared and levels of MPO, phosphorylation of IRS1, PKCθ, IKK, and JNK were analyzed by western blotting. (B) Expression of 3-chlorotyrosine in WAT was analyzed by using immunohistochemistry, magnification: 20×. (E) Schematic diagram of the relationship between for HOCl-induced phosphorylation of IRS at Ser307 and insulin resistance.[](https://www.ncbi.nlm.nih.gov/mesh/D012694)
Body weight and metabolic characteristics in WT mice and Mpo − / − mice after 16 weeks of HFD. Values are mean±s.e.m., n=7 for each group. Six-week-old male mice were fed a high-fat diet for 16 weeks. Blood was collected after 6 h of fasting and metabolic parameters were measured from serum.[](https://www.ncbi.nlm.nih.gov/mesh/D004041)
*P<0.05.
|
# Introduction
[Dexamethasone](https://www.ncbi.nlm.nih.gov/mesh/D003907) Improves Heat Stroke-Induced Multiorgan Dysfunction and Damage in Rats
# Abstract
*In the **Abstract** section:*
Dexamethasone (DXM) is known as an immunosuppressive drug used for inflammation contr[ol. In the pr](https://www.ncbi.nlm.nih.gov/mesh/D003907)es[ent](https://www.ncbi.nlm.nih.gov/mesh/D003907) study, we attempted to examine whether DXM administration could attenuate the hypercoagulable state and the overproduction o[f p](https://www.ncbi.nlm.nih.gov/mesh/D003907)ro-inflammatory cytokines, improve arterial hypotension, cerebral ischemia and damage, and vital organ failure in a rat model of heat stroke. The results indicated that all the rats suffering from heat stroke showed high serum levels of tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β), accompanied with increased prothrombin time, activated partial thromboplastin time and D-D dimer, and decreased protein C. During the induction period of heat stroke, plasma levels of blood urea nitrogen (BUN), creatinine, glutamic oxaloacetic transaminase (SGOT), glutamic p[yruv](https://www.ncbi.nlm.nih.gov/mesh/D014508)i[c transa](https://www.ncbi.nlm.nih.gov/mesh/D009584)minase ([SGPT), and](https://www.ncbi.nlm.nih.gov/mesh/D003404) alkaline phosphatase (ALP), were consistently increased. High striatal levels of glycerol, glutamate, and lactate/pyruvate were simultaneously detected. On the contra[ry, the ](https://www.ncbi.nlm.nih.gov/mesh/D005990)me[an arteri](https://www.ncbi.nlm.nih.gov/mesh/D018698)al pre[ssure, ](https://www.ncbi.nlm.nih.gov/mesh/D019344)p[lasma le](https://www.ncbi.nlm.nih.gov/mesh/D019289)vels of interleukin-10 (IL-10), and local cerebral blood flow at the striatum were all decreased. Importantly, intravenous administration of DXM substantially ameliorated the circulatory dysfunction, systematic inflammation, h[ype](https://www.ncbi.nlm.nih.gov/mesh/D003907)rcoagulable state, cerebral ischemia and damage during the induction period of heat stroke. These findings demonstrated that DXM may be an alternative therapy that can ameliorate heat stroke victims by attenuat[ing](https://www.ncbi.nlm.nih.gov/mesh/D003907) activated coagulation, systemic inflammation, and vital organ ischemia/injury during heat stroke.
## 1. Introduction
*In the **1. Introduction** section:*
A clinical diagnosis of heat stroke suggests that body hyperthermia (over 42 °C) associated with a systemic inflammatory response leads to multiple organ dysfunction, in particular, neurological abnormalities after exposure to high temperature. Several lines of evidence indicate that animals share with humans almost the same heat stroke syndromes. In rodents, heat stress leads to arterial hypotension, hyperpyrexia, and hypercoagulable state, and excessive activated inflammation may contribute to multiple organ failure (including cerebral, hepatic and renal ischemia, injury, and dysfunction) in heat stroke.
Steroidal anti-inflammatory drugs have been shown to decrease the generation of leukotrienes and prostaglandins by inhibiting the secretion of phospholipase A2 and the release of arachidonic acid. Glucocorticoids (GCs) are known to be potent inhibitors of cytokine production and to exert a protective effect against lipopolysaccharide-induced death. In addition, GCs has been shown to be of benefit in the treatment of human and animal-spinal cord injury or cerebral ischemia. Our previous results revealed that systemic pretreatment with exogenous GCs, such as DXM, before exposure to heat stress, but not immediate treatment at onset of heat stroke, could increase the survival time via reduction of serum interleukin-1β (IL-1β) in rat heat stroke. However, there are fewer studies showing the immediate treatment with DXM at the onset of heat stroke, and it will be more meaningful if survival prolongation after heat stroke attacks is demonstrated.
After all, it is not practical to give pretreatment in clinical practice. Meanwhile, there is less attention to evaluate effects of DXM on heat stroke-induced pathophysiological changes, especially for the hypercoagulable state, various cytokines levels, and multiple organ dysfunction. The objective of this study was to observe firstly whether immediate treatment with different doses of DXM has efficacy to elongate survival time, and improve heat stroke-induced circulatory shock, cerebral ischemia and damage in rats. Furthermore, we also investigated whether the ameliorative effects of acute treatment with DXM were associated with inhibition of the hypercoagulable state, multiple organ dysfunction, and changes of systemic cytokines levels after heat stroke induction.[](https://www.ncbi.nlm.nih.gov/mesh/D015289)
## 2. Results and Discussion
*In the **2. Results and Discussion** section:*
## 2.1. DXM Improves Survival during Heat Stroke in a Dose-Dependent Manner
*In the **2.1. DXM Improves Survival during Heat Stroke in a Dose-Dependent Manner** section:*
We see from Table 1 that in anesthetized rats treated with normal saline (0.9% NaCl solution) 70 min after the onset of heat exposure (Ta = 43 °C) followed by room temperature (Ta = 24 °C) exposure, the value for survival time is found to be 24 ± 3 min (n = 8). However, immediate treatment with DXM 4, 6, and 8 mg/kg b.w. (i.v.) at the onset of heat stroke increases the survival time in a dose-dependent manner to a new value of 104 ± 9, 204 ± 25, and 268 ± 27 min, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
Values are the means ± S.E. of 8 rats per group. Groups 2–5 exposed to 43 °C had heat exposure withdrawn at the onset of heatstroke. * p < 0.05 in comparison with group 2; † p < 0.05 in comparison with group 3; ‡ p < 0.05 in comparison with group 4. (one-way ANOVA, followed by Duncan’s test). Group 1 was killed about 480 min at the end of the experiments with an overdose of urethane.[](https://www.ncbi.nlm.nih.gov/mesh/D014520)
Effects of heat exposure (HE; ambient temperature of Ta = 43 °C for 70 min) on survival time values in different groups of rats.
## 2.2. DXM Ameliorates Arterial Hypotension, Cerebral Ischemia and Damage during Heat Stroke
*In the **2.2. DXM Ameliorates Arterial Hypotension, Cerebral Ischemia and Damage during Heat Stroke** section:*
As shown in Figure 1, fifteen minutes after the onset of heat stroke in normal saline (0.9% NaCl solution) treated group, all the values of mean arterial pressure (MAP) and cerebral blood flow (CBF) were significantly decreased as compared with those of normothermic controls. On the other hand, the values of extracellular concentrations of glutamate, glycerol, and lactate/pyruvate ratio in corpus striatum were significantly greater than those of the normothermic controls. Treatment with an i.v. dose of DXM (8 mg/kg) 70 min after the start of heat exposure (or at the time of the onset of heat stroke) significantly attenuates the heat stroke-induced arterial hypotension, cerebral ischemia, and increased levels of glutamate, glycerol, and lactate/pyruvate ratio in corpus striatum.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## 2.3. DXM Attenuates Heat Stroke-Induced Hypercoagulable State
*In the **2.3. DXM Attenuates Heat Stroke-Induced Hypercoagulable State** section:*
Figure 2 summarizes the plasma levels of prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen degradation products (FDP), protein C, and D-D dimer for normothermic controls, normal saline-treated heat stroke rats, and DXM-treated heat stroke rats. It can be seen from Figure 2 that PT, aPTT, FDP, and D-D dimer values during heat stroke for rats treated with normal saline (1 mL/kg b.w., i.v.) are all significantly higher at 85 min after the start of heat exposure than those of the normothermic controls. On the contrary, the value for plasma of protein C is significantly lower than that of the normothermic controls. In turn, administration with DXM (8 mg/mL b.w., i.v.) at 70 min after initiation of heat exposure (or immediately at the onset of heat stroke) appreciably attenuates the heat stress-induced increased plasma levels of PT, aPTT, FDP, and D-D dimer as well as the decreased plasma levels of protein C.[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
Effects of heat exposure (43 °C) on colonic temperature (Tco), MAP, heart rate (HR), CBF and the extracellular concentrations of glutamate, glycerol, and lactate/pyruvate ratio of the corpus striatum in normothermic control rats (open circles), 0.9% NaCl solution-treated (filled circles, 8 mk/kg b.w., i.v.) or DXM-treated rats (open triangles). The dotted line indicates time of heat stroke onset and drug injection. *
p < 0.05, compared with normothermic control rats (ANOVA followed by Duncan’s test); †
p < 0.05, compared with saline-treated rats (ANOVA followed by Duncan’s test).[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
Effects of heat exposure (43 °C) on plasma levels of prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen degradation products (FDP), D-D dimer, and protein C in normothermic control rats (white bar), saline-treated (black bar) or dexamethasone-treated rats (grey bar, 8 mk/kg b.w., i.v.). *
p < 0.05, in comparison with normothermic control rats; †
p < 0.05, in comparison with saline-treated rats (ANOVA followed by Duncan’s test). The values were obtained 85 min after the initiation of heat exposure (or 15 min after the onset of heat stroke) in heat stroke rats or the equivalent time in normothermic controls. Bars are each the mean ± S.E. of 8 rats for each groups.[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
## 2.4. DXM Protects from Hepatic and Renal Dysfunction during Heat Stroke
*In the **2.4. DXM Protects from Hepatic and Renal Dysfunction during Heat Stroke** section:*
Plasma levels of blood urea nitrogen (BUN), creatinine, glutamic oxaloacetic transaminase (SGOT), glutamic pyruvic transaminase (SGPT), and alkaline phosphatase (ALP) for normothermic controls, normal saline-treated heat stroke rats, and DXM-treated heat stroke rats are summarized in Figure 3. It can be seen from the figure that the plasma levels of BUN, creatinine, SGOT, SGPT, and ALP for heat stroke rats treated with normal saline (1 mL/kg b.w., i.v.) are significantly higher at 85 min after the start of heat exposure than these are for normothermic controls. Acute treatment with DXM (8 mg/mL b.w., i.v.) at 70 min after initiation of heat exposure (or immediately at the onset of heat stroke) significantly attenuates the heat stress-induced increased plasma levels of BUN, creatinine, SGOT, SGPT, and ALP.[](https://www.ncbi.nlm.nih.gov/mesh/D014508)
Effects of heat exposure (43 °C) on plasma levels of BUN, creatinine, SGOT, SGPT, and ALP in normothermic control rats (white bar), saline-treated (black bar) or dexamethasone-treated rats (grey bar, 8 mk/kg b.w., i.v.). *
p < 0.05, in comparison with normothermic control rats; †
p < 0.05, in comparison with saline-treated rats (ANOVA followed by Duncan’s test). The values were obtained 85 min after the initiation of heat exposure (or 15 min after the onset of heat stroke) in heat stroke rats or the equivalent time in normothermic controls. Bars are each the mean ± S.E. of 8 rats for each groups.[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
## 2.5. DXM Reduces both IL-1β and Tumor Necrosis Factor-α (TNF-α) Increase but Enhances Interleukin-10 (IL-10) during Heat Stroke
*In the **2.5. DXM Reduces both IL-1β and Tumor Necrosis Factor-α (TNF-α) Increase but Enhances Interleukin-10 (IL-10) during Heat Stroke** section:*
The serum IL-1β and TNF-α, and IL-10 levels for normothermic controls, NS-, and DXM-treated heat stroke rats are summarized in Figure 4. It can be seen from the figure that the serum IL-1β and TNF-α levels in NS-treated heat stroke rats are all significantly higher at 85 min after the start of heat stress than those of the normothermic controls. The acute treatment with DXM (8 mg/mL b.w., i.v.) at 70 min after initiation of heat exposure (or immediately at the onset of heat stroke) significantly attenuates the heat stroke-induced increased serum levels of IL-1β and TNF-α. In NS-treated heat stroke rats, the serum level of IL-10 is maintained at an extremely low level. However, the serum of IL-10 is greatly elevated in heat stroke rats treated with an i.v. dose of DXM (8 mg/mL b.w.).[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
Effects of heat exposure (43 °C) on serum levels of IL-1β, TNF-α and IL-10 in normothermic control rats (white bar), saline-treated (black bar) or dexamethasone-treated rats (grey bar, 8 mk/kg b.w., i.v.). *
p < 0.05, in comparison with normothermic control rats; †
p < 0.05, in comparison with saline-treated rats (ANOVA followed by Duncan’s test). The values were obtained 85 min after the initiation of heat exposure (or 15 min after the onset of heat stroke) in heat stroke rats or the equivalent time in normothermic controls. Bars are each the mean ± S.E. of 8 rats for each groups.[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
## 2.6. Discussion
*In the **2.6. Discussion** section:*
During heat stroke, rodents display hyperthermia, arterial hypotension, intracranial hypertension, cerebral ischaemia, neuronal damage and overproduction of inflammatory cytokines. The present results, as well as our previous results revealed that all heat-stressed animals displayed systemic inflammation and activated coagulation, evidenced by increased TNF-α, IL-1β, PT, aPTT, and D-D dimer, and decreased IL-10 and protein C. Biochemical markers evidenced cellular ischemia and injury/dysfunction: plasma levels of BUN, creatinine, SGOT, SGPT, and ALP, and striatal levels of glycerol, glutamate, and lactate/pyruvate ratio, were all elevated during heat stroke. In contrast, the values of mean arterial pressure and striatal levels of local blood flow were all significantly lower during heat stroke. Our present results further demonstrate the circulatory dysfunction, systemic inflammation, hypercoagulable state, and cerebral ischemia and multiple organ dysfunction occurring during heat stroke. However, these heat stroke-induced pathophysiologic changes can all be significantly suppressed by acute treatment with DXM in a dose-dependent manner at the onset of heat stroke.[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
Our previous results indicated that pretreatment with DXM (4 mg/kg, i.v.) could attenuate the heat stroke-induced damage; however, acute treatment with DXM (4, 6, 8 mg/kg, i.v.) immediately at the onset of heat stroke can apparently increase the survival time in a dose-dependent manner in the present study. This has more meaningful application for emergency treatment in clinics for heat stroke survival. As mentioned above, it was found that rodents share with humans almost the same heat stroke reactions. These findings demonstrate that rodent heat stroke models can nearly mirror the full spectrum of human heat stroke.
Experimental heat stroke fulfills the empirical triad used for the diagnosis of classical human heat stroke. After onset of heat stroke, rats revealed ischemia and injury in several cerebral regions, especially in the corpus striatum. Both of our previous and present results displayed that the local CBF of striatum decreased significantly, but the values of striatal neuronal ischemic and injury index were sharply increased in rats with heat stroke (as shown in Figure 1). However, acute immediate treatment with DXM revealed appreciable decrements of neuronal damage, ischemic and hypoxic indexes in rats of heat stroke.[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
In addition, heat stroke rats showed hepatic failure (evidenced by levels of SGOT, SGPT and alkaline phosphatase) and renal failure (evidenced by increased plasma levels of BUN and creatinine) and hypercoagulable state. Kew et al. indicated that hepatic and renal failure may be related to tissue ischemia (due to circulatory shock) and thermal injury. As shown in the present results, heat stroke rats indeed displayed vital organ failure, and these pathophysiological changes were consistent with our previous studies. Nevertheless, immediate administration of DXM greatly diminished the severity of cerebral, hepatic and renal failure in rats of heat stroke. So far, there has been no study that tried to investigate the effects of glucocorticoids on blood coagulation state, and hepatic and renal functions in rats during heat stroke.
The present study has been focused on whether acute treatment with DXM can improve the heat stroke-induced hypercoagulable state and vital organ dysfunctions, and positive and meaningful findings were obtained.[](https://www.ncbi.nlm.nih.gov/mesh/D003404)
The serum concentrations of inflammatory cytokines including TNF-α and IL-1β are overproduced in human victims and in rodents with heat stroke. Evidence has demonstrated that expression of cytokines correlates well with the severity of heat stroke. The hypotension, intracranial hypertension and cerebral ischemia that occurred during heat stroke can be mimicked by intravenous administration of infusion of IL-1β, but prevented by prior antagonism of IL-1β receptors. Indeed, our previous studies have also shown that heat stroke induces systemic overproduction of TNF-α and IL-1β in rodents, as well as shown in the present results that an increase of serum IL-1β and TNF-α levels is observed in heat stroke rats. Furthermore, the present study shows that acute immediate treatment with DXM diminishes the heat stroke-induced elevation in serum levels of TNF-α and IL-1β.
Meanwhile, both arterial hypotension and cerebral ischemic damage are prevented and survival of heat stroke rats is improved following acute immediate administration of DXM at the onset of heat stroke. The present results further indicate that acute treatment with DXM causes a significant increase in the serum level of IL-10 during heat stroke. Studies found that IL-10 has important anti-inflammatory and immunosuppressive properties through attenuation of proinflammatory cytokines. In our present study, acute immediate administration of DXM may improve arterial hypotension and cerebral ischemia and prevent damage by increasing IL-10 but suppressing levels of TNF-α and IL-1β.[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
It has been shown that the glutamate and lactate/pyruvate ratio are well-known markers of cellular ischemia, whereas glycerol is a marker of how severely cells are affected by ongoing pathology. Indeed, as shown in our previous and present results, cerebral ischemia induced by heat stroke is associated with an increased production of glycerol, lactate/pyruvate ratio and glutamate in the brain as well as a decreased level of MAP in the periphery. It has been reported that the increased glutamate in the brain during the rat heat stroke also mediated the development of neuronal damage. Cerebral glutamate overload resulting from arterial hypotension and intracranial hypertension might be responsible for the occurrence of central nervous system syndromes associated with heat stroke. Systemic administration of glutamate receptor antagonists could protect against ischemic neuronal injury in experimental heat stroke.
In addition, recent studies reveal the excessive accumulation of cytotoxic free radicals in the brain and oxidative stress occurred during heat stroke. Evidence had accumulated to suggest that heat stroke-induced cerebral ischemia and neuronal damage might be associated with an increased production of free radicals. Pretreatment with hydroxyl radicals scavengers, such as α-tocopherol, prevented production of hydroxyl radicals, reduced lipid peroxidation and ischemic neuronal damage in several brain areas (corpus striatum, hypothalamus and cortex) of rats exposed to heat stroke and prolonged subsequent survival. After the onset of heat stroke, the cessation or reduction of blood flow to the brain induced neuronal damage. This neurotoxic cascade involved overproduction of glutamate in the brain.
In this study, our findings have indicated that the improvement of animal survival and the cerebral neuronal damage by acute DXM administration are also associated with the amelioration of cerebral glutamate, suggesting that DXM may exert its neuroprotective effect via attenuation of glutamate overload. To delineate whether the DMX effect also involves the removal of other deterioration factors or neurotoxins (e.g., oxidative substances, lipid peroxidation products), further investigations into the precise mechanisms of DXM-mediated neuroprotective effect will be required.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
The current choice for treatment of heat stroke is immediate administration of DXM in a dose dependent manner. However, DXM efficacy was tested in a baboon model of heat stroke, despite dexamethasone treatment prior to heat stress and during cooling, protection against the lethal effects of heat stroke were not realized. Similarly, the treatment of synthetic GC for human and animal-spinal cord injury or cerebral ischemia has been shown to be beneficial, but some studies have demonstrated that administration of steroids do not improve mortality or mobility after an acute ischemic brain stroke or in acute spinal cord injury. Xue et al. indicated that perhaps short-term and high dose of GCs are cardioprotective, albeit long-term and inadequate dose of glucocorticoid exposure may cause quite different consequences of deleterious cardiac function. In the aspect of heat stroke, although DXM caused sustained elevation of plasma interleukin-6 levels and decreased complement system activation with similar mortality rates between control and treated animals in a baboon model of heat stroke, our present findings suggest that rats treated with the synthetic GC, DXM, show improved heat stroke tolerance, as illustrated by attenuation of hypotension, cerebral ischemia, and neuronal damage and a prolongation to survival time.
Perhaps GC efficacy is thought to be at least partially dependent on appropriate dosage or heat severity, as metyrapone (an inhibitor of corticosterone synthesis) is without effect on cytokine mRNA expression, except at high heat loads in which it induces increased TNF-α mRNA expression. Clearly, whether permissive actions of GCs are sufficient for cytokine regulation and heat stroke protection or stress-induced levels are required is currently unknown, but the mechanism of protection of DXM in the present study appears to be at least partially mediated through the inhibition of IL-1-β TNF-α, but not promotion of IL-10 actions. Furthermore, more studies are required in this area to discuss and determine the potential benefit of GC therapy as a heat stroke prevention and treatment strategy.[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
## 3. Materials and Methods
*In the **3. Materials and Methods** section:*
## 3.1. Experimental Animals
*In the **3.1. Experimental Animals** section:*
Adult male Sprague-Dawley rats weighing between 280 and 330 g were obtained from the Animal Resource Center, National Science Council of Republic of China (Taipei, Taiwan, ROC). Between experiments the animals were housed individually at an ambient temperature of 24 ± 1 °C with a 12 h light-dark cycle, with the lights being switched on at 0600 h. Animal chow and water were allowed ad libitum. All experiments were approved by the Animal Ethics Committee of the Chia-Nan University of Pharmacy and Science, Tainan, Taiwan (approbated No. CN-IACUC-96032). Animal care and experiments were conducted according to the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and the guidelines of the Animal Welfare Act.
## 3.2. Surgery and Physiological Parameter Monitoring
*In the **3.2. Surgery and Physiological Parameter Monitoring** section:*
The femoral artery and vein of rats, under urethane anesthesia, were cannulated with polyethylene tubing (PE50) for blood pressure monitoring from the right femoral artery, blood samplings (for cytokines assay from right femoral vein, and for biochemical measurements from left femoral artery) and drug administration into left femoral vein. The animals were positioned in a stereotaxic apparatus (Kopf model 1460, Grass. Instrument. Quincy, MA, USA) to allow insertion of probes for measurement of CBF. Physiological monitorings included Tco, MAP, HR, and CBF in the corpus striatum.[](https://www.ncbi.nlm.nih.gov/mesh/D014520)
## 3.3. Induction of Heat Stroke and Experimental Design
*In the **3.3. Induction of Heat Stroke and Experimental Design** section:*
Rats under anesthesia were randomly assigned to one of the following five groups. One group of rats was exposed to an ambient temperature (Ta) of 24 °C for at least 90 min to reach thermal equilibrium before they were tested and used as normothermic controls. They were treated with 0.9% NaCl solution (1 mL/kg, i.v.) at 70 min after the start of the experiments. Their Tco was maintained at about 36 °C using an electric thermal mat before the start of experiments. The second group of rats with heat stroke received 0.9% NaCl solution (1 mL/kg, i.v.)
70 min after heat exposure. Heat stroke was induced by exposing the animals to an ambient temperature of 43 °C (with a relative humidity of 60% in a temperature-controlled chamber). The instant at which MAP and local CBF began to sharply decrease from their peak levels was taken as the onset of heat stroke, as shown in Figure 1. Our pilot results displayed that the interval between the start of heat exposure and onset of heat stroke was found to be 70 ± 3 min (n = 8). Accordingly, in the following heat stroke groups, all rats were exposed to 43 °C for exactly 70 min and after the onset of heat stroke, and allowed to recover at room temperature (24 °C).
The other three groups of rats with heat stroke respectively received DXM 4, 6, and 8 mg/mL/kg i.v. at the onset of heat stroke (70 min after the start of heat exposure). Each group of animals was subjected to: (i) measurement of survival time; (ii) measurement of Tco, MAP, HR, CBF; and striatal concentration of glutamate, glycerol, and lactate/pyruvate; (iii) measurement of plasma levels of PT, aPTT, protein C, D-D dimer, and FDP; (iv) measurement of plasma levels of BUN, creatinine, SGOT, SGPT and ALP; (v) measurement of serum levels of interleukin-1β (IL-1β), TNF-α, and IL-10. Colon temperature was monitored continuously by a thermocouple, while MAP and HR were monitored with a pressure transducer. Adequate anesthesia was maintained to abolish the corneal reflex and pain reflexes induced by tail pinch throughout the course of all experiments (about 8 h) following a single dose of urethane (1.4 g/kg b.w., i.p.). At the end of the experiments, control rats were killed with an overdose of urethane.[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## 3.4. Measurement of CBF
*In the **3.4. Measurement of CBF** section:*
Local CBF in the corpus striatum (SBF) was monitored with a Laserflo BPM2 laser Doppler flowmeter (Vasametics, St. Paul, NM, USA). A 24 gauge stainless steel needle probe (diameter, 0.58 mm; length, 40 mm) was inserted into the right corpus striatum using the coordinates: A, interaural 9.7 mm; L, 2.0 mm from mid-line; and H, 4.5 mm from the top of the skull.
## 3.5. Measurements of Extracellular Ischemia and Damage Markers in Brain
*In the **3.5. Measurements of Extracellular Ischemia and Damage Markers in Brain** section:*
After cannulation of vessels, the animal’s head was mounted on a stereotaxic apparatus with the nose bar positioned 3.3 mm below the horizontal line. Following a midline incision, the skull was exposed and a burr hole was made in the skull for the insertion of a dialysis probe (4 mm in length, CMA/12, Carnegie Medicine, Stockholm, Sweden). The microdialysis probe was stereotaxically implanted into the corpus striatum according to the atlas and coordinates of Paxinos and the coordinates of Paxinos and Watson (1982). As the methods described previously, an equilibrium period of 2 h without sampling was allowed after probe implantation. The dialysis probe was perfused with Ringer’s solution (147 mM Na+, 2.2 mM Ca2+, 4 mM K+, pH 7.0) at 2 μL/min using a CMA/100 microinfusion pump.
Dialysates were collected every 10 or 20 min in a CMA140 fraction collector. Aliquots of dialysates (5 μL) were injected onto a CMA600 Microdialysis Analyzer (Carnegie Medicine) for measurement of lactate, glycerol, pyruvate and glutamate. Four analytes can be analyzed per sample and the result is displayed graphically within minutes. The thermal experiments were started after showing stabilization in four consecutive samples.[](https://www.ncbi.nlm.nih.gov/mesh/D012964)
The lactate/pyruvate ratio is a well-known marker of cell ischemia, that is, an inadequate supply of oxygen and glucose. Glycerol is a marker of how severely cells are affected by the ongoing pathology. Glutamate is released from neurons during ischemia and initiates a pathological influx of calcium leading to cell damage. It is an indirect marker of cell damage in the brain.[](https://www.ncbi.nlm.nih.gov/mesh/D019344)
## 3.6. Biochemical Measurements
*In the **3.6. Biochemical Measurements** section:*
For biochemical determination, blood samples at 85 min after the start of heat exposure (or 15 min after the onset of heat stroke) were drawn by arterial femoral cannulation. The plasma levels of activated partial thromboplastin time, prothrombin time, and D-D dimer were measured by automated coagulation instruments (SYSMEX CA-1500, Kobe, Japan). The plasma levels of SGOT, SGPT, and alkaline phosphatase were determined by spectrophotometry (HITACHI 7600, Tokyo, Japan). For determination of plasma protein C, plasma was prepared as described previously, protein C in the sample was activated by specific venom activator. The resulting protein C activator was assayed in a kinetic test by measuring the increase in absorbance at 405 nm.
The reagents for the determination of protein C activity were provided by Berichrom Protein C (Dade Behring Marburg GmbH, Marburg, Germany).
## 3.7. Measurement for Serum Cytokines
*In the **3.7. Measurement for Serum Cytokines** section:*
Blood samples were taken at 85 min after the start of heat exposure (or 15 min after the onset of heat stroke) for determination of IL-1β, TNF-α, and IL-10 levels. For measurement of serum cytokines, 5 mL of blood was drawn from the femoral vein of rats. The amounts of the cytokines including IL-1β, TNF-α, and IL-10 in serum were determined by using a double-antibody sandwich enzyme-linked immunoabsorbant assay (ELISA, R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. This assay employs the quantitative colorimetric sandwich ELISA technique. Optical densities were read on a plate reader set at 450 nm for IL-1β, TNF-α, and IL-10.
The concentration of these cytokines in the serum samples was calculated from the standard curve multiplied by the dilution factor and was expressed as pg/mL.
## 3.8. Data Analysis
*In the **3.8. Data Analysis** section:*
Data are presented as means ± S.E. Repeated-measures ANOVA is conducted to test the treatment by time interactions and the effect of treatment over time on each score. The Duncan multiple-range test is used for post hoc multiple comparisons among means. A p-value less than 0.05 is calculated as statistical significance.
## 4. Conclusions
*In the **4. Conclusions** section:*
In summary, the prolongation of survival time in rats with immediate DXM therapy adopted at the onset of heat stroke was found to be associated with augmentation of both arterial blood pressure and cerebral blood flow, as well as reduction of cerebral ischemia, and vital organ damage, activated coagulation, and systemic inflammation during heat stroke. Our present results have shown a convincingly significant dose-dependent therapeutic effect of DXM administered immediately at the onset of heat stroke. Altogether, our data support that DXM may exert its therapeutic benefits by suppressing both cytokine overproduction and hypercoagulable state during heat stroke.[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
# Author Contributions
*In the **Author Contributions** section:*
Tsai-Hsiu Yang conceived the experiments, funded the project and wrote the manuscript. Chia-Chyuan Liu operated the animals, assessed biochemical and physiological parameters, and interpreted the data. Mei-Fen Shih and Ying-Hsiu Lai collected blood samples and performed the ELISA. Yi-Szu Wen provided DXM and finalized the manuscript.[](https://www.ncbi.nlm.nih.gov/mesh/D003907)
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
|
# Introduction
[Tocotrienol](https://www.ncbi.nlm.nih.gov/mesh/D024508) Rich [Palm Oil](https://www.ncbi.nlm.nih.gov/mesh/D000073878) Extract Is More Effective Than Pure [Tocotrienols](https://www.ncbi.nlm.nih.gov/mesh/D024508) at Improving Endothelium-Dependent Relaxation in the Presence of Oxidative Stress
# Abstract
*In the **Abstract** section:*
Oxidative endothelial dysfunction is a critical initiator of vascular disease. Vitamin E is an effective antioxidant but attempts to use it to treat vascular disorders have been disappointing. This study investigated whether tocotrienol[s, the le](https://www.ncbi.nlm.nih.gov/mesh/D014810)ss abundant components of vitamin E compared to tocopherols, might be more effective at preserving endothelial function. Superoxide gener[ated by hypo](https://www.ncbi.nlm.nih.gov/mesh/D024508)xanthine/xanthine oxidase or rat a[orta was ](https://www.ncbi.nlm.nih.gov/mesh/D014810)measured usin[g lucigenin](https://www.ncbi.nlm.nih.gov/mesh/D024505)-enhanced chemiluminescence. The effect of α-tocopherol, α-, δ[-, and γ-t](https://www.ncbi.nlm.nih.gov/mesh/D013481)ocotrienols and a tocotrienol rich palm oil extract (tocomin) on levels of s[uperoxide](https://www.ncbi.nlm.nih.gov/mesh/C033472) was assessed. Endothelial function in rat [aorta was as](https://www.ncbi.nlm.nih.gov/mesh/D024502)se[ssed in the presence of th](https://www.ncbi.nlm.nih.gov/mesh/C013649)e auto-[oxidant pyr](https://www.ncbi.nlm.nih.gov/mesh/D024508)ogallo[l. Whils](https://www.ncbi.nlm.nih.gov/mesh/D000073878)t all of t[he comp](https://www.ncbi.nlm.nih.gov/mesh/C570135)ounds displayed[ antioxida](https://www.ncbi.nlm.nih.gov/mesh/D013481)nt activity, the tocotrienols were more effective when superoxide was produced by hypoxanthine/xan[thine oxid](https://www.ncbi.nlm.nih.gov/mesh/D011748)ase whereas tocomin and α-tocopherol were more effective in the is[olated aorta](https://www.ncbi.nlm.nih.gov/mesh/D024508). Tocomin and α-tocopherol[ restored ](https://www.ncbi.nlm.nih.gov/mesh/D013481)endothelial function in the presence of oxidant stress [but α-,](https://www.ncbi.nlm.nih.gov/mesh/C570135) δ-, [and γ-tocotr](https://www.ncbi.nlm.nih.gov/mesh/D024502)ienols were ineffective. The protective effe[ct of t](https://www.ncbi.nlm.nih.gov/mesh/C570135)ocomi[n was replic](https://www.ncbi.nlm.nih.gov/mesh/D024502)ated when the tocotrienols were present with, but not without, α-toco[pherol. Tocotrienol rich t](https://www.ncbi.nlm.nih.gov/mesh/C013649)ocomin is more effective than α-tocopherol a[t reduc](https://www.ncbi.nlm.nih.gov/mesh/C570135)ing oxidative stress and [restoring en](https://www.ncbi.nlm.nih.gov/mesh/D024508)dothelium-dependent relaxation in rat[ aortae and ](https://www.ncbi.nlm.nih.gov/mesh/D024502)al[though α-, ](https://www.ncbi.nlm.nih.gov/mesh/D024508)δ-, an[d γ-toc](https://www.ncbi.nlm.nih.gov/mesh/C570135)otrienols effectively sc[avenged supe](https://www.ncbi.nlm.nih.gov/mesh/D024502)roxide, they did not improve endothelial function.[](https://www.ncbi.nlm.nih.gov/mesh/C013649)
## 1. Introduction
*In the **1. Introduction** section:*
Vitamin E, in addition to the four isoforms of tocopherol, contains four isoforms of tocotrienol. While there has been extensive investigation of the biological activity of the tocopherols, there has been much less attention paid to the tocotrienols. There is, however, emerging evidence that the tocotrienols have molecular targets distinct from those of the tocopherols that may result in new therapeutic opportunities. There are now a number of studies demonstrating cardioprotective actions of tocotrienols. For example, γ-tocotrienol is known to inhibit HMG-CoA reductase and therefore to decrease cholesterol synthesis.
Further, extracts of palm oil, a rich source of tocotrienols, have been demonstrated to activate the NO-cGMP pathway and, as a consequence, to decrease myocardial reperfusion injury perhaps due to scavenging of peroxynitrite. The antioxidant activity of tocotrienols may also contribute to protective actions in the vasculature. For example, Newaz et al. demonstrated that treatment of spontaneously hypertensive rats with γ-tocotrienol increased NOS activity and lowered arterial pressure, and γ-tocotrienol has also been shown to reduce oxidative stress and inflammation in rats with streptozotocin- (STZ-) induced diabetes. Further Norsidah et al. reported that a palm oil extract rich in tocotrienols, when orally administered to rats with hyperhomocysteinemia, reduced aortic oxidative stress and increased the plasma level of NO metabolites. In addition, Muharis et al. recently demonstrated that a palm oil fraction rich in tocotrienols restored endothelium-dependent relaxation in arteries in rats with STZ-induced type 1 diabet[es but it](https://www.ncbi.nlm.nih.gov/mesh/D014810) is not clear whether this may have be[en consequ](https://www.ncbi.nlm.nih.gov/mesh/D024505)ent to a lowering of glucose[ levels as ](https://www.ncbi.nlm.nih.gov/mesh/D024508)reported by Budin et al..
There is evidence that the beneficial vascular effects [of tocotrie](https://www.ncbi.nlm.nih.gov/mesh/D024505)nols may extend to man given the report that 2-mo[nth treatmen](https://www.ncbi.nlm.nih.gov/mesh/D024508)t with tocotrienols improves pulse wave velocity[ in healthy ](https://www.ncbi.nlm.nih.gov/mesh/D024508)males.[](https://www.ncbi.nlm.nih.gov/mesh/D024505)
The mechanism(s) of the beneficial effects of tocotrienols have not been well investigated nor, to the best of our knowledge, has there been any examination of the vascular actions of individual tocotrienol isomers. Therefore, the aims of this study were to compare the antioxidant activity of α-, δ-, and γ-tocotrienols with α-tocopherol and tocomin, a palm oil extract rich in tocotrienols (tocotrienol rich fraction: 40%, and palm olein: 38%) but also containing some α-tocopherol (11%). Given the antioxidant activity of these compounds we were further interested to investigate their capacity to protect NO-mediated vascular relaxation as an indication of whether they may be effective in preventing endothelial dysfunction in vascular diseases involving oxidant stress, for example, as a result of diabetes. It has been reported that tocotrienols are incorporated into cellular membranes more rapidly than tocopherol and that this may contribute to greater antioxidant efficacy. We therefore hypothesized that the tocotrienols would more effectively preserve endothelium-dependent relaxation in the presence of oxidative stress.[](https://www.ncbi.nlm.nih.gov/mesh/D024508)
## 2. Materials and Methods
*In the **2. Materials and Methods** section:*
## 2.1. Animals
*In the **2.1. Animals** section:*
Male Wistar rats 6–8 weeks of age (240–280 g) (Animal Resource Centre, Perth, WA, Australia) were used in the study. All procedures were approved by the Animal Experimentation Ethics Committee of RMIT University and conformed to the National Health and Medical Research Council of Australia code of practice for the care and use of animals for scientific purposes (AEC approval numbers 1309 and 1211).
## 2.2. Isolation of Aorta
*In the **2.2. Isolation of Aorta** section:*
The rats were killed by CO2 inhalation, followed by decapitation. The thoracic aorta was isolated and immediately placed in ice-cold Krebs bicarbonate solution (118 mM NaCl, 4.7 mM KCl, 1.18 mM MgSO4, 1.2 mM KH2PO4, 25 mM NaHCO3, 11.1 mM D-glucose, and 1.6 mM CaCl2). The aorta was then cleared of fat and connective tissue and cut into 2-3 mm long segments. The aortic rings were mounted between two stainless steel wires, one of which was linked to an isometric force transducer (model FT03, Grass Medical Instruments, Quincy, MA, USA) connected to a MacLab/8 (model MKIII, AD Instrument Co., Sydney, Australia), and the other end anchored to a glass rod submerged in a standard 10 mL organ bath. The organ bath was filled with Krebs-bicarbonate solution.
The bath medium was maintained at 37°C, pH 7.4, and continuously aerated with 95% O2 and 5% CO2. Aortic rings were equilibrated for 45 minutes at a resting tension of 1 g and then were precontracted with an isotonic, high potassium physiological salt solution (KPSS, 122.7 mM KCl, in which K+ ions replaced Na+ ions in the solution) for 20 minutes to achieve maximal contraction. After reequilibration, the rings were submaximally contracted with phenylephrine (PE, 0.01–0.3 M) and endothelial integrity was tested by a single concentration of acetylcholine (ACh, 10−5 M). Where relaxation was greater than 80% of the precontraction, the endothelium was considered to be intact and the aortic ring was included in the study. Some additional segments of the thoracic aortae were used to measure superoxide production.[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
## 2.3. Superoxide Generation Using Hypoxanthine/Xanthine Oxidase
*In the **2.3. Superoxide Generation Using Hypoxanthine/Xanthine Oxidase** section:*
Superoxide production was also measured by lucigenin enhanced chemiluminescence using hypoxanthine plus xanthine oxidase as a generator of oxygen radicals. Krebs-HEPES buffer (300 μL) containing lucigenin (5 mmol/L) and appropriate treatments were placed into a 96-well OptiPlate, followed by the addition of 1 unit/mL xanthine oxidase. A background reading was performed after which hypoxanthine (10−4 M) was added to all wells and superoxide production was measured. Superoxide inhibition was quantified by subtracting the superoxide reading from the background reading and expressing them as a percentage of the counts in the presence of the control.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
## 2.4. Superoxide Generation by Aorta
*In the **2.4. Superoxide Generation by Aorta** section:*
Superoxide production in the thoracic aorta was measured using lucigenin enhanced chemiluminescence based on methods described by Leo et al. with the following modification. The thoracic aorta was isolated, cleared of fat and connective tissue, and cut into 2-3 mm long segments in Krebs-HEPES buffer (composition (mM): NaCl 99.90, KCl 4.7, KH2PO4 1.0, MgSO4·7H2O 1.2, D-glucose 11.0, NaHCO3 25.0, CaCl2·2H2O 2.5, Na HEPES 20.0, pH 7.4). Aortic ring segments were incubated at 37°C for 45 min in Krebs-HEPES buffer in the presence of NADPH (100 mmol/L) as a substrate for NADPH oxidase and either alone or in the presence of tocomin, α-tocopherol, or α-, δ-, or γ-tocotrienol. In addition superoxide was measured in the presence of diphenylene iodonium (DPI, 5 mmol/L), a flavoprotein inhibitor that inhibits NADPH oxidase, as a positive control. 300 μL of Krebs-HEPES buffer containing lucigenin (5 mmol/L) and the appropriate treatments were placed into a 96-well OptiPlate, and superoxide production was measured and quantified.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
## 2.5. Vascular Function Experiments
*In the **2.5. Vascular Function Experiments** section:*
Cumulative concentration response curves to ACh (0.1 nM–10 mM) and sodium nitroprusside (SNP, 0.1 nM–10 mM) were determined using aortic rings contracted with phenylephrine (10−8 to 10−7 M) to 40%–60% of maximal contraction. Oxidative stress was generated by the addition of pyrogallol (30 μM) as previously described. Pyrogallol is well established to auto-oxidise to generate superoxide which then impairs endothelium-dependent relaxation by inactivating NO. Responses to ACh and SNP were also tested in the presence of pyrogallol by exposing the aortae for 20 minutes to tocotrienol rich tocomin (10−6–10−4 mg/mL), α-tocopherol (10−4–10−2 mg/mL), or tocotrienol isomers (α-, δ-, or γ-tocotrienol 10−3–10−1 mg/mL) to determine the effect of tocomin, α-tocopherol or α-, δ-, or γ-tocotrienol on endothelium-dependent and -independent relaxation in the presence of oxidative stress and if there is any potency difference between tocomin, α-tocopherol, and α-, δ-, or γ-tocotrienol.[](https://www.ncbi.nlm.nih.gov/mesh/D000109)
Responses to ACh and SNP were also tested in the presence of pyrogallol plus various combinations of α-tocopherol and tocotrienol isomers to replicate tocomin (10% δ-tocotrienol: 20% α-tocotrienol: 50% γ-tocotrienol: 20% α-tocopherol) and other tocotrienol combinations (α+γ)-tocotrienols and (α+δ+γ)-tocotrienols at a concentration of 10−4 mg/mL. These experiments were conducted to determine whether an interaction between α-tocopherol and the tocotrienols is necessary to improve endothelium-dependent relaxation in the presence of oxidative stress.[](https://www.ncbi.nlm.nih.gov/mesh/D000109)
## 2.6. Reagents
*In the **2.6. Reagents** section:*
All drugs were purchased from Sigma Aldrich except for acetylcholine perchlorate (BDH Chemicals, Poole, Dorset, UK), tocomin, and α-, δ-, and γ-tocotrienols (Carotech, Malaysia). All drugs were dissolved in distilled water, with the exception of tocomin, α-tocopherol, and α-, δ-, and γ-tocotrienols that were dissolved in 0.1% DMSO. A mixture of α-tocopherol and α-, δ-, and γ-tocotrienols which resembles tocomin was prepared as per the tocomin MSDS (10% δ-tocotrienol: 20% α-tocotrienol: 50% γ-tocotrienol: 20% α-tocopherol). Various tocotrienol combinations were also prepared using the following proportions: (α+γ)-tocotrienols (20% α-tocotrienol: 50% γ-tocotrienol: 30% DMSO) and (α+δ+γ)-tocotrienols (10% δ-tocotrienol: 20% α-tocotrienol: 50% γ-tocotrienol: 20% DMSO).[](https://www.ncbi.nlm.nih.gov/mesh/D000109)
## 2.7. Statistical Analyses
*In the **2.7. Statistical Analyses** section:*
All results are expressed as mean ± SEM, where n represents the number of animals per group. Concentration-response curves from the rat-isolated aortae were constructed and fitted to a sigmoidal curve using nonlinear regression (Graphpad Prism version 6.0, Graphpad Software, San Diego, CA, USA) to calculate the sensitivity of each agonist (pEC50). Maximum relaxation (R max) to ACh was measured as a percentage of the precontraction to phenylephrine. Group pEC50 and R max values were compared using a one-way ANOVA with post hoc analysis using Sidak's test as appropriate. p < 0.05 was considered statistically significant.[](https://www.ncbi.nlm.nih.gov/mesh/D000109)
Superoxide levels from rat aortic rings are expressed as average counts per second ± SEM normalized to dry tissue weight. Results were compared by one-way ANOVA with a post hoc Dunnett's test. p < 0.05 was considered statistically significant.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
Results from superoxide production and antioxidant capacity using hypoxanthine/xanthine oxidase are expressed as a percentage of the counts in the presence of the control (0.1% Krebs-HEPES buffer). The level of superoxide inhibition at each concentration was compared to vehicle for each compound using 1-way ANOVA with post hoc multiple comparisons using Dunnett's test (Prism version 6.0). p < 0.05 was considered statistically significant.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
## 3. Results
*In the **3. Results** section:*
## 3.1. Superoxide Scavenging Capacity of Tocomin, α-Tocopherol, and α-, δ-, and γ-Tocotrienols Using Hypoxanthine/Xanthine Oxidase
*In the **3.1. Superoxide Scavenging Capacity of Tocomin, α-Tocopherol, and α-, δ-, and γ-Tocotrienols Using Hypoxanthine/Xanthine Oxidase** section:*
Superoxide production induced by the presence of hypoxanthine/xanthine oxidase is shown in Figure 1. α-tocopherol (1A) caused an approximately 50% reduction in superoxide at a concentration of 10−2 mg/mL (Figure 1(a)). At the same concentration, all of the tocotrienol isomers caused approximately 80% reductions in superoxide (Figures 1(c), 1(e), and 1(g)). Tocomin (Figure 1(i)) caused a 50% inhibition of superoxide similar to α-tocopherol but at a concentration 10–100 times lower.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
## 3.2. Superoxide Scavenging Capacity of Tocomin, α-Tocopherol, and α-, δ-, and γ-Tocotrienol in Rat Aorta
*In the **3.2. Superoxide Scavenging Capacity of Tocomin, α-Tocopherol, and α-, δ-, and γ-Tocotrienol in Rat Aorta** section:*
As observed with the hypoxanthine/xanthine oxidase assay all of the compounds of interest were able to significantly reduce superoxide levels but the potency and efficacy were quite different (Figures 1(b), 1(d), 1(f), 1(h), and 1(j)). Interestingly, whereas α-tocopherol and tocomin produced relatively greater inhibition of aorta-derived superoxide, the tocotrienol isomers were less effective at the same concentrations in the hypoxanthine/xanthine oxidase assay. The relative potency between α-tocopherol and tocomin remained the same in this assay with tocomin being effective at approximately 100-fold lower concentration.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
## 3.3. Vascular Function
*In the **3.3. Vascular Function** section:*
The effect of pyrogallol-induced oxidative stress and the acute addition of varying concentrations of α-tocopherol, the tocotrienols, and tocomin is shown in Figures 2 and 3. Endothelium-dependent relaxation in response to ACh was significantly inhibited in the presence of pyrogallol-induced oxidative stress with a significant decrease in R max without affecting pEC50 (Table 1). Both α-tocopherol (Figure 2(a), 10−2 mg/mL) and tocomin (Figure 3(a), 10−4 mg/mL) were able to significantly improve endothelium-dependent relaxation in the presence of pyrogallol; however, tocomin improved endothelium-dependent relaxation at a concentration 100 times lower compared to α-tocopherol (Table 1). None of the tocotrienol isomers (α-, δ-, and γ-tocotrienols) improved endothelium-dependent relaxation even at concentrations 100 times higher than that of α-tocopherol (Table 1). Endothelium-independent relaxation to SNP was not affected by pyrogallol, α-tocopherol, tocomin, or the tocotrienols (Table 1).[](https://www.ncbi.nlm.nih.gov/mesh/D011748)
The effect of pyrogallol-induced oxidative stress and the acute addition of varying combinations of α-tocopherol and α-, δ-, and γ-tocotrienols (10−4 mg/mL) is shown in Figure 4. Both tocomin and the mixture of (T3(α+δ+γ) + (α-TC)) (Figure 4, 10−4 mg/mL) significantly improved endothelium-dependent relaxation in the presence of pyrogallol (Table 2 and Figure 4). Other preparations in the absence of α-TC, that is, T3(α+γ) and (α+δ+γ)-tocotrienols (10−4 mg/mL), did not improve endothelium-dependent relaxation (Figure 4 and Table 2). Endothelium-independent relaxation was not affected by the presence of pyrogallol, α-tocopherol, or α+δ+γ-tocotrienols (Table 2).[](https://www.ncbi.nlm.nih.gov/mesh/D011748)
## 4. Discussion
*In the **4. Discussion** section:*
This study demonstrated that the tocotrienol isomers were more effective at scavenging superoxide radicals produced by hypoxanthine/xanthine oxidase in comparison to those generated by isolated aortic segments in the presence of NADPH. Tocomin and α-tocopherol restored endothelial function in the presence of oxidative stress but α-, δ-, and γ-tocotrienols were ineffective. α-tocopherol was less effective than the tocotrienol isomers at similar concentrations when superoxide was generated by hypoxanthine/xanthine oxidase but more effective against superoxide generated by vascular tissue. Tocomin, an extract of palm oil containing predominantly tocotrienols but with some tocopherol, was effective in both assays at 100-fold lower concentrations than α-tocopherol. Consistent with their relatively lower antioxidant activity in isolated vascular tissue, the tocotrienol isomers failed to improve endothelium-dependent relaxation in the presence of oxidant stress.
Surprisingly tocomin was the most effective compound at improving endothelium-dependent relaxation and this effect could be replicated by a mixture of α-tocopherol and α-, δ-, and γ-tocotrienols, suggesting that the tocotrienol isomers provide more effective vasoprotection when acting together in combination with α-tocopherol.[](https://www.ncbi.nlm.nih.gov/mesh/D024508)
In the present study the antioxidant capacity of α-tocopherol, tocomin, and α-, β-, or γ-tocotrienols was examined using hypoxanthine/xanthine oxidase to generate superoxide in a tissue-free system or superoxide was produced by NADPH oxidase in segments of rat isolated aorta in the presence of NADPH. We have used both of these assays previously when testing the antioxidant activity of flavonols as a tool to predict efficacy as vasoprotectants in vascular disease. Xanthine oxidase (XO) is located on blood vessel walls and is an important enzyme that catalyzes the conversion of hypoxanthine to xanthine as a part of purine metabolism producing superoxide (O2 −) and hydrogen peroxide as a byproduct. XO induced free radical production has been implicated in the pathogenesis of diabetes related vascular complications. α-tocopherol and α-, δ-, and γ-tocotrienol were able to scavenge O2 − at concentrations as low as 10−3 mg/mL whereas tocotrienol rich tocomin was able to achieve the same effect at concentrations as low as 10−5 mg/mL. Tocopherol and tocotrienols have been demonstrated to exert their antioxidant activity by physically quenching superoxide.
Our findings suggest that α-tocopherol is 10 times more potent than α-, δ-, and γ-tocotrienols at scavenging hypoxanthine induced O2 − is somewhat surprising given the report by Yoshida et al. that tocopherol and tocotrienol isomers have a similar antioxidant activity when tested in homogenous solutions. A further surprising observation was that the tocotrienol isomers were less effective at scavenging superoxide derived from the aortic segments as this suggests a limited ability to access the tissue derived reactive oxygen species. This is in contrast to previous observations that tocotrienols are rapidly incorporated into cell membranes, a contributing factor to their antioxidant efficacy.[](https://www.ncbi.nlm.nih.gov/mesh/D024502)
The relative antioxidant efficacy of the compounds under examination was different when aortic segments provided the source of superoxide. Tocomin, containing a mixture of tocotrienol isomers and α-tocopherol, was more effective than the individual isomers at reducing oxidative stress whereas in the hypoxanthine/xanthine oxidase assay the opposite situation was observed. As noted above this may indicate an increase in activity when the tocotrienol isomers are combined or perhaps there is also an interaction with α-tocopherol.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
Our next aim was to investigate whether the compounds could effectively improve endothelium-dependent relaxation impaired by the presence of oxidative stress. Endothelium-derived nitric oxide (NO) rapidly reacts with O2 − (rate constant 2 × 1010 M/sec), which reduces its relaxant activity. Superoxide dismutase (SOD) also reacts rapidly with O2 − (rate constant 1-2 × 1010 M/sec) and in doing so enhances NO bioavailability and may enhance endothelium-dependent relaxation. However, antioxidant capacity alone does not guarantee the ability to enhance endothelium-dependent relaxation. For example, the well-known antioxidant ascorbate (vitamin C) does not enhance endothelium-dependent relaxation in arteries when endogenous O2 − levels are enhanced by inhibiting SOD.
This is probably due to the relatively slow rate of reaction between ascorbate and O2 − (2 × 105 M/sec) since exogenous SOD did enhance relaxation. Therefore one of the aims of this study was to determine whether the tocotrienols scavenged O2 − rapidly enough to enhance endothelium-dependent relaxation in the presence of basal O2 − levels and when high concentrations of O2 − were generated by pyrogallol. Surprisingly none of the tocotrienols were effective at improving endothelium-dependent relaxation, even at concentrations that decreased detection of superoxide generated by vascular tissue. By contrast, the less effective antioxidant α-tocopherol did improve ACh-induced relaxation. Significantly, tocomin was the compound that most effectively improved endothelium-dependent relaxation in the presence of pyrogallol-induced oxidative stress.
These observations make an interesting comparison to reports that a tocotrienol rich extract was able to acutely improve impaired endothelium-dependent relaxation in aortae removed from spontaneously hypertensive rats or rats with type 1 diabetes caused by STZ. A third component of tocomin, palm olein consisting mainly of triglycerides, was unlikely to account for the protective actions as it was reported to be without effect in the study by Muharis et al..[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Thus we speculated that the combination of multiple tocotrienol isomers and/or the additional presence of α-tocopherol was necessary to preserve endothelium-dependent relaxation. By testing the preparations with the same proportion of α-, δ-, and γ-tocotrienols and α-tocopherol present in tocomin, we determined that only the preparation containing α-tocopherol plus α-, δ-, γ-tocotrienol preserved endothelial function in the presence of oxidative stress. This data suggests an important interaction between α-tocopherol and tocotrienols to promote protection of vascular function. The mechanism of this positive interaction between α-tocopherol and the tocotrienols is worthy of further investigation.[](https://www.ncbi.nlm.nih.gov/mesh/D024508)
We propose that the capacity of tocomin to preserve endothelium-dependent relaxation is by rapidly eliminating superoxide as has been previously reported with superoxide dismutase preservation of relaxation in the presence of oxidative stress. Whilst tocopherols and tocotrienols have been reported to suppress signaling processes, for example, through the inhibition of NF-κB and STAT, the rapid effect seen in this study seems more likely due to antioxidant activity.[](https://www.ncbi.nlm.nih.gov/mesh/C570135)
## 5. Conclusion
*In the **5. Conclusion** section:*
It has been suggested that tocotrienols may have superior antioxidant activity to tocopherols, and we did find that to be true when superoxide is generated by hypoxanthine/xanthine oxidase in vitro. By contrast α-, δ-, and γ-tocotrienols were largely ineffective in improving NO mediated, endothelium-dependent relaxation in the presence of oxidative stress. However, tocomin, an extract from palm oil rich in tocotrienols and with a minor component of α-tocopherol, was found to be the most effective compound tested. The efficacy of tocomin could be replicated by the presence of α-tocopherol with α-, δ-, and γ-tocotrienols but not by the combined presence of the 3 tocotrienols alone. Thus the combination of tocotrienol isomers and tocopherol may prove to be an effective approach to the preservation of endothelial function where there is disease-induced oxidative stress such as in diabetes and hypertension.[](https://www.ncbi.nlm.nih.gov/mesh/D024508)
## Conflict of Interests
*In the **Conflict of Interests** section:*
The authors declare that there is no conflict of interests regarding the publication of this paper.
Superoxide generated by hypoxanthine (100 μM)/xanthine oxidase (0.01 U/mL) or rat aorta in the presence of NADPH: tocomin ((a) and (b)), α-tocopherol ((c) and (d)), α-tocotrienol ((e) and (f)), δ-tocotrienol ((g) and (h)), and γ-tocotrienols ((i) and (j)). Data is expressed as mean ± SEM. *Significantly different to control p < 0.05. ***Significantly different to control p < 0.001. Dunnett's multiple comparisons test.[](https://www.ncbi.nlm.nih.gov/mesh/D013481)
Endothelium-dependent relaxation in rat aortae in the presence of pyrogallol (P). Cumulative concentration-response curves to ACh in the absence (control) or presence of pyrogallol with varying concentrations of α-tocopherol (a), α-tocotrienol (b), δ-tocotrienol (c), and γ-tocotrienols (d). Data is expressed as mean ± SEM. *Significantly different to control p < 0.05. #Significantly different to pyrogallol p < 0.05 Sidak's multiple comparison test.[](https://www.ncbi.nlm.nih.gov/mesh/D011748)
Endothelium-dependent and -independent relaxation in rat aortae in the presence of pyrogallol (P): cumulative concentration-response curves to ACh (a) and SNP (b) in the absence (control) or presence of pyrogallol with varying concentrations of tocomin. Data is expressed as mean ± SEM. *Significantly different to control p < 0.05. #Significantly different to pyrogallol p < 0.05 Sidak's multiple comparison test.[](https://www.ncbi.nlm.nih.gov/mesh/D011748)
Endothelium-dependent relaxation in rat aortae in the presence of pyrogallol (P): cumulative concentration-response curves to ACh in the absence (control) or presence of pyrogallol with varying combinations of α-tocopherol (TC) and α-, δ-, and γ-tocotrienols (T3) (10−4 mg/mL). Tocotrienol isomers and α-tocopherol were present in the proportions found in tocomin (i.e., α-T3- 20%, δ-T3 10%, γ-T3 50%, and α-TC 20%). Data is expressed as mean ± SEM. *Significantly different to control p < 0.05. #Significantly different to pyrogallol p < 0.05 Sidak's multiple comparison test.[](https://www.ncbi.nlm.nih.gov/mesh/D011748)
#Significantly different to control p < 0.05.
∗Significantly different to pyrogallol p < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011748)
Sidak's multiple comparison test.
The effect of tocomin, α-tocopherol (TC), and α-, δ-, and γ-tocotrienols (T3) on ACh-induced endothelium-dependent and SNP-induced endothelium-independent relaxation of rat aortae in the presence of pyrogallol-induced oxidative stress.[](https://www.ncbi.nlm.nih.gov/mesh/C570135)
Tocotrienol isomers and α-tocopherol were present in the proportions found in tocomin (i.e., is α-T3- 20%, δ-T3 10%, and γ-T3 50% and α-TC 20%).[](https://www.ncbi.nlm.nih.gov/mesh/D024508)
#Significantly different to control p < 0.05.
∗Significantly different to pyrogallol p < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D011748)
Sidak's multiple comparison test.
The effect of various combinations of α-tocopherol (TC) and α-, δ-, and γ-tocotrienols (T3) on ACh-induced endothelium-dependent and SNP-induced endothelium-independent relaxation of rat aortae in the presence of pyrogallol- (P-) induced oxidative stress.[](https://www.ncbi.nlm.nih.gov/mesh/D024502)
|
# Introduction
Chemical Swarming: Depending on Concentration, an Amphiphilic [Ruthenium Polypyridyl Complex](https://www.ncbi.nlm.nih.gov/mesh/D011725) Induces Cell Death via Two Different Mechanisms
# Abstract
*In the **Abstract** section:*
Abstract
The crystal structure and in vitro cytotoxicity of the amphiphilic ruthenium complex [3](PF6)2 are reported. Complex [3](PF6)2 contains a Ru−S bond that is stable in the dark in cell‐growing medium, but is photosensi[tive. Upon blue‐light irrad](https://www.ncbi.nlm.nih.gov/mesh/D011725)iation, complex [3](PF6[)2 releas](https://www.ncbi.nlm.nih.gov/mesh/D011725)es the cholesterol–thioether ligand 2 and an aqua ruthenium complex [1](PF6)2. Although ligand 2 and complex [1](PF6)2 are by themsel[ves not c](https://www.ncbi.nlm.nih.gov/mesh/D011725)ytotoxic, comp[lex [3](PF6)2 was une](https://www.ncbi.nlm.nih.gov/mesh/D002784)xpectedly found t[o be as cytotoxic as cisplatin i](https://www.ncbi.nlm.nih.gov/mesh/D011725)n the dark, that is, with microm[olar effe](https://www.ncbi.nlm.nih.gov/mesh/D011725)ctive concentrations (EC50), against six h[uman canc](https://www.ncbi.nlm.nih.gov/mesh/D011725)er cell lines (A375, A431, A549, MCF‐7, MDA‐MB[‐231, and](https://www.ncbi.nlm.nih.gov/mesh/D002945) U87MG). Blue‐light irradiation (λ=450 nm, 6.3 J cm−2) had little influence on the cytotoxicity of [3](PF6)2 after 6 h of incubation time, but it increased the cytotoxicity of the complex by a factor 2 after longer (24 h) incubation. Exploring the [unexpecte](https://www.ncbi.nlm.nih.gov/mesh/D011725)d biological activity of [3](PF6)2 in the dark elucidated an as‐yet unknown bifaceted mode of action that depended on concentration, and thus, on the aggregation state of the[ compound](https://www.ncbi.nlm.nih.gov/mesh/D011725). At low concentration, it acts as a monomer, inserts into the membrane, and can deliver [1]2+ inside the cell upon blue‐light activation. At higher concentrations (>3–5 μm), complex [3](PF6)2 forms supramolecular aggregates that induce non‐apoptotic cell death by permeabilizing cell membranes and extracting lipids and membrane pr[oteins.](https://www.ncbi.nlm.nih.gov/mesh/D011725)(https://www.ncbi.nlm.nih.gov/mesh/D008055)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Collective behavior belongs to the most successful evolutionary models in nature; a single bee prick only kills the most sensitive victims, whereas an attack by swarming bees kills an Asian giant hornet by producing heat.1 Can this concept of biological swarming be applied to chemistry? Herein, we demonstrate how the chemical modification of a poorly toxic drug‐like compound with a nontoxic lipophilic ligand leads to a strong biological response mediated either by individual molecules or by their supramolecular assemblies.
The drug‐like structure of interest is a ruthenium–polypyridyl complex: [Ru(tpy)(bpy)(OH2)]2+ ([1]2+; tpy=2,2′;6′,2“‐terpyridine, bpy=2,2′‐bipyridine). For most metallodrugs, specific DNA and/or protein interactions have been proposed as the mode of action.2 In principle, aqua complex [1]2+ is a strong electrophile and its binding to DNA3 and proteins,4 which has been thoroughly investigated in the past, suggested that such compounds may be used as an anticancer agent.3a, 3b, 4 However, Reedijk et al. demonstrated that [Ru(tpy)(bpy)Cl]Cl, which in water hydrolyzes into [1]2+, is poorly cytotoxic.3b Probably, complex [1]2+ loses its ability to bind to biomolecules and become cytotoxic before it even enters the cell, by undergoing quick ligand‐exchange reactions with nucleophiles present in media (Figure S2 in the Supporting Information); thus forming an inactive complex. Exchanging the labile aqua or chloride ligand with a much more strongly bound ligand, L (e.g., a thioether‐, a nitrile‐, or pyridine‐based ligand), may prevent such undesired reactions in the dark.
In addition, ruthenium complexes such as [Ru(tpy)(bpy)(L)]2+ are photochemically active because visible‐light irradiation leads to ligand‐exchange reactions that do not occur in the dark.4, 5 Such photosubstitution reactions have been proposed as a way to trigger the toxicity of anticancer metallodrugs with spatial and temporal resolution.5a, 5g, 5h, 6 Likewise, photosubstitution of the protecting monodentate ligand L in [Ru(tpy)(bpy)(L)]2+ may activate the complex by producing [1]2+ inside a cell.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
Herein, we report on reactivity and cytotoxicity studies with compound [3](PF6)2, which is a conjugate of [1](PF6)2 and the thioether–cholesterol ligand 2 (Figure 1). For metallodrugs, ligands can, in principle, be utilized to alter drug‐like parameters, for example, solubility and/or stability, or to introduce functional moieties, such as cancer‐cell targeting groups or linkers to a drug carrier. In [3](PF6)2, the cholesteryl group was initially proposed as a lipid bilayer anchor to deliver the complex to cancer cells by using liposomes.7 However, cholesterol is also lipophilic, which is expected to dramatically change the partition coefficient (log P)7b–7d and localization of [3]2+ compared with that of [1]2+. The biological properties of [3](PF6)2 were studied herein in the absence of any liposome drug‐delivery system. As shown below, the amphiphilic character of [3](PF6)2 resulted in an unexpectedly high and nonselective cytotoxicity profile against a range of human cancer cell lines.
Chemical biological investigations and in vitro light irradiation experiments characterized a cell death mechanism that depended on concentration.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
A) Formation of the active compound [1](PF6)2 through light expulsion of the thioether ligand [3](PF6)2. B) Chemical structure of the investigated thioether–cholesterol ligand 2. C) Displacement ellipsoid plot (50 % probability level) of cationic [3]+, as observed in its crystal structure at 110(2) K. Only one of the two crystallographically independent molecules is shown. Counteranions, hydrogen atoms, and disorder have been omitted for clarity. Characteristic bond lengths [Å]: Ru1−N1=2.090(5), Ru1−N2=2.072(4), Ru1−N3=2.081(5), Ru1−N4=1.976(5), Ru1−N5=2.065(5), Ru1−S1=2.3639(14).[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
## Results and Discussion
*In the **Results and Discussion** section:*
Compound [3](PF6)2 was synthesized according to previous reports.7a Single crystals suitable for X‐ray structure determination were grown by vapor diffusion of diethyl ether into a solution of the compound in ethyl acetate. The structure confirmed the coordination of ligand 2 to ruthenium through the sulfur atom (Figure 1 C). The photochemical release of 2 by the blue‐light irradiation of [3]2+ (λ=455 nm, 10.5 mW cm−2) was studied in the absence of cells, but under the conditions used for in vitro toxicity experiments,8 that is, in Opti‐MEM complete cell‐growing medium (see composition in the Supporting Information). Within 8 min of irradiation (5.0 J cm−2), a bathochromic shift of the metal‐to‐ligand charge transfer (MLCT) absorption band of the complex was observed, from λ max=454 nm for [3]2+ to λ max=480 nm, which was characteristic for [1]2+ (Figure S3 in the Supporting Information). According to ESI‐MS results, the signal for [3]2+ at m/z 519.6 indeed disappeared and an intense signal at m/z 571.3 appeared for [2+Na]+ (calcd m/z 571.4; Figure S4 in the Supporting Information); this demonstrates that the photochemical release of ligand 2 (Figure 1 A) also occurs in the medium.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
The cytotoxicity of [1](PF6)2, 2, and [3](PF6)2 was investigated on six different human cancer cell lines derived from photodynamic therapy (PDT)‐relevant malignant tissues (skin, lung, brain, and breast; see the Experimental Section for more details). Briefly, 24 h after seeding, the cells were incubated with the compounds for 6 or 24 h, the media was refreshed, and blue‐light irradiation was performed (λ=455 nm, 10 min, 6.3 J cm−2). Following irradiation, the cells were incubated for an additional 48 h, and then counted by using the sulforhodamine B (SRB) assay (Figure 2 A).9 When possible, the effective concentration (EC50) leading to 50 % lower cell population compared to a drug‐free control was determined in µm (Table 1). In the dark, neither 2 nor the aqua complex [1](PF6)2 showed any cytotoxicity.
Surprisingly, however, complex [3](PF6)2 was found very cytotoxic, that is, with micromolar EC50 values for all cancer cell lines tested, whereas cisplatin showed expected increased cytotoxicity against faster proliferating cells (e.g., A375) and decreased cytotoxicity against a slower proliferating cell line (e.g., MDA‐MB‐231).10 Meanwhile, the results of blue‐light photocytotoxicity studies (Figure 2 and Table S1 in the Supporting Information) were intriguing. After 6 h of incubation, the cytotoxicity was barely influenced by irradiation, whereas after 24 h of incubation the activity increased by a factor of two. Taken together, the unspecific toxicity of [3](PF6)2 in the dark suggested that this compound might not interfere with cell proliferation through DNA binding, whereas the time‐dependent, light‐enhanced cytotoxicity suggested an adenosine triphosphate (ATP)‐dependent internalization process.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
A) Logarithmic dose–response curve for A549 lung cancer cells after treatment for 24 h with 2 (inverted triangles), [Ru(tpy)(bpy)Cl]Cl (triangles), [1](PF6)2 (squares), and [3](PF6)2 (filled dots) in the dark. The empty dots data points and the dashed curve represent the dose–response curve for cells treated with [3](PF6)2 and irradiated with blue light (λ=455 nm, 10 min, 6.3 J cm−2). B) Effective concentrations (EC50 in μm with 95 % confidence interval) in the dark and after blue‐light irradiation (λ=450 nm, 10 min, 6.3 J cm−2) of [3](PF6)2 on A549 cancer cells after 6 and 24 h of incubation.[](https://www.ncbi.nlm.nih.gov/mesh/C517923)
Cytotoxicity of [1](PF6)2, 2, and [3](PF6)2 in the dark given as effective concentrations (EC50) in μm with a confidence interval of 95 %.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
MALDI‐MS was used to obtain qualitative information about drug uptake. An immediate advantage of MALDI‐MS compared with the more usual inductively coupled plasma (ICP) MS technique is the ease of sample preparation.11 At several time points of drug incubation (1, 6, and 24 h), MS spectra were measured directly from thoroughly washed cells. Furthermore, the detection limit for [1](PF6)2 and [3](PF6)2 was determined to be below the applied concentration range (i.e., below 1 μm, data not shown). Unlike for ICP‐MS, partial speciation was possible with this technique.
MALDI‐MS detects monocationic species,12 and in the particular case of [1]2+ leads to the detection of a unique signal at m/z 490.1,13 which can be assigned to [Ru(tpy)(bpy‐H)]+ (Figures S6 and S7 in the Supporting Information). Because MALDI‐MS utilizes UV laser ionization (λ exc=355 nm), the ionization of [3](PF6)2 was followed by photosubstitution of ligand 2 and detection of the same signal at m/z 490.1; in other words, it was impossible to distinguish [3]2+ from [1]2+. However, MALDI‐MS conditions are much milder than those used for ICP‐MS analysis, that is, cells were not destroyed before the measurement and molecules were not atomized during ionization. As a consequence, it was possible to compare the proportion of cell‐based signals to that of ruthenium‐based signals (e.g., m/z 490.1), which were characterized by their unique isotope patterns.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
The ratio of ruthenium to cell signal in MALDI‐MS was used to compare drug uptake (see Figure 3 and Table S2 in the Supporting Information for details). As shown in Figure 3, incubation with complex [1](PF6)2 resulted in negligible ruthenium signals, whereas incubation with [3](PF6)2 resulted in a significant, time‐dependent increase in the ratio of Ru/lipid signals. The observed difference between [1](PF6)2 and [3](PF6)2 is in good correlation with their different hydrophobicities,14 and indicates different drug–cell interactions. The less hydrophobic ruthenium–polypyridyl complex [1]2+ was seemingly washed away from the cells, which suggested that [1](PF6)2 did not enter the cells, and/or interacted minimally with them. This result correlates with the low toxicity of this compound.
In contrast, the cholesterol‐containing compound [3](PF6)2 was already found in higher quantities in the cells after 1 h of incubation, which indicated increased drug uptake compared with [1](PF6)2.[](https://www.ncbi.nlm.nih.gov/mesh/D012428)
Ratio of Ru‐based/cell‐based signals observed by MALDI‐MS after different cell treatment conditions. Black bars correspond to samples kept in the dark, white bars to irradiated cells (λ=454 nm, 10 min, 6.3 J cm−2). Incubation times before refreshing of medium and irradiation are indicated.[](https://www.ncbi.nlm.nih.gov/mesh/D012428)
For cells treated with [3](PF6)2, the influence of light irradiation, which is accompanied by the in vitro formation of [1]2+, is indicated by the white bars in Figure 3. The ruthenium‐based MALDI‐MS intensity significantly decreased after light irradiation (see also Figure S7 in the Supporting Information in comparison with Figure S6). If the ruthenium species remained inside the cell after irradiation, and thus, were not washed away, a similar value for the ruthenium signals would have been expected under dark and irradiated conditions. Surprisingly, 85 % of the ruthenium‐based signals disappeared when light irradiation was performed after 6 h of incubation, and roughly 50 % when it was performed after 24 h of incubation. These results implied that after 6 h of incubation the light‐induced ligand exchange reaction shown in Figure 1 a released [1]2+ outside the cell into the medium, which was washed away before the MALDI‐MS analysis was performed.
Therefore, compound [3]2+ must stick initially in the biological membrane with the cholesterol ligand inserted in the outer leaflet of the cell membrane, and the ruthenium ion pointing into the media. The higher ruthenium intensity observed when light irradiation was performed after a longer (24 h) incubation time further implied that [3](PF6)2 flipped in a time‐dependent manner into the inner leaflet of the cell membrane, probably in a similar but slower manner to that for cholesterol.15 These observations are consistent with the negligible effect of light irradiation after 6 h. It is only when the ruthenium complex has flip‐flopped towards the cytosol (i.e., after 24 h) that light irradiation, and with it the intracellular formation of [1]2+ (the “bee”), leads to additional cytotoxicity by one of the (unknown) intracellular interactions with DNA or proteins.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
The unspecific cytotoxicity in the dark, however, cannot be explained with this model. A more collective mode of action, similar to that of a swarm of bees, was found to be responsible for the cytotoxicity of [3](PF6)2 at higher concentrations in the dark. First, standard chemical biological experiments were conducted to analyze the type of cell death. A DNA ladder experiment (Figure S9 in the Supporting Information) showed that [3](PF6)2 acted independently of caspases.16 Second, cell cycle analysis (Figure S10 in the Supporting Information) confirmed that [3](PF6)2 did not interfere with proliferation,9, 17 and third, dye‐exclusion assays18 implied a concentration‐dependent loss of cell‐membrane integrity (Figure S11 in the Supporting Information). All these observations indicated that a necrotic form of cell death was induced by [3](PF6)2 at 10 μm in the dark.
This hypothesis was confirmed by a flow cytometric annexin V‐propidium iodide assay (Figure S12 in the Supporting Information).19 In addition, optical microscopy imaging revealed an unusual cell‐size modulating activity of [3](PF6)2. With increasing concentrations, the cells shrunk to the size of the nuclei, without losing the outer cell membrane (Figure 4). In addition, significant debris was observed at concentrations of [3](PF6)2 higher than 10 μm. These debris particles could be stained with the protein‐binding dyes SRB9 (Figure 4 c) and trypan blue18b (Figure S11D in the Supporting Information), but not with the DNA‐specific dye DAPI.20, 4[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
Micrographs of A549 cells (40×). 4′,6‐Diamidino‐2‐phenylindole (DAPI) staining (of DNA) of A) untreated cells and B) cells treated with [3](PF6)2 (25 μm). C) Micrograph of cells after treatment with [3](PF6)2 (25 μm), fixation (trichloroacetic acid (TCA), 10 %), and SRB staining (to visualize proteins).[](https://www.ncbi.nlm.nih.gov/mesh/C007293)
Stimulated Raman scattering (SRS) microscopy experiments21 were conducted to look for the presence of lipids in those debris particles. Briefly, SRS signals were measured in living cells by utilizing the lipid‐specific CH2 stretching vibration at =2850 cm−1 (an example is depicted in Figure 5 B). The investigated cells showed the typical lipid distribution: a bright signal for the cell membrane, endoplasmic reticulum, Golgi apparatus and endosomal compartments, but no signal in the lipid‐poor nuclear region.22 The debris particles, which formed after treatment with high concentrations of [3](PF6)2, were already visible in the bright‐field image (Figure 5 A), and were between 1.0–2.5 μm in size. In the lipid‐sensitive SRS experiments, this debris gave a strong resonance signal (indicated with arrows in Figure 5 B). Altogether, these results implicated that the debris particles were lipid–protein aggregates extracted from the cell membrane, which explained why [3](PF6)2 induced a cell‐line‐unspecific, DNA‐independent cell death above 5 μm.
Detergents, such as sodium dodecyl sulfate (SDS), Triton‐X 100, or cetyltrimethylammonium bromide (CTAB), are also known for their ability to extract lipid–protein aggregates from cell membranes and are visible as debris particles (Figure S8 in the Supporting Information).23 Thus, we hypothesized that [3](PF6)2, which consists of a charged ruthenium polypyridyl head group and a lipophilic tail, might behave as a metal–organic surfactant capable of aggregating above a certain concentration and affecting the cell membrane.5[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
A) Bright‐field micrograph of A549 cells (32×) treated with [3](PF6)2 (10 μm). B) Stimulated Raman scattering (SRS) image of the same cell at =2850 cm−1, primarily selective for lipids. Blue arrows indicate some of the lipid‐containing debris particles. The intensity scale for the SRS signal (8‐bit) is given as inset. The scale bar is 20 μm.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
To confirm this hypothesis, the critical aggregate concentration (CAC) of [3](PF6)2 was measured by dynamic light scattering (DLS).23 The CAC of a compound is the concentration at which further addition of amphiphilic molecules does not change the monomer concentration. Above this concentration, the monomer molecules are in equilibrium with supramolecular aggregates of finite size.24 A CAC of 3.5 (±0.5) μm was found for [3](PF6)2 (Figures S13 and S14 in the Supporting Information), which is a prototypical characteristic of molecular detergents.25 Thus, complex [3](PF6)2 is able to form aggregates of typically 68 (±10) nm (z av, according to DLS), which, in contact with cells, are susceptible to form mixed assemblies that also contain cellular lipids and proteins, in analogy to non‐metalated surfactants.23, 26[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
Additional experiments were conducted to gain a deeper understanding of these interactions. First, the time evolution of the A549 cell population treated with [3](PF6)2 was studied and compared with those for cisplatin, staurosporine, Triton‐X, and SDS. As shown in Figure S5 in the Supporting Information, after compound withdrawal, the cell population initially treated with [3](PF6)2 recovered, as observed for Triton‐X 100 or SDS. In contrast, for cisplatin‐ or staurosporine‐treated cells, no recovery was observed after withdrawing the drug‐loaded media. The recovery of cell proliferation for cells treated with [3](PF6)2 correlated well with cell cycle analysis by means of flow cytometry (Figure S10 in the Supporting Information), in which no difference between the control and the treated cell populations was found 24 h after media refreshment.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
In a separate experiment, the dependence of the cytotoxicity (EC50) on the cell population was measured. As shown in Figure S15 in the Supporting Information, the EC50 of [3](PF6)2 decreased with an increased number of cells, whereas changing the cell density did not influence the EC50 of cisplatin. Thus, an increased number of cell membranes diminishes the toxicity of [3](PF6)2. Taken together, these facts are consistent with the hypothesis of a thermodynamically driven mode of action for high concentrations of [3](PF6)2 in the dark.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
## Conclusion
*In the **Conclusion** section:*
Our understanding of the cytotoxic activity of [3](PF6)2 relies on two separate modes of action (Figure 6). At low concentrations relative to the CAC, monomers of [3](PF6)2, like single bees, insert quickly into the outer leaflet of the cell membrane. The flip‐flop to the inner leaflet, which leads to internalization, occurs in a slower manner. After a short (6 h) incubation time, the light‐induced release of membrane‐impermeable species [1]2+ occurs outside the cell without biological consequences. After a prolonged incubation time (24 h), the same photoreaction leads to the release of [1]2+ inside the cell, where an unknown target is reached that may coordinate to ruthenium, eventually resulting in a more lethal signal.
This first mode of action is similar to that of recently reported ruthenium compounds.27 At concentrations above the CAC of (3.5±0.5) μm, however, complex [3](PF6)2 most likely behaves like a swarm by forming aggregates. When the ratio between these aggregates and the cell membrane lipids is high enough, thermodynamic forces lead to the generation of holes in the cell membrane (Figure S11 in the Supporting Information) and, at the highest concentrations tested (10–25 μm), to lipid–protein extraction of the cell membrane and formation of ternary aggregates containing cell lipids, membrane proteins, and [3](PF6)2 (Figures 4 and 5 and Figure S8 in the Supporting Information). Because it is simply based on the lipid/detergent equilibrium, this second mode of action is neither cell‐line specific, nor enhanced by light irradiation. On the contrary, light irradiation transforms [3]2+ back to ligand 2 and the aqua complex [1]2+, that is, it destroys the amphiphilic character of [3]2+ and its ability to form toxic aggregates.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
Proposed mode of action of [3](PF6)2. Treatment with concentrations above the CAC leads to lipid–protein extraction, and eventually to necrotic cell death. Treating the cells with concentrations below the CAC leads to insertion of the compound in the cell membrane, followed by a time‐dependent internalization. A significant photon‐enhanced effect was only observed after 24 h incubation, that is, after internalization of [3]2+, and thus, when the resulting active complex [1]2+ could interact with an (unknown) intracellular target.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
Such a bifaceted mode of action is unprecedented among metallodrugs, which are usually supposed to target nuclear DNA, mitochondrial membranes, or proteins. The complex behavior of compound [3](PF6)2 is due to the combination of its amphiphilic character and light sensitivity. This type of compound opens up two separate roads for drug delivery. First, as a monomer they can be used as prodrugs to release the toxic part (herein [1]2+) by light irradiation.
Second, as aggregate‐forming molecules they may be used as light‐sensitive drug carriers that, similar to a Trojan horse, can release a lipophilic load under visible‐light irradiation, which annihilates its amphiphilic character. In addition, these results should serve as a warning for the design of future metallodrugs. Increasing lipophilicity is often used as a way to increase cellular uptake.28 However, we showed herein that, by combining a charged metal‐based head with a fatty tail, self‐aggregation became possible, which radically changed not only how much of the compound penetrated into the cell, but also the mode of action of the compound. Finally, the possibility of extracting lipid–protein aggregates with an inorganic surfactant may offer significant advantages in the field of proteomics and lipidomics, due to the unique methods (e.g., MALDI‐MS) available for detecting inorganic compounds.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
## Experimental Section
*In the **Experimental Section** section:*
## General
*In the **General** section:*
1H NMR spectra were recorded by using a Bruker DPX‐300 spectrometer; chemical shifts are indicated in ppm relative to tetramethylsilane (TMS). Electrospray mass spectra were recorded on a Finnigan TSQ‐quantum instrument by using an ESI technique. DLS was performed with a Nanosizer instrument from Malvern operating at λ irr=633 nm. A Tecan M1000 PRO plate reader was used for fluorescence and absorbance measurements in 96‐well plates. Images and data were processed with Origin Pro, FCS Express, Prism 5.0, ChemDraw, Gimp 2.0, and Microsoft Excel software.[](https://www.ncbi.nlm.nih.gov/mesh/D006859)
## Synthesis and crystal growth
*In the **Synthesis and crystal growth** section:*
The synthesis of all described ligands and complexes was performed as reported previously. 3β‐(2‐{2‐[2‐(Methylthio)ethoxy]ethoxy}ethoxy)cholesterol (2; CAS‐Nr: 1373125‐91‐7) was described by Bahreman et al.5d [Ru(tpy)(bpy)(Cl)]Cl and [1](PF6)2 were synthesized according to previous reports,5b, 29 and [3](PF6)2 was prepared as described by Askes et al.7a Single crystals of [3](PF6)2 were obtained by slow recrystallization through the vapor diffusion of diethyl ether into a solution of the complex in ethyl acetate. Long and thin crystals with a ruby color were obtained that were suitable for X‐ray crystal structure determination.[](https://www.ncbi.nlm.nih.gov/mesh/D002782)
## Single‐crystal X‐ray crystallography
*In the **Single‐crystal X‐ray crystallography** section:*
All reflection intensities were measured at 110(2) K by using a SuperNova diffractometer (equipped with an Atlas detector) with CuKα radiation (λ=1.54178 Å) under the program CrysAlisPro (Version 1.171.36.32 Agilent Technologies, 2013). The program CrysAlisPro (Version 1.171.36.32 Agilent Technologies, 2013) was used to refine the cell dimensions and for data reduction. The structure was solved with the program SHELXS‐201330 and was refined on F 2 with SHELXL‐2013 (Sheldrick, 2015). Analytical numeric absorption corrections based on a multifaceted crystal model were applied by using CrysAlisPro (Version 1.171.36.32 Agilent Technologies, 2013). The temperature of the data collection was controlled by using the system Cryojet (manufactured by Oxford Instruments).
The H atoms were placed at calculated positions by using the instructions AFIX 13, AFIX 23, AFIX 43, or AFIX 137 with isotropic displacement parameters with values 1.2 or 1.5 times Ueq of the attached C atoms.
The structure was partly disordered. The fragment C52B→C59B/C52′→C59’ (C53B excluded) was disordered over two orientations, and the occupancy factor of the major component of the disorder was refined to 0.561(7). The contribution of two disordered ethyl acetate solvent molecules was removed from the final refinement by using SQUEEZE5f (details are provided in the CIF file). The absolute configuration was established by anomalous dispersion effects in diffraction measurements on the crystal. The Flack parameter refined to 0.009(7).[](https://www.ncbi.nlm.nih.gov/mesh/C007650)
Crystal data for [3](PF6)2: M r=1329.34; orange–red plates; 0.48×0.38×0.02 mm3; monoclinic; P21 (no. 4); a=37.5901(3), b=10.59806(7), c=16.68304(13) Å; β=100.6353(8)°; V=6532.05(9) Å3; Z=4; D x=1.352 g cm−3; μ=3.389 mm−1; T min−T max: 0.341–0.935; 77 205 reflections were measured up to a resolution of (sin θ/λ)max=0.62 Å−1; 22 930 reflections were unique (R int=0.0464), of which 21 312 were observed [I>2σ(I)]; 1552 parameters were refined by using 223 restraints; R1/wR2 [I>2σ(I)]: 0.0387/0.1017; R1/wR2 (all reflns): 0.0422/0.1046; S=1.035; residual electron density was found between −0.47 and 0.64 e Å−3.[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
CCDC‐1430105 contains the supplementary crystallographic data for this paper. These data are provided free of charge by The Cambridge Crystallographic Data Centre.
## Stability assays
*In the **Stability assays** section:*
In a mixture of [D6]DMSO/PBS (7:1): Complex [3](PF6)2 was dissolved in a 7:1 mixture of [D6]DMSO/phosphate‐buffered saline (PBS) mixture and immediately subjected to 1H NMR spectroscopy measurements (300 MHz, 128 scans). The tube was placed in an incubator set at 37 °C, and additional spectra were measured after 26, 49, and 73 h in the dark. No changes could be measured (see Figure S1 in the Supporting Information), which proved that [3](PF6)2 was thermally stable under such conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D004121)
In Opti‐MEM: The thermal stability of complexes [Ru(tpy)(bpy)Cl]Cl, [1](PF6)2, and [3](PF6)2 under cell culture conditions (Opti‐MEM complete, 37 °C) was investigated by measuring the evolution of the UV/Vis spectra in the dark by using a Tecan M1000PRO reader (Figure S2 in the Supporting Information). After 6 h, the first two compounds had reacted with medium components, whereas [3](PF6)2 was essentially unchanged.[](https://www.ncbi.nlm.nih.gov/mesh/C517923)
## Cell culturing: Reagents and cells
*In the **Cell culturing: Reagents and cells** section:*
Cells (A‐375, human malignant melanoma; A‐431, human epidermoid carcinoma; A549, human lung carcinoma; MCF7, human mammary gland adenocarcinoma; MDA‐MB‐231, human mammary gland adenocarcinoma; U‐87 MG, human glioblastoma grade IV) were distributed by the European Collection of Cell Cultures (ECACC), and purchased through Sigma Aldrich. Dulbecco's modified Eagle medium (DMEM, with and without phenol red, without glutamine), 200 mm glutamine‐S (GM), TCA, glacial acetic acid, SRB, and tris(hydroxylmethyl)aminomethane (tris base) were purchased from Sigma Aldrich. Fetal calf serum (FCS) was purchased from Hyclone. Penicillin and streptomycin were purchased from Duchefa and were diluted to a 100 mg mL−1 solution of penicillin/streptomycin (P/S). Trypsin and Opti‐MEM® (without phenol red) were purchased from Gibco® Life Technologies.
Trypan blue (0.4 % in 0.81 % sodium chloride and 0.06 % potassium phosphate dibasic solution) was purchased from BioRad. Plastic disposable flasks and 96‐well transparent plates were obtained from Sarstedt. The 96‐well black plates were from Greiner Bio‐one (5665–5090). Eight‐chamber microscope slides, Nunc® Lab‐Tek® II Chamber slideTM systems, were purchased from Sigma Aldrich. RNAse A and proteinase K (from Tritirachium album) were purchased from Sigma Aldrich.
The 1‐kb Plus DNA ladder was purchased from Fisher Scientific. Tris‐acetate–ethylenediaminetetraacetic acid (EDTA; TAE) buffer (pH 8) was prepared as a 50× stock solution and diluted to 1× prior to usage. DNA‐loading buffer was prepared as a 10 mL stock solution by using glycerol (3 mL), a solution of EDTA (1 mL, 0.5 m, pH 8), and methylene blue (5 mg). Cell lysis buffer (100 mm Tris‐HCl, pH 8, 20 mm EDTA, 0.8 % SDS) was freshly prepared. Mass spectrometry experiments were performed on a Synapt G2‐Si MALDI‐TOF mass spectrometer (Waters Corporation, Milford, MA), equipped with a λ=355 nm laser.
Before the measurements, the instrument was calibrated by using red phosphorus (Acros Organics). A 0.5 m solution of 2,5‐dihydroxybenzoic acid (DHB; Sigma) in methanol was used as a matrix. Mass spectra were acquired in positive mode.[](https://www.ncbi.nlm.nih.gov/mesh/D010637)
## General cell culturing
*In the **General cell culturing** section:*
Cells were purchased and upon receipt were cultured for working and frozen stocks. Each cell line was cultured in DMEM complete, with phenol red, supplemented with 8.0 % v/v FCS, 0.2 % v/v P/S, and 0.9 % v/v GM. Cells were cultured in 25 cm2 flasks and were split into a new passage at 70–80 % confluence (approximately 3× per week). Flasks were incubated at 37 °C with a CO2 level of 7.0 %. Media was changed every second day.
For all irradiation experiments, Opti‐MEM® media, supplemented with 2.5 % FCS, 0.2 % v/v P/S, and 1 % v/v GM (later on called OMEM complete) was used. Cells were passaged for 4–8 weeks. SRB was purchased from Alfa Aesar GmbH&Co under the tradename Kiton Red S.[](https://www.ncbi.nlm.nih.gov/mesh/D010637)
## ‐LED irradiation setup
*In the **‐LED irradiation setup** section:*
A 96 light‐emitting diode (LED) array, described in depth by Hopkins et al. ,8 allowed the irradiation of a 96‐well plate at 37 °C and, in parallel, maintained dark control under otherwise identical conditions. The light of the LED array had a wavelength of λ max=(454±22) nm, and at a voltage of 28.9 V the power at the bottom of each well was (10.5±0.7) mW cm−2.
## Cell‐free light irradiation of [3](PF6)2 in 96‐well plates
*In the **Cell‐free light irradiation of [3](PF6)2 in 96‐well plates** section:*
In a 96‐microtiter black plate, solutions of [3](PF6)2 (25 μm) in Opti‐MEM complete media (200 μL) were irradiated in triplicate at 37 °C by using the LED setup. After 20, 15, 12, 10, 8, 6, 4, 2, and 0 min of irradiation, the UV/Vis spectrum of each well was measured by using a Tecan M1000pro plate reader (Figure S3 in the Supporting Information). The samples were submitted to ESI‐MS measurements to confirm the light‐induced release of ligand 2 from the complex upon irradiation in Opti‐MEM complete (Figure S4 in the Supporting Information). Overall, under cell‐growing conditions, 8 min or 5 J cm−2 of blue‐light irradiation was enough to almost fully activate 5 nmol of the Ru complex [3]2+.[](https://www.ncbi.nlm.nih.gov/mesh/D012428)
## Dark cytotoxicity and phototoxicity on human cancer cell lines
*In the **Dark cytotoxicity and phototoxicity on human cancer cell lines** section:*
The cytotoxicity of compounds 2, [Ru(tpy)(bpy)(Cl)]Cl, [1](PF6)2, and [3](PF6)2 was evaluated by using the SRB microculture colorimetric assay. In short, exponentially growing cells were seeded in Opti‐MEM® (without phenol red, w/2.5 % FCS, P/S, and GM) into 96‐well plates at t=0 at the appropriate cell densities (A375=7000 cells/well, A431=8000 cells/well, A549=5000 cells/well, MCF‐7=8000 cells/well, MDA‐MB‐231=12 000 cells/well, U87Mg=6000 cells/well) to prevent confluence of the cells during the experiment. At t=6 or 24 h, the cells were treated with serial dilutions of each compound in Opti‐MEM, depending on the expected EC50 value after 6 or 24 h of incubation. For 2, [Ru(tpy)(bpy)(Cl)]Cl, and [1](PF6)2, the concentration series was 1.50, 7.50, 15.00, 30.00, 75.00, and 150.00 μm; for [3](PF6)2 it was 0.25, 1.25, 2.50, 5.00, 12.50, and 25.00 μm. The final DMSO concentration per well never exceeded 0.75 %, which was nontoxic to the cells.31 After 6 or 24 h of incubation with the drug‐loaded media, the media was aspirated and replaced by fresh, warm (37 °C) media; the plates were kept on a heating mat during media refreshing to avoid a significant drop in temperature for the cells, and to insure a temperature of at least 33 °C during light irradiation.
Light irradiation was performed for 10 min (λ=454 nm, 6.3 J cm −2) by using the LED‐based, 96‐well plate irradiation setup described by Hopkins et al.8 A duplicate plate was treated the same way, but without light irradiation and was further referred to as the dark control.[](https://www.ncbi.nlm.nih.gov/mesh/C517923)
The percentages of surviving cells relative to compound‐free wells were determined 72 h after the beginning of drug exposure, that is, at t=96 h, by using the SRB assay.9 Briefly, cells were fixed by using cold TCA (10 % w/v) and maintained at 4 °C for 4–48 h. Once fixed, TCA was removed from the wells, plates were gently washed 5× with water, air dried, stained by using 100 μL SRB (0.6 % w/v SRB in 1 % v/v acetic acid) for 30–45 min, washed with approximately 300 μL acetic acid (1 % v/v) 5× times, air dried, and the dye was then solubilized by using 10 mm tris base. The absorbance at λ=510 nm was read by using a M1000 Tecan Reader. The SRB absorbance data were used to evaluate the viable cell population in Excel and GraphPad Prism. The absorbance data from three wells (technical replicates, n t=3) for each cell line and concentration were averaged.
Relative cell populations were calculated by dividing the average absorbance of the irradiated wells by the average absorbance of the dark control. Three biological replicates (n b=3) of each treatment and cell line were completed. The averages of the biological replications were plotted as relative cell population versus log (concentration in μm) with standard error of each concentration. For each cell line, the EC50 was calculated by fitting the logarithmic dose–response curves through nonlinear regression with a fixed Y maximum (100 %) and minimum (0 %) relative cell population, and a variable Hill slope; this resulted in the simplified two parameter Hill slope equation by using PRISM 5.0.[](https://www.ncbi.nlm.nih.gov/mesh/C022027)
## MALDI‐MS
*In the **MALDI‐MS** section:*
Qualitative uptake experiments were performed through MALDI‐MS experiments. Cells of cell‐line A549 were seeded in Opti‐MEM® (without phenol red, w/2.5 % FCS, P/S, and GM) in an eight‐chamber glass slide (25 000 cells/well). Treatment with solutions of [1](PF6)2, [3](PF6)2, or [Ru(tpy)(bpy)(Cl)]Cl for 24 h, 6 h, 1 h, or 1 min was performed 24–48 h after seeding. After incubation with the drug, the supernatant media was aspirated, the cells were washed gently 3× with PBS, fresh media was added, and the cells were irradiated with blue light by using the same 96‐LED array as that used for irradiating 96‐well plates (λ=455 nm, 10 min, 6.3 J cm−2, 37 °C). An identical eight‐chamber slide was prepared, but left in the dark as a control.
After removing the chamber (on top of the glass slide) and drying the cell monolayer under ambient conditions, a 0.5 m solution of DHB matrix in methanol was applied by means of a pipette, and the samples were submitted to the MALDI SYNAPT G2‐Si mass spectrometer.[](https://www.ncbi.nlm.nih.gov/mesh/D010637)
To analyze the data, the individual signal height of the cell and the drug‐specific signals were first measured. Because the mass spectra from the untreated cell culture (control) showed a different pattern, depending on the treatment (light irradiated vs. dark probe, compare Figure S6 A vs Figure S7 A* in the Supporting Information), several lipid signals were chosen to decrease a potential bias due to irradiation. After summing up the height of the cell‐ and drug‐specific signals, the ratio between the heights was calculated to analyze the distribution of the investigated sample, and to eventually be able to determine the uptake indirectly. In Table S2 in the Supporting Information, the list of chosen signals and assignments, as far as possible, is given.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
## Microscopic investigation of living cells in the presence of surfactants in the dark
*In the **Microscopic investigation of living cells in the presence of surfactants in the dark** section:*
Cells were seeded in a 96‐well plate according to the cytotoxicity assay. After 24 h of incubation, the medium was removed and the cells were treated with increasing concentrations of the indicated drugs in Opti‐MEM complete. Optical microscopy images were recorded after 24 h of drug incubation in the dark (37 °C, 7 % CO2) at the indicated magnifications.[](https://www.ncbi.nlm.nih.gov/mesh/D002245)
Dye‐exclusion assay: The supernatant media was removed, the cells were treated with a diluted solution of trypan blue (0.25 %), and the cells submitted to microscopic investigations (see Figure S11 in the Supporting Information).[](https://www.ncbi.nlm.nih.gov/mesh/D014343)
DAPI staining: The supernatant media was removed, the cells were stained for 10 min with a solution of DAPI (0.01 mg mL−1), and then the cells were imaged.[](https://www.ncbi.nlm.nih.gov/mesh/C007293)
## DNA‐laddering experiment
*In the **DNA‐laddering experiment** section:*
Approximately 500 000 cells of the cell‐line A549 were seeded in cell culture flasks (25 cm2) and grown in DMEM (10 % FCS, 0.2 % P/S, 0.9 % GM). After 24 h, the supernatant media was removed and the nonconfluent cell monolayer was reloaded with substance‐loaded medium (or a blank fresh medium as a control). After 24–72 h, the supernatant medium was collected and the cell monolayer was washed with PBS. The combined media and PBS were centrifuged (1500 rpm, 5 min, Eppendorf 5702 centrifuge). The pellet of dead cells was gently suspended in PBS (1 mL) and centrifuged again (1500 rpm, 5 min, 278 K, Eppendorf Centrifuge 5418).
PBS was removed and lysis buffer (30 μL, 0 °C, 10 min) was added. Then RNAse (100 μg mL−1, 10 μL) was added and the cells were incubated for 10 min on ice followed by a prolonged incubation at 37 °C for 2 h. To finally digest the cell proteins, the cell pellet was treated with protein kinase K (10 μL) at 52 °C for 12 h. The extract was mixed with DNA‐ladder dye (10 μL) and analyzed by gel electrophoresis (2 % agarose loaded with 10 μL ethidium bromide (10 mg mL−1), 150 mV, 2 h, TAE buffer). The DNA bands were analyzed (see Figure S9 in the Supporting Information) by using a UV transilluminator (BioRad).[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Cell cycle investigation by flow cytometry
*In the **Cell cycle investigation by flow cytometry** section:*
Approximately 500 000 cells of the cell‐line A549 were seeded in cell culture flasks (25 cm2) and grown in DMEM complete. After 24 h, the supernatant media was removed and the nonconfluent cell monolayer reloaded with substance‐loaded medium (or a blank fresh medium as a control). After 24–72 h of drug incubation, the supernatant medium was submitted to the DNA‐laddering experiment (see above), whereas the living cells were washed and harvested with trypsin. After centrifugation (1500 rpm, 5 min, Eppendorf Centrifuge 5702), the supernatant was withdrawn, the cells resuspended with PBS (1 mL), and centrifuged (1500 rpm, 5 min, Eppendorf Centrifuge 5418). After additional washing with PBS, the cells were fixed by adding ice‐cold ethanol (70 %) dropwise to the cells stored on ice.
For thorough fixation and permeabilization, the cells were stored at least for 24 h at −20 °C. Thereafter, the cells were washed with PBS buffer (with Mg2+ and Ca2+, containing 1 % bovine serum albumin (BSA), and 0.1 % NaN3, 3×1 mL, 1000 rpm, Eppendorf Centrifuge 5418). Several cell suspensions were adjusted to the same cell concentration (ca. 100 000 cells mL−1), gently suspended in staining buffer (PBS buffer containing BSA, RNAse, NaN3, and PI analogue reported by Darzynkiewicz et al.17), and incubated for 30 min at room temperature in the dark. Analyses were performed by using a Beckman Coulter Quanta machine; collecting data from the FL2 channel. The cell population of interest was selected by plotting EV versus FL1.
For each cell cycle distribution 100 000 events were collected. The distribution was calculated by using FCSexpress software by applying the method of Dean and Jett.32 Representative histograms and compiled results are shown in Figure S10 in the Supporting Information.[](https://www.ncbi.nlm.nih.gov/mesh/D010710)
## SRS microscopy
*In the **SRS microscopy** section:*
The cell lipids were visualized by utilizing SRS microscopy. Cells of the cell‐line A549 were seeded in Opti‐MEM® (without phenol red, w/2.5 % FCS, P/S, and GM) in an eight‐chamber glass slide (5000 cells/well). After 24 h, cells were treated with solutions of [3](PF6)2 (10 μm) for 24 h. After incubation with the drug, the supernatant media was aspirated, the cells were washed gently with PBS, and fresh media was added. Before SRS measurements, the media was aspirated, the chamber (on top of the glass slide) was removed, and a cover glass was mounted.
The experimental setup for SRS imaging was mainly as described previously.21d In brief, laser light at λ=1064.4 nm (80 MHz, 8 ps) was intensity‐modulated at 3.636 MHz with an acousto‐optic modulator and overlapped with λ=816.7 nm light for imaging at =2850 cm−1. A Zeiss laser scanning microscope with a 32× objective (C‐Achroplan W, NA=0.85) was used to image samples with non‐descanned detection in forward scattering mode at 512×512 pixels with a pixel dwell time of 177 μs. Time‐averaged laser powers on the sample were 10 mW for the pump beam and 20 mW for the Stokes beam. The signal was amplified with a homebuilt transimpedance amplifier before demodulation in a lock‐in amplifier (SR844 Stanford Research Systems). The X‐phase output at a sensitivity of 1 mV was supplied to the ZEN microscopy software for image recording and processing.[](https://www.ncbi.nlm.nih.gov/mesh/D008055)
## Determination of the CAC
*In the **Determination of the CAC** section:*
The CAC was measured by using a fixed‐angle light‐scattering technique.25a, 33 Duplicates of a concentration series of [3](PF6)2 (0 to 10 μm) in distilled water were prepared and submitted to DLS measurements (Malvern Nanosizer, λ=633 nm). The attenuation factor was fixed to 11. For the analysis, the intensity of the scattered light (in kilocounts per second) was plotted against the concentration of [3](PF6)2 (see Figure S13 in the Supporting Information), according to an application note from Malvern.25b[](https://www.ncbi.nlm.nih.gov/mesh/D011725)
## Supporting information
*In the **Supporting information** section:*
|
# Introduction
Chemosensitizing [indomethacin](https://www.ncbi.nlm.nih.gov/mesh/D007213)-conjugated [dextran](https://www.ncbi.nlm.nih.gov/mesh/D003911)-based micelles for effective delivery of [paclitaxel](https://www.ncbi.nlm.nih.gov/mesh/D017239) in resistant breast cancer therapy
# Abstract
*In the **Abstract** section:*
Multidrug resistance (MDR) against chemotherapeutic agents has become the major obstacle to successful cancer therapy and multidrug resistance-associated proteins (MRPs) mediated drug efflux is the key factor for MDR. Indomethacin (IND), one of the non-steroidal anti-inflammatory agents, has been demonstrated to increase cytotoxic effects of anti-tumor[ agents as M](https://www.ncbi.nlm.nih.gov/mesh/D007213)RP[ su](https://www.ncbi.nlm.nih.gov/mesh/D007213)bstrates. In this study, dextran-g-indomethacin (DEX-IND) polymeric micelles were designed to delivery paclitaxel (PTX) for the treatment of MDR tumors. The DEX-[IND polymer could effe](https://www.ncbi.nlm.nih.gov/mesh/D003911)ct[ively e](https://www.ncbi.nlm.nih.gov/mesh/D003911)ncapsulate PTX with high loading content and DE[X-IND/PTX ](https://www.ncbi.nlm.nih.gov/mesh/D017239)mi[cel](https://www.ncbi.nlm.nih.gov/mesh/D017239)les present a small size distribution. [Compare](https://www.ncbi.nlm.nih.gov/mesh/D003911)d with free PTX, the release of PTX fro[m D](https://www.ncbi.nlm.nih.gov/mesh/D017239)EX-IND/PTX micelles could be pr[olonged](https://www.ncbi.nlm.nih.gov/mesh/D003911) [to ](https://www.ncbi.nlm.nih.gov/mesh/D017239)48 h. Cellular uptake test showed that the internalization of DE[X-I](https://www.ncbi.nlm.nih.gov/mesh/D017239)ND/PTX micelles b[y d](https://www.ncbi.nlm.nih.gov/mesh/D017239)rug-se[nsitive](https://www.ncbi.nlm.nih.gov/mesh/D003911) [MCF](https://www.ncbi.nlm.nih.gov/mesh/D017239)-7/ADR cells was significantly higher than free PTX benefiting from the inhibitory effect of I[ND on M](https://www.ncbi.nlm.nih.gov/mesh/D003911)R[Ps.](https://www.ncbi.nlm.nih.gov/mesh/D017239) In vitro cytotoxicity test further demonstrated that DEX-IND/PTX micelles coul[d e](https://www.ncbi.nlm.nih.gov/mesh/D017239)nhance the cytotoxicity of PTX against MCF[-7/](https://www.ncbi.nlm.nih.gov/mesh/D007213)ADR tumor cells. In vivo pharmacokinetic results showed that DE[X-IND/P](https://www.ncbi.nlm.nih.gov/mesh/D003911)T[X m](https://www.ncbi.nlm.nih.gov/mesh/D017239)icelles had longer systemic circulation time[ an](https://www.ncbi.nlm.nih.gov/mesh/D017239)d slower plasma elimination rate in comparison to PTX. The anti-tumor effica[cy test](https://www.ncbi.nlm.nih.gov/mesh/D003911) [sho](https://www.ncbi.nlm.nih.gov/mesh/D017239)wed that DEX-IND/PTX micelles exhibited greater tumor growth-inhibition effects on MDR tumor-bearin[g m](https://www.ncbi.nlm.nih.gov/mesh/D017239)ice, with good correlation between in vitro[ and in](https://www.ncbi.nlm.nih.gov/mesh/D003911) [viv](https://www.ncbi.nlm.nih.gov/mesh/D017239)o. Overall, the cumulative evidence indicates that DEX-IND/PTX micelles hold significant promise for the treatment of MDR tumors.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Multidrug resistance (MDR) is a great obstacle for cancer chemotherapy, which leads to the poor treatment outcomes [,]. The overexpression of drug efflux transporters on the cell surface has been confirmed based on the clinical and experimental studies []. The most commonly reported efflux membrane transporter multidrug resistance-associated proteins (MRPs) are extensively overexpressed in various tumor cells and actively pump the broad spectrum of chemotherapeutics outward from the cells [,].
Several chemotherapeutics can be served as substrates for MRPs [,,,]. The anti-tumor agent paclitaxel (PTX) is widely used for the treatment of various solid tumors via promoting polymerization of tubulin dimers to form microtubules and stabilizing microtubules by preventing depolymerization [], but it is also a substrate for MRPs []. The abnormal increase of drug efflux and decrease of intracellular drug concentration lead to PTX resistance. In addition, it has several therapeutic limitations, including irreversible nephrotoxicity, neurotoxicity, and cardiotoxicity [].[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
Nano-drug delivery system provides potential solutions to some limitations, such as increased drug solubility, target site distribution or reduced drug-induced toxicity [,]. Dextran (DEX), as hydrophilic moieties, has been widely used as drug carrier and it has no surface charge, which can reduce aggregate with negatively charged serum proteins and increase the nonspecific cellular uptake []. Indomethacin (IND), one of non-steroidal anti-inflammatory agents, has been demonstrated to suppress MDR pump and glutathione-S-transferase activities, and then reduce MRPs-mediated efflux of chemotherapeutics []. IND sensitizes the drug-resistant tumor cells to PTX by inhibiting multi-drug resistance protein 1 (MRP1) promoter activity and then reduces the expression of MRP1
[].[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
Here, we tried to combine polymeric micelle with chemosensitizer to exert synergetic oncotherapy. Dextran-indomethacin (DEX-IND) polymeric micelles were prepared to encapsulate hydrophobic chemotherapy PTX. So far as we know, PTX-loaded DEX-IND (DEX-IND/PTX) micelles have never been used for oncotherapy, nor have their in vivo behavior been systematically investigated. In this study, we assessed the therapeutic effects of DEX-IND/PTX micelles in MDR tumor-bearing mouse model. In addition, we systematically assessed their characteristics, cytotoxicity, and pharmacokinetic profiles.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Materials and methods
*In the **Materials and methods** section:*
## Materials
*In the **Materials** section:*
Paclitaxel (PTX) was purchased by Jingyan Chemicals Corporation (Shanghai, China); Dextran (DEX, Mn = 10 KDa), Indomethacin (IND), Dicyclohexylcarbodiimide (DCC) and 4-Dimethylaminopyridine (4-DMAP) were purchased from Shanghai Aladdin Bio-chem Technology Co. Limited (Shanghai, China); Pyrene and 4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide (MTT) were purchased from Sigma-Aldrich Co (St Louis, MO, USA); All other solvents and reagents were chemical grade.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
## Animal
*In the **Animal** section:*
BALB/c nude (20 ± 2 g) mice and female Sprague Dawley (200 ± 20 g) rats were purchased from Shanghai slack laboratory animal co., Ltd and fed under standard laboratory conditions. This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of Wenzhou medical university. All surgery was performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering.[](https://www.ncbi.nlm.nih.gov/mesh/D010424)
## Synthesis of DEX-IND polymer
*In the **Synthesis of DEX-IND polymer** section:*
DEX-IND polymer was synthesized via the esterification reaction between hydroxy group of DEX and carboxyl group of IND in the presence of DCC and DMAP. Briefly, 2.148 g IND (10% to number of D-glucose units in Dex), 3.708 g DCC and 0.244 g DMAP (1:3:0.3, mol:mol:mol) were dissolved in 20 mL DMSO and stirred at 50°C for 1 h to activate the carboxylic acid of IND. Then, 10 g DEX was added and stirred at 50°C the protection of nitrogen for 2 days. After the reaction, the mixture was transferred into a dialysis membrane (MWCO 7.0 kDa) to dialyze against pure water 48 h with frequent exchanges of pure water. The final solution was filtered through 0.8 μm filter and lyophilization to achieve DEX-IND polymer.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Characterization of DEX-IND polymer
*In the **Characterization of DEX-IND polymer** section:*
The obtained DEX-IND polymer was confirmed by 1H NMR spectra using a Bruker (AVACE) AV-500 spectrometer. 20 mg·mL-1 DEX, IND and DEX-IND were respectively dissolved in dimethylsulfoxide-d6 and measured.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
The critical micelle concentration of DEX-IND was determined by fluorescence measurement using pyrene as a probe []. Briefly, 1 mL pyrene solution was added into brown volumetric flask and acetone was removed with the nitrogen flow. Different concentration DEX-IND solution (ranged from 10−4 to 10−1 mg·mL-1) were added into each flask to reach the final pyrene concentration (6.0 × 10−7 M). The fluorescence intensity was measured using fluorescence spectrophotometer (RF-5301PC, Japan), and the intensity ratio of the first peak (I1, 374 nm) to the third peak (I3, 385 nm) was calculated to determinate CMC value.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Preparation and characterization of DEX-IND/PTX micelles
*In the **Preparation and characterization of DEX-IND/PTX micelles** section:*
DEX-IND/PTX micelles were prepared via the dialysis method. Briefly, 10 mg DEX-IND was dissolved in 10 mL pure water under magnetic stirring at room temperature. Then, the 1 mg·mL-1 PTX ethanol solution was added into DEX-IND solution (PTX: DEX-IND = 10%, w/w), and stirred for 10 min. The mixed solution was dialyzed (MWCO 7.0 kDa) against pure water for 24 h with frequent exchange of pure water. After dialysis, the mixed solution was centrifuged at 5,000 rpm for 10 minutes to remove unencapsulated PTX, and the DEX-IND/PTX micelles were obtained.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
The morphological examinations were performed by transmission electron microscopy (TEM, Hitachi, Tokyo, Japan). The samples were dropped on copper grids and stained with 2% (w/v) phosphotungstic acid for viewing. The size and polydispersity index (PDI) were measured using dynamic light scattering (DLS).[](https://www.ncbi.nlm.nih.gov/mesh/D003300)
## Evaluation of stability
*In the **Evaluation of stability** section:*
The stability of DEX-IND/PTX micelles was evaluated at 4°C. At pre-determined times, DEX-IND/PTX micelle solution was taken and the mean size and PDI were recorded by DLS.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Determination of drug-encapsulation efficiency and drug loading
*In the **Determination of drug-encapsulation efficiency and drug loading** section:*
PTX content was measured using high performance liquid chromatography (HPLC). DEX-IND/PTX micelles were diluted in methanol solution to dissociate the micelles, and the PTX was measured. The content of PTX was assayed with C18 column (250 mm × 4.6 mm, 5 μm), and acetonitrile/water (45:55, v/v) was used as the mobile phase. The column temperature and the detection wavelength were set as 35°C and 240 nm with flow rate at 1.0 mL·min-1 []. Encapsulation efficiency and drug loading were calculated using eqs below:[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
## In vitro PTX release
*In the **In vitro PTX release** section:*
In vitro drug release profiles of DEX-IND/PTX micelles were investigated using the dialysis method. 2 mL DEX-IND/PTX micelle solution was transferred into a dialysis membrane (MWCO 7.0 kDa), and then immersed in 30 mL PBS solution (including pH 7.4 and pH 5.0) containing 2 M sodium salicylate and incubated at 37°C with constant shaking at 70 rpm []. At pre-determined time intervals (1, 2, 4, 6, 8, 10, 12, 24, 36 and 48 h), the samples were collected and replaced with fresh medium. PTX content was measured using HPLC. All drug-release tests were repeated thrice.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Cell culture
*In the **Cell culture** section:*
MCF-7 Cells and resistant human breast carcinoma cells (MCF-7/ADR) were purchased from Nanjing Kaiji Biotech. Ltd. Co. (Nanjing, China). Cells were cultured in RPMI-1640 medium with 10% (v/v) fetal bovine serum (FBS) and 1% penicillin-streptomycin in a humidified atmosphere at 37°C with 5% CO2.[](https://www.ncbi.nlm.nih.gov/mesh/D010406)
## Cellular uptake
*In the **Cellular uptake** section:*
The cellular uptake test of DEX-IND micelles was then investigated in vitro. MCF-7 cells were seeded in 24-well plates at 3×104 cells per well, and incubated for 24 h. Then, the cells were exposed to a medium containing DEX-IND/RITC micelles, and further incubated for 1 and 8 h. After washed with PBS, the cells were observed using a confocal microscopy.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
MCF-7 and MCF-7/ADR cells were seeded in 24-well plates at 3×104 cells per well, and incubated for 24 h. Then, the cells were exposed to PTX, free PTX+IND or DEX-IND/PTX micelles (0.5 μmol·mL-1) for further incubation (1 and 8h). Then, cells were lysed with RIPA buffer to release the intracellular PTX, free PTX+IND or DEX-IND/PTX micelles. The intracellular concentrations of PTX were determined by HPLC method as mentioned above. Uptake was expressed as the amount (nmol) of PTX associated with a unit weight (1 mg) of cellular protein.
Protein contents of cell lysate were measured using BCA protein assay reagent kit.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
## Cytotoxicity
*In the **Cytotoxicity** section:*
MCF-7 and MCF-7/ADR cells were used to determine the cytotoxicity of DEX-IND/PTX micelles via the MTT assay. Cells were seeded in 96-well plates at 1×104 cells per well, and incubated for 24 h. Then, PTX and DEX-IND/PTX micelles with serial concentrations were added into the cell medium and cultured for 48 h. At pre-determined times, 10 μL of MTT (5 mg·mL-1, 5% MTT) was added and incubated for further 4 h, and then 150 μL DMSO was added to dissolve MTT formazan. The absorbance was measured at 570 nm in a microplate reader (Bio-Rad, USA) and the viability was expressed as the percentage of the control. The test was repeated thrice.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## In vivo pharmacokinetics study
*In the **In vivo pharmacokinetics study** section:*
The in vivo pharmacokinetic study was performed in male Sprague-Dawley rats (200 ± 20 g), and pharmacokinetic parameters were calculated via the software of Drug and Statistics (2.0). The rats were fasted overnight with free to water before conducting the study. The experimental protocols and animal care were approved by the Committee for Animal Experiments of Wenzhou Medical University. In this study, the rats were randomly divided into two groups, including PTX and DEX-IND/PTX group (n = 6). Rats were administered intravenously at a dose of 10 mg·kg-1 PTX solution or DEX-IND/PTX micelles solution, respectively.
At designated intervals (0.25, 0.5, 1, 2, 4, 6, 8, 12, 24 and 48 h), blood samples (300 μL) were drawn from orbit and immediately placed into heparinized tubes. The obtained blood samples were centrifugation at 4000 rpm for 10 min, and then stored at -20°C for further analysis. To determine PTX concentration, methanol and acetonitrile were used for both protein precipitation and PTX extraction []. The sample mixture was vortexed for 10 min, and then centrifuged for 15 min at 10,000 rpm. The supernatant was transferred and evaporated under the nitrogen flow.
The extraction residual was redissolved in the mobile phase solution and injected for analysis. The analysis was performed on Agilent-C18 column (250 mm × 4.6 mm, 5 μm) with a security guard column (C18, 10 × 4 mm, 5 mm); mobile phase: acetonitrile/water (45:55, v/v); detection wavelength: 240 nm; flow rate: 1.0 mL·min-1; column temperature: 35°C. The linear standard curve presented good linearity over the concentration range of 0.1–20 μg·mL-1. The standard curve in plasma: A = 43.175C - 1.743 (R2 = 0.9993). The average recovery was (101.54 ± 1.271)% and the coefficients of variation within and between days were 3.58% and 4.72%, respectively.[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
## In vivo anti-tumor efficacy study
*In the **In vivo anti-tumor efficacy study** section:*
In vivo anti-tumor efficacy of DEX-IND/PTX micelles was investigated in male BALB/c nude mice transplanted with MDR tumor cells and pharmacological intervention began when the tumor volume grew to approximately 100 mm3. Mice were randomly divided into four groups (n = 5), and received saline solution, 10 mg·kg-1 PTX, 10 mg·kg-1 DEX-IND/PTX micelles and 10 mg·kg-1 DEX-IND/PTX micelles, respectively, for 7 days consecutively. The tumor volume ((((longest diameter)^2)*(shortest diameter)))/2) and body weight were monitored every 4 days. At the end of the experiment, mice were sacrificed and tumors were weighted individually.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Statistics
*In the **Statistics** section:*
All data, expressed as means ± SEM, were from at least three separate experiments. Statistical analysis was conducted using Student’s t-test with *p < 0.05 as indicative of statistically significant differences.
## Results and discussion
*In the **Results and discussion** section:*
## Synthesis and characterization of DEX-IND polymer
*In the **Synthesis and characterization of DEX-IND polymer** section:*
DEX-IND polymer was successfully synthesized via the esterification between the carboxyl group of IND and the hydroxyl group of DEX. The synthesis route is presented in Fig 1A. IND was used as a hydrophobic chain and DEX was used as hydrophilic moieties. The structure of DEX-IND conjugates was confirmed via 1H NMR spectra and the results are presented in Fig 1B. The characteristic peaks of IND could be observed in 1H NMR spectrum of DEX-IND conjugates.
Based on this, it is evidence that the DEX-IND conjugates are successfully synthesized in this study.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
Preparation and characterization of DEX-IND micelles.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
(A) Synthetic route of DEX-IND polymer. (B) 1H NMR spectra. (C) Negative-stain transmission electron microscopy of DEX-IND and DEX-IND/PTX micelles. (D) Characteristics of DEX-IND and DEX-IND/PTX micelles.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
The obtained DEX-IND polymer could spontaneously form micelles in aqueous medium. The CMC is one of the important characteristics for amphiphilic materials and represents the self-assembly ability to form micelles. Low CMC value means that polymer can form micelles under highly diluted condition. The aggregation behavior of DEX-IND was determined using fluorescence method using pyrene as a probe. The CMC value of DEX-IND polymer was 34.2 μg·mL-1.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Preparation and characterization of DEX-IND/PTX micelles (cont.)
*In the **Preparation and characterization of DEX-IND/PTX micelles (cont.)** section:*
The DEX-IND/PTX micelles were prepared through solvent diffusion method. Ethanol has been removed by dialysis method. Amphiphilic DEX-IND polymer could spontaneously form micelles in aqueous medium and encapsulate hydrophobic PTX. Drug encapsulating efficiency of DEX-IND/PTX micelles was 80.8 ± 1.1%, and drug loading was 7.14 ± 0.25% during 10% drug feeding amount (Table 1). The obtained DEX-IND/PTX micelles were examined using TEM and DLS.
Fig 1C showed that DEX-IND and DEX-IND/PTX micelles both had a uniform spherical shape observed by TEM, and their sizes were 68.3 ± 4.63 nm and 64.1 ± 3.81 nm measured by DLS, respectively. The particle size of DEX-IND/PTX micelles was smaller than DEX-IND micelles, which was associated with the hydrophobic interaction between the hydrophobic chains (IND) and free PTX becoming stronger after PTX loading.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
Characteristics of DEX-IND and DEX-IND/PTX micelles.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
To investigate in vitro stability of DEX-IND/PTX micelles at 4°C, micellar size and PDI were detected at different periods of time (1, 2, 3, 4, 5, 6 and 7d). Fig 2A and 2B showed that micellar size remained nearly unchanged within a week, and PDI increased slightly over the same period, which provided the strong evidence that DEX-IND/PTX micelles could keep good colloidal stability and be suitable to be stored at 4°C.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
Stability and in vitro release of DEX-IND/PTX micelles.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
(A, B) In vitro stability of DEX-IND/PTX micelles at 4°C, including size and PDI. (C) In vitro release profiles of free PTX and DEX-IND/PTX micelles in pH 7.4 PBS, and DEX-IND/PTX micelles in pH 5.0 PBS. Data represent mean ± standard deviation (n = 3).[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## In vitro PTX release from DEX-IND/PTX micelles
*In the **In vitro PTX release from DEX-IND/PTX micelles** section:*
In vitro drug release profiles of DEX-IND/PTX micelles were investigated by dialysis method in pH 7.4 and pH 5.0 PBS. As shown in Fig 2C, free PTX is released quickly, more than 90% within 12h. In contrast, the release of PTX can be maintained for more than 48 h. The release of DEX-IND/PTX micelles exhibits a biphasic release pattern, including rapid release during the initial 12 h and slow release later on (12–48 h).[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Cellular uptake of DEX-IND micelles
*In the **Cellular uptake of DEX-IND micelles** section:*
The cellular uptake test of DEX-IND micelles were investigated in MCF-7 cells. In this study, rhodamine B isothiocyanate (RITC) was used to label DEX-IND micelles. Fig 3A presents the cellular images of cells after incubation with DEX-IND/RITC micelles for 1 and 8 h. The results showed that DEX-IND/RITC micelles could be internalized into MCF-7 cells in a time-dependent manner.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
Cellular uptake of DEX-IND micelles.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
(A) Fluorescence images of MCF-7 cells were incubated with DEX-IND/RITC micels for 1 and 8 h, respectively. (B) Comparison of the cellular uptake of PTX, PTX+IND or DEX-IND/PTX in MCF-7 cells and MCF-7/ADR cells. Data represent mean ± standard deviation (n = 3). *P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
Then, cellular accumulation efficiency was quantitatively analyzed using HPLC method to determine intracellular concentrations of PTX in drug-sensitive MCF-7 cells and drug-resistant MCF-7/ADR cells after 12 h incubation with free PTX, free PTX+IND or DEX-IND/PTX micelles. The results showed that the internalization of free PTX by drug-resistant MCF-7/ADR cells was significantly decreased in comparison to that by drug-sensitive MCF-7 cells. In contrast, the internalization of free PTX+IND and DEX-IND/PTX micelles by drug-resistant MCF-7/ADR cells was similar to that by drug-sensitive MCF-7 cells (Fig 3B). Therefore, DEX-IND micelles could provide more efficient cellular uptake both in MCF-7 and MCF-7/ADR cells. This result can be explained by the following reasons: 1) more efficient endocytosis of PTX can be accomplished with the help of micellar carriers; 2) MRPs-mediated efflux can be reversed by IND released from DEX-IND/PTX micelles.
Free PTX could be internalized into tumor cells via molecular diffusion mechanism, while PTX encapsulated in micelles could be internalized into tumor cells via endocytosis, which was a high-efficiency route for drugs going through cell membrane.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
## In vitro antitumor activity
*In the **In vitro antitumor activity** section:*
Cytotoxicities of PTX, and DEX-IND/PTX micelles against MCF-7 and MCF-7/ADR cells were addressed using MTT assay. As shown in Fig 4A and 4B, DEX-IND conjugates showed negligible toxicity with DEX-IND concentration ranging from 1–600 μg·mL-1 in MCF-7 and MCF-7/ADR cells. Fig 4C showed that PTX, PTX + IND and DEX-IND/PTX micelles could inhibit the proliferation of MCF-7 cells in a dose-dependent manner. In contrast, PTX + IND and DEX-IND/PTX micelles both led to the higher cytotoxicity in comparison to PTX in MCF-7/ADR cells, which indicated that IND in micelles probably suppress MRPs-mediated efflux to some extent and increase PTX accumulation in cells. Here, it was shown that IND could enhance the cytotoxicity of PTX in MCF-7/ADR cells, which was consistent with cellular uptake test that IND could retard the efflux of PTX and then enhance its cytotoxicity.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
In vitro anti-tumor activity of DEX-IND/PTX micelles.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
(A, B) Cytotoxicity of DEX-IND micelles without PTX encapsulation in MCF-7 and MCF-7/ADR cells for 24 h (n = 3). (C, D) The viability of MCF-7/ADR cells after incubation with PTX, PTX + IND and DEX-IND/PTX micelles for 24 h. Data represent mean ± standard deviation (n = 3). *P < 0.05.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## In vivo pharmacokinetics
*In the **In vivo pharmacokinetics** section:*
The plasma concentration-time profiles of PTX and DEX-IND/PTX micelles following single dose administration are showed in Fig 5. As presented, PTX plasma concentration reduced quickly at initial 10 h of intravenous administration, leading to short t1/2, approximately 7.462 h, and little PTX could be measured in the plasma after 12 h. As expected, DEX-IND micelles could significantly increase the blood circulation of PTX, and appreciable PTX could still be detected in rats treated with DEX-IND/PTX micelles at 24 h after administration. Compared with PTX, DEX-IND/PTX micelles showed prolonged blood circulation (t1/2, 12.894 h). The concentration-time data was analyzed by the non-compartmental model. The pharmacokinetic parameters are presented in Table 2.
Compared with PTX, the area under concentration curve (AUC0-∞) in DEX-IND/PTX micelles was significantly increased from 99.71 ± 19.347 ng·mL-1·h-1 to 300.069 ± 89.089 ng·mL-1·h-1. The mean residence time (MRT) of DEX-IND/PTX micelles (13.234 ±1.175 h) was 1.8-fold increase for PTX (7.136 ± 1.06 h).[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
The concentration versus time curve of PTX and DEX-IND/PTX after intravenous administration.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
Data represent mean ± standard deviation (n = 5).
Plasma pharmacokinetic parameters of PTX after intravenous administration of Taxol and DEX-IND/PTX micelles in rats (n = 5).[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
AUC, area under the concentration-time curve; MRT, mean residence time; CL, clearance rate.
The stability of drug delivery system is of great significance because it is a prerequisite for the successful delivery to target tissues. After systemic administration, drug delivery system will encounter plasma proteins before it reaches target tissues. It was reported that serum proteins easily interacted with drug carriers and thus affect their stability and tissue disposition. Therefore, the integrity of drug delivery system in the presence of blood components is of great significance to efficient drug delivery to the target tissues. In vivo pharmacokinetic results showed that DEX-IND micelles could enhance circulation time of PTX via slowing PTX clearance from body, which ensure more PTX distributed into tumor cells via enhanced permeability and retention (EPR) effect.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## In vivo antitumor activity
*In the **In vivo antitumor activity** section:*
The in vivo antitumor efficacy of DEX-IND/PTX micelles was investigated in MDR tumor-bearing mice. The anti-tumor efficacy of DEX-IND/PTX micelles was compared with PTX and saline. The changes of tumor volume were plotted. As shown in Fig 6A, both PTX and DEX-IND/PTX micelles could effectively suppress tumor growth. After 24 d, tumor volumes in tumor-bearing mice treated with PTX and DEX-IND/PTX micelles were significantly smaller than those treated with saline (vs saline, P < 0.05).
After 48 d, tumor volumes in tumor-bearing mice treated with DEX-IND/PTX micelles were significantly smaller than those treated with PTX (vs PTX, P < 0.05), but were no significant difference between two different dose DEX-IND/PTX micelles (P > 0.05). The tumor volume in tumor-bearing mice treated with 20 mg·mL-1 DEX-IND/PTX micelles was 1.81-fold smaller than those treated with 20 mg·mL-1 PTX. The synergetic effect of micellar passive targeting and enhanced anti-tumor activity with IND could be the main reason for the significant inhibition on tumor growth in tumor-bearing mice treated with DEX-IND/PTX micelles.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
In vivo antitumor activities of PTX and DEX-IND/PTX after intravenous administration tumor-bearing mice.[](https://www.ncbi.nlm.nih.gov/mesh/D017239)
(A) Mice tumor volume changes within 48 days. (B) Mice body weight changes within 48 days. Data represent mean ± standard deviation (n = 5).
The toxicity of Dex-Ind/DOX micelles was next assessed through bodyweight changes. As shown in Fig 6B, PTX led to 30% bodyweight reduction, which was involved with its severe drug-related toxicity. In contrast, DEX-IND/PTX micelles could significantly decrease PTX toxicity during systemic circulation, which benefited from encapsulated PTX in DEX-IND micelles leading to the reduced exposure of normal tissues to it and enhanced passive accumulation of PTX in tumor sites via EPR effect. Therefore, DEX-IND/PTX micelles could reduce the undesirable side effects and then improve the reduction of bodyweight.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
Co-delivery of anti-tumor agents together with MRPs inhibitors has become more popular to overcome MDR. In some studies, anti-tumor agents were encapsulated into drug delivery systems and chemosensitizers were administered as free solutions. However, chemosensitizers are easily distributed to all tissues during administered in a free form, which induces the non-specific activity of chemosensitizers. Therefore, it is of great importance to delivery chemosensitizers to the site of action. In this study, DEX-IND/PTX micelles were prepared to achieve co-delivery of anti-tumor agent and chemosensitizer, and the in vivo anti-tumor result showed that DEX-IND/PTX micelles could effectively suppress tumor growth with reduced toxicity to normal tissues.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Conclusion
*In the **Conclusion** section:*
DEX-IND was synthesized successfully with low CMC in this study. The DEX-IND could spontaneously form nanosized micelles in aqueous medium and encapsulate the hydrophobic anti-tumor agent PTX. The PTX release from DEX-IND micelles could be maintained for more than 48h. DEX-IND/PTX micelles were effective for suppressing both drug sensitive and resistant MCF-7 cells. The assay of anti-tumor activity indicated that DEX-IND/PTX micelles could increase anti-tumor activity in comparison to commercial PTX.
Overall, the results indicated that DEX-IND/PTX micelles were a promising potential candidate for oncotherapy.[](https://www.ncbi.nlm.nih.gov/mesh/D003911)
## Supporting information
*In the **Supporting information** section:*
# References
*In the **References** section:*
|
# Introduction
Neural changes associated with cerebellar tDCS studied using MR spectroscopy
# Abstract
*In the **Abstract** section:*
Anodal cerebellar transcranial direct current stimulation (tDCS) is known to enhance motor learning, and therefore, has been suggested to hold promise as a therapeutic intervention. However, the neural mechanisms underpinning the effects of cerebellar tDCS are currently unknown. We investigated the neural changes associated with cerebellar tDCS using magnetic resonance spectroscopy (MRS). 34 healthy participants were divided into two groups which received either concurrent anodal or sham cerebellar tDCS during a visuomotor adaptation task. The anodal group underwent an additional session involving MRS in which the main inhibitory and excitatory neurotransmitters: GABA and glutamate (Glu) were measured pre-, during, and post anodal cerebell[ar t](https://www.ncbi.nlm.nih.gov/mesh/D005680)DCS, [but witho](https://www.ncbi.nlm.nih.gov/mesh/D018698)ut[ th](https://www.ncbi.nlm.nih.gov/mesh/D018698)e behavioural task. We found no significant group-level changes in GABA or glutamate during- or post-tDCS compared to pre-tDCS levels, however, [ther](https://www.ncbi.nlm.nih.gov/mesh/D005680)e wa[s large d](https://www.ncbi.nlm.nih.gov/mesh/D018698)egree of variability across participants. Although cerebellar tDCS did not affect visuomotor adaptation, surprisingly cerebellar tDCS increased motor memory retention with this being strongly correlated with a decrease in cerebellar glutamate levels during tDCS across participants. This work provides novel in[sights re](https://www.ncbi.nlm.nih.gov/mesh/D018698)garding the neural mechanisms which may underlie cerebellar tDCS, but also reveals limitations in the ability to produce robust effects across participants and between studies.
## Introduction (cont.)
*In the **Introduction (cont.)** section:*
Numerous studies have shown a facilitatory effect of anodal cerebellar transcranial direct current stimulation (tDCS) on both motor and cognitive behavioural tasks (Galea et al. ; Grimaldi et al. ; Cantarero et al. ). For instance, Galea et al. () applied anodal cerebellar tDCS during visuomotor adaptation and found anodal cerebellar tDCS led to faster adaptation, relative to either primary motor cortex (M1) anodal tDCS or sham tDCS (Galea et al. ). This effect on motor adaptation/learning has been replicated in visuomotor adaptation (Hardwick and Celnik ; Block and Celnik ; Doppelmayr et al. ; Leow et al. ), force-field adaptation (Herzfeld et al. ), locomotor adaptation (Jayaram et al. ), saccade adaptation (Panouilleres et al. ; Avila et al. ), motor skill learning (Cantarero et al. ), and language prediction tasks (Miall et al. ).
As a result, it has been suggested that cerebellar tDCS is not only a useful tool to understand cerebellar function but also as a possible clinical technique to restore cerebellar function in patients suffering from cerebellar-based disorders (Grimaldi et al. ). However, there are also inconsistencies regarding the impact of cerebellar tDCS with several studies reporting cerebellar tDCS to have little or no effect on motor learning (Conley et al. ; Minarik et al. ; Jalali et al. ) or large variability between- and within-subjects (Dyke et al. ). Therefore, understanding the underlying causes of this variability is essential.
Previous work has investigated the neural changes associated with M1 anodal tDCS using a range of MRI techniques (Stagg et al. ; Kim et al. ; Antal et al. ; Hunter et al. ; Kunze et al. ). For example, magnetic resonance spectroscopy (MRS) revealed that M1 anodal tDCS caused a decrease in gamma-aminobutyric acid (GABA), with the magnitude of this decrease being correlated with improvements in both sequence learning (Stagg et al. ) and force-field adaptation (Kim et al. ), but they did not report any significant change in Glu or any correlation between the change in Glu and motor behaviour.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Despite this work relating to M1 tDCS, no previous research has attempted to use MRS to investigate the neural changes observed with cerebellar tDCS. Given the abundance of GABA and Glu within the cerebellar cortex (Waddell et al. ), we predicted that these were the two metabolites most likely to be affected by anodal cerebellar tDCS. Therefore, using MRS, the changes in GABA and Glu were quantified within the right cerebellar cortex directly underneath the anodal electrode pre, during and post tDCS. We sought to understand if there is any detectable change in GABA or Glu in response to cerebellar tDCS and if their alteration could predict individual differences in the effect of cerebellar tDCS on visuomotor adaptation performance.
According to previous findings, we hypothesised that a reduction in GABA induced by cerebellar anodal tDCS would be positively correlated with the degree of visuomotor adaptation.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Materials and methods
*In the **Materials and methods** section:*
## Participants
*In the **Participants** section:*
34 healthy young individuals participated in this study (mean age: 22 ± 2 years; 11 male) and were divided into two groups of 17: anodal (23 ± 5 years; 8 male) and sham (19 ± 2 years; 3 male). All were naïve to the behavioural task, self-assessed right handed, had normal/corrected vision, and reported to have no history of any neurological condition. The study was approved by the Ethical Review Committee at the University of Birmingham and was in accordance with the declaration of Helsinki. Written informed consent was obtained from all participants after screening for suitability for MR imaging and brain stimulation. Participants were recruited through online advertising and received monetary compensation.
All participants first completed a behavioural task, testing visuomotor adaptation during active or sham TDCS. At the end of the behavioural session, 29 of the 34 participants reported their attention, fatigue, and quality of sleep using a questionnaire with a scale from 1 to 7. They also reported whether they believed they had received active or sham stimulation, and their hours of sleep during the previous night (Table 1). After completing the behavioural task, all 17 participants from the anodal group underwent a session of MRS, with concurrent tDCS. The sham group were not imaged.
## Transcranial direct current stimulation (tDCS)
*In the **Transcranial direct current stimulation (tDCS)** section:*
For the behavioural session, anodal tDCS (DC-Stimulator, NeuroConn, Germany) was delivered through a pair of rubber electrodes (4 × 4 cm2) within two 5 × 5 cm2 pads soaked in a saline solution (Wagner et al. ) and attached to the head with Coban self-adhesive tape. The anodal electrode was placed over the right cerebellar cortex, 3 cm lateral to the inion. The cathodal electrode (reference) was placed over the right buccinator muscle (Galea et al. ) as it has been shown to be an effective montage for cerebellar stimulation (Rampersad et al. ). At the onset of stimulation, current was increased in a ramp-like fashion over a period of 10 s.
For the behavioural study, in the anodal group, a 2 mA current (current density J = 0.08 mA/cm2) was applied for 25 min. In the sham group, tDCS was ramped up over period of 10 s, remained on for 10 s before being ramped down and switching off. Participants during the behavioural task were blinded to whether anodal or sham was applied (Table 1).
For the MR session, 1.8 mA anodal tDCS was delivered (J = 0.07 mA/cm2) through a pair of rubber electrodes (5 × 5 cm2). The electrodes were attached to each participant’s head, in the same position as the behavioural session, using EEG paste and Coban self-adhesive tape. Electrodes were connected to an MR-compatible tDCS machine (DC-Stimulator-MR, NeuroConn, Germany). Ideally 2 mA stimulation would have been used, however, high impedance (> 55 kΩ) within the MRI-compatible tDCS equipment meant this was not possible.
To avoid MR image artefacts, the tDCS current was set at 0 mA for the pre-and post-stimulation data acquisition, rather than switching the tDCS device off. This was because the tDCS device employed two filters to prevent leakage of radio-frequency electromagnetic fields into the MRI faraday cage, which operated only when the tDCS device was active. Participants were informed of when the stimulation was turned on and were instructed not to fall asleep during the scans.
## Behavioural protocol
*In the **Behavioural protocol** section:*
Participants were seated at a table, with their chin supported by a rest (Fig. 1a), in front of a computer monitor (30-inch; 1280 × 1024 pixel resolution; 105 cm from chin rest). A Polhemus motion tracking sensor (Colchester, VT, USA) was attached to their right index finger and their arm was placed underneath a horizontally suspended wooden board, which prevented direct vision of the arm (Fig. 1a). The visual display consisted of a 1-cm diameter starting box, a green cursor (0.25 cm diameter) representing the position of the subject’s index finger, and a circular white target (0.33 cm diameter). Targets appeared in 1 of 8 positions (45° apart) arrayed radially at 8 cm from the central start position. Targets were selected pseudo-randomly so that every set of eight consecutive trials (one epoch) included all eight target positions.
Participants controlled the green cursor on the screen by moving their right index finger across the table top (Fig. 1a). At the beginning of each trial, participants were asked to move their index finger to the start position and a target then appeared. Participants were instructed to make a fast ‘shooting’ movement through the target such that online corrections were effectively prevented. At the moment the cursor passed through the invisible boundary circle (an invisible circle centred on the starting position with an 8 cm radius), the cursor was hidden and the intersection point was marked with a static yellow square to denote the terminal (endpoint) error. In addition, a small square icon at the top of the screen changed colour based on movement speed.
If the movement was completed within 100–300 ms, then it remained white. If the movement was slower than 300 ms, then the box turned red (too slow). Importantly, the participants were reminded that spatial accuracy was the main goal of the task. After each trial, subjects moved back to the central start position, with the cursor only reappearing once they were within 2 cm of its location.
Visuomotor adaptation task. a Experimental set up; participants sat behind a table facing a vertically orientated screen placed 105 cm in front of them. b Task protocol: Following 2 baseline blocks (each 96 trials: pre 1–2), an abrupt 30° VR was applied to the screen cursor and was maintained across 3 blocks (adapt 1–3). Cerebellar tDCS (anodal/sham) was applied from pre 2 until adapt 3 (pink). Following this, retention was examined by removing visual feedback (grey) for the final 3 blocks (post 1–3)
## Visuomotor adaptation
*In the **Visuomotor adaptation** section:*
The aim of the behavioural experiment was to replicate the findings of Galea et al. (). Therefore, participants were exposed to 8 blocks of 96 trials (12 epochs of all 8 targets). The first 2 blocks acted as baseline and consisted of veridical feedback with (pre1) and without (pre2) online visual feedback (Fig. 1b). During the no visual feedback trials, participants were instructed to continue to strike through the visible target, but received no visual feedback either during or at the end of their movement. Following this, participants were exposed to 3 blocks of trials (adapt 1–3) in which an abrupt 30° counter clockwise (CCW) visual rotation was applied.
Finally, to assess retention, three blocks (post-1–3) were performed without visual feedback. tDCS was applied from the start of pre2 and throughout the adaptation blocks, lasting 25 min (Fig. 1b).
## MRS acquisition
*In the **MRS acquisition** section:*
The anodal group also participated in a MRS session in which data was acquired pre-, during and post-25 min of cerebellar tDCS (Fig. 2) on a Philips Achieva 3T system (Philips Medical Systems, Best, The Netherlands) with a 32-channel radio frequency head receive-coil. The aim of this session was to measure tDCS-induced changes in GABA and Glu concentrations within the cerebellum. Three orthogonal T2-weighted localiser scans (34 slices, 4 mm thickness, and 1 mm gap, voxel size = 0.8 mm × 1.1 mm, 40 s duration) were collected to allow precise manual localisation of the 2 cm × 2 cm × 2 cm MRS single voxel in the posterior part of the cerebellum underneath the electrode.
A high-resolution T1-weighted had been acquired in a different session (sagittal, 175 slices, voxel size 1 × 1 × 1 mm, TR/TE = 8.4/3.8 ms, NSA = 1, 10.40 min duration). A cod liver oil capsule was placed on the top right corner of the electrode. As this could be seen in the localizer images, it was used as a marker to aid the placement of the MRS voxel (Fig. 3a).[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Graphical representation of MRS session using voxel localiser scans (T2) and MEGA-PRESS pulse sequence. MRS data was acquired pre-, during, and post-tDCS (lasting 25 min each) performed sequentially within the same individually localised voxel
MRS voxel localisation. a A single 2 × 2 × 2 cm voxel size was located manually in the posterior part of the right cerebellum underneath the anodal electrode. A cod liver oil capsule (yellow arrow) was situated at the top left edge of the electrode to assist with voxel localisation. Three sets of data were acquired: pre-, during and post- cerebellar tDCS; example MRS spectra are shown for three participants including N-acetylaspartate (NAA) and highlighting the (b) GABA and c GLX metabolite signals[](https://www.ncbi.nlm.nih.gov/mesh/C000179)
A GABA signal was measured from the proton spin coherence resonance at 3.0 ppm, accomplished by J-difference editing after scanning using a MEscher–GArwood-Point RESolved Spectroscopy (MEGA-PRESS) (Mescher et al. ) sequence with a pulse repetition time (TR) of 2000 ms, echo time (TE) of 68 ms and total duration 25 min. We produced an average GABA spectrum from a total of 512 spectral acquisitions each with a bandwidth of 2150 Hz, sampled at 2048 data points, and with prior water suppression using variable power radio-frequency pulses with optimized relaxation delays (VAPOR) (Tkac et al. ) at 4.68 ppm.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
To achieve an edited GABA spectral signal without contamination from macromolecules (MM), the two frequency selective 180° RF pulses (Gaussian pulses with duration of 16.5 ms) in the MEGA-PRESS sequence were applied with the centre of the frequency band interleaving between 1.9 ppm (edit-On) and 1.5 ppm (edit-Off) (Henry et al. ), across the 512 spectral acquisitions. The edit-Off spectra were subtracted from the edit-On spectra resulting in a spectrum with an unequivocal GABA signal. The acquired edit-Off spectra were also separately analysed to measure concentrations of other metabolites including GLX (Glu + Glutamine (Gln)). Typical spectra identifying GABA and GLX from three participants are shown in Fig.
3b, c. Additional unsuppressed water scans were also acquired to allow corrected metabolite signal quantification. Both metabolites were expressed relative to water concentration.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
This study required three separate scans to measure GABA pre-, during- and post- tDCS in a single voxel. To examine the temporal stability and reproducibility of the GABA signal measurements in three subsequent scans, we carried out three test scans on a phantom containing 18 mM of GABA. We found the GABA signal to be highly consistent across the scans. All spectra were aligned and the measured concentration from all three scans were similar: GABA:H2O = mean ± standard deviation (stdev) = (1.2 ± 0.11) × 10−3. The small stdev confirms the stability of our GABA measurements during in vitro conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Data analysis
*In the **Data analysis** section:*
## Visuomotor adaptation task
*In the **Visuomotor adaptation task** section:*
Data and statistical analysis was performed using MATLAB (The Math Works, USA) and SPSS (IBM, USA). Index finger position (X & Y position) data was collected at 120 Hz. For each trial, angular hand direction (°) was calculated as the difference between the angular hand position and angular target position at the point when the cursor intersected the 8 cm invisible circle centred on the starting position. During veridical feedback (pre1, Fig. 1b), the goal was for hand direction error to be 0°. However, with the visuomotor transformation (adapt 1–3), hand direction had to compensate; that is, for the − 30° (CCW) visuomotor rotation, a hand direction of + 30° relative to the target was required.
Positive values indicate a CW direction, whereas negative values indicate a CCW direction. In addition, reaction time (RT: difference between the target appearing and the participant moving out of the start position) and movement time (MT: difference between reaction time and movement end) were calculated for each trial. We removed any trial in which hand direction, RT or MT exceeded 2.5 standard deviations above the group mean. This accounted for 1.2% of trials. Epochs were created by binning 8 consecutive movements, 1 towards each target.
The angular hand direction (°) of anodal and sham groups was compared for each block of baseline using separate 2-tailed independent t tests. For adaptation and retention, separate repeated-measures ANOVAs compared groups (anodal/sham) across blocks (3). Finally, for reaction time (RT) and movement time, two separate repeated-measures ANOVAs compared groups (anodal/sham) across all 8 blocks (Pre 1–2, Adapt 1–3, Post 1–3). The threshold for all statistical comparisons was P < 0.05. Effect sizes are reported as partial eta squared for ANOVA and Cohen’s d for t tests.
All data are presented as mean ± standard error of the mean, unless otherwise specified.
## MRS analysis
*In the **MRS analysis** section:*
Spectroscopy data was analysed using TARQUIN version 4.3.4 (Wilson et al. ). First, pre-processing was carried out including inspection and removal of corrupted spectra arising from motion or technical problems. Then, raw data were Fourier-transformed to a spectrum of 2048 data points, the signal was smoothed by a 3 Hz Lorentzian filter, phased and referenced to water signal at 4.7 ppm. Random drift due to scanner instability or subject motion was corrected by aligning the water peak before fitting a Lorentzian–Gaussian (Voigt) line shape model. The amount of drift was plotted and used to assess the quality of acquisition.
Scans with less than 10 Hz drift were taken to have acceptable spectra. However, high drift was not the only criterion used to remove data; quality control was performed based on a flat baseline, the shape of the GABA peak in the average spectrum and the smoothness of the residual between the actual data and the fitted model. Signal to noise ratio (SNR) or Cramér–Rao bound (CRLB) were not recommended to be used as quality control in TARQUIN due to the small GABA signal SNR (according to TARQUIN forum discussions). As a result, four subjects were removed from analysis due to an unreliable spectrum and/or poor fitting in one of the three acquisitions (pre-, during, or post-tDCS).[](https://www.ncbi.nlm.nih.gov/mesh/D014867)
A basis set predefined in TARQUIN was initially constructed based on known peak positions (Voigt function). This basis set was fit to the average spectrum allowing peak amplitudes, widths, and frequencies to be optimized (Wilson et al. ). The basis set was then updated with the newly determined frequencies and peak widths and this process of basis set refinement was repeated until fitting resulted in negligible adjustment to the basis set. To detect GABA, all edit-On and edit-Off spectra were averaged separately and then subtracted from each other, but GLX (Glu + Gln) was measured from the average of edit off spectra and Glu extracted from GLX using the predefined basis set in TARQUIN. The reason for using edit off is to avoid subtraction artefacts from the misalignment of the edit-on and edit-off spectra (Evans et al. ).[](https://www.ncbi.nlm.nih.gov
/mesh/D005680)
Next, the T1-image of each participant was co-registered to their T2-image using Statistical Parametric Mapping (SPM12) (Friston et al. ) and the quality of registration was checked by plotting joint histograms of co-registered T1 vs. T2 images, and by inspection of land marks (specifically on the cerebellum). Then segmentation of the T1 image was carried out using the FMRIB automated segmentation tool (FAST) (Zhang et al. ) to calculate the relative volume of each tissue type; grey matter, white matter (WM) and cerebral spinal fluid (CSF) within the MRS voxel. The amplitude of GABA and Glu were corrected for the proportion of GM volume in the voxel by multiplying by (Stagg et al. ; Kim et al. ). Finally, the percentage change ratios for both metabolites for pre- versus during-tDCS, and pre- versus post-tDCS scans were calculated by (100 × (during-pre)/pre) and (100 × (post-pre)/pre) respectively (Stagg et al. ).[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
To assess the modulation of metabolites in response to cerebellar tDCS, repeated-measures ANOVAs compared concentrations of each metabolite pre-, during, and post- tDCS. A Bonferroni correction for multiple comparisons meant the threshold for statistical comparisons was set at P < 0.016. All data are presented as mean ± standard error of the mean, unless otherwise specified.
Finally, we examined whether changes in GABA and Glu could predict visuomotor performance. Therefore, partial correlations were carried out between: (1) the change in GABA:H2O ratio during tDCS with both total adaptation and retention; (2) the change in GABA:H2O ratio change post-tDCS and retention. In both cases, we controlled for Glu:H2O ratio change because Glu is precursor for GABA synthesis. Correlations were also carried out between (3) the change in Glu:H2O ratio during tDCS and total adaptation and retention, and (4) the change in Glu:H2O ratio change post-tDCS and retention. A Bonferroni correction for multiple comparisons meant the threshold for statistical comparisons was set at P < 0.008.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Results
*In the **Results** section:*
## Visuomotor adaptation (cont.)
*In the **Visuomotor adaptation (cont.)** section:*
The performance of 17 anodal and 17 sham participants were compared across all blocks. Both groups behaved similarly during baseline with no significant differences in hand direction between groups during either pre1 (anodal: 1.20 ± 0.22, sham 1.83 ± 0.32; t(32) = − 1.4, p = 0.1, d = 0.08; Fig. 4) or pre2 (anodal: 2.24 ± 0.33, sham: 1.53 ± 0.34; t(32) = 0.9, p = 0.4, d = 0.2). For adaptation, we found no significant differences between the anodal and sham groups. Specifically, there was a significant main effect for blocks (F(2,32)
= 205.6, p < 0.005, ɳ2 = 0.86), but no significant main effect for group (F(1, 32) = 2.3, p = 0.14, ɳ2 = 0.07) or block–group interaction (F(1,32) = 0.63, p = 0.43, ɳ2 = 0.02; Fig. 4). Based on these results (total adaptation: anodal = 20.84 SD = 2.3, sham = 19.44 SD = 2.98), a power analysis revealed (d = 0.53, power = 0.8) that group sizes of 45 participants would be required to observe a significant result. For retention, we found an unexpected difference between groups whereby the anodal group retained significantly more than the sham group. Specifically, there was a significant main effect for blocks (F(2,32) = 114.9, p < 0.005, ɳ2 = 0.78) and group (F(1,32) = 4.7, p = 0.037, ɳ2 = 0.13), but no significant block–group interaction (F(1,32) = 0.6, p = 0.44, ɳ2 = 0.02; Fig. 4).
For RT, there were no significant main effect for group (anodal: 0.43 ± 0.04, sham: 0.39 ± 0.05; F(1,32) = 2.02, p = 0.2, ɳ2 = 0.06), blocks (F(2,32) = 2.5, p = 0.1, ɳ2 = 0.07), or block–group interaction (F(1,32) = 1.2, p = 0.3, ɳ2 = 0.04). Similarly, for MT there were no significant main effect for group (anodal: 0.22 ± 0.08, sham: 0.24 ± 0.08, F(1,32) = 3.3, p = 0.08, ɳ2 = 0.09) or block–group interaction (F(1,32) = 0.4, p = 0.8, ɳ2 = 0.01), but a significant main effect for blocks (F(2,32) = 9.9, p < 0.005, ɳ2= 0.24).
Influence of cerebellar tDCS on visuomotor adaptation. Epoch data (average across 8 trials) for angular hand direction (˚) for the anodal (blue) and sham cerebellar tDCS groups. Positive values indicate CW hand direction. The inset bar graphs indicate mean hand direction for the anodal and sham groups during adaptation (adapt 1–3) and retention (post-1–3). Solid lines, mean; shaded areas/error bars, S.E.M
## MRS
*In the **MRS** section:*
## tDCS did not consistently modulate GABA or glu
*In the **tDCS did not consistently modulate GABA or glu** section:*
In the anodal group, we measured metabolites within the right posterior cerebellar cortex underneath the anodal electrode at three time-points: pre-, during and post-25 min of anodal cerebellar tDCS. First, grey matter tissue fraction was not significantly different across the three time-points (F(2,24) = 0.95, p = 0.4, ɳ2 = 0.07). Crucially, there was no significant change in either GABA:H20 (F(2,24) = 0.56, p = 0.58, ɳ2 = 0.04; Bonferroni-corrected threshold p = 0.016; Fig. 5a) or Glu:H2O (F(2,24)
= 4.2, p = 0.02, ɳ2 = 0.26; Fig. 5b) across the three time points.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Cerebellar tDCS dependent changes in GABA and Glu. The average a GABA:H2O and b Glu:H2O ratio pre-, during and post-cerebellar tDCS. Change (%) in c GABA:H2O; d Glu:H2O during and post-cerebellar tDCS relative to baseline (pre-tDCS). The box-plot limits represent the 25th and 75th data percentiles and the middle line represents the median. The error bars represent the range of data[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## tDCS-induced changes in Glu:H2O during tDCS were inversely correlated with retention
*In the **tDCS-induced changes in Glu:H2O during tDCS were inversely correlated with retention** section:*
Given the large between-subject variability (e.g. from ~ 90% increase to a 100% decrease for GABA:H2O and from ~ 20% decrease to ~ 40% increase for Glu:H2O during vs. pre; Fig. 5c, d), we went on to examine whether changes in GABA and Glu could predict visuomotor adaptation performance across participants. There was no significant correlation between the change in the GABA:H2O ratio during tDCS with total adaptation (r = − 0.40, p = 0.15; Bonferroni-corrected threshold p = 0.008, Fig. 6a) or total retention (r = − 0.19, p = 0.49), nor between the change in the GABA:H2O ratio post-tDCS with total retention (r = − 0.07, p = 0.81). In addition, there was no significant correlation between the change in the Glu:H2O ratio during tDCS with total adaptation (r = − 0.08, p = 0.78), but there was a significant correlation with total retention (r = − 0.74, p = 0.004, Fig. 6b). There was also no significant correlation between the change in the Glu:H2O ratio post-tDCS and total retention (r = − 0.29, p = 0.32).[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Finally, in a purely explorative nature we observed that cerebellar tDCS led to enhanced performance during the late phase (adapt 3) of adaptation. As previous work has suggested that this part of adaptation is more cerebellar-dependent (McDougle et al. ), we asked whether this performance was correlated with changes in either the GABA:H2O or Glu:H2O ratio during tDCS. There was a significant negative correlation between the change in GABA:H2O ratio during tDCS and late (adapt 3) adaptation (r = − 0.66, p = 0.014), but not in Glu:H20 (r = 0.20, p = 0.50). Although exploratory, this provides subtle evidence that participants who showed a decrease in GABA during cerebellar tDCS also displayed greater late adaptation.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Correlations between MRS and visuomotor adaptation. a There was no significant correlation between the changes in the GABA:H20 ratio during cerebellar tDCS and total adaptation. The red line represents the sham group’s mean performance during total adaptation (shaded area = SD across group). b A significant negative correlation was observed between changes in the Glu:H2O ratio during cerebellar tDCS and total retention. The red line represents the sham group’s mean performance during total retention (shaded area = SD across group)[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Self-reported ratings of attention, fatigue, and sleep
*In the **Self-reported ratings of attention, fatigue, and sleep** section:*
There were no significant differences between groups for the self-reported ratings of attention, fatigue and quality of sleep (Table 1).
Self-reported rate of attention, fatigue, quality of sleep (1 is poorest and 7 is the maximal), perceived tDCS as active (1) or sham (0) and sleep hours. All the values are averaged and compared using independent t tests between the groups, and presented as mean ± standard deviation (SD)
## Discussion
*In the **Discussion** section:*
This study revealed no statistically significant behavioural differences between anodal and sham cerebellar tDCS groups during visuomotor adaptation, and no consistent change in GABA and Glu in response to concurrent cerebellar tDCS. However, surprisingly, we found cerebellar tDCS led to an improvement in motor memory retention which was strongly correlated with a decrease in Glu during tDCS.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Cerebellar tDCS did not significantly improve visuomotor adaptation, but enhanced retention
*In the **Cerebellar tDCS did not significantly improve visuomotor adaptation, but enhanced retention** section:*
Although participants showed a clear ability to adapt to the novel visuomotor rotation, the expected significant enhancement of adaptation by anodal cerebellar tDCS, that had been shown in various studies (Galea et al. ; Hardwick and Celnik ; Block and Celnik ; Leow et al. ), was not observed here. Despite our sample size being in the same range of previously published tDCS papers, a recent study indicates this could be significantly under powered (Minarik et al. ). Minarik et al. () showed that with a suggested tDCS effect size of 0.45, the likelihood of observing a significant result with 14 participants per group was approximately 20%. In fact, a power analysis based on our results revealed that we achieved an effect size of 0.53, suggesting group sizes of 45 participants would have been required to observe a significant difference between the anodal and sham tDCS groups. In accordance with this, some previous work indicates that there is substantial variation in the behavioural effect of cerebellar tDCS across participants (Jalali et al. ).
Unexpectedly, the anodal group showed greater motor memory retention in comparison to sham tDCS. Although in force-field adaptation it has been shown that cerebellar tDCS influences both the formation of motor memory and its retention (Herzfeld et al. ), this effect of stimulation has not been previously shown in similar visuomotor adaptation tasks (Galea et al. ; Jalali et al. ). At present, we have no clear reason why we observed a positive effect of cerebellar tDCS on memory retention during a visuomotor adaptation task; however, we return to this question when discussing the strong correlation observed between retention and Glu across participants.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
## No significant detectable change in GABA or glu in response to cerebellar tDCS
*In the **No significant detectable change in GABA or glu in response to cerebellar tDCS** section:*
Similar to the behavioural results, there was no consistent group effect of tDCS on GABA or Glu measured within the cerebellum either during or after stimulation. This is in contrast to several previous studies that have shown a significant decrease in GABA in response to M1 tDCS (Stagg , Stagg et al. ; Stagg et al. ; Kim et al. ; Bachtiar et al. ), but similar to reports of no significant changes in Glu being observed following M1 tDCS (Stagg et al. ; Kim et al. ).
However, as this is a different brain region with different stimulation duration/intensity it is very difficult to make comparisons. It is possible that cerebellar tDCS simply does not cause consistent between-subject changes in GABA and Glu as it is also shown for M1 tDCS (Tremblay et al. ). Alternatively, as each MRS measurement represented the average of 25 min of acquisition we may have been unable to capture any fast or short-lasting changes in these metabolites.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## No correlation between changes in GABA and adaptation, but online cerebellar tDCS reductions in Glu were correlated with motor retention
*In the **No correlation between changes in GABA and adaptation, but online cerebellar tDCS reductions in Glu were correlated with motor retention** section:*
Although there were no consistent metabolite concentration changes during or post- cerebellar tDCS, we observed large inter-subject variability. Therefore, we examined whether changes in GABA and Glu could predict visuomotor performance with cerebellar tDCS. Our findings demonstrated no significant correlation across participants between changes in GABA during stimulation and total adaptation, however there was an, exploratory, significant negative correlation with the late phase of adaptation. This latter result might confirm the finding by McDougle et al. , who reported that the late phase of adaptation is more cerebellar-dependant (McDougle et al. ). As this correlation was specific to GABA, and not Glu, it might suggest a role for GABA in the online effects of cerebellar tDCS
during visuomotor adaptation, however, further investigation is required. Specifically, during the current task it is likely that adaptation involved cerebellar-dependent sensorimotor recalibration but also the use of explicit strategies (Taylor et al. ). Importantly, it has recently been shown that cerebellar tDCS increases implicit learning only when strategic re-aiming is suppressed during adaptation (Leow et al. ). Therefore, it is possible that a stronger relationship between cerebellar tDCS changes in GABA and visuomotor adaptation performance would be observed when using a task that minimised the use of strategies.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Surprisingly, we also found that participants who showed decreases in Glu during cerebellar tDCS within the MRS session showed greater levels of motor memory retention post-tDCS during the behavioural session. At present, it is difficult to explain this correlation. One possibility is that a decrease in Glu reflects a decrease in glutamatergic input into the cerebellar cortex (from mossy fibres and/or granule cells) and would, therefore, lead to reduced activity of Purkinje cells. This would reduce cerebellar brain inhibition (CBI) and enhance M1 function. It is known that excitation of M1 facilitates retention, potentially retaining or consolidating what has been learnt by the cerebellum (Galea et al. ; Sami et al. ).
Although extremely interesting, as the positive effect of cerebellar tDCS on retention was unexpected and contrary to previous literature, we believe it is crucial that future work attempts to replicate the negative correlation between levels of Glu and motor retention.[](https://www.ncbi.nlm.nih.gov/mesh/D018698)
A major limitation of this study was the lack of a sham tDCS MRS session. As we compared changes in GABA and Glu across MRS scans, it is possible that inter-individual variability in these measures simply reflected either unreliable GABA quantification or natural variations in neurotransmitter levels at rest. Although our small phantom study suggested that we could measure GABA across three scans with little variability, this does not mean that our in vivo measurements of GABA and Glu did not suffer from inter-scan variability. Therefore, to confirm the significant correlation between the changes in Glu and memory retention, future work should include a sham tDCS condition which would enable tDCS-dependent changes in MRS signal to be dissociated from natural changes occurring at rest.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
Lastly, not having any measures of participant alertness during data acquisition leaves the possibility that some of the heterogeneity in the observed MRS results could have been driven by variability in the alertness of the participants while lying in the scanner even though according to their report, none of the participants fell asleep. MRS data from a sham tDCS condition would have been useful in assessing this possibility- revealing the GABA and Glu changes which might simply be associated with lying in the scanner for this kind of duration.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
## Conclusion
*In the **Conclusion** section:*
In summary, we found no statistically significant behavioural differences between anodal and sham cerebellar tDCS groups during visuomotor adaptation, and no consistent change in GABA and Glu in response to concurrent cerebellar tDCS. However, cerebellar tDCS did lead to an improvement in motor memory retention which was strongly correlated with a decrease in Glu during tDCS. Thus, this work provides novel insights regarding the neural mechanisms that could underlie cerebellar tDCS. Although interesting, these effects are incompatible with previous literature highlighting the need for replication and limitations in the ability to produce robust cerebellar tDCS effects across participants and between studies.[](https://www.ncbi.nlm.nih.gov/mesh/D005680)
# Compliance with ethical standards
*In the **Compliance with ethical standards** section:*
## Conflict of interest
*In the **Conflict of interest** section:*
Authors have no conflict of interest, financial or otherwise.
# References
*In the **References** section:*
|
# Introduction
Impact of the [NO](https://www.ncbi.nlm.nih.gov/mesh/D009569)-Sensitive Guanylyl Cyclase 1 and 2 on Renal Blood Flow and Systemic Blood Pressure in Mice
# Abstract
*In the **Abstract** section:*
Nitric oxide (NO) modulates renal blood flow (RBF) and kidney function and is involved in blood pressure (BP[) regulation](https://www.ncbi.nlm.nih.gov/mesh/D009569) p[re](https://www.ncbi.nlm.nih.gov/mesh/D009569)dominantly via stimulation of the NO-sensitive guanylyl cyclase (NO-GC), existing in two isoforms, NO-GC1 and NO-GC2. Here, we used isoform-sp[ec](https://www.ncbi.nlm.nih.gov/mesh/D009569)ific knockout (KO) mice and i[nv](https://www.ncbi.nlm.nih.gov/mesh/D009569)estigated their contribution to renal hemod[yn](https://www.ncbi.nlm.nih.gov/mesh/D009569)amics under normotensive and angiotensin II-induced hypertensive conditions. Stimulation of the NO-GCs by S-nitrosoglutathione (GSNO) reduced BP in normotensive and hypertensive wildtype (WT) and NO-GC2-K[O ](https://www.ncbi.nlm.nih.gov/mesh/D009569)mice mor[e efficiently than i](https://www.ncbi.nlm.nih.gov/mesh/D026422)n [NO-G](https://www.ncbi.nlm.nih.gov/mesh/D026422)C1-KO. NO-induced increase of RBF in normotensive mice did not d[if](https://www.ncbi.nlm.nih.gov/mesh/D009569)fer between the genotypes, but the res[pe](https://www.ncbi.nlm.nih.gov/mesh/D009569)ctive increase under hypertensive conditions was impaired in NO-GC1-KO. Similarly, inhibition of endogenous NO increased BP and reduced RBF to a lesser extent in NO-GC1-[KO](https://www.ncbi.nlm.nih.gov/mesh/D009569) than in NO-GC2-KO. These findings indicate N[O-](https://www.ncbi.nlm.nih.gov/mesh/D009569)GC1 as a target of NO to normalize RBF in hypertensi[on](https://www.ncbi.nlm.nih.gov/mesh/D009569). As these effec[ts](https://www.ncbi.nlm.nih.gov/mesh/D009569) were not completely abolished in[ N](https://www.ncbi.nlm.nih.gov/mesh/D009569)O-GC1-KO and renal c[yc](https://www.ncbi.nlm.nih.gov/mesh/D009569)lic guanosine monophosphate (cGMP) levels were decreased in both NO-GC1-KO and NO-GC2[-K](https://www.ncbi.nlm.nih.gov/mesh/D009569)O, the results sug[gest an additional contributio](https://www.ncbi.nlm.nih.gov/mesh/D006152)n [of N](https://www.ncbi.nlm.nih.gov/mesh/D006152)O-GC2. Hence, NO-GC1 plays a pre[do](https://www.ncbi.nlm.nih.gov/mesh/D009569)minant role [in](https://www.ncbi.nlm.nih.gov/mesh/D009569) the regulation of BP and RBF, especially in hypertension. [Ho](https://www.ncbi.nlm.nih.gov/mesh/D009569)wever, renal [NO](https://www.ncbi.nlm.nih.gov/mesh/D009569)-GC2 appears to compensate the loss of NO-GC1, and is able to regulate renal hemodynamics under physiologi[ca](https://www.ncbi.nlm.nih.gov/mesh/D009569)l conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 1. Introduction
*In the **1. Introduction** section:*
To ensure renal function, renal blood flow (RBF) is kept constant during a wide range of systemic blood pressure levels due to adjustment of renal vascular resistance. An important regulator of vascular resistance is the nitric oxide (NO)/cyclic guanosine monophosphate (cGMP) signaling cascade that acts in smooth muscle cells and blunts the effects of vasoconstrictor factors. A key enzyme in this pathway is the NO-sensitive guanylyl cyclase (NO-GC), which is stimulated by binding of NO to its prosthetic heme group and catalyzes, subsequently, the synthesis of the second messenger cGMP.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
The NO-GC exists in two distinct isoforms, the NO-GC1 that corresponds to the α1β1 heterodimer and the NO-GC2 that refers to the α2β1 enzyme. Since isoform-selective inhibitors do not exist, studies in knockout (KO) mice lacking the NO-GC1 or the NO-GC2 isoform are essential to elucidate their physiological roles. Previous studies have suggested that the NO-GC1 is the predominant isoform in the systemic vasculature. The NO-GC2 plays a less important role as it accounts for a very low portion of total cGMP formation induced by NO (about 5% in the aorta). Yet, NO-GC2 is able to contribute to vascular relaxation induced by endothelial NO, as shown in NO-GC1-KO mice.
Additional deletion of NO-GC2 in smooth muscle cells resulted in hypertension by reducing vascular responsiveness to NO, which confirms the importance of NO-GC2 as a target for endothelial NO in the systemic vasculature.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
In the kidney, NO modulates vascular resistance and has a profound impact on autoregulation of renal blood flow. Deficiency of NO causes dysfunction of renal blood flow and promotes renal failure and hypertension. In this context, a recent study has shown that activation of NO-GC by cinaciguat under NO-deficient conditions can normalize blood pressure and reduce renal vasoconstriction to improve renal blood flow autoregulation. In addition, other studies indicate NO-GC as an emerging target for the treatment of renal disorders.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Similarly, inhibition of the cGMP-degrading phosphodiesterase 5 (PDE5) enhances cGMP levels, which results in an increase in RBF, improved renal vascular function, and lower blood pressure in experimental models of hypertension.[](https://www.ncbi.nlm.nih.gov/mesh/D006152)
In order to get more insights into how the NO/cGMP pathway regulates renal hemodynamics, we investigated the relative contribution of the NO-GC1 and NO-GC2 isoforms on renal blood flow, renal vascular function, and blood pressure regulation by using NO-GC1-KO and NO-GC2-KO mice.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 2. Results
*In the **2. Results** section:*
## 2.1. Renal Blood Flow under Basal Conditions
*In the **2.1. Renal Blood Flow under Basal Conditions** section:*
In order to investigate the impact of NO-GC isoforms on renal hemodynamics, we measured renal blood flow in unconscious WT, NO-GC1-KO, and NO-GC2-KO mice. Interestingly, under anesthesia, invasively measured mean arterial blood pressure (MAP) was significantly increased in both NO-GC1-KO (86 ± 2 mmHg) and NO-GC2-KO (84 ± 2 mmHg) compared to WT mice (77 ± 1 mmHg) indicating that both isoforms contribute to systemic vascular resistance (Figure 1A). Nevertheless, renal blood flow was unaltered in NO-GC1-KO (0.51 ± 0.04 mL/min) and NO-GC2-KO (0.53 ± 0.03 mL/min) compared to WT (0.59 ± 0.03 mL/min) mice (Figure 1B). These observations suggest preserved renal blood flow autoregulation in NO-GC1-KO and NO-GC2-KO mice under basal conditions.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 2.2. Effects of NO-sensitive guanylyl cyclase (NO-GC)1 and NO-GC2 on Renal Blood Flow in Response to NO Stimulation
*In the **2.2. Effects of NO-sensitive guanylyl cyclase (NO-GC)1 and NO-GC2 on Renal Blood Flow in Response to NO Stimulation** section:*
To further investigate the role of NO-GC1 and NO-GC2 in renal hemodynamics, we applied S-nitrosoglutathione (GSNO) (0.01–0.1–1.0 mg/kg body weight (BW)) intravenously, and measured changes in blood pressure and renal blood flow under baseline conditions. GSNO induced a concentration dependent decrease in blood pressure in WT, NO-GC1-KO, and NO-GC2-KO mice (Figure 2A,B). The highest GSNO concentration decreased blood pressure to hypotensive values in all three groups (WT: 49 ± 4; NO-GC-1-KO: 58 ± 4; NO-GC2-KO 43 ± 4 mmHg). GSNO-induced stimulation of the NO-GC2, the minor isoform in the vasculature, resulted in an attenuated blood pressure reduction in NO-GC1-KO mice compared to WT and NO-GC2-KO mice (Figure 2A,B). Interestingly, GSNO-induced increase of renal blood flow did not differ between the three groups and did not further increase at the highest GSNO concentration, most likely due to hypotension (Figure 2C).
To exclude any pathophysiological effects of hypotension and to unmask the impact of the NO-GC isoforms on renal blood flow, we tested the effect of GSNO in mice infused with angiotensin II. During angiotensin II infusion, blood pressures increased, and did not differ between the three groups (WT: 134 ± 7; NO-GC1-KO: 150 ± 7; NO-GC2-KO: 136 ± 3 mmHg; Figure 2D). Similar to the effects described above, GSNO-induced (1 mg/kg BW) stimulation of both NO-GC isoforms in WT or NO-GC1 in NO-GC2-KO mice resulted in a significantly greater blood pressure reduction compared to GSNO-induced NO-GC2 stimulation in NO-GC1-KO mice (WT: 53 ± 3; NO-GC1-KO: 34 ± 2; NO-GC2-KO: 53 ± 4 mmHg; Figure 2E). In accordance with attenuated blood pressure reduction in NO-GC1-KO mice, GSNO-induced increase of renal blood flow was significantly smaller in mice lacking the NO-GC1 isoform, suggesting a major role of the NO-GC1 isoform in renal blood flow regulation (Figure 2F).[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 2.3. Inhibition of NO Production by L-NAME Affects Renal Blood Flow in NO-GC1-KO and NO-GC2-KO Mice
*In the **2.3. Inhibition of NO Production by L-NAME Affects Renal Blood Flow in NO-GC1-KO and NO-GC2-KO Mice** section:*
Infusions of the nitric oxide synthase inhibitor N(G)-Nitro-l-arginine methyl ester (L-NAME) (30 mg/kg BW, iv) were administrated to inhibit endogenous NO production. L-NAME increased MAP to a level that did not significantly differ between the three genotypes (WT: 108 ± 3; NO-GC1-KO: 104 ± 6; NO-GC2-KO: 114 ± 3 mmHg; Figure 3A). Yet, the increase of MAP induced by L-NAME was significant attenuated in NO-GC1-KO mice compared to WT and NO-GC2-KO mice (Figure 3B). Concomitant to the increase in blood pressure, renal blood flow decreased in the presence of L-NAME in all three genotypes (Figure 3C,D). These findings indicate that both NO-GC1 and NO-GC2 mediate the effects of endogenous NO on renal blood flow.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 2.4. Contribution of NO-GC1 and NO-GC2 to Renal Vascular Relaxation
*In the **2.4. Contribution of NO-GC1 and NO-GC2 to Renal Vascular Relaxation** section:*
To determine the relative contribution of NO-GC1 and NO-GC2 to renal vascular resistance independent from systemic blood pressure changes, we measured vasorelaxation ex vivo in isolated perfused kidneys. To characterize the contribution of the NO-GC isoforms in the signaling of endogenous NO, carbachol was applied at a maximally effective concentration (30 µM). Vasorelaxation induced by carbachol was significantly reduced in kidneys of NO-GC1-KO mice compared to WT and NO-GC2-KO mice (Figure 4A).[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Similarly, renal vasorelaxation in response to exogenous NO, GSNO, was significantly attenuated in NO-GC1-KO compared to WT and NO-GC2-KO mice. In addition, GSNO induced renal vasorelaxation was also reduced in NO-GC2-KO compared WT mice (Figure 4B). Our findings suggest that NO-GC1 is the major target of NO in renal vasculature, but also demonstrate a substantial role of NO-GC2 in renal vasorelaxation. To determine if cGMP produced by the membrane-bound guanylyl cyclase GC-A compensates, in part, to the reduced renal vasorelaxation towards NO, ANP-induced renal relaxation was examined in all three genotypes. As shown in Figure 4C, renal vasorelaxation in response to ANP did not differ between WT, NO-GC1-KO, and NO-GC2-KO mice.
This finding indicates that no quantifiable compensation takes place on the level of GC-A or of downstream cGMP effectors.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 2.5. NO-Stimulated cGMP Formation in Kidney Homogenates and Cortical Slices
*In the **2.5. NO-Stimulated cGMP Formation in Kidney Homogenates and Cortical Slices** section:*
To assess the amount of the NO-GCs in the kidney, we measured NO-stimulated GC activity in whole kidney homogenates of WT and KO mice deficient of NO-GC1 or NO-GC2. In the NO-GC1-KO mice, the residual NO-GC activity is due to NO-GC2 content and amounted to 20% of the NO-GC activity in WT (Figure 4D). Yet, the portion of NO-GC2 was too low to detect a reduction of the NO-GC content in the NO-GC2-KO kidneys.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
We further studied the contribution of the NO-GCs to cGMP formation by measuring cGMP formation in renal cortical slices (Figure 4E). In untreated conditions, cGMP formation was found to be reduced in both NO-GC-KO lines by about 50% compared to WT (NO-GC1-KO: 0.26 ± 0.03; NO-GC2-KO: 0.30 ± 0.06; WT: 0.49 ± 0.06 pmol cGMP/mg protein). In response to carbachol (30 µM, 3 min), which induces eNOS-mediated NO formation, cGMP levels increased, but remained significantly reduced in the renal cortical slices of NO-GC1-KO and NO-GC2 KO mice compared to WT mice (NO-GC1-KO: 0.5 ± 0.09; NO-GC2-KO: 1.7 ± 0.46; WT: 3.5 ± 0.43 pmol cGMP/mg protein). Accordingly, administration of exogenous NO by the NO donor DEA-NO (100 µM, 3 min) induced an attenuated cGMP increase in renal slices of NO-GC1-KO and NO-GC2-KO mice compared to WT (NO-GC1 KO: 3 ± 0.56; NO-GC2 KO: 5.1 ± 0.74; WT: 8.4 ± 1.22 pmol cGMP/mg protein) mice. In summary, these experiments demonstrate the contribution of both NO-GC isoforms to cGMP formation in response to NO.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 2.6. Blood Pressure Monitoring in Freely Moving NO-GC1-KO and NO-GC2-KO Mice
*In the **2.6. Blood Pressure Monitoring in Freely Moving NO-GC1-KO and NO-GC2-KO Mice** section:*
Finally, we studied the role of NO-GC1 and NO-GC2 in blood pressure regulation via radio telemetry in WT and KO mice. Baseline blood pressure did not differ between freely moving WT, NO-GC1-KO, and NO-GC2-KO mice (127 ± 3 vs. 132 ± 4 vs. 123 ± 3 mmHg, n = 6). Yet, comparable to the results seen in anesthetized mice, blood pressure reduction induced by the NO donor sodium nitroprusside (SNP) was significantly diminished in conscious NO-GC1-KO mice compared to WT and NO-GC2-KO (Figure 5A). This finding confirms NO-GC1 as the major NO-GC isoform mediating the vasodilatory effect of NO. In contrast to SNP, which acts by stimulating both NO-GCs, the general NOS inhibitor, L-NAME, acts by offsetting NO-GC activity.
In contrast to the results with SNP, NOS inhibition by L-NAME increased blood pressure in all three genotypes (Figure 5B). Interestingly, the maximal blood pressure effect induced by L-NAME treatment was significantly smaller in NO-GC1-KO than in WT and NO-GC2-KO mice. These differences were not observed when the ganglionic blocker hexamethonium (30 mg/kg BW) has been co-administrated. This suggests that NO-GC1 and NO-GC2 regulate blood pressure not only through vascular but most probably also through central nervous effects (Figure 5C,D).[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 3. Discussion
*In the **3. Discussion** section:*
In the present study, we examined the impact of NO-GC1 and NO-GC2 on renal vascular function and blood pressure regulation. NO-GC is the main target mediating the modulatory effects of nitric oxide in the kidney. Recent studies demonstrated that the NO-GC plays an important role in the pathogenesis of chronic kidney disease through both hemodynamic and direct pro-fibrotic effects. However, the two distinct NO-GC isoforms have indistinguishable enzymatic properties and isoform-specific inhibitors are lacking. Therefore, not much attention has been paid to the role of NO-GCs in renal hemodynamics.
In contrast, many pharmacological or genetic approaches have been used to study the different NO synthase isoforms, eNOS, nNOS, or iNOS in kidney. By analyzing KO mice deficient for each of the NO-GC isoforms, we were able to demonstrate a predominant role of the NO-GC1 isoform in regulating renal hemodynamics. In addition, we could also show a relative contribution of NO-GC2 to the regulation of blood pressure and RBF under normotensive condition.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
In agreement with our previous study, NO-GC1 was found to be the major NO-GC isoform expressed in the kidney, accounting for about 80% of the NO-stimulated GC activity. In relation to NO-GC1, the amount of NO-GC2 appears negligible. Accordantly, NO-GC activity measured in kidneys of NO-GC2-KO mice did not differ from WT. Nevertheless, here we demonstrate that cGMP levels in renal cortical slices were reduced in both KO lines under baseline and NO-stimulated conditions. These results suggest a participation of both isoforms in the NO/cGMP signaling in the kidney.
In line with this observation, both NO-GC isoforms were found to participate in NO-dependent vasorelaxation in isolated perfused kidneys. Yet, the NO-GC2 was not sufficient to fulfil a WT-like effect, most likely due to a lower cGMP formation. Taken together, our data demonstrate the participation of both NO-GC isoforms in mediating the renal vasodilator action of endothelial NO.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Previously, it has been reported that an enhanced vasorelaxation mediated by the membrane-bound GC-A in response to ANP was able to partly compensate the reduced vasorelaxation in aortas of NO-GC1-KO mice. In the present study, ANP-induced renal vasorelaxation did not differ between WT and KO mice, and we excluded any differences on GC-A levels or downstream to cGMP formation in kidneys of NO-GC1-KO or NO-GC2-KO mice.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Despite the influence of both NO-GCs on renal vasorelaxation, baseline renal blood flow was neither altered in NO-GC1-KO nor in NO-GC2-KO mice. By contrast, blood pressure was increased in both KO lines under anesthesia. This observation indicates a sufficient autoregulatory response in the absence of either NO-GC1 or NO-GC2, and is in accordance with previous studies showing normal baseline blood flow despite hypertension in eNOS KO mice.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
To directly address the contribution of the NO-GC isoforms to blood pressure and renal blood flow, we stimulated the NO-GCs by applying exogenous NO (GSNO). Stimulation of the NO-GCs by GSNO induced a prominent blood pressure reduction in WT and NO-GC2-KO mice, which was attenuated in NO-GC1-KO. The lacking effect of NO to efficiently lower blood pressure in NO-GC1-KO mice is consistent with a lower content of NO-GC2 in the systemic vasculature, and consequently, lower increase of cGMP. The prominent vasodilating effect of NO induced an increase of renal blood flow which did not differ between the three genotypes. This finding suggests that an increase in renal blood flow under normotensive or hypotensive conditions is most likely mediated by a very low amount of cGMP, which can be supplied interchangeably either by NO-GC1 or by NO-GC2.
Similarly, application of the NO-GC activator cinaciguat (BAY 58-2667), which induces a rather low cGMP increase, decreases BP without affecting renal blood flow in healthy rats.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Under hypertensive conditions induced by Ang II infusion, it becomes more obvious that the cGMP increase by the NO-GC2 is not capable to lower BP to the same extent as the cGMP increase induced by the NO-GC1. Accordingly, NO-GC2 failed to increase renal blood flow, as shown in hypertensive NO-GC1-KO mice. In comparison, NO led to a similar increase of renal blood flow in hypertensive WT and NO-GC2-KO mice. Thus, in hypertensive conditions, cGMP-formed by the NO-GC1 is needed to regulate BP and renal blood flow.[](https://www.ncbi.nlm.nih.gov/mesh/D006152)
The contribution of the NO-GC isoforms mediating the effects of endothelial NO in mice was examined by inhibition of endogenous NO production by L-NAME. As demonstrated in many studies, L-NAME increases blood pressure, and yet lowers renal blood flow by shifting the balance of the regulatory mechanisms toward a stronger vasoconstriction. In the eNOS-KO mice, L-NAME did not alter blood pressure or baseline blood flow, indicating that the presence of eNOS is mandatory for the L-NAME effects. Compared to this observation, L-NAME alters blood pressure and renal blood flow in NO-GC1-KO and NO-GC2-KO mice by inhibiting NO stimulation of the residual isoform (NO-GC2 or NO-GC1, respectively). Thus, the effects obtained with L-NAME in our KO mice reveal the contribution of the residual NO-GC isoform in the regulation of blood pressure and renal blood flow.
Inhibition of NO production increased blood pressure and reduced renal blood flow to a lower extent in NO-GC1-KO mice compared to WT and NO-GC2-KO mice.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Taken together, our study shows that NO-GC1 is the predominant target of endothelial NO in the systemic and renal vasculature. Nevertheless, we could also uncover a substantial contribution of the NO-GC2 to mediating NO effects on blood pressure and renal blood flow in vivo. Thus, the present study highlights the NO-GC as a master regulator of renal hemodynamics and blood pressure and therefore as an interesting new drug target in hypertension and kidney disease.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 4. Material and Methods
*In the **4. Material and Methods** section:*
## 4.1. Animal Models
*In the **4.1. Animal Models** section:*
Experiments were performed with male wild-type C57Bl/6J, NO-GC1-KO, and NO-GC2-KO mice (10–14 weeks old). The KO mice were generated and genotyped as described previously. Mice experiments were approved by the responsible authority (Landesamt fuer Natur-, Umwelt- und Verbraucherschutz Nordrhein-Westfalen; reference: 87-51.04.2010.A039 (01.05.2010) and 8.87-50.10.34.08.216 (01.11.2008) and performed according to the guidelines from Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 4.2. Acute Blood Pressure Response and Changes in Renal Blood Flow
*In the **4.2. Acute Blood Pressure Response and Changes in Renal Blood Flow** section:*
Acute pressor responses and renal blood flow to L-NAME and GSNO were measured in anesthetized WT, NO-GC1-KO, or NO-GC2-KO mice, as described previously. In brief, mice were anesthetized intraperitoneally (ip) with ketamine (100 mg·kg−1) and xylazine (5 mg·kg−1), and decapitated at the end of the experiment. Mean arterial blood pressure was monitored continuously through a catheter placed in the right common carotid artery. Basal fluids, GSNO and L-NAME were administered via a catheter placed in the right jugular vein. Ang II (200 ng/kg/min) was applied continuously throughout the experiment via a second catheter placed in the left jugular vein.
For measuring RBF, a small incision was made on the left flank and the left renal artery was dissected. An ultrasonic flowmeter interfaced with a 5 mm V-shaped probe was then placed around the left renal artery (MA0.5PSB and TS420 Flowmeter, Transonic Systems Inc., Ithaca, NY, USA). After a stabilization period of 30 min, the effects of GSNO and L-NAME were tested under normotensive conditions. Therefore, GSNO was administrated in increasing doses (0.01, 0.1, 1.0 mg/kg BW) at 5 min intervals. L-NAME was administrated at a dose of 0.03 mg/kg BW. Intra-arterial pressure and renal blood flow were monitored continuously using the PowerLab data acquisition system and LabChart software (ADInstruments, Colorado Springs, CO, USA).
Changes in BP or RBF were recorded as the delta of BP- or RBF-increase or decrease in relation to their baseline values determined before the application of either GSNO or L-NAME.[](https://www.ncbi.nlm.nih.gov/mesh/D019331)
## 4.3. Isolated Perfused Kidneys
*In the **4.3. Isolated Perfused Kidneys** section:*
Kidneys of WT, NO-GC1-KO, or NO-GC2-KO were isolated and perfused with Krebs–Henseleit buffer as described previously. Changes in perfusion pressure reflected changes in vascular resistance of renal vessels. Immediately after preparation, a bolus of 60 mM KCl was injected to test the viability of the preparation followed by a stabilization period of 30 min. To assess renal vasodilation, kidneys were preconstricted with norepinephrine (1 µM; Sigma Aldrich, Taufkirchen, Germany). Concentration–response curves induced by GSNO (Alexis Corp., Enzo Life Sciences AG, Lausen, Germany) and ANP (Sigma Aldrich) were recorded in presence of diclofenac (3 µM; Sigma Aldrich) and L-NAME (300 µM; Sigma Aldrich).
Vasodilation induced by carbachol (30 µM) was tested in the presence of diclofenac (3 µM). Renal relaxation is expressed as a percentage pressor response of the preconstricted kidney, which was set as 100%.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 4.4. Blood Pressure Measurements in Freely Moving WT, NO-GC1-KO, and NO-GC2-KO Mice
*In the **4.4. Blood Pressure Measurements in Freely Moving WT, NO-GC1-KO, and NO-GC2-KO Mice** section:*
In order to measure the blood pressure effects of the NO donor, SNP, or the NO synthase inhibitor, L-NAME, in conscious unrestrained mice, radio telemetry catheters (Data Sciences International, PA-C10, s’Hertogenbosch, The Netherlands) were implanted as described previously. For catheter implantation, mice were anesthetized intraperitoneally with ketamine and xylazine (100 and 10 mg/kg, respectively), and the left common carotid artery was dissected. The artery was cannulated, and the catheter was advanced to the point where the small notch on the tubing resided at the vessel opening. Finally, the catheter was fixed, and the transmitter placed subcutaneously. After radio telemetry catheter implantation, mice were allowed to recover for seven days to re-establish normal circadian rhythms.
For habituation, before the experiment, mice were trained daily (ip. injection of an equal amount of saline) for three consecutive days. Additionally, ip. injections were always performed between 8.00 and 10.00 a.m. Blood pressure levels were recorded continuously with measuring every 20 min for 10 s intervals. Thirty minutes before and 2 h after administration of SNP or L-NAME, blood pressure levels were recorded every 20 s for 10 s intervals. SNP (30 µg/kg BW), L-NAME (50 mg/kg BW), or a combination of hexamethonium (30 mg/kg BW) and SNP (30 µg/kg BW) or L-NAME (30 mg/kg BW) were administrated ip.
once per mouse every 24 h for three days. The mean of three measurements per mouse was analyzed.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 4.5. Measurement of cGMP Content and NO-Stimulated GC-Activity in Renal Cortical Slices
*In the **4.5. Measurement of cGMP Content and NO-Stimulated GC-Activity in Renal Cortical Slices** section:*
NO-stimulated GC activity was determined in kidney homogenates in the presence of 100 µM DEA-NO (2-(N,N-diethylamino)-diazenolate-2-oxide, Alexis Corp.), as described previously. For measuring cGMP changes ex vivo, cortical slices (250 µm) were cut with a vibratome (NVSLM1 from WPI) and equilibrated for 10 min in tempered (37 °C), oxygenated (with 95% O2, 5% CO2) Krebs–Henseleit buffer as described previously. To increase cGMP, cortical slides were incubated with carbachol (30 µM) or DEA-NO (100 µM) for 3 min. cGMP levels of equilibrated untreated slices were taken as baseline measurements. After incubation, slices were snap frozen in liquid nitrogen, homogenized in 70% ice-cold ethanol using a glass/glass homogenizer, and then centrifuged (20,000× g, 15 min, 4 °C). Supernatants were dried at 95 °C and the cGMP content was measured in duplicate by Radioimmunoassay (RIA).
Protein content was determined in pellets used for standardizing the different samples.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
## 4.6. Statistical Analysis
*In the **4.6. Statistical Analysis** section:*
All data are expressed as mean ± SEM (n = number of animals). Student’s t-test was used to compare means of two groups with Gaussian distribution. Differences between dose–response curves were analyzed by one-way or two-way ANOVA for repeated measurements, followed by Bonferroni’s multiple comparison post hoc test. Data of two groups with no Gaussian distribution were analyzed by the Mann–Whitney U test. Probability levels of p < 0.05 were considered statistically significant.
# Author Contributions
*In the **Author Contributions** section:*
Evanthia Mergia, Manuel Thieme, Henning Hoch, Mina Yakoub, Georgios Daniil and Lydia Hering performed experiments. Evanthia Mergia, Manuel Thieme, Henning Hoch and Johannes Stegbauer analyzed the data. Evanthia Mergia and Johannes Stegbauer wrote the manuscript. Doris Koesling, Lars Christian Rump and Christina Rebecca Scherbaum revised the manuscript. Johannes Stegbauer and Evanthia Mergia planed the study.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# Abbreviations
*In the **Abbreviations** section:*
ANP atrial natriuretic peptide BP blood pressure BW body weight cGMP cyclic guanosine monophosphate DEA-NO DEA NONOate GSNO S-nitrosoglutathione L-NAME N(G)-Nitro-L-arginine methyl ester MAP mean arterial blood pressure NO nitric oxide NO-GC NO-sensitive guanylyl cyclase (synonym: sGC) RBF renal blood flow SEM standard error of the mean SNP sodium nitroprusside[](https://www.ncbi.nlm.nih.gov/mesh/D006152)
# References
*In the **References** section:*
Increased blood pressure but normal renal blood flow in anesthetized NO-sensitive guanylyl cyclase1 knockout (NO-GC1-KO) and NO-GC2-KO mice. (A) Blood pressure was significantly increased in NO-GC1-KO (n = 16) and NO-GC2-KO (n = 17) compared to WT (n = 54) mice; (B) Renal blood flow was not different in unconscious WT (n = 33), NO-GC1-KO (n = 16) and NO-GC2-KO (n = 16) mice. Data represent means ± standard error of the mean (SEM); * p < 0.05 vs. wildtype (WT). One-way standard error of the mean (ANOVA) followed by Bonferroni’s multiple comparison post hoc test.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Predominant role of NO-GC1 in GSNO-induced blood pressure reduction and renal blood flow increase in anesthetized mice. (A,B) In normotensive mice, GSNO induced blood pressure reduction, expressed as absolute blood pressure values and blood pressure reduction, was attenuated in NO-GC1-KO (n = 3) compared to WT (n = 8) and NO-GC2-KO (n = 5) mice; (C) In normotensive mice, GSNO-induced increase in renal blood flow did not differ between the three genotypes (n = 3–8); (D–F) In Ang II-infused mice (200 ng/kg/min), GSNO-induced blood pressure reduction, expressed as absolute blood pressure values and blood pressure reduction, and renal blood flow increase were significantly attenuated in NO-GC1-KO (n = 3) compared to WT (n = 5) and NO-GC2-KO (n = 4) mice. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. NO-GC1-KO; # p < 0.01, ## p < 0.001 vs. NO-GC2-KO. Two-way ANOVA followed by Bonferroni’s multiple comparison post hoc test.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Inhibition of NO production by L-NAME affects blood pressure and renal blood flow in NO-GC1-KO and NO-GC2-KO mice. (A) In anesthetized mice, administration of L-NAME increased blood pressure to the same level; (B) L-NAME-induced blood pressure increase was significantly attenuated in NO-GC1-KO (n = 5) and NO-GC2-KO (n = 7) compared to WT (n = 9) mice; (C) Renal blood flow was significantly decreased by L-NAME in WT (n = 7), NO-GC1 (n = 4) and NO-GC2-KO (n = 6) mice; (D) L-NAME-induced reduction in renal blood flow was significantly reduced in NO-GC1-KO compared to NO-GC2-KO mice (n = 4–7). * p < 0.05; # p < 0.01 vs. NO-GC2-KO. One-way ANOVA or two-way ANOVA followed by Bonferroni’s multiple comparison post hoc test.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Impaired NO-mediated renal vasorelaxation in NO-GC1-KO and NO-GC2-KO mice. (A) Endothelial-dependent renal vasorelaxation by carbachol (30 µM) was significantly reduced in isolated perfused kidneys of NO-GC1-KO compared to WT mice; (B) Smooth muscle cell-dependent vasorelaxation induced by the NO donor GSNO was significantly attenuated in isolated perfused kidneys of NO-GC1-KO (n = 8) and NO-GC2-KO (n = 13) compared to WT (n = 11) mice; (C) Atrial natriuretic peptide (ANP)-induced renal vasorelaxation did not differ between WT (n = 6), NO-GC1-KO (n = 10) and NO-GC2-KO (n = 5) mice; (D) Renal NO-GC activity was significantly decreased in NO-GC1-KO (n = 6) compared to WT (n = 19) and NO-GC2-KO (n = 7) mice; (E) Renal cGMP levels at baseline, in response to carbachol (30 µM) or DEA-NO (100 µM) were significantly decreased in NO-GC1-KO (10 slices of n = 3) and NO-GC2-KO (11 slices of n = 3) compared to WT (10 slices of n = 3) mice. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. WT; # p < 0.05, ## p < 0.01 vs. NO-GC1-KO-mice. Two-way ANOVA followed by Bonferroni’s multiple comparison post hoc test or one-way ANOVA followed by Bonferroni’s multiple comparison post hoc test.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
Predominant role of NO-GC1 in NO-mediated blood pressure changes in conscious mice. (A) Blood pressure reduction induced by the NO donor sodium nitroprusside (SNP) (30 µg/kg BW) was significantly diminished in conscious NO-GC1-KO (n = 4) mice compared to WT (n = 4) and NO-GC2-KO (n = 4) mice measured by radio telemetry; (B) Attenuated blood pressure increase by L-NAME (30 mg/kg BW) in NO-GC1-KO mice compared to WT and NO-GC2-KO mice; (C,D) Co-administration of hexamethonium (30 mg/kg BW) enhanced blood pressure reduction by GSNO and blood pressure increase by L-NAME in NO-GC1-KO (n = 6) compared to WT (n = 6) and NO-GC2-KO (n = 6) mice. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. WT; # p < 0.05, ## p < 0.01, ### p < 0.001 vs. NO-GC2-KO-mice. Two-way ANOVA followed by Bonferroni’s multiple comparison post hoc test.[](https://www.ncbi.nlm.nih.gov/mesh/D009569)
|
# Introduction
New Insight into the Octamer of TYMS Stabilized by Intermolecular [Cys43-Disulfide](https://www.ncbi.nlm.nih.gov/mesh/D003545)
# Abstract
*In the **Abstract** section:*
Thymidylate synthase (TYMS) is an essential enzyme for the de novo synthesis of deoxythymidine monophosphate (dTMP) and has been a primary target for cancer chemo[therapy. Although the physic](https://www.ncbi.nlm.nih.gov/mesh/D013938)al[ str](https://www.ncbi.nlm.nih.gov/mesh/D013938)ucture of TYMS and the molecular mechanisms of TYMS catalyzing the conversion of deoxyuridine monophosphate (dUMP) to dTMP have been the subject of thorough studie[s, its oligomeric structur](https://www.ncbi.nlm.nih.gov/mesh/C007267)e [rema](https://www.ncbi.nlm.nih.gov/mesh/C007267)ins u[ncle](https://www.ncbi.nlm.nih.gov/mesh/D013938)ar. Here, we show that human TYMS not only exists in dimer form but also as an octamer by intermolecular Cys43-disulfide formation. We optimized the expression conditions of recombinant h[uman TYMS using](https://www.ncbi.nlm.nih.gov/mesh/D003545) the Escherichia coli system. Using high-performance liquid chromatography–tandem mass spectrometry (HPLC–MS/MS), we have shown that purified TYMS has catalytic activity for producing dTMP. In the absence of reductant β-mercaptoethanol, SDS-PAGE and size exclusion c[hrom](https://www.ncbi.nlm.nih.gov/mesh/D013938)atography (SEC) showed that th[e size of the TYM](https://www.ncbi.nlm.nih.gov/mesh/D008623)S [pro](https://www.ncbi.nlm.nih.gov/mesh/C032259)tein is about 35 kDa, 70 kDa, and 280 kDa. When the Cys43 was mutated to Gly, the band of ~280 kDa and the peak of the octamer disappe[are](https://www.ncbi.nlm.nih.gov/mesh/D003545)d. Therefore, TYMS[ wa](https://www.ncbi.nlm.nih.gov/mesh/D005998)s determined to form an octamer, depending on the presence of Cys43-disulfide. By measuring steady-state parameters for the monomer, dimer, and [octamer, we fou](https://www.ncbi.nlm.nih.gov/mesh/D003545)nd the kcat of the octamer was increased slightly more than the monomer. On the basis of these findings, we suggest that the octamer in the active state might have a potential influence on the design of new drug targets.
## 1. Introduction
*In the **1. Introduction** section:*
Classical thymidylate synthase (TYMS), encoded by the thyA gene, is highly conserved in most eukaryotes, including humans [,]. It catalyzes the transfer of a methylene group from the cofactor 5,10-methylenetetrahydrofolate (mTHF) to its substrate deoxyuridine monophosphate (dUMP) and forms deoxythymidine monophosphate (dTMP), yielding 7,8-dihydrofolate (DHF) as a secondary product [,,]. A second class of thymidylate synthases, flavin-dependent thymidylate synthases (FDTSs)
[,,], is encoded by the thyX gene and has been found primarily in prokaryotes and viruses [,]. FDTSs utilize a noncovalently bound flavin adenine dinucleotide (FAD) prosthetic group to catalyze the redox chemistry and use mTHF only as a methylene donor. Several organisms, including human pathogens, rely solely on thyX for thymidylate synthesis. Recent studies further showed the catalytic mechanism of TYMS and FDTS
[,], which are essential enzymes for DNA replication and frequently targeted by chemotherapeutic and antibiotic drugs [,]. However, drug resistance has become an increasing concern due to long-term use [,,,]. Therefore, researchers continue to search for effective and specific inhibitors of TYMS to overcome the resistance problem.[](https://www.ncbi.nlm.nih.gov/mesh/C013123)
Extensive knowledge of the structure and properties of the target protein could contribute to formulating more efficient strategies for drug development. Many studies have reported that TYMS exists as a dimer–monomer equilibrium, whose two residues R175 and R176 form part of the dUMP binding site, and the TYMS dimer form can adopt active and inactive conformation [,,]. There is evidence that the TYMS dimer interface plays an important role in TYMS–mRNA recognition, perhaps by controlling a conformational change of the protein that exposes the mRNA binding site [,,].
In addition, Chu et al. thought the dimer obligates catalytic function, while the monomer is believed to play a crucial role in TYMS–mRNA regulation []. Considering these different insights in TYMS structure and function, it is of undeniable importance to further investigate the oligomeric form of the TYMS protein, which contributes to the design of compounds that bind at the oligomer interface of TYMS. Such compounds could overcome drug resistance problems [].[](https://www.ncbi.nlm.nih.gov/mesh/C007267)
The aim of this study was to determine the oligomeric state of TYMS and reconstitute the dTMP synthesis system in vitro. We optimized the overexpression conditions of TYMS, such as the host strain, the inducer concentration, temperature, and culture medium. TYMS catalytic activity for producing dTMP was assessed by mass spectrometry. More importantly, we used SDS-PAGE and size exclusion chromatography (SEC) to analyze the oligomeric state of TYMS. The data showed the full functionality of TYMS on DNA biosynthesis and demonstrated that TYMS coexists in an octamer–dimer–monomer equilibrium and that Cys43 disulfide contributes to octamer formation.
In conclusion, our study demonstrated that the octamer exists in an active state by measuring steady-state parameters of different oligomeric form.[](https://www.ncbi.nlm.nih.gov/mesh/D013938)
## 2. Results
*In the **2. Results** section:*
## 2.1. TYMS Overexpression and Purification
*In the **2.1. TYMS Overexpression and Purification** section:*
To optimize the overexpression condition of the target protein, five different Escherichia coli strains (Tuner (DE3), BL21 (DE3), C41 (DE3), C43 (DE3), and BL21 (DE3)-pLysS) and bacteria concentrations with added isopropyl-β-d-thiogalactoside (IPTG) were initially used to screen. The results showed that 0.8 OD600 is optimal for pLysS (Figure S1A), C43 (Figure S1B), and C41 (Figure S1C), while 0.6 OD600 is optimal for BL21 (Figure S1D) and Tuner (Figure S1E). Then, comparing all of the expression levels of the optimal bacterial density of the different strains, we found 0.8 OD600 for pLysS is the optimal expression level (Figure S1F). Additionally, the concentration of IPTG, the temperature, and four different types of media were also screened. The optimal induction conditions for TYMS was found to be 0.4 mM IPTG and LB medium at 20 °C, after the cells reached 0.8 OD600 for pLysS (Figure S1G–I).
Hence, we chose the recombinant 0.4 mM IPTG-LB-20 °C-0.8 OD600-pLysS for the large-scale expression condition.[](https://www.ncbi.nlm.nih.gov/mesh/D007544)
After confirming the optimal system, the overexpressed TYMS was examined with Western blot and then purified with a nickel column twice. The samples were subjected to SDS-PAGE, followed by Coomassie brilliant blue staining after incubating at 100 °C with a loading buffer containing β-mercaptoethanol. Figure S2A shows that enrichment of TYMS via Ni-NTA chromatography yielded a large amount of full-length proteins. After the proteins were purified twice, they were diluted equivalently and subjected to SDS-PAGE followed by Coomassie brilliant blue staining (Figure S2B,C) to detect the purity of the proteins, which was found to be more than 90%. Finally, according to the BSA standard curve (Figure S2D), the concentration of purified TYMS was determined to be >60 mg/mL.[](https://www.ncbi.nlm.nih.gov/mesh/D009532)
## 2.2. Reconstitution of TYMS-Mediated dTMP Synthesis
*In the **2.2. Reconstitution of TYMS-Mediated dTMP Synthesis** section:*
The catalytic mechanism of classical thymidylate synthases is presented in Figure 1A. To formally test the functionality of such a pathway and to provide a tool to investigate its mechanistic features, we reconstituted the entire process of reductive methylation with defined components. Using the dUMP substrate, we performed reductive methylation to measure the formation of dTMP using HPLC–MS/MS. Multiple reaction monitoring (MRM) was used to determine base ion mass transitions of dU (229.1 to 113.1) and dT (243.1 to 127.1) (Figure 1B,C). Standard curves were built to quantify dU and dT modification (Figure 1D,E).
In the absence of TYMS, no dT was detectable. However, full reconstitution of the TYMS with the dUMP and mTHF substrates generated a substantial amount of dU (Figure 1F). Together, these results demonstrate that TYMS-mediated reductive methylation of dUMP generates dTMP.[](https://www.ncbi.nlm.nih.gov/mesh/C007267)
## 2.3. TYMS Formed Octamer by Intermolecular Cys43-Disulfide
*In the **2.3. TYMS Formed Octamer by Intermolecular Cys43-Disulfide** section:*
To determine the existence of the homologous dimer, we used Coomassie brilliant blue and Western blot to analyze the status of the TYMS protein. Surprisingly, after TYMS incubated with a loading buffer (no β-mercaptoethanol) at 37 °C, we found that TYMS showed three bands with molecular weights ~35 kDa, ~70 kDa, and ~280 kDa, respectively (Figure 2A,B). Compared to the conditions at 37 °C, after TYMS incubated with a loading buffer (no β-mercaptoethanol) at 100 °C, the bands of ~70 kDa and ~280 kDa weakened (Figure 2C,D). While TYMS was incubated with a loading buffer (containing β-mercaptoethanol) at 100 °C, the band of ~280 kDa disappeared and the band of ~70 kDa weakened. This phenomenon made us speculate that TYMS may exist as an octamer; the proportion of each oligomeric form in different conditions are shown in Figure S3.
To investigate the active or inactive conformation of TYMS, we incubated TYMS with dUMP and mTHF and subjected them to SDS-PAGE followed by Western blot (Figure 2E) and Coomassie brilliant blue (Figure 2F) staining. The results showed that the bands of ~70 kDa and~280 kDa remain unchanged, compared to samples without dUMP and mTHF. It is obvious from this result that the presence of ligands does not influence the oligomeric state of the enzyme.[](https://www.ncbi.nlm.nih.gov/mesh/C004692)
To further investigate the oligomeric form of TYMS, we performed a calibration SEC experiment with five protein standards, including myosin (212.0 kDa), beta-galactosidase (116.0 kDa), bovine serum albumin (67.0 kDa), ovalbumin (43.0 kDa), and ribonuclease A (13.7 kDa). The established standard curve allowed a more reliable estimation of the protein molecular weight for the SEC column. As shown in the Figure 2G, the lg (Mr) value is plotted as a function of the retention volume. The experimental standard curve was well-fitted by the equation y = 2.46 − 0.016x with R2 = 0.996. TYMS (6 mg/mL) was subjected to SEC and collected at a retention volume of 8.9 mL, 13.53 mL, and 16.6 mL (Figure 2H). Meanwhile, TYMS (3 mg/mL) was subjected to SEC in the presence or absence of dUMP and mTHF (Figure 2I,J).
The results indicated that the octamer is independent of the concentration and substrates of TYMS. Then, the collected samples were subjected to Western blot and Coomassie brilliant blue. (The Monomer was detected with SDS-PAGE (12% acrylamide), while the dimer and octamer were found with NATIVE-PAGE (7.5% acrylamide)) (Figure 2K). According to the standard curve equation and the Western Blot marker, the molecular weight of TYMS was estimated to be ~35 kDa, ~70 kDa, and ~280 kDa, which was about twice and 8-fold as much as the theoretical molecular weight of monomeric TYMS (35 kDa), indicating that the TYMS protein existed in oligomeric forms as a dimer and octamer by intermolecular Cys-disulfide.[](https://www.ncbi.nlm.nih.gov/mesh/C007267)
To determine which Cys site contributes to the form of the octamer, we examined the effect of Cys43, Cys180, and Cys210 disulfide on the TYMS octamer by mutating these to Gly (depicted as red in Figure 3A), due to Cys195 being an active site and Cys199 contributing to the dimerization interface [,,]. TYMS (Cys43Gly), TYMS (Cys180Gly), and TYMS (Cys210Gly) were expressed and purified with Ni-NTA chromatography twice, and then they were detected by Western blot (Figure 3B), Coomassie brilliant blue without β-mercaptoethanol (Figure 3C), and SEC (Figure 3D). The band of octamer at ~280 kDa with a peak at 22.25 mL in TYMS (Cys43Gly) disappeared, while there were no changes in TYMS (Cys180Gly) and TYMS (Cys210Gly), which demonstrate the Cys43 residue is essential for the octamer.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## 2.4. Analysis of Kinetic Properties of All Oligomeric Forms
*In the **2.4. Analysis of Kinetic Properties of All Oligomeric Forms** section:*
To further analyze the effect of the octamer, we detected the steady-state parameters of all the oligomeric forms by measuring the formation of 7,8-dihydrofolate (DHF). The reaction process curves for dUMP and mTHF of the monomer, dimer, and octamer are shown in Figure 4A–F, respectively. The Km values for dUMP were increased by 3-fold for the dimer and 1.32-fold for the octamer compared with the monomer (Figure 4G). The Km values for mTHF were increased by 2.9-fold for the dimer and 1.6-fold for the octamer, relative to for the monomer (Figure 4H). The Vmax and kcat values for dUMP of the dimer were decreased slightly, while the octamer increased more than the monomer, with the same results for mTHF, suggesting that the octamer of TYMS is in the active state.
The experimental standard curve of DHF was well-fitted by the equation y = 0.00394 + 0.00446x with R2 = 0.999 (Figure 4I). HPLC–MS/MS was carried out to demonstrate the enzyme activity of the monomer, dimer, and octamer of TYMS (Figure S4).[](https://www.ncbi.nlm.nih.gov/mesh/C010920)
## 3. Discussion
*In the **3. Discussion** section:*
Recent research on the only de novo source of synthesized dTMP points towards a mechanism involving TYMS. Despite the classic nature of this pathway, experimental evidence that directly links TYMS activity with mTHF and dUMP is insufficient and the oligomeric form of TYMS remains unclear. This work aimed to analyze the oligomeric form of TYMS and reconstitute TYMS-mediated dTMP synthesis.[](https://www.ncbi.nlm.nih.gov/mesh/D013938)
In line with spectrophotometric assay [], our work provides mass spectrum evidence for a direct productive action of TYMS with mTHF and dUMP, confirming dTMP generation (Figure 1). Reaction efficiency of dUMP to dTMP reaches at least 20%, suggesting the TYMS we purified has enzyme activity. Many studies have reported TYMS exists as the dimer, which has two distinct states: one is the active state in the crystal structures of TYMS-nucleotide-(anti) folate ternary complexes []; the other one is an inactive state in sulfate-containing conditions []. The cavity in the dimer interface could serve as an allosteric site used to regulate the conformational switching between the active and inactive states [,].
In addition, TYMS performs at least two different functions with specific interaction regions: the dimer obligates catalytic function, while both the monomer and the dimer are believed to play crucial roles in TYMS–mRNA recognition and regulation [,,,]. In this study, we found TYMS not only formed a dimer, but also an octamer by intermolecular Cys43-disulfide (Figure 3). The octamer is a higher homologous aggregation and in an active conformation, and the Vmax and kcat were increased slightly (<1.5-fold) for the octamer, while the Km was decreased (<1.5-fold) (Figure 4). To the best of our knowledge, this is the first detailed report on the octamer of TYMS in an active conformation. Still, the role of the octamer is unclear.
Since dimers have such an important function, it is possible that the octamer structure of TYMS may potentially affect the activity of TYMS so as to provide new drug targets to overcome resistance problems, as well as the synthesis of other nucleotides, so that the appropriate balance of the four nucleotides required for DNA synthesis is maintained. It should be noted that this study only examined the molecular weight of TYMS using SDS-PAGE and SEC. Our results lack crystallographic data. Despite its preliminary character, this study can clearly indicate TYMS exists in octamer–dimer–monomer equilibrium.[](https://www.ncbi.nlm.nih.gov/mesh/C013123)
In summary, this work serves as the first comprehensive evaluation of the oligomeric structure and the activities of TYMS in vitro. Further, it provides a foundation for further inquiry into the role of this very interesting DNA synthesis and repair enzyme.
## 4. Materials and Methods
*In the **4. Materials and Methods** section:*
## 4.1. Plasmid Construction
*In the **4.1. Plasmid Construction** section:*
The human TYMS (NM_001071.2) sequence was amplified by polymerase chain reaction (PCR) from human cDNA (reverse transcription from total mRNA) with primers (PET28A-TYMS-HindIII-Forward and PET28A-TYMS-XhoI-Reverse). TYMS (Cys43Gly), TYMS (Cys180Gly), and TYMS (Cys210Gly) were amplified by Fusion PCR [] with primers PET28A-TYMS-HindIII-Forward, TYMS-43-Forward, TYMS-43-Reverse, TYMS-180-Forward, TYMS-180-Reverse, TYMS-210-Forward, TYMS-210-Reverse, and PET28A-TYMS-XhoI-Reverse (the sequence of the primers are shown in Table S1). The fragments were cloned into the pET-28a (+) vector using the restriction recognition site for Hind III and Xho I, carrying a N-terminal 6× His tag. DNA sequencing was used to verify the sequences of the constructed vector.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
## 4.2. Expression Screening
*In the **4.2. Expression Screening** section:*
To find out an optimized expression condition for TYMS, the vector pET-28a-6His-TYMS was transformed into Tuner (DE3), BL21 (DE3), C41 (DE3), C43 (DE3), BL21 (DE3)-pLysS. When IPTG (MedChemExpress, Monmouth, NJ, USA) was added, we optimized bacterial density, including concentrations of 0.3 OD600, 0.5 OD600, 0.8 OD600, 1.1 OD600, and 1.3 OD600, for different host strains. Furthermore, for the best host cell and its optional bacteria concentration, it was cultured in the nutrient-rich medium 32Y and several modified media, including LB, 2× TYE, and 4× TY (compositions of culture media are shown in Table S2) []. Then, the appropriate concentration of IPTG containing 0.1 mM, 0.2 mM, 0.4 mM, and 0.8 mM was investigated. Finally, the inducing temperature was optimized at 20 °C, 25 °C and 30 °C.[](https://www.ncbi.nlm.nih.gov/mesh/D006639)
## 4.3. Protein Purification
*In the **4.3. Protein Purification** section:*
Cells were harvested by centrifugation at 5000× g for 10 min at 4 °C and washed twice with ice-cold PBS. Then, 2 g cell pellets were suspended with 10 mL of PBS, which contained 1 mM MgCl2, 10 mM imidazole (MedChemExpress), and protease inhibitor cocktail, and DNase at final concentrations of 1 mM, 20 mM, 1 mg/mL, 1 tablet/50 mL, and 100 U/mL, respectively. Resuspended cells were broken using an ultrasonic cell disruptor (NOISE ISOLATING CHAMBER JY 92-IIN) on ice (on 5 s, off 5 s). Lysed cells were subjected to centrifugation at 100,000× g for 1 h at 4 °C to obtain the supernatant. The supernatant containing TYMS was then loaded onto a Ni-NTA column pre-equilibrated with a binding buffer (PBS buffer, 150 mM NaCl, 10% (v/v) glycerol, 20 mM imidazole, pH 7.4).
The resins were then washed eight times with the binding buffer containing 20 mM imidazole to remove nonspecifically bound proteins, and the bound proteins were eluted with an elution buffer (PBS buffer, 150 mM NaCl, 10% (v/v) glycerol, 400 mM imidazole, pH 7.4). To obtain more purified TYMS, we purified this protein twice using the Ni-NTA column as described above. Purified TYMS was concentrated using an Amicon Ultrafree centrifugal filter (Millipore Corporation, Billerica, MA, USA) with a cutoff of 10 kDa. The concentration buffer contained 20 mM Tris-Base, 150 mM NaCl, and 10% (v/v) glycerol at pH 7.4. Protein concentration was determined using the BCA assay according to manufacturer’s instructions (Pierce, Rockland, IL, USA).[](https://www.ncbi.nlm.nih.gov/mesh/D000077330)
## 4.4. SDS-PAGE, Western Blot, and Coomassie Brilliant Analysis
*In the **4.4. SDS-PAGE, Western Blot, and Coomassie Brilliant Analysis** section:*
After the purification, Coomassie brilliant blue was used to examine the purity of the proteins. Purified proteins were series diluted (0-, 2-, 4-, 8-, 16-, 32-, 64-, 128-, 256-fold) and then incubated with a loading buffer with or without β-mercaptoethanol at 37 °C or 100 °C for 10 min. Then, these sample were subjected to SDS-PAGE to determine the protein purity and oligomeric form. After that, 10 µM TYMS was incubated with 2 mM dUMP and 2 mM mTHF in a 40 µL reaction system at 37 °C and then subjected to SDS-PAGE to confirm the active or inactive state of TYMS.[](https://www.ncbi.nlm.nih.gov/mesh/C004692)
Western blot was used to verify the protein expression, TYMSs were subjected to SDS-PAGE, the SDS gel was washed with a transfer buffer, and then the proteins were transferred from the gel onto a nitrocellulose membrane with a constant current of 250 mA for 2 h. The membrane was blocked with 1% nonfat milk powder in PBST (PBS containing 0.05% (v/v) Tween-20). Mouse anti-His-tag antibody (Sigma, St. Louis, MO, USA) was used as the primary antibody at a 1:5000 dilution in blocking solution. Goat-anti-mouse was the secondary antibody (Sigma), which was diluted at 1:3000, and protein bands were detected on photographic films using an enhanced chemiluminescent substrate.[](https://www.ncbi.nlm.nih.gov/mesh/C032259)
## 4.5. Activity Detection
*In the **4.5. Activity Detection** section:*
TYMS activity assays were carried out in 100 µL reaction volume containing a reaction buffer (10 mM boracic acid pH 6.0, 150 mM NaCl), 200 µM mTHF, 1 µM dUMP, and 10 µM TYMS. After incubation at 37 °C for 3 h, CIAP was added at a final concentration of 1 U and the reaction was incubated at 37 °C for 4 h. The samples were then subjected to HPLC–MS/MS analysis of deoxyuridine (dU) and deoxythymidine (dT). Quantification was performed using an HPLC system (Waters, Milford, MA, USA) coupled to an API 5500 triple quadrupole (ABSciex, Framingham, MA, USA) operating in positive electrospray ionization mode. The chromatographic separation was performed at 25 °C with the use of a C18 reverse-phase column (150 × 2.1 mm; 5 μm particle size; Thermo Fisher).
The mobile phase consisted of A (water and 0.1% formic acid) and B (methanol and 0.1% formic acid) solutions []. The following conditions were employed during chromatography: 0.4 mL/min flow, 0–1 min, 1% B; 1–2 min, to 20% B; 2–3 min, to 20% B, 3–4 min, 1% B. To minimize potential salt and other contaminants in the ESI source, a time segment was set to direct the first 0.5 min of column elute to waste. For mass spectrometry detection, the multiple reaction monitoring was implemented using the following mass transitions: 243.1/127.1 (dT) and 243.1/127.1 (dU).[](https://www.ncbi.nlm.nih.gov/mesh/C032688)
## 4.6. Size Exclusion Chromatography
*In the **4.6. Size Exclusion Chromatography** section:*
Purified TYMS by Ni-NTA column was subjected to SEC at a flow rate of 1.0 mL/min on a Superdex-200 HiLoad 10/300 column (GE Healthcare, Pittsburgh, PA, USA) that had been pre-equilibrated with HEPES buffer (10 mM HEPES, 150 mM NaCl, 10% glycerol, pH 7.4). The eluent was collected in constant volumes of 500 μL and examined by ultraviolet absorption at 280 nm. Columns are often calibrated using five standard samples, including myosin (212.0 kDa), beta-galactosidase (116.0 kDa), bovine serum albumin (67.0 kDa), ovalbumin (43.0 kDa), and ribonuclease A (13.7 kDa).[](https://www.ncbi.nlm.nih.gov/mesh/C088321)
The components separated by SEC were concentrated using an Amicon Ultrafree centrifugal filter (Millipore Corporation, Billerica, MA, USA) with a cutoff of 10 kDa in concentration buffer. The dimer and octamer separated by SEC were subjected to NATIVE-PAGE (7.5%) and the monomer was subjected to SDS-PAGE (12%), and then were analyzed by using Western blot and Coomassie brilliant blue. For NATIVE-PAGE, samples were run at constant voltage (100 V before the indicator to spacer gel and then switched to 120 V).[](https://www.ncbi.nlm.nih.gov/mesh/C032259)
## 4.7. Reaction Kinetics Detection
*In the **4.7. Reaction Kinetics Detection** section:*
Enzyme activity was determined by measuring the formation of dihydrofolate, which was monitored at 340 nm after the addition of the enzyme to the reaction assay [,]. Measurements were made at pH 6.0 and 37 °C in the reaction buffer (10 mM boric acid, 150 mM NaCl) for 5 min. To determine the Km(dUMP) for different oligomeric forms, varying concentrations of dUMP (0–50 µM) were used with constant concentrations of the enzyme (0.75 µM) and mTHF (100 µM). Km (mTHF) was determined with varying concentrations of mTHF (0–100 µM) and were used with constant concentrations of the enzyme (0.75 µM) and dUMP (50 µM).[](https://www.ncbi.nlm.nih.gov/mesh/C010920)
# Supplementary Materials
*In the **Supplementary Materials** section:*
The supplementary materials are available online at .
# Author Contributions
*In the **Author Contributions** section:*
Li. W. and Q.Y. co-designed the research. D.X. performed most experiments and wrote the paper, Lu. W. detected the oligomeric forms using SEC, Q.X. constructed the plasmids, X.W., performed some western blot analysis, L.Z. carried out the Language polishing.
# Conflicts of Interest
*In the **Conflicts of Interest** section:*
The authors declare no conflict of interest.
# References
*In the **References** section:*
HPLC–MS/MS assay measures the reductive methylation activity of thymidylate synthase (TYMS) in vitro. (A) Deoxyuridine monophosphate (dUMP) and 5,10-methylenetetrahydrofolate (mTHF) as cosubstrates generate deoxythymidine monophosphate (dTMP) by TYMS, yielding 7,8-dihydrofolate (DHF) as a secondary product; (B,C) Base ion mass transitions for LC–MS-MS analysis of dU and dT standard. The multiple reaction monitoring (MRM) transitions were monitored as follows: 229.1 to 113.1 (dU); 243.1 to 127.1 (dT); (D,E) HPLC–MS-MS standards curves of dU and dT (F) LC–MS-MS profiles of nucleosides derived from TYMS treatment. The upper LC–MS-MS profile shows nucleoside dU and dT standards. The lower LC–MS-MS profile shows dT generation in vitro reaction system of TYMS with dU as substrate.[](https://www.ncbi.nlm.nih.gov/mesh/C007267)
The determination of protein dimers and octamers. (A) Coomassie brilliant blue of concentrated proteins incubated without β-mercaptoethanol (β-ME) for 10 min at 37 °C: (1) undilutedly concentrated proteins; (2) double-diluted proteins; (3) 4-times diluted proteins; (4) 8-times diluted proteins; (5) 16-times diluted proteins; (6) 32-times diluted proteins; (7) 64-times diluted proteins. The MW of monomer, dimer, and octamer are ~35 kDa, ~70 kDa, and ~280 kDa, respectively. (B) Coomassie brilliant blue of eight diluted proteins incubated without β-mercaptoethanol (β-ME) for 10 min at 37 °C. (C) Coomassie brilliant blue of concentrated proteins incubated without β-mercaptoethanol (β-ME) for 10 min at 100 °C: (1) undilutedly concentrated proteins; (2) double-diluted proteins; (3) 4-times diluted proteins; (4) 8-times diluted proteins; (5) 16-times diluted proteins; (6) 32-times diluted proteins; (7) 64-times diluted proteins. (D) Coomassie brilliant blue of 8-times diluted proteins incubated without mercaptoethanol for 10 min at 100 °C. (E) Western blot and (F) Coomassie brilliant blue: (1) TYMS was directly incubated with a loading buffer (containing β-mercaptoethanol) for 10 min at 100 °C; (2) TYMS reacting with dUMP and mTHF for 1 h at 37 °C, and then incubated with a loading buffer (containing β-mercaptoethanol) for 10 min at 100 °C; (3) TYMS was directly incubated with a loading buffer (without β-mercaptoethanol) for 10 min at 37 °C; (4) TYMS reacting with dUMP and mTHF for 1 h at 37 °C, and then incubated with a loading buffer (without β-mercaptoethanol) 10 min in 37 °C; (5) TYMS was directly incubated with a loading buffer (without β-mercaptoethanol) for 10 min at 100 °C; (6) TYMS reacting with dUMP and mTHF for 1 h at 37 °C, and then incubated with a loading buffer (without β-mercaptoethanol) for 10 min at 100 °C. (G) Standard curve for molecular weight estimation and (H) size exclusion chromatography (SEC). Peak 1, 2, 3 represent a symmetrical peak eluted at a retention volume of ~8.9 mL, ~13.53 mL, and ~16.6 mL, indicating that TYMS existed in octamer, dimer, and monomer form. (I) Size exclusion chromatography of protein that was diluted to 3 mg/mL. (J) Size exclusion chromatography of protein that was diluted to 3 mg/mL with dUMP and mTHF. (K) The Western blot and Coomassie brilliant blue of components separated by SEC. The monomer was subjected to SDS-PAGE with 12% acrylamide, while the dimer and octamer were subjected to NATIVE-PAGE with 7.5% acrylamide.[](https://www.ncbi.nlm.nih.gov/mesh/C004692)
Determination of the oligomeric state of TYMS by mutated Cys. (A) Available TYMS structures with Cys43, Cys180, and Cys210 sites depicted as red. The data were derived from PDB (Protein Data Bank) and the ID is 1YPV; (B,C) Western blot and Coomassie brilliant blue of TYMS with Cys43Gly, Cys180Gly, or Cys210Gly (without β-mercaptoethanol); (D) size exclusion chromatography of TYMS with Cys43Gly, Cys180Gly, or Cys210Gly.[](https://www.ncbi.nlm.nih.gov/mesh/D003545)
The reaction process curves and steady-state parameters for different oligomeric forms of TYMS. (A) Reaction process curves of dUMP for TYMS monomer, the left panel showing the substrate saturation curve, and the right panel showing the Lineweaver–Burk double-reciprocal plot; (B) reaction process curves of dUMP for TYMS dimer; (C) reaction process curves of dUMP for TYMS octamer; (D) reaction process curves of mTHF for TYMS monomer; (E) reaction process curves of mTHF for TYMS dimer; (F) reaction process curves of mTHF for TYMS octamer; (G) steady-state parameters of dUMP for different oligomeric forms; (H) steady-state parameters of mTHF for different oligomeric forms; (I) standard curve for ultraviolet absorption value at OD340 of DHF.[](https://www.ncbi.nlm.nih.gov/mesh/C007267)
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.