pmcid stringlengths 6 6 | title stringlengths 9 374 | abstract stringlengths 2 4.62k ⌀ | fulltext stringlengths 167 106k | file_path stringlengths 64 64 |
|---|---|---|---|---|
544556 | The effect of short-duration sub-maximal cycling on balance in single-limb stance in patients with anterior cruciate ligament injury: a cross-sectional study | Background It has previously been shown that an anterior cruciate ligament (ACL) injury may lead to impaired postural control, and that the ability to maintain postural control is decreased by fatigue in healthy subjects. To our knowledge, no studies have reported the effect of fatigue on postural control in subjects with ACL injury. This study was aimed at examining the effect of fatigue on balance in single-limb stance in subjects with ACL injury, and to compare the effects, and the ability to maintain balance, with that of a control group of uninjured subjects. Methods Thirty-six patients with unilateral, non-operated, non-acute ACL injury, and 24 uninjured subjects were examined with stabilometry before (pre-exercise) and immediately after (post-exercise) short-duration, sub-maximal cycling. In addition, the post-exercise measurements were compared, to evaluate the instantaneous ability to maintain balance and any possible recovery. The amplitude and average speed of center of pressure movements were registered in the frontal and sagittal planes. The paired t-test was used for the intra-group comparisons, and the independent t-test for the inter-group comparisons, with Bonferroni correction for multiple comparisons. Results No differences were found in the effects of exercise between the patients and the controls. Analysis of the post-exercise measurements revealed greater effects or a tendency towards greater effects on the injured leg than in the control group. The average speed was lower among the patients than in the control group. Conclusions The results of the present study showed no differences in the effects of exercise between the patients and the controls. However, the patients seemed to react differently regarding ability to maintain balance in single-limb stance directly after exercise than the control group. The lower average speed among the patients may be an expression of different neuromuscular adaptive strategies than in uninjured subjects. | Background The anterior cruciate ligament (ACL) is the most commonly injured ligament in the knee. The risk of future joint problems, in the form of functional limitations, secondary lesions, and arthrosis, is increased following such an injury. Secondary effects commonly seen after an ACL injury include defective neuromuscular function with reduced strength and functional performance, a different movement and activation pattern, defective proprioception and impaired postural control [ 1 ]. Impaired postural control has been reported after acute [ 2 ], and chronic ACL injury [ 3 - 5 ], as well as after ACL reconstruction [ 6 - 8 ]. Higher amplitude values [ 2 - 5 ] and longer reaction time when subjected to perturbations [ 4 , 6 , 7 ] have been observed among patients compared to controls. Studies have also shown that patients with better subjective function have lower amplitude values [[ 8 ], Ageberg E, Roberts D, Holmström E, Fridén T: Balance in single-limb stance in individuals with anterior cruciate ligament injury – relation to knee laxity, proprioception, muscle strength, and subjective function. Manuscript submitted]. The present study was initiated by the clinical knowledge that although patients with ACL injury have had extensive neuromuscular training and function well during daily life and (modified) physical activities, they experience a decreased ability to maintain balance during weight-bearing on the injured leg in demanding situations while fatigued. This may be related to an increased risk of further injuries. Fatigue is caused by a combination of different physiological mechanisms occurring at both the central and peripheral levels [ 9 ], affecting afferent neuromuscular pathways, observed as proprioceptive deficiency [ 10 - 12 ], and efferent neuromuscular pathways, seen, for example, as a delay in muscle response [ 13 , 14 ]. Thus, muscular fatigue leads to a decline in work performance, which may also include effects on postural control. A decreased ability to maintain balance in bilateral stance [ 15 - 17 ], and single-limb stance [ 15 , 18 - 20 ] after fatiguing exercise (i.e., higher values after exercise) has been reported in uninjured subjects, and it has been suggested that individuals are therefore at increased risk of injury when fatigued [ 15 , 19 , 20 ]. Studies of balance in single-limb stance are of importance and of interest since these movement patterns resemble the stance phase, and since many knee injuries occur during weight-bearing on one leg [ 21 ]. To our knowledge, no studies evaluating the effect of fatigue on postural control in subjects with ACL injury have been reported. The main purposes of this study were: 1) to examine the effect of short-duration, sub-maximal exercise performed on a cycle ergometer, on postural control, measured by stabilometry in single-limb stance on a force platform, in individuals with ACL injury in comparison with that of a control group of uninjured subjects, and 2) to explore the patients' instantaneous ability to maintain balance in single-limb stance after exercise, in comparison with that of the control group. Furthermore, the patients were compared to the control group in order to verify previous findings that postural control is affected in both legs by a unilateral ACL injury [ 2 - 5 ]. No comparisons were, therefore, made between the injured and uninjured legs. Our hypothesis was that the patients with ACL injury would be more affected by exercise than the uninjured subjects, since fatigue has been shown to reduce postural control in healthy subjects, and since postural control may already be impaired due to the injury. Methods Patients Thirty-six patients (18 men and 18 women) were included in the study. Inclusion criteria were: 1) age between 15 and 35 years, 2) unilateral, non-operated, non-acute ACL deficiency with or without associated lesions of other structures of the knee, 3) an uninjured contralateral extremity, back and neck, and 4) no history of neurological disease, vestibular or visual disturbance. Their mean age was 26 years (SD 5 years), mean height 174 cm (SD 9 cm), and mean body mass 72 kg (SD 13 kg). Their median activity level before injury was 6.5 (range 3 to 9) and on the test occasion 4 (range 1 to 9) according to the Tegner activity level scale [ 22 ]. The mean time elapsed from injury to the test occasion was 3.8 years (SD 3, range 0.5 to 11 years). The patients had undergone an extensive neuromuscular training program [ 23 ] after the injury under the supervision of physical therapists, with a mean duration of 7 months (SD 5 months). A visual analog scale graded from 0 to 100 mm was used for subjective evaluation of extremity function, where 0 was "as if the knee had been recently injured" and 100 was "perfect" [ 24 ]. The patients' mean value and median value on this scale were 68 mm and 59 mm (range 12–95 mm), respectively. Control group The measurements of twenty-four uninjured volunteers (11 men and 13 women) from a previous study [ 18 ], with no history of neurological disease, major orthopedic lesion, vestibular or visual disturbance, constituted control values. Their mean age was 24 years (SD 3 years), mean height 176 cm (SD 8 cm), and mean body mass 71 kg (SD 13 kg). Their median activity level was 5 (range 2 to 9) according to the Tegner activity level scale [ 22 ]. The subjects in the control group were chosen in order to have the same distribution in age, sex, and physical activity as the patients [ 25 ]. No significant difference was found between the groups in age, height, body mass or activity level. The Research Ethics Committee at Lund University approved the study. All subjects gave their written informed consent to participate in the study. Assessment Stabilometry Balance in single-limb stance was tested by means of a strain gauge force plate (33 × 38 cm) with the subject barefoot in a standardized position [ 5 , 26 , 27 ] (Figure 1 ). This measurement was performed before (pre-exercise) and immediately after fatiguing exercise (post-exercise). The foot was placed pointing straight forward in relation to reference lines in the frontal and sagittal planes (origin of coordinates). The other leg was flexed 90° at the hip and knee joints with both arms hanging relaxed at the sides. The subjects were instructed to stand as motionless as possible, looking straight ahead at a point on the wall 65 cm away; they were allowed to practice maintaining this position for about 20 s before three measurements on each leg were made, with the subjects standing alternately on their right and left leg. The test order between legs was randomized regarding injured/uninjured leg in the patient group (injured leg n = 20, uninjured n = 16), and regarding right/left leg in the control group (right leg n = 13, left n = 11). No differences were observed in the stabilometric measurements between these randomization groups. Accordingly, the assessment included three measurements made on each leg, giving a total of six measurements pre- and post-exercise, respectively. These six measurements lasted for approximately 3 1/2 minutes, with about 10 seconds between each measure. The median value of the three measurements on each leg was used to compare pre- and post-exercise values. Decreasing values in the three measurements have been observed in a previous study, indicating a learning effect [ 26 ]. Some degree of recovery may, therefore, occur during the three post-exercise measurements. For this reason, the first and third of the three post-exercise measurements on each leg were used, to evaluate the instantaneous value of the ability to maintain postural control (first measurement) and the possible recovery (third measurement). Movements of the center of pressure (CP) in the frontal plane (FP) and sagittal plane (SP) were recorded for 25 s at a sampling frequency of 20 Hz. A computer program (Viewdac 2.1, Keithley Instruments, Inc., Cleveland, Ohio, USA), was used to analyze the following variables: 1) average speed of CP movements in mm·s -1 ; and 2) number of movements exceeding 10 mm from the mean value of CP (DEV 10) , giving a total of four variables (two variables in each plane). The mean value of CP is the average distance (mm) of the CP from the reference lines, and DEV 10 is the number of movements exceeding 10 mm from the mean value of CP. DEV 10 (n) reflects the deviation of CP (i.e., displacement of CP), and the average speed (mm·s -1 ) reflects the amplitude and frequency of CP movements. Figure 2 shows raw data from a stabilometry test. Average speed and DEV 10 were used in the present study, since our previous studies have shown that these variables are reliable [ 18 , 26 ], and sensitive in detecting differences between patients and uninjured subjects [ 2 ], and sensitive in detecting the effects of exercise [ 18 ]. We expected to find higher values after exercise [ 18 ]. Short-duration sub-maximal exercise Short-duration, sub-maximal exercise was performed on a cycle ergometer. The subjects' heart rate was continuously recorded during the entire test. Borg's scale for Rating of Perceived Exertion (RPE scale) was used to assess the subjective effort level during exercise [ 28 ]. On this scale, numbers ranging from 6 to 20 are matched with descriptors (e.g., 6 = No exertion at all , 13 = Somewhat hard , 15 = Hard , 17 = Very hard , 19 = Extremely hard , and 20 = Maximal exertion ). The RPE scale was designed to increase linearly with exercise intensity and heart rate for work on a bicycle ergometer, and correlates closely with several physiological variables, including heart rate and blood lactate concentration [ 28 ]. A linear relationship exists between heart rate and oxygen consumption with increasing rate of work. A given percentage of the maximum oxygen consumption (VO 2max ) results in a higher percentage of the maximum heart rate (HR max ); e.g., 75% of VO 2max represents an intensity of 86% of HR max [ 29 ]. The maximum heart rate can be estimated from the following equation: maximum heart rate (beats/min) = 220 – age (years) [ 29 ]. Effects of fatigue are likely to occur after a few minutes of sub-maximal exercise [ 9 ]. The rate of pedaling was kept constant at 60 revs/min. The level of exercise was calculated so as to be similar to that perceived during a general exercise session. The workload (W) was set individually, depending on the sex and physical condition of each subject, with the aim of reaching a heart rate above 60% of the predicted HR max [ 30 ] in all subjects. Cycling was stopped when the subjects had reached a heart rate exceeding 60% of the predicted HR max , perceived the exercise as hard or very hard (values 14–17 of the RPE scale), and had reached steady-state heart rate, i.e., after approximately 5 min. Statistical analysis The average of the right and left legs; i.e., (right+left)/2, was used for statistical analysis in the control group, since there were no clinically or statistically significant differences between the legs. The use of the mean value of both legs when performing parametric statistics can be questioned, since this may affect the data variability. It cannot, however, be excluded that a dominance of one or the other side exists, which is difficult to define [ 25 ], and therefore it is hard to determine which leg to use in comparison with the patients. For this reason we used the average of the right and left legs. However, the results were confirmed using the right and left legs separately as the control leg. The median value of the three measurements was used to compare pre- and post-exercise values. In addition, the first and third of the three post-exercise measurements were compared, to evaluate the instantaneous ability to maintain postural control in single-limb stance (first measurement) and the possible recovery (third measurement). We used the paired t-test for the intra-group comparisons, and the independent t-test for the inter-group comparisons, with Bonferroni correction for multiple comparisons. The present study is of exploratory character, and the level of correction for multiple comparisons was chosen with regard to this. For each stabilometric variable, five separate t-tests were performed in comparisons between pre- and post-exercise values for the injured leg and the control group: 1) injured leg pre-exercise vs. post-exercise, 2) control group pre-exercise vs. post-exercise, 3) injured leg vs. control group pre-exercise, 4) injured leg vs. control group post-exercise, and 5) effects of exercise (post-exercise minus pre-exercise) injured leg vs. control group. These five t-test were also performed in the analysis of possible differences between pre- and post-exercise values for the uninjured leg and the control group. Since five comparisons were made in the above-mentioned analyses, the alpha level was set at 0.05/5 = 0.01. For each stabilometric variable, three separate t-tests were performed in comparisons between post-exercise measurements 1 and 3 for the injured leg and the control: 1) injured leg measurement 1 vs. measurement 3, 2) control group measurement 1 vs. measurement 3, and 3) effects of exercise (measurement 3 minus measurement 1) injured leg vs. control group. These three t-tests were also performed in the analysis of possible differences between post-exercise measurements 1 and 3 for the uninjured leg and the control group. Since three comparisons were made in the above-mentioned analyses, the alpha level was set at 0.05/3 = 0.02. The statistical analyses were performed using the program package SPSS 11.0 (SPSS Inc., Chicago, Illinois, USA). Results Fatiguing exercise All subjects exceeded the 60% value of the predicted HR max ; the mean level being 82% (SD 6%, range 66 to 92%) among the patients and 81% (SD 7%, range 68 to 99%) among the controls. The median power output produced by the patients and the control group at the end of fatiguing exercise was 125 W (range 75 to 200 W) and 150 W (range 100 to 200 W), respectively, and the mean value of perceived exertion, rated according to the RPE scale, was 15.8 (SD 1.1) and 15.4 (SD 0.9), respectively. The final heart rate attained among the patients and the control group was 159 beats/min (SD 11 beats/min) and 159 beats/min (14 beats/min), respectively, and the heart rate after the stabilometric assessment, approximately 3 1/2 minutes after exercise, was 112 beats/min (SD 14 beats/min) and 117 beats/min (SD 16 beats/min), respectively. No significant differences were found between the patients and controls with regard to the above-mentioned variables. Average speed of CP movements Higher values were noted post- than pre-exercise in the FP and SP in the injured and uninjured legs, but only in the FP in the control group (Table 1 ). No differences were noted between the groups regarding the effects of exercise (mean difference of post-exercise minus pre-exercise values) (Table 2 ). Figures 3 and 4 show the pre-and post-exercise values for the injured leg and the control group. A lower value was observed in the third than in the first of the post-exercise measurements on the injured leg in the FP, but no differences were noted on the uninjured leg or in the control group (Table 3 ). The injured leg of the patients was more affected by exercise directly after cycling than the legs of the control group in the FP (Table 4 ). Figures 7 and 8 show the first and third of the post-exercise measurements on the injured leg and in the control group. Lower values were observed pre-exercise in the SP in the injured and uninjured legs of the patients than in the control group (Table 5 ). Number of movements exceeding 10 mm from the mean value of CP A higher value was found post- than pre-exercise in the uninjured leg in the FP, and the post-exercise value tended to be higher in the injured leg and in the control group (Table 1 ). No differences were found between pre- and post-exercise values in the SP (Table 1 ), or between the groups regarding the effects of exercise (mean difference of post-exercise minus pre-exercise values) (Table 2 ). Figures 5 and 6 show the pre-and post-exercise values for the injured leg and the control group. The third of the post-exercise measurements was lower than the first in the injured leg in both planes, but no differences were found for the uninjured leg or in the control group (Table 3 ). No differences were noted between the groups regarding the effects of exercise directly after cycling (Table 4 ). Figures 9 and 10 show the first and third of the post-exercise measurements on the injured leg and in the control group. No differences were found between the injured leg and the control group, or between the uninjured leg and the control group (Table 5 ). Discussion Short-duration, sub-maximal exercise on a cycle ergometer resulted in increased average speed in both planes, and in the amplitude of CP movements (DEV 10) in the FP during balance in single-limb stance among the patients with ACL injury. In the intra-group comparisons, three of four variables showed higher values post- than pre-exercise in the uninjured leg, and two of four variables were higher post-exercise in the injured leg. In the control group, one of four variables was higher post- than pre-exercise (Table 1 ). However, no differences in the effects of fatigue (mean difference of post-exercise minus pre-exercise values) were found in the inter-group comparisons (Table 2 , and Figures 3 , 4 , 5 , 6 ). The variables were more sensitive in detecting the effects of exercise in the FP than in the SP. The primary motions of the knee joint occur in the SP, and the joint has limited capacity to make postural adjustments in the FP due to anatomical constraints, whereas the hip joint and ankle are involved in postural corrections in both the FP and SP during weight-bearing [ 31 ]. Since many injuries to the knee occur during weight-bearing on one leg [ 21 ]; i.e., in a closed kinetic chain including the hip joint and ankle, it is of interest to examine postural control in both the FP and SP in individuals with ACL injury. The results of a previous study [ 18 ] and the present one indicate that measurements in the FP may be more sensitive and revealing in detecting effects of exercise than measurements in the SP. It has been demonstrated that afferent information has an effect on the neuromuscular function of both the ipsilateral and contralateral limb muscles [ 32 ], which may explain why more variables were higher post- than pre-exercise not only in the injured leg, but also in the uninjured one, than in the control group. Several studies have reported bilateral defects in postural control after an ACL injury or reconstruction [ 2 - 7 ], which may be due to central nervous system modifications following the loss of knee mechanoreceptors after the injury [ 33 , 34 ]. Another explanation may be that the patients had inherently poor balance, which might have contributed to the original injury. This has been reported by Tropp et al. [ 35 ], where soccer players with abnormal stabilometric values (defined as a value exceeding 2 SD of the mean value in a control group), ran a higher risk of sustaining an ankle injury than players with normal values. In a previous study [ 26 ], we observed decreasing values in the three measurements, indicating a learning effect. In another study [ 36 ], fatigue was shown to interfere with this learning process, which is in agreement with the results that we found on the uninjured leg and in the control group. However, the injured leg reacted differently from the uninjured one, and the control group when the first and third of the post-exercise measurements were compared. It was assumed that the first measurement could provide us with the instantaneous value of the ability to maintain postural control in single-limb stance. The results showed that the third measurement was lower, or tended to be lower, than the first in the injured leg, regarding average speed and DEV 10 in both planes. No such effect was, however, found in the uninjured leg or in the control group (Table 3 ). The inter-group comparisons for these post-exercise measurements showed greater effects in the injured leg than in the control group in average speed in the FP, and a tendency towards greater effects in the other three variables (Table 4 and Figures 7 , 8 , 9 , 10 ). This finding indicates that balance standing on one leg may be improved during the recovery period, and that a learning process may be needed in the injured leg after exercise. A different strategy in the injured leg than in the uninjured one has been reported in individuals with ACL injury [ 37 ]. In that study, Di Fabio et al. [ 37 ], found that postural responses, measured with external perturbations while standing on a force platform, could be unilaterally restructured and preprogrammed to compensate for the injury. Mechanoreceptors in the ACL contribute to the neuromuscular control of the muscle tonus around the knee joint via the reflex arc (i.e., reflex from joint afferents to the muscle spindles via the gamma motoneurons), and therefore to the stabilization of the knee joint [ 32 ]. Decreased proprioception [ 11 , 12 ], increased joint laxity in the knee joint [ 14 , 38 ], and a delay in muscle response in leg muscles [ 13 , 14 ] have been described after fatiguing exercise. In these studies, uninjured subjects were tested. The activity of joint receptors, muscle spindles and Golgi tendon organs may be reduced by fatigue, resulting in proprioceptive deficiency in muscle receptors and loss of muscular reflexes responsible for joint stability [ 10 ]. Since this afferent information is important for the maintenance of postural control [ 32 ], this may lead to decreased muscle response and poorer ability to maintain balance. The increase in joint laxity following fatigue has been suggested to be due to reduced muscle tone [ 38 ], viscoelastic changes in the collagenous tissues of the knee and fatigued muscle stabilizers [ 14 ], and results in inadequate ligament mechanoreceptor feedback, which is required to elicit the muscular reflexes responsible for joint stability [ 10 ]. It has been suggested that muscle receptors are the primary determinant of joint position sense, and capsular receptors may have a secondary role [ 12 , 32 ]. Therefore, the decreased proprioceptive ability following fatigue has been proposed to be due to the decrease in muscle receptor activity [ 11 , 12 ]. Since defects in proprioception [ 39 ], impaired postural control [ 2 - 5 ], increased joint laxity [ 32 ], and a delay in muscle reaction time [ 4 , 40 , 41 ] are present already in an unfatigued state in individuals with ACL injury, they may, at least theoretically, be more affected by fatigue than uninjured subjects. Although we found effects of exercise after a short period of cycling above 60% of the predicted HR max , it is possible that greater effects of exercise on balance in single-limb stance may be seen after longer durations of exercise than in the present study. It is also possible that larger effects of exercise may be reflected in more challenging measures of postural control, such as dynamic balance tests. Since, to our knowledge, this is the first study on the effects of fatigue on postural control in patients with ACL injury, the clinical relevance of our results remains unclear. More research is needed to further study whether this may be related to an increased risk of further injuries. The lower average speed and lack of difference in DEV 10 in the patients compared to the control group, indicate sway movements at a lower speed with retained amplitudes to be neuromuscular adaptive strategies, rather than more rapid, smaller adjustments (Table 5 ). These strategies may be the result of decreased proprioception [ 39 ], and a delay in muscle reaction time [ 4 , 40 , 41 ], which has been reported after an ACL injury, and thus, these strategies may be needed to generate sufficient afferent impulses to obtain dynamic stabilization of the knee joint. Another possible explanation may be that the patients had all undergone neuromuscular training, which may have affected the strategies of maintaining balance in single-limb stance compared with the control group who had not undergone such training. The clinical relevance of the fact that the patients' post-exercise values approached those of the control group, remains, however, unclear. More research is needed to elucidate this further. Conclusions The results of the present study showed no differences in the effects of exercise between the patients and the controls. However, the injured leg was more affected or tended to be more affected directly after exercise than the control group, which indicates that patients with ACL injury react differently regarding their ability to maintain balance in single-limb stance after short-duration, sub-maximal cycling, than a control group of uninjured subjects. The patients used sway movements at a lower speed with retained amplitudes, which may be an expression of neuromuscular adaptive strategies. Competing interests The author(s) declare that they have no competing interests. Authors' contributions EA participated in the design of the study, participated in collecting the data, performed the statistical analysis, and drafted the manuscript. DR participated in collecting the data. EH participated in the progress and revision of the manuscript. TF participated in the design of the study, and in the progress and revision of the manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544556.xml |
497041 | Assessing harmful effects in systematic Reviews | Background Balanced decisions about health care interventions require reliable evidence on harms as well as benefits. Most systematic reviews focus on efficacy and randomised trials, for which the methodology is well established. Methods to systematically review harmful effects are less well developed and there are few sources of guidance for researchers. We present our own recent experience of conducting systematic reviews of harmful effects and make suggestions for future practice and further research. Methods We described and compared the methods used in three systematic reviews. Our evaluation focused on the review question, study designs and quality assessment. Results One review question focused on providing information on specific harmful effects to furnish an economic model, the other two addressed much broader questions. All three reviews included randomised and observational data, although each defined the inclusion criteria differently. Standard methods were used to assess study quality. Various practical problems were encountered in applying the study design inclusion criteria and assessing quality, mainly because of poor study design, inadequate reporting and the limitations of existing tools. All three reviews generated a large volume of work that did not yield much useful information for health care decision makers. The key areas for improvement we identified were focusing the review question and developing methods for quality assessment of studies of harmful effects. Conclusions Systematic reviews of harmful effects are more likely to yield information pertinent to clinical decision-making if they address a focused question. This will enable clear decisions to be made about the type of research to include in the review. The methodology for assessing the quality of harmful effects data in systematic reviews requires further development. | Background Systematic reviews are important tools for evidence-based health care. They are certainly one of the reasons for the progress that has been made in obtaining reliable evidence on the beneficial effects of interventions. A recent study of the medical literature, using Medline and the Cochrane Library, showed that the number of systematic reviews published has increased dramatically, from a single publication in the years 1966 to 1970, to 23 in 1981 to 1985, and 2467 in 1996 to 2000 [ 1 ]. Most of the systematic reviews focused on efficacy or effectiveness. However, to make a balanced decision about any intervention it is essential to have reliable evidence on the harms as well as the benefits. Although the coverage of harmful effects has increased over time, only 27% of the reviews published between 1996 and 2000 included any information about safety, and only 4% focused primarily on the safety of the intervention reviewed [ 1 ]. This is perhaps unsurprising as many authors of systematic reviews restrict inclusion to randomised controlled trials (RCTs) to minimise bias, and harmful effects are often inadequately assessed and/or reported in RCTs [ 2 , 3 ]. Another important reason for the relative lack of reliable evidence on harmful effects is that RCTs are not always suitable to evaluate them and other types of study design need to be considered [ 4 ]. The methodology for conducting systematic reviews of beneficial effects from RCTs is well established, whereas the methods for systematically reviewing randomised or observational data on harmful effects are less well developed and less often used. Only 1.25% of 3604 publications cited in the 2001 edition of Side Effects of Drugs (SEDA-24) were systematic reviews [ 5 ]. At present researchers, like us, who conduct systematic reviews have limited sources of guidance, such as the suggestions offered by the Cochrane Collaboration [ 6 ]. Fortunately, research into the methodology of incorporating harmful effects data in systematic reviews is on the increase, from which we expect more sources of guidance to emerge. It is not uncommon, even among experienced reviewers, to assume that the objective of a systematic review of harmful effects should encompass all known and previously unrecognised harmful effects and that data from all types of study design should be sought. We have re-visited three systematic reviews of drug interventions in which we had reviewed harmful effects, to evaluate our own recent experience, identify areas for improvement and to share our ideas with other researchers undertaking reviews. Methods We used three reviews for this study on the basis that they had been completed recently (between 2001 and 2003) and that one of us had been the lead reviewer of harmful effects in each review. The reviews were conducted as Health Technology Assessments for the National Coordinating Centre for Health Technology Assessment (NCCHTA) on behalf of the National Institute for Clinical Excellence (NICE). The reviews, in order of completion, were: nicotine replacement therapy (NRT) and bupropion sustained release (SR) for aiding smoking cessation [ 7 ], atypical antipsychotics for schizophrenia [ 8 ], and newer antiepileptic drugs for epilepsy in adults [ 9 ]. We described and compared the methods used in each review and the problems we encountered in applying those methods. We focused our evaluation on the review objectives, the inclusion criteria for study design and the quality assessment of the primary studies. We do not report on the matter of searching for studies about harmful effects which presents another challenge to those who conduct systematic reviews [ 10 , 11 ], because exploratory work following from the reviews described here is underway and preliminary results are reported elsewhere [ 12 , 13 ]. Results The main components of the three systematic reviews of harmful effects are described in Table 1 . Our evaluation highlighted the following aspects of the methodology that could have been improved on and others that require further development. Table 1 Description of the assessment of harmful effects in the three systematic reviews Review Schizophrenia Smoking cessation Epilepsy Intervention evaluated 8 atypical antipsychotics. NRT and bupropion SR. 7 newer antiepileptic drugs. Objective / Scope Review commissioned by the HTA programme and an update of the HTA report commissioned by NICE. The objective regarding harmful effects was to determine the incidence of specific rare adverse events to populate an economic model. Scope provided by NICE: To review all known or unknown harmful effects that might be associated with the interventions. Scope provided by NICE: To include adverse effects in a review of RCTs of clinical effectiveness in adults with epilepsy. The reviewers undertook a supplementary review of serious, rare and long-term harmful effects. Serious was defined by WHO criteria [25], long-term as longer than 6 months, and rare as defined by the authors of primary studies. Study designs included Randomised trials RCTs of atypical antipsychotics versus alternative drug treatment or placebo in schizophrenia. An existing Cochrane review was used as a source of summary data on adverse effects from RCTs of effectiveness [26]. Studies that assessed safety as the primary objective were included in the review of primary studies of harmful effects. This included RCTs that investigated aspects of clinical pharmacology that might impact on the drugs' tolerability and safety. The five most commonly reported adverse effects were extracted from RCTs as part of the review of clinical effectiveness in epilepsy. RCTs in indications other than epilepsy and dose comparisons were eligible for inclusion in the supplementary review of harmful effects. Non-randomised studies Cohort studies and case series with 2000 or more participants or at least 2 years follow-up, and case-control studies of any size or duration. Uncontrolled trials, prospective and retrospective observational studies, data from adverse events monitoring systems (e.g. UK yellow card scheme) and case reports. Non-randomised controlled trials, cohort and case-control studies, prospective case series and other uncontrolled trials, and open-label extension phases of trials. More than 300 participants had to be exposed or follow-up more than 6 months unless the study objective was to investigate a specific adverse effect. Prescription event monitoring [27], and post-marketing surveillance reports were also included. Studies identified 6477 items screened, 924 articles retrieved, and 223 studies included: 171 RCTs, 13 cohort studies, 1 case-control study, 38 case series. 1280 items screened, 353 articles retrieved, and 123 studies included: 25 RCTs, 4 non-randomised controlled trials, 30 uncontrolled trials, before/after studies or cohort studies, 1 case-control study, 9 surveillance studies, 1 survey, 53 case reports or case series. 108 RCTs were included in the review of effectiveness, selected from 4211 items screened and 887 articles retrieved. In the supplementary review of harmful effects 3884 items were screened, 227 articles retrieved, and 77 studies included: 2 RCTs, 2 non-randomised controlled trials, 26 uncontrolled trials, 14 open-label phases, 25 cohort studies, 1 case-control study, 4 prescription event monitoring studies, 3 post-marketing surveillance studies. Quality assessment Quality checklists for various study designs provided in CRD Report 4 were used [28]. The quality checklist for RCTs provided in CRD Report 4 [28], and checklists for the other study designs published elsewhere [29], were used. Published checklists were used as a starting point. Questions were amended and others added to capture information specifically on the reliability of harmful effects data. Findings of review Very few studies with useful data were found, so the economic model could not be populated with incidence rates of the adverse events of interest. Primarily the findings merely reflected the accepted side-effect profiles for NRT and bupropion SR. The review did not identify any previously unknown harmful effects. The supplementary review of harmful effects did identify reports of potential adverse effects not reported in the RCTs of clinical effectiveness. However, these were mostly effects already documented in tertiary sources. There was insufficient evidence to attribute causality of other reported effects to the test drugs. Review objectives The schizophrenia review objective appeared to be appropriate in seeking to determine the incidence of named outcomes that were considered by health economists to be most likely to lead to a change in prescribed treatment [ 14 ]. The objectives of the smoking cessation and epilepsy reviews were very broad in comparison. Given that the side-effect profiles of the drugs for smoking cessation were well established, with details available in various published standard reference texts [ 15 , 16 ], it would have been more efficient to focus the review effort on a clear question, such as the significance of seizures for bupropion SR and the cardiovascular effects of nicotine in NRT. The objective of the review of harmful effects of the antiepileptic drugs did not target clinical decision-making; the supplementary review of harmful effects might have been of real use to decision makers if we had focused on a crucial clinical question such as the safety of the drugs in pregnancy. Study designs All three reviews included study designs other than RCTs to assess harmful effects. The types of non-randomised studies included for each review reflected differences in the reviews' objectives, our judgment as reviewers as to where the most useful data were likely to be found, and was to some extent pragmatic in terms of the time available to complete the reviews. The reviews with the broad objectives included more non-randomised studies and more diverse study designs. The schizophrenia and epilepsy reviews specified a minimum size and duration of study to be included (see table) in an attempt to add data over and above what was available from the largest and longest RCTs. Doing this did involve some indeterminable risk of missing important information. The review of observational studies carried out in the schizophrenia review was necessary because the pre-determined harmful effects of interest were known to be under-reported in RCTs [ 8 ]. The inclusion of non-randomised studies in the smoking cessation review might have targeted observational data on specific questions about harmful effects had we reviewed beforehand the RCTs that were summarised briefly in the Cochrane review. Similarly in the epilepsy review all the adverse events (not just the most common) reported in the RCTs of clinical effectiveness should have been reviewed before moving on to observational studies. Applying the inclusion criteria Once the inclusion criteria for study design had been defined, applying them was problematic. Reports of primary studies rarely described the study design in sufficient detail. Many of the studies included in the schizophrenia review purported to be cohort studies but on closer examination were in fact large case series involving more than one intervention. Some of the 'cohort study' data on bupropion SR included in the smoking cessation review had actually been derived retrospectively from RCTs. How exactly the 'cohorts' had been established in studies of epilepsy was often unclear in terms of the source population, eligibility criteria, and selection, or was simply not reported. Had we, in all three reviews, only included reports of studies fitting textbook definitions of particular study designs, virtually all of the primary study reports we identified would have been excluded. The inclusive approach we took turned out to be unrewarding. In the smoking cessation review, in addition to difficulties with the study design inclusion criteria, application of the criterion to only include studies in which assessment of adverse events was the primary objective was problematic because it involved a high degree of subjective judgment. Quality assessment We encountered problems when applying published checklists in our reviews of harmful effects. The response to some questions depended on the outcome of interest, for example, follow-up may have been adequate for the assessment of the primary (usually a beneficial) outcome of the study but not for the collection of data on harmful effects. We also found that published checklists omit key features such as how harmful effects data were recorded. In the epilepsy review we were in a position to learn from the earlier reviews and spent time clarifying the questions in the checklists so that they would provide information relevant to the reliability of the harmful effects data. We also added items pertinent to reports of harmful effects such as how and when events were recorded and whether the time at which they occurred during the study was reported. Although this informed approach was a step in the right direction, the major hindrance to applying checklists in all three reviews was inadequate reporting of the basic design features of the primary studies. Once the quality criteria had been applied there remained the challenge of interpreting the results. In our reviews we described the evidence identified and tabulated the response to each checklist question for each primary study. This generated lengthy summaries that had limited utility. Even comparing validity within study designs (not across them) we found it impossible to synthesise the information as all the included studies had methodological flaws and features that could not be assessed due to inadequate reporting. Reaching a decision about which studies were likely to give the most reliable results was not straightforward. Discussion Our experience of reviewing harmful effects mirrors that of other researchers in that a significant investment of effort failed to yield significant new information [ 6 , 17 ]. A focused review question is standard practice for assessing beneficial outcomes in systematic reviews and should also be so when reviewing harms. Researchers conducting reviews need to make sure that they address a well-formulated question about harms that are likely to impact on clinical decisions. Focusing a review question about harmful effects will not necessarily mean restricting it to specific adverse events but may mean, for example, addressing a particular issue such as long-term effects, drug interactions, or the incidence of mild effects of importance to patients. If the aim of the research is to look for previously unrecognised harmful effects, analysis of primary surveillance data may be more appropriate than a systematic review [ 18 ]. Researchers also need to be aware that scopes set by external commissioning bodies, despite having consulted with national professional and patient organisations, may not be a suitable question to address in a systematic review. The wisdom of broad and non-specific questions about harmful effects should be questioned because the resources, especially time, needed to do this comprehensively are usually insufficient. It is important to realise that an unquestioning belief that observational studies are the best source of harmful effects data simply because they are not RCTs can be a pitfall. It is essential to think carefully about the review question before widening the inclusion criteria to include non-randomised study designs. Some harmful effects, such as very rare events or those emerging in the long-term, are unlikely to be addressed adequately in RCTs. But, even if observational studies are appropriate to the review question researchers should be prepared for the difficulty of interpreting observational study data outweighing the anticipated benefits. The importance of quality assessment of RCTs in systematic reviews of effectiveness is well established [ 19 ], but debate continues over the usefulness of checklists and scales. Quality assessment of other study designs in systematic reviews is far less well developed [ 20 ]. Although the feasibility of creating one quality checklist to apply to various study designs has been explored [ 21 ], and research has gone into developing an instrument to measure the methodological quality of observational studies [ 22 ], and a scale to assess the quality of observational studies in meta-analyses [ 23 ], there is as yet no consensus on how to synthesise information about quality from a range of study designs within a systematic review. Our appraisal of our reviews has shown that these difficulties are compounded when reviewing data on harms. It is essential that quality assessment is able to discriminate poor from better quality studies of harmful effects. Levels of evidence hierarchies have several shortcomings. The hierarchy of evidence is not always the same for all harmful or beneficial outcomes. For example, an RCT with adequate internal validity but limited sample size or follow-up may be a less reliable source of information about relatively uncommon harmful effects emerging in the long-term than a large well-conducted cohort study with many years of follow-up. Another problem with ranking evidence in a hierarchy is that different dimensions of quality get condensed into a single grade, resulting in a loss of information. Furthermore, the dimensions included in current hierarchies may not be the most important in terms of reflecting the reliability of a particular study's findings [ 24 ]. Researchers need to clarify a priori what exactly they need to glean from their quality assessment of the primary studies in their own review of harmful effects and it may be necessary to differentiate clearly between internal and external validity. We suggest that further research is needed to collate, assimilate and build on the existing information relevant to systematically reviewing primary studies for harmful effects of health care interventions. This should include a review of the literature pertinent to the methodology of incorporating evidence of harmful effects in systematic reviews; a description and categorisation of the methods used in systematic reviews published to date, and any evidence from methodological research on which they are based; and the development of quality assessment methods. Conclusions Appraisal of our recent experience highlighted some of the problems inherent in conducting systematic reviews of harmful effects of health care interventions. Such reviews need to address a well-formulated question to facilitate clear decisions about the type of research to include and how best to summarise it, and to avoid repeating what is already known. The review question about harmful effects needs to be relevant to clinical decision-making. A systematic review of the methodology pertinent to systematic reviews of harmful effects is warranted. Competing interests None declared. Authors' contributions AMB, HMM and NFW conducted the review work described. NFW conceived of the study. NFW and HMM drafted the manuscript. All authors contributed to, read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC497041.xml |
544957 | Association between monocyte Fcγ subclass expression and acute coronary syndrome | Background Atherosclerosis lesions contain abundant immunoglobulins complexed with oxidized LDL (OxLDL) that are endocytosed by macrophages to form foam cells. While recent evidence supports a role for the macrophage scavenger receptor pathway in 75–90% of OxLDL uptake, in vitro evidence suggests another potential uptake pathway could involve autoantibody binding to IgG subclass-specific Fc receptors. Objective and Methods To address this mechanism from an in vivo standpoint, the objective of this study was to utilize flow cytometry to prospectively determine monocyte Fcγ (FcR) I, II, and III receptor expression levels in patients with acute coronary syndrome (ACS, n = 48), diabetes mellitus (DM, n = 59), or neither (C, n = 88). Results Increased FcR I expression was found in the ACS versus DM groups [geometric mean, (95% CI) = 2.26 (2.07, 2.47) versus 1.83 (1.69, 1.98) (p < 0.001)] and versus C [1.90 (1.78, 2.03) (p = 0.005)]. Similar relationships were found with both the FcR II receptor [ACS mean = 4.57 (4.02, 5.19) versus DM 3.61 (3.22, 4.05) (p = 0.021) and versus C 3.86 (3.51, 4.24) (p = 0.09)] and FcR III receptor [ACS mean = 1.55 (1.44, 1.68) versus DM 1.36 (1.27, 1.46) (p = 0.038) and versus C 1.37 (1.30, 1.45) (p = 0.032)]. There was no difference between DM and C groups in FcR I, II or III expression. Conclusions This in vivo data supports a possible second OxLDL-autoantibody macrophage uptake mechanism through an Fc receptor-mediated pathway and a potential relationship between atherosclerotic plaque macrophage FcR levels and ACS. | Introduction Atherosclerosis is a chronic inflammatory process that results from hyperlipidemia and complex interactions involving other genetic and environmental factors. OxLDL plays a central role in the atherogenic process through generation of highly immunogenic neodeterminants for the immune system [ 1 ]. Natural autoantibody titer to a number of these epitopes and extent of immune complex formation may correlate with plaque size and rate of progression, and plaques have been shown to contain OxLDL/autoantibody immune complexes [ 2 - 5 ]. It is clear that both innate and adaptive immunity can modulate lesion progression and composition, and most studies to date have indicated a proatherogenic influence of the immune system on this process [ 1 , 4 ]. Recent evidence supports the macrophage scavenger receptors SR-A and CD36 as a mechanism responsible for up to 90% of uptake of OxLDL that leads to foam cell formation with no other scavenger receptors compensating for their absence in a knockout mouse model [ 6 ]. Earlier evidence involving in vitro incubation of both human monocyte-derived macrophages and the monocytic cell line THP-1 with human LDL-rabbit anti-apo B immune complexes demonstrated a potential role for the FcγRI receptor in its uptake [ 7 ]. A second in vitro study also suggested a potential Fc receptor role through inhibition of immune complex uptake when Fab or F(ab') 2 fragments were substituted for an intact anti-apo B antibody [ 8 ]. Though findings from the latter two studies may have been partly explained by contributions from the scavenger pathway, it is reasonable to speculate the Fc receptor pathway maybe playing a small but important role as well [ 9 - 11 ]. Immune complexes with modified lipoproteins have recently emerged as an important coronary artery and macrovascular disease risk factor in DM [ 12 , 13 ]. Evidence supports an increased content of macrophages in the atherosclerotic lesions of persons with DM that is thought to be due to altered levels of cytokines [ 12 ]. Furthermore, while DM itself does not increase levels of LDL, the small dense LDL particles found in type 2 DM are more atherogenic because they are more easily glycated and are thought to be more susceptible to oxidation [ 14 , 15 ]. In recent work our group has shown FcγRII expression to be increased in the platelets of patients experiencing an acute atherothrombotic event, or who are healthy with two or more atherosclerosis risk factors [ 16 ]. Non-acutely ill diabetes patients have significantly elevated expression levels and this may play a role in the increased sensitivity of their platelets to activation by subendothelial collagen [ 16 - 19 ]. We speculate that Fc expression levels and activity on macrophages and platelets may represent another link between the immune system and atherosclerosis progression and plaque disruption. In view of the controversy regarding the mechanism of cholesterol uptake by monocyte-macrophages in atherosclerosis and diabetes [ 20 , 21 ] and the previous lack of in vivo data to help elucidate any role of the Fc receptors in this process, we have prospectively determined IgG-binding receptor expression levels for each Fcγ receptor subclass on the monocytes of three groups: (1) patients admitted to the hospital with ACS, (2) well patients with no history of heart disease but one or more atherosclerosis risk factors (ARF's) that included DM, and (3) control patients (with no history of ACS or DM). Materials and Methods All 195 patients were randomly chosen for study participation from a larger group who fit study inclusion criteria and gave written informed consent. Forty-eight patients in the study had heart disease (HD) and were within 2 hours of onset of an ACS (myocardial infarction or unstable angina), 59 were DM outpatients (both type 1 and type 2 were included) with no known history of HD, and an additional 88 outpatients without HD or DM were randomly chosen as controls (C). The number and nature of ARF's was documented for each group (Table 1 ). Table 1 Demographic characteristics at enrollment Characteristic Group 1 (ACS) Group 2 (DM) Group 3 (C) p-value 1 Total patients (% with MI Group 1) 48 (52) 59 88 Mean age (years) 56 55 56 0.86 2 Male (%) 35 (73) 24 (41) 55 (63) 0.002 Positive family history (%) 3 6 (13) 21 (36) 21 (24) 0.018 Current cigarette smoker (%) 16 (33) 10 (17) 5 (6) <0.001 Hypertension (%) 29 (60) 29 (49) 20 (23) <0.001 Abnormal lipid profile present (%) 24 (50) 12 (21) 17 (19) <0.001 Diabetes mellitus present (%) 15 (31) 59 (100) 0 (0) <0.001 1. Based on chi-square test. 2. Based on ANOVA F-test 3. One patient in group 2 had missing data with positive family history (Abbreviations: ACS: acute coronary syndrome; DM: diabetes mellitus; C: control; MI: myocardial infarction) Blood was collected in 3.8% trisodium citrate and divided into 50 μl aliquots to which 5 μl of a saturating concentration of anti-FcR I (32.2), anti-FcR II (IV.3), anti-FcR III (3G8), or a negative class-specific control antibody (MOPC-141, Sigma) was added. Following a 15 minute incubation, 5 μl of FITC-sheep anti-mouse antibody (Sigma) was added and a second 15 minute incubation done. The pellet was washed twice before 5 μl of phycoerythrin-conjugated anti-CD14 (Becton Dickinson) was added and incubated for 15 minutes. The solution was diluted with 1 ml ammonium chloride lysing solution and incubated for 10 minutes or until the solution was clear, and the pellet washed twice before flow cytometry to determine relative receptor expression levels was carried out according to the manufacturer's specifications (Becton Dickinson). Monocytes were readily identifiable from other blood cells by their forward and side scatter properties along with CD14 expression. Following establishment of saturating concentrations for each antibody, mean inter and intra-individual coefficients of variation for each of the three Fc receptors were calculated in the antibody labeling assay employing blood samples from five healthy laboratory volunteers who met control patient criteria and FcR I values found to be 3.2 and 9.7%, FcR II 11.7 and 16.1% and FcR III 4.1 and 26.9% respectively. The ratios of FcR I, FcR II, FcR III and MOPC-141 antibody expression were calculated for each patient and the logarithm of the ratios were used to analyze the results. An analysis of variance (ANOVA) was performed to compare the groups of (1) these 3 ratios with the 3 groups of patients, (2) the 3 ratios with the total number of major ARFs, (3) the 3 ratios with each ARF, such as hypertension, diabetes, and smoking, and (4) the 3 ratios with MI, and unstable angina in group 1. The overall p-values were based on the ANOVA F-test. If the overall F-test p-value < 0.05, the LSD method (least significant difference) [ 22 ] was used for multiple comparison. The geometric means and the associated 95% confidence intervals were calculated to summarize the data. Patient baseline data between groups was analyzed using the chi-square test. To perform statistical analysis, SAS software, version 8.2, was used (SAS Institute, Inc., Cary, NC). Results Significantly increased FcR I expression was found in ACS patients compared with DM patients [geometric mean FcR I expression, (95% CI) = 2.26 (2.07, 2.47) versus 1.83 (1.69, 1.98) (p < 0.001)] and compared with C [1.90 (1.78, 2.03) (p = 0.005)] (Table 2 , Figure 1 ). Similar relationships between the three groups were found to exist employing antibodies specific to the FcR II receptor: ACS geometric mean (CI) = 4.57 (4.02, 5.19) versus DM 3.61 (3.22, 4.05) (p = 0.021) and versus C 3.86 (3.51, 4.24) (p = 0.09) and the FcR III receptor: ACS geometric mean (CI) = 1.55 (1.44, 1.68) versus DM 1.36 (1.27, 1.46) (p = 0.038) and versus C 1.37 (1.30, 1.45) (p = 0.032). There was no difference between DM and C groups in FcR I, II or III expression (p = 0.73, 0.66, and 0.99 respectively). Table 2 Mean monocyte FcR expression in 195 study patients FcR I FcR II FcR III N Geometric mean (95% CI) p-value Geometric mean (95% CI) p-value Geometric mean (95% CI) p-value Unadjusted analysis Groups <0.001 2 0.024 2 0.021 2 Group 1 (ACS) 48 2.26 (2.07, 2.47) group 1 vs. 2: 0.001 3 4.57 (4.02, 5.19) group 1 vs. 2: 0.021 3 1.55 (1.44,1.68) group 1 vs. 2: 0.038 3 Group 2 (DM) 59 1.83 (1.69, 1.98) group 1 vs. 3: 0.005 3 3.61 (3.22, 4.05) group 1 vs. 3: 0.09 3 1.36 (1.27, 1.46) group 1 vs. 3: 0.032 3 Group 3 (C) 88 1.90 (1.78, 2.03) group 2 vs. 3: 0.73 3 3.86 (3.51, 4.24) group 2 vs. 3: 0.66 3 1.37 (1.30, 1.45) group 2 vs. 3: 0.99 Adjusted analysis 1 Groups 0.010 2 0.061 2 0.10 2 Group 1 (ACS) 48 2.33 (2.14, 2.55) group 1 vs. 2: 0.007 3 4.57 (4.00, 5.22) group 1 vs. 2: 0.050 3 1.58 (1.45,1.71) group 1 vs. 2: 0.10 3 Group 2 (DM) 59 1.95 (1.78, 2.13) group 1 vs. 3: 0.09 3 3.71 (3.24, 4.24) group 1 vs. 3: 0.47 3 1.41 (1.30, 1.53) group 1 vs. 3: 0.22 3 Group 3 (C) 88 2.05 (1.87, 2.25) group 2 vs. 3: 0.56 3 4.11 (3.59, 4.71) group 2 vs. 3: 0.38 3 1.44 (1.32, 1.56) group 2 vs. 3: 0.43 3 1. Adjusted by smoking and hypertension status 2. p-value based on F-test 3. p-value based on Tukey-Kramer test Figure 1 Monocyte Fcγ receptor subclass expression levels of 48 patients with heart disease (HD) and 59 patients with DM compared with 88 control patients with neither HD nor DM. HD patients display significantly increased expression levels of all 3 subclasses versus controls. * p = 0.002, 0.037, and 0.014 for FcR I, FcR II, and FcR III respectively when HD group is compared to control group. There were no statistically significant associations with increased expression of any Fc receptor and gender, family history of premature coronary disease, diabetes, or abnormal lipid profiles. Current cigarette smoking significantly increased expression of FcR I, 2.24 (2.01, 2.50), compared to absence of current cigarette smoking, 1.91 (1.82, 2.01) (p = 0.010) (Table 2 adjusted analysis). FcR II was significantly increased among the patients with hypertension, 4.29 (3.88, 4.74) compared to those without hypertension 3.73 (3.44, 4.05) (p = 0.034). There was a slight association between age and FcR I, II or III expression (p = 0.042, 0.050, 0.022 respectively). Expression levels in younger (age < 45) and older (age > 55) groups were higher than the middle age group. No difference was found in FcR expression with respect to ARF number (with the lone exception of increased FcR III in patients with 2 or more ARFs compared with less than two), or between the ACS subgroups of acute MI and unstable angina. When FcR expression is compared between all diabetes and non-diabetes patients in the 3 groups, there is no difference in monocyte FcR I, II, or III expression. Discussion It can be speculated from this in vivo data that phagocytosis of OxLDL-autoantibody immune complexes by plaque-associated macrophage through an Fc-mediated pathway could be a second uptake mechanism in addition to that involving the scavenger receptors. The potential clinical implication behind these findings is that while marrow and blood monocyte scavenger receptors SR-A and CD36 have not demonstrated inter-individual variability in their basal expression levels (prior to initial uptake of OxLDL or differentiation to macrophages) [ 23 , 24 ], the variable expression of Fcγ receptors found in this series of patients maybe playing a role in the extent of OxLDL immune complex uptake by atherosclerosis plaques. The fact we were able to document relatively increased surface expression of all three receptor classes in patients with ACS, along with increased FcR I in those who smoked and FcR II in those with hypertension, supports this hypothesis. Any precise pathophysiological implication behind these findings, though, or any cause and effect relationship between monocyte Fc expression and ACS is presently uncertain. No difference was noted in expression of any Fc receptor between the diabetes and control groups. Given that the control group was the largest of the three and that there was a difference noted in the expression levels of all three receptors between the two smaller groups, it can be concluded that increasing the sample size of either of the two groups with similar expression levels could possibly lead to an increase in the difference between them, but this difference is still unlikely to be of any significance compared with that between the ACS group and the other two groups. There was an interesting trend in both the unadjusted and adjusted analyses (Table 2 ) in which the control groups had consistently higher monocyte Fc expression levels compared with the diabetes groups. One potential explanation for this would be the hypothesis that those with diabetes may have reduced FcR expression levels (with a possible consequent decreased uptake in oxidized LDL) compared with non-diabetes subjects in response to relatively higher levels of molecular mediators that support atherosclerosis progression and a pro-inflammatory, pro-thrombotic environment. This may reflect a biological ying-yang type of response that leads to an attempt at dampening the effects of molecular players capable of contributing to atherothrombotic events. An issue pertinent to this study would be the possible effects of inflammation on monocyte FcR expression levels. In diabetes patients it would be reasonable to speculate a significant number of activated cells in the circulation would be unlikely since in most patients with atherosclerosis the inflammatory reaction is circumscribed to the vessel wall. Overexpression of monocyte Fc receptors may have been a possibility in ACS however. Figure 1 shows the ratio of ACS/control mean FcR expression levels to be very similar between Fc receptor subtypes. The extent of increased expression associated with inflammation-associated monocyte activation has been shown to be variable between FcR subtypes [ 25 ]. In this respect FcR II and III represent the receptors primarily involved in the inflammatory response in vivo. Since all three receptors had uniform increases in expression levels in ACS compared with controls (with the FcR I ratio being the highest of the three), it may be reasonable to attribute a relatively minimal effect of acute inflammation to the ACS data. Circulating monocyte activation may also have returned relatively close to baseline as a consequence of blood being drawn around 2 hours after symptom onset in most cases. The average circulation time of blood monocytes in response to an inflammatory stimulus may fall to as little as 30 minutes [ 26 ]. Observational studies of this nature have certain limitations related to their design and patient selection. As an example, selection bias needs to be considered in any study involving a population of volunteers associated with an atherosclerosis prevention trial (Groups 2 and 3). Even though the ethnic composition of the groups was the same, the implication of this selection bias is that the results are not generalizable to the population at large and this is attested to by the demographic characteristics in Table 1 . Through bridging innate and adaptive immune processes, macrophages play an important role in the progression of atherosclerosis and mediating plaque disruption that is considered to be the inciting event in the majority of coronary thrombi [ 27 , 28 ]. In this respect there is continual migration of monocytes between neighboring endothelial cells as well as two-way migration of monocytes between blood and OxLDL-containing foam cells when there is separation of endothelial cells associated with the fatty streak [ 29 ]. This observation and the inherent difficulty in isolating monocytes from tissues compared with serum both served as justification for utilizing blood monocyte Fc expression as a surrogate for plaque macrophage expression [ 30 ]. The overall effect of the humoral immune response on atherogenesis is likely to be complex. Of note, for example, is that the FcγRII receptor has an inhibitory role on B cells that are rarely seen in plaques, while it mediates phagocytosis and release of inflammatory mediators from cells of the myeloid lineage when cross-linked by immune complexes [ 31 , 32 ]. Thus FcR binding by opsonized OxLDL could induce either negative or positive regulation of immune cell responses. Elucidation of the immune mechanisms involved in atherogenesis will continue to evolve and lead to new insights into the molecular pathways associated with disease progression. Ultimately these insights will contribute towards the full explanation behind the clinical diversity of atherosclerosis expression in patients who appear to have equal risk. Competing Interests The authors declare that they have no competing interests. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544957.xml |
512293 | Family-based clusters of cognitive test performance in familial schizophrenia | Background Cognitive traits derived from neuropsychological test data are considered to be potential endophenotypes of schizophrenia. Previously, these traits have been found to form a valid basis for clustering samples of schizophrenia patients into homogeneous subgroups. We set out to identify such clusters, but apart from previous studies, we included both schizophrenia patients and family members into the cluster analysis. The aim of the study was to detect family clusters with similar cognitive test performance. Methods Test scores from 54 randomly selected families comprising at least two siblings with schizophrenia spectrum disorders, and at least two unaffected family members were included in a complete-linkage cluster analysis with interactive data visualization. Results A well-performing, an impaired, and an intermediate family cluster emerged from the analysis. While the neuropsychological test scores differed significantly between the clusters, only minor differences were observed in the clinical variables. Conclusions The visually aided clustering algorithm was successful in identifying family clusters comprising both schizophrenia patients and their relatives. The present classification method may serve as a basis for selecting phenotypically more homogeneous groups of families in subsequent genetic analyses. | Background Schizophrenia is a severe mental illness which tends to run in families. Moreover, schizophrenia is a complex disorder with multiple environmental as well as genetic predisposing effects. Previous studies have shown that many neuropsychological functions are impaired in schizophrenia patients, and, to a lesser degree, also in their unaffected relatives [ 1 - 3 ]. Consequently, the continuous traits derived from neuropsychological tests have been suggested as one type of endophenotypes of schizophrenia to be included in genetic analyses [ 4 - 9 ], for a review see Egan and Goldberg [ 10 ]. Identifying more homogeneous subgroups of families with a similar pattern of cognitive test performance would further refine the data to be included in these analyses. Recently, cluster analysis of verbal learning and memory tests was used to divide patients with schizophrenia into subtypes. Categorization by these cognitive traits resulted in meaningful subgroups of schizophrenia [ 11 ]. In another study, extended neuropsychological test data of patients with schizophrenia were included in a hierarchical and iterative partitioning cluster analysis [ 12 ]. Four clusters were identified, ranging from good performance to profound global dysfunction. In Sautter et al. [ 13 ] an exploratory study comparing clustering of neuropsychological test performance in schizophrenia patients with familial history to those without was performed. In their analysis, patients with family history fell into three distinct clusters, while only one homogeneous cluster was found for the non-familial group. However, only patients were included in the analyses of these studies. As schizophrenia is likely to be a multifactorial disorder with low penetrance, the inclusion of relatives in the clustering analyses would be a powerful way to reveal subgroups based on the endophenotype of interest. In the present study, we report a new visually aided clustering approach aimed at identifying clusters of multiply affected families with schizophrenia on the basis of performance in neuropsychological tests. In the clustering process, each family was represented by the test scores of its affected and unaffected members, and the closeness the families was defined by the maximum pairwise distance between the members of the families. To our knowledge, this is the first study in which the clustering has been applied to families instead of solely to affected subjects with schizophrenia. Methods Subjects and data collection From a general population cohort of people born between 1940 and 1976 inclusive in Finland, a northern European country with approximately 5 million inhabitants, we identified 33,731 individuals with a diagnosis of schizophrenia, schizoaffective disorder or schizophreniform disorder. Data on the diagnosis were derived from three nation-wide computerized health care registers covering the years 1969 to 1998: the Hospital Discharge Register, the Free Medicine Register, and the Pension Register. Linking the personal identification numbers of the affected subjects to the National Population Register database allowed us to identify their family members and to construct pedigrees. Information on families with at least two members with schizophrenia, schizoaffective disorder or schizophreniform disorder, and at least two members with no diagnosis of psychiatric disorder was received from the aforementioned registers for 895 families from the whole of Finland. A blood sample for subsequent genetic analyses was drawn from 2295 subjects of 643 families. All available case note records were collected for those with a diagnosis of schizophrenia, schizoaffective disorder or schizophreniform disorder in any of the three registers. Two psychiatrists independently assessed the lifetime diagnoses for each case, according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) [ 14 ]. One of the assessors also completed the Operational Criteria Checklist for Psychotic Illness (OPCRIT) [ 15 ]. The collection of blood samples complied with the Declaration of Helsinki and its amendments. The protocol was accepted by the Ethics Committee of the National Public Health Institute, and the study was approved by the Ministry of Social Affairs and Health. Of those multiply affected families who already had given the blood samples, a subsample was targeted for collection of more detailed phenotypic information. This sample was selected randomly based on the data from the registers and the OPCRIT process. All subjects from the families gave a written informed consent for the study protocol comprising a diagnostic research interview and neuropsychological testing. Both patients and their family members were interviewed using the Structured Clinical Interview for DSM-IV (SCID-I for axis I disorders and SCID-II for axis II disorders) [ 16 ]. All the interviewers were trained in a similar manner for the use of these instruments. The final consensus diagnoses were based on the data collected from the records, the OPCRIT process, and the SCID interview. A total of 281 subjects from 54 families fulfilled the inclusion criteria and thus included at least two siblings with schizophrenia, schizoaffective disorder or schizophreniform disorder, and at least two siblings without these disorders. Altogether 16 patients were excluded because of being too psychotic ( n = 6), having a current substance use diagnosis ( n = 6), or being mentally retarded ( n = 4). Of the family members to whom no psychiatric diagnosis was assigned for their lifetime, 6 were excluded because of high age, or for a defect in vision or hearing. The final sample thus comprised 165 subjects with a psychiatric diagnosis and 94 unaffected family members from 54 families. Of the 165 subjects with a diagnosis, altogether 82 subjects had schizophrenia, while 13 subjects suffered from schizoaffective disorder, 10 from schizophreniform disorder and 12 from bipolar disorder. A nonpsychotic disorder was assigned to 48 individuals. The 94 unaffected subjects did not get any current or lifetime psychiatric diagnosis. In 51 families, at least one of the patients included in the analysis suffered from pure schizophrenia. In the remaining three families, at least one subject with schizoaffective or schizophreniform disorder was included. All families from which the subjects for the present study were drawn, represent familial schizophrenia, as in each of them there were at least one sibling with a diagnosis of pure schizophrenia, plus at least one other sibling with schizophrenia, schizoaffective disorder or schizophreniform disorder. Test procedures A neuropsychological test battery was administered to all subjects in fixed order by well-trained examiners either after the interview during the same day, or the following day. All examiners were psychologists or advanced psychiatric nurses extensively trained and supervised with the test battery. Experienced psychologists scored all the tests. Auditory attention was assessed with the Digit Span Forward task, and verbal working memory with the Digit Span Backward task of the Wechsler Memory Scale-Revised (WMS-R) [ 17 ]. According to Finnish normative data, the test-retest reliability coefficients of the Span subtests vary with age from 0.74 to 0.82 [ 18 ]. The Visual Span forward subtest of the WMS-R [ 17 ] was used to assess visual attention. The backward condition of the span task was used for measuring visual working memory. According to Finnish normative data, the test-retest reliability coefficients of the Visual Span subtests vary with age from 0.72 to 0.80 [ 18 ]. The Logical Memory story A, of the WMS-R [ 17 ], immediate and delayed, was used to assess recall and retention in a story format. Visual memory was measured by the Visual Reproduction subtest of the WMS-R [ 17 ], immediate and delayed. In Finnish normative data, the test-retest reliabilities of these subtests have varied with age from 0.84 to 0.91, and 0.31 to 0.34, respectively. Verbal learning and memory were assessed with the California Verbal Learning Test (CVLT) [ 19 ] which examines recall and recognition of word lists over a number of trials. The present study reports the following variables derived from the test: verbal learning (total recall over 5 trials), semantic clustering, and recognition memory (discriminability). No reliability data for Finnish subjects exist, but the split-half reliability of the CVLT is 0.77 to 0.86, according to the test manual [ 19 ]. Controlled Oral Word Association test (COWA) [ 20 ] was used to assess verbal fluency. The quantity of words the subject produces in one minute, both with words beginning with a designated letter (S,K), and within a category (animals), was assessed. No reliability data for Finnish subjects are available. Four subtests of the Wechsler Adult Intelligence Test – Revised (WAIS-R) [ 21 ] were used. Verbal abilities were measured with the Vocabulary and Similarities subtests. Vocabulary is considered the best single measure of general ability [ 22 ]. The Similarities subtest is a task of abstraction and concept formation. The Block Design and Digit Symbol subtests have a motor component as the trials are timed. The former is a measure of visuospatial reasoning and abstraction. The latter subtest measures psychomotor performance. According to Finnish normative data, the test-retest reliabilities for Vocabulary, Similarities, Block Design, and Digit Symbol are 0.89–0.95, 0.69–0.88, 0.78–0.83, and 0.82–0.86, respectively, depending on age [ 23 ]. Clustering and statistical analyses Notation and imputation of missing values The variables used in cluster analysis included 17 neuropsychological test variables together with the age and the sex of the subjects. With a total of M = 19 variables and N = 259 subjects, the data formed an M × N matrix x = ( x ik ), where x ik is the the value of the i th variable for the k th subject. However, there were 85 (1.7 %) missing values as not all test results were obtained for all subjects. The missing values were handled by the following procedure, which replaces an individual's missing value with an estimate obtained from a linear fit between the test with the missing value and the test that correlates with it most and that also has the individual's test result available. 1. Pairwise correlations were calculated between all test variables using only subjects with results available for both tests. We denote such correlation between the tests i and j by c ij . 2. Given a missing value in the test i for the subject k , we found the test j = j 0 which had the highest value of |c ij | among the tests with the value x jk available and set where the coefficients a and b were found by computing linear regression of the test j 0 on the test i using only subjects with results available for both tests. Cluster analysis The families were clustered using a complete-linkage clustering algorithm. Each variable was normalized by subtracting the mean value and dividing by the standard deviation. The normalization was done to ensure that each variable contributes equally to the clustering procedure. Denote by x k = ( x 1 k ,..., x Mk ) the normalized data for subject k and define the distance between two clusters C r and C s by d rs = max{|| x k - x l || : x k ∈ C r and x l ∈ C s }, that is, d rs is the maximum pairwise distance between members of the two clusters. Here ||·|| denotes the euclidean distance, . In the sense of this distance measure, two clusters are close when all subjects in both clusters are close. Clustering was carried out using the following algorithm. 1. Initial clusters are defined by the families. 2. The two clusters with the smallest inter-cluster distance d rs are merged into one larger cluster. 3. Steps 2 and 3 are repeated until a desired number of clusters remains. In Figure 1 , two steps of the above procedure are demonstrated. Three clusters are depicted by the green solid lines. The two nearest clusters are combined (the dashed green line). Their inter-cluster distance d rs is shown by the solid red line. The inter-cluster distance between the two remaining clusters is shown by the dashed red line. Note that by using a different inter-cluster distance measure, such as the minimum pairwise distance, a different merging order would result (see the Discussion). Figure 1 Visualization of clustering. Two merging steps of the clustering algorithm (see the text). Visualization of clusters We introduce a visualization technique that helps in identifying candidate clusters and also gives an overall picture of the main differences between the produced clusters as measured by all variables simultaneously. The method gives information about the dynamics of the clustering process and the characteristics of the candidate clusters. The upper part of Figure 2 presents the data matrix as what is called the "color histogram" [ 24 ] or "the data image" [ 25 ]. The rows correspond to variables and the columns correspond to subjects. The values of the neuropsychological tests and other variables are visualized using color coding. The color of a variable changes from blue to red as its value increases. To better utilize the dynamic range of coloring, 5 % of the highest values and 5 % of the lowest values were set to the 95th percentile and the 5th percentile of the test results, respectively. The variables were then shifted and scaled to the interval [0,1] and color coded (0 = blue, 1 = red). On the last line of the color histogram, clusters are depicted with different colors. Figure 2 Visualization of clustering result. A visualization of the data and the cluster solution using the data image and the dendrogram. On the last line of the data image clusters 1, 2 and 3 are depicted with different colors. Both the subjects and the neuropsychological tests were ordered by a complete linkage clustering algorithm (see the text). To further improve the visual impression of the clusters, the neuropsychological test variables (the rows of the data image) were ordered using essentially the same procedure that was used in clustering the families. The initial clusters were now the individual variable vectors x i = ( x i 1 ,..., x iN ) and the pairwise distance between two clusters C r and C s was defined as d rs = max{1/| c ij | : i ∈ C r , j ∈ C s }, where c ij denotes the correlation between the variables i and j . Thus, at each step, the algorithm merged clusters with the highest correlating variables. The lower part of Figure 2 visualizes the actual clustering process using the dendrogram. The history (vertical direction) of the mergings is shown from the beginning (one family in each cluster) to the end (all families in one cluster). By simultaneously exploring the two images, a reasonable value for the number of clusters can be found and the characteristics of the cluster solution visualized in a useful manner. It is also helpful to monitor the inter-cluster distance measure for possible large jumps which indicate that two distant clusters are being merged (Figure 3 ). Figure 3 Merging distances. The inter-cluster distance (merging distances) as a function of the number of clusters. The vertical line indicates the suggested three cluster result, after which there is a clear jump in the merging distances. Validation of cluster result The clusters were obtained by treating families as single objects whose dissimilarity was measured by the pairwise test performance differences between the family members. One may therefore ask whether the clusters found still appear to be distinct groups when viewed simply as sets of individual subjects. We examined this question by dividing repeatedly the 54 families into three random clusters that had the same number of families as in the proposed three cluster solution and by computing, for each of the three pairs of the generated clusters, the ratio BW r , s = B r , s /( W s + W r ), where B r , s is the mean distance between subjects from clusters r and s (in the 19-dimensional space) and W r is the mean distance between subjects within cluster r . The statistic BW r , s takes on a large value if the distance between the subjects from the different clusters is large compared to the distance between the subjects within the clusters themselves indicating that the two clusters are separated in the 19-dimensional space defined by the variables used. If the values of BW r , s for the proposed three clusters are significantly higher than for a random partition we take this as evidence that the clusters found indeed constitute meaningful groups also at the level of individual subjects. Further, after the cluster analysis, the proposed family clusters were examined for differences on demographic and neuropsychological measures. In addition, the patients included in the clusters were examined for the differences in clinical variables as evaluated by the OPCRIT (premorbid social adaptation, response to neuroleptic treatment, chronicity, age of onset) of the disorder. In comparing the demographic and clinical variables, the Chi-square test, or t-test, both two-tailed, were applied. The differences in the quantitative neuropsychological measures were analyzed using the linear mixed effects (LME) model, which takes into account the dependence between the subjects, who, a priori, came from the same families. Thus, family was included as a random effect in all models with age and sex as the fixed effects. In addition, post hoc models were conducted with education years as an added fixed effect, a known confounder for cognitive functions. In all these analyses, the probability level < 0.05 indicated statistical significance. Analyses were performed using the S-Plus statistical software, version 3.4 [ 26 ]. Results The cluster solution Three clusters of families were successfully identified from the study sample. The first cluster comprised 94 subjects from 17 families, the second cluster 50 subjects from 12 families, and the third cluster 115 and 25. Adding more neuropsychological test variables or leaving out the sex or the age of the subjects had little effect on the solution. The data image (Figure 2 ) indicated that the overall performance of the subjects was higher in the first cluster than in the second, and that the performance in the third cluster was between the other two. The three clusters were therefore identified as consisting of subjects that were relatively well-performing, impaired and intermediate, respectively. A three cluster solution is supported by the homogeneity of the within-cluster test performance patterns of the proposed groups (Figure 2 ). As shown by the dendrogram, the two-cluster solution would combine the impaired and the intermediate clusters, and the four-cluster result would divide the well-performing cluster into two subclusters one of which is very small, consisting only of six families. Stopping the merging process even earlier does not appear to suggest any interesting alternative cluster solutions. Note also the jump in the distance function of Figure 3 after 3 clusters. In Figure 4 the three family clusters are further visualized by classic metric multidimensional scaling (MDS) [ 27 , 28 ]. Thus, with a total of 19 variables, the 54 families comprising the three clusters are represented as points in the 19-dimensional euclidean space so that the pairwise distances between the points match the original distances between the families (maximum pairwise euclidean distances between subjects in the families). The two-dimensional projection of the 19-dimensional space, although capturing only 27.0% of the total variation, shows the two most important directions (the directions with the highest variance) and provides evidence on the success of the clustering process itself, i.e., in the sense of the distance measured used, clustering does appear to produce three separate classes of families. Further visualization of the clusters, including an animation, is provided by the supplementary material to this article (See Additional file 1 , 2 , 3 , 4 and 5 . Figure 4 Multidimensional scaling visualization. A two-dimensional visualization using multidimensional scaling (MDS) of the families in the three clusters found. The similarity measure employed in MDS was the same one that was used in the family clustering procedure, with the natural modification that the distance between a family and itself was set to zero. The horizontal and vertical axes are the directions with the highest and the second highest variance, respectively. In each of the three pairwise comparisons of the statistic BW r , s the randomly generated clustering solution almost always had a smaller value than the proposed clustering solution (the fraction of opposite results in 10 000 trials was less than 0.01). This lends support to the visual impression that the three clusters are separate groups when viewed as subsets of individual subjects. Results were similar when the family structure was ignored and the random clusters were generated allowing subjects from the same family to be assigned to different clusters. Demographic and clinical characteristics The demographic characteristics of the clusters of families are shown in Table 1 , and Table 2 shows the clinical characteristics of the subjects with schizophrenia, schizoaffective disorder or schizophreniform disorder. The three clusters did not differ by age or sex distribution. The well-performing cluster had significantly more years of education than the two others ( p < 0.001 in contrasts versus both other clusters). Overall, the clusters did not differ in clinical characteristics, except that the well-performing cluster showed better premorbid adaptation than the intermediate cluster (p = 0.04). The age of onset did not differ between the clusters (mean 25.9, SD 7.8, mean 24.7, SD 7.6, mean 23.7, SD 7.6 in clusters 1, 2, and 3, respectively, all p -values > 0.20). The impaired cluster did not include any patients with schizoaffective disorder, bipolar disorder or other affective psychotic disorders, while in the well-performing and intermediate clusters, these diagnoses were assigned to 14% and 11% of the subjects, respectively. About 36% of family members in all three clusters were unaffected. Table 1 Demographic characteristics Demographic characteristics of the family members in the well-performing (Cluster 1), impaired (Cluster 2), and intermediate (Cluster 3) clusters. Cluster 1 ( n = 94, 17 families) Cluster 2 ( n = 50, 12 families) Cluster 3 ( n = 115, 25 families) Mean SD Mean SD Mean SD Sex (F/M) 48/46 24/26 50/65 Age 48.8 9.4 52.3 12.6 48.5 11.3 Education years 11.3 a , b 3.0 9.3 2.4 9.9 2.3 a Contrast versus cluster 1 and 2, p < 0.001, t-test, two-tailed. b Contrast versus cluster 1 and 3, p < 0.001, t-test, two-tailed. Table 2 Clinical characteristics Clinical characteristics of the affected family members in the well-performing (Cluster 1), impaired (Cluster 2), and intermediate (Cluster 3) clusters Cluster 1 ( n = 35) Cluster 2 ( n = 27) Cluster 3 ( n = 43) p Yes No Yes No Yes No 1 vs. 2 * 1 vs. 3* 2 vs. 3* Poor premorbid social adjustment 14 21 13 14 27 16 ns 0.04 ns Response to neuroleptics 30 5 21 6 34 9 ns ns ns Chronic course of the disorder 15 20 17 10 24 19 ns ns ns * Chi square, two-tailed. Neuropsychological variables The impaired cluster scored lowest in all measured neuropsychological variables, and the intermediate cluster showed consistently worse performance than the well-performing one (Table 3 ). The differences between the family clusters in the neuropsychological variables were tested by the within-family linear mixed effect models. In these models, the impaired cluster was found to achieve significantly lower scores than both other clusters in almost all traits (Table 4 ). The only variable not reaching statistical significance in differentiating any of the clusters was auditory attention. Table 3 Neuropsychological test performance Means and Standard Deviations in neuropsychological test performance (raw scores) among the family members in the well-performing (Cluster 1), impaired (Cluster 2), and intermediate (Cluster 3) clusters Cluster 1 ( n = 94) Cluster 2 ( n = 50) Cluster 3 ( n = 115) Mean SD Mean SD Mean SD Auditory attention 6.7 2.1 5.8 1.9 6.3 1.8 Verbal working memory 6.1 2.3 4.5 2.0 4.7 1.7 Visual attention 8.6 2.1 6.8 1.9 7.6 1.8 Visual working memory 7.8 2.1 5.3 2.7 6.9 2.0 Story recall immediate 20.1 7.9 11.7 6.8 14.5 7.1 Story recall delayed 16.6 8.0 8.0 6.9 11.2 6.8 Visual reproduction imm 32.1 6.9 23.5 11.4 27.6 8.4 Visual reproduction del 27.4 9.9 16.2 12.8 21.2 10.5 Vocabulary 41.8 12.1 27.8 14.3 33.1 11.6 Similarities 24.7 4.6 18.5 6.7 21.1 5.3 Digit Symbol 39.4 16.5 26.9 13.6 34.4 13.8 Block design 28.6 11.6 17.7 12.5 22.3 10.8 Verbal learning 45.2 12.2 30.9 11.5 36.6 11.7 Semantic clustering 13.2 8.4 6.7 4.8 8.1 6.6 Recognition 93.2 6.0 81.1 16.6 87.2 10.3 Verbal fluency 29.9 11.5 22.0 10.4 25.2 9.3 Verbal fluency, animals 20.2 6.2 14.2 5.2 16.2 5.0 Table 4 Differences in neuropsychological test performance Differences in neuropsychological test performance between the well-performing (Cluster 1), impaired (Cluster 2), and intermediate (Cluster 3) clusters. Linear mixed effects models with family as a random effect, and sex and age as fixed effects Cluster 2 vs. Cluster 1 Cluster 1 vs. Cluster 3 Cluster 2 vs. Cluster 3 Coeff SD Wald p Coeff SD Wald p Coeff SD Wald p Auditory attention -0.81 0.43 -1.87 0.07 0.40 0.35 1.12 0.28 -0.41 0.41 -1.00 0.32 Verbal working m -1.52 0.46 -3.31 0.002 1.40 0.38 3.74 < 0.001 -0.11 0.44 -0.27 0.80 Visual attention -1.65 0.39 -4.23 < 0.001 1.00 0.31 3.19 0.002 -0.65 0.37 -1.74 0.08 Visual working m -2.31 0.46 -4.98 < 0.001 0.91 0.38 2.42 0.02 -1.40 0.44 -3.17 0.003 Story recall imm -8.18 1.65 -4.95 < 0.001 5.71 1.33 4.29 < 0.001 -2.47 1.56 -1.58 0.11 Story recall del -8.38 1.59 -5.27 < 0.001 5.37 1.28 4.20 < 0.001 -3.01 1.50 -2.00 0.05 Visual reprod imm -7.82 1.56 -5.02 < 0.001 4.49 1.23 3.65 < 0.001 -3.33 1.49 -2.23 0.03 Visual reprod del -10.20 2.03 -5.03 < 0.001 5.93 1.60 3.71 < 0.001 -4.26 1.93 -2.21 0.03 Vocabulary -14.39 2.39 -6.03 < 0.001 8.64 1.92 4.50 < 0.001 -5.74 2.29 -2.51 0.02 Similarities -9.95 2.43 -4.09 < 0.001 3.51 0.84 4.17 < 0.001 -2.45 1.00 -2.45 0.02 Digit Symbol -10.01 2.66 -3.91 < 0.001 4.42 2.13 2.08 0.04 -5.66 2.54 -2.23 0.03 Block design -8.48 2.76 -3.07 < 0.004 6.64 1.96 3.38 0.001 -3.31 2.31 -1.43 0.16 Verbal learning -13.74 2.11 -6.53 < 0.001 8.22 1.69 4.87 < 0.001 -5.52 2.03 -2.72 0.009 Semantic clust -6.20 1.21 -5.12 < 0.001 4.92 0.97 5.10 < 0.001 -1.28 1.17 -1.09 0.28 Recognition -11.21 1.82 -6.14 < 0.001 5.79 1.45 3.98 < 0.001 -5.41 1.77 -3.07 0.004 Verbal flu -7.50 2.21 -3.40 0.001 4.59 1.79 2.57 0.01 -2.90 2.08 -1.40 0.17 Verbal flu, anim -5.80 1.00 -5.78 < 0.001 4.01 0.80 5.03 < 0.001 -1.79 0.96 -1.87 0.07 Effect of education As the clusters differed significantly from each other in education years, we conducted post hoc linear mixed effects models with family as the fixed effect, and age, sex and education years as the random effects (data not shown). This did not eliminate the significant differences in cognitive functioning between the well-performing and the impaired cluster. In contrasts between the well-performing and intermediate cluster, all other differences remained significant, except in the scores of Visual immediate recall, Digit Symbol, and Verbal fluency, which lost their significance. Between the intermediate and the impaired cluster, scores in Vocabulary and Digit Symbol were no longer significantly different after controlling for education years. Discussion We report on the application of a visually aided clustering algorithm to data based on performance in a set of neuropsychological test measures, these being potential endophenotypic traits in schizophrenia. We were able to successfully detect three separate family clusters comprising both schizophrenia patients and their family members. In the impaired cluster, the families scored significantly worse than those in the other two. The well-performing cluster received the highest scores in each cognitive test, and the intermediate cluster scored consistently between the other two. However, the clusters of families did not differ from each other in age, sex distribution, and, regarding the affected subjects, in the age of onset, or in most of the other clinical features. The well-performing cluster was significantly more educated than the two others, but controlling for education years did not change the main results. We tested the differences in the diagnostic class distributions (including those with no diagnosis), and although the differences did not reach statistical significance, we find it interesting that none of the subjects with schizoaffective disorder, bipolar disorder, or other affective psychotic disorders ended up into the impaired cluster. We consider this as supporting the validity of particularly the poor cluster, which seems to represent a subsample of core schizophrenia with the most defected cognitive functioning. This cluster included the same proportion of unaffected subjects than the other two clusters, and based on the clustering algorithm, these family members without any psychiatric diagnoses during their lifetime performed generally poorly, too. Global verbal memory, including the story recall from the WMS-R [ 17 ] and verbal learning from the CVLT [ 19 ], were among the measures that differentiated well the clusters. This is in line with results by Heinrichs and Zakzanis [ 29 ], who found the best effect sizes in these functions in differentiating schizophrenia patients from controls. However, against a background of global dysfunction, any selective impairments such as those in verbal memory, are only relative [ 29 ]. The present study suggests that it is possible to characterize families with convergent cognitive performance using variables from several domains of cognition, such as attention, verbal memory, executive functioning, and intelligence. In efforts aiming at sample homogeneity, the best method may be using multiple endophenotypic measures. In part, our results are also comparable to those by Erlenmeyer-Kimling et al [ 30 ], who found that impairments in multiple cognitive measures best predicted future schizophrenia in high risk subjects. Our results suggest that molecular genetic analyses could benefit from prior appliance of our method, revealing meaningful family subgroups in a representative sample of familial schizophrenia. It would allow the resources to be targeted primarily for gene hunting projects among more homogeneous groups of families. Our new approach to combining data visualization and clustering appears to offer a valuable tool for identifying clusters in family-based data. Applying hierarchical clustering and the data image interactively helps to identify a reasonable value for the number of clusters in the cluster solution. By ordering the variables in the data image suitably, one gains useful insight into the test performance characteristics of the subjects in the clusters. As suggested in Palmer et al. [ 31 ], there may be a group of schizophrenia patients with no observed global impairment in cognition. One result of the present cluster analysis was the detection of a group of schizophrenia families with clearly better performance than the families in the two other clusters. This finding, together with those of previous studies, warrants further research for detecting putative factors protecting the cognitive development of these patients and their family members. Interestingly, attention, as measured by the simple auditory attention task (verbal span forwards), did not differentiate the clusters. The mean score in this task was also below the national normative mean in all clusters. This may indicate a fundamental impairment of attention in schizophrenia [ 32 ], observed also in patients and family members who otherwise perform well. When education was controlled for, it was further found that score in Digit Symbol, a test measuring information processing speed, and verbal fluency, a measure of executive function, did not any more separate the clusters. These results are in line with those in Weickert et al. [ 33 ], who found a selective impairment in executive function and attention in a group of schizophrenia patients defined as cognitively preserved. The present study is the first one in which cluster analysis of neuropsychological test variables has been conducted among a representative sample of familial schizophrenia comprising both affected and unaffected family members. The sample of the present study was randomly selected from a nationwide familial schizophrenia cohort. However, the results may not be generalizable to families with only one patient with the disorder. The similar patterns in neuropsychological performance in the clusters may be due to a variety of familial environmental effects, which are difficult to define ex post facto. Furthermore, our set of neuropsychological measures did not cover all those cognitive domains that previous studies have suggested as valid cognitive endophenotypes. However, it has been demonstrated in twin and in family studies [ 3 , 7 , 34 ], that the cognitive traits from our test selection measuring attention, working memory, verbal memory and visual memory do show genetic effects. Furthermore, in the present sample, the included test variables discriminated the affected and unaffected subjects, both in the whole sample and within clusters (data not shown). In the absence of a control sample, the present study could not test the possibility that the same clustering solution would emerge in normal families from the population. However, to our knowledge, such family clustering studies have not been conducted. In a study by Horan and Goldstein [ 35 ], a cluster analysis was conducted both in a patient and in a non-psychotic patient control group. The clustering solutions in these groups did not resemble each other, suggesting a specific pattern in the schizophrenic population. It is known that family members of schizophrenia patients tend to perform worse than subjects from control populations [ 1 - 3 ], and particularly those in multiply affected families [ 8 , 34 ]. Indeed, the aim of the present study was to explore the clustering of families in multiply affected families with schizophrenia. Thus the generalizability of the results may be limited to such samples representing about one fifth of all schizophrenia cases [ 36 ]. Clearly, the choice of the inter-cluster distance measure can greatly influence the merging process and hence the cluster solutions obtained. The maximum pairwise distance between subjects adopted in our analysis assigns a small distance between clusters only if all subjects in the clusters are close to each other in their test performance. We also experimented with the minimum pairwise distance but the results were poor. An explanation for this emerges by studying the minimum distance measure along the first few principal component directions of the normalized test results (the directions of largest variance). It turns out that, along these directions, most families have a member with nearly average performance and whose test results therefore closely match those of many members of other families. The variance of the distribution of the pairwise minimum distances is small and modest changes in the test results can lead to significantly different cluster solutions. The mean of pairwise distances between two clusters would be a compromise between the maximum and the minimum distances, but it turned out to behave much like the minimum distance and was therefore not used. Conclusions The new approach which combines clustering and data visualization was effective in identifying homogeneous subgroups of schizophrenia families with convergent cognitive test performance. Our results emerging from a sample of familial schizophrenia patients are in line with previous studies in which two extreme clusters have consistently emerged, characterized by a well-performing and a dysfunctional group of subjects, and at least one intermediate [ 11 , 12 , 37 ]. Our results agree with those in Sautter et al. [ 13 ], in which neuropsychological data of familial schizophrenia patients formed three clusters with respect to the level of performance. The fact that our findings, after including both affected and unaffected subjects agree with prior evidence, suggest further use of the cognitive traits as valid endophenotypes to be used in genetic linkage analyses. This method seems valid for partitioning the schizophrenia families by a relevant phenotypic category, resulting in more homogeneous subgroups. The method and results of the present study may be exploited in selecting whole families for subsequent analyses using the actual genetic marker data. Competing interests None declared. Authors' contributions FH and LH made the cluster analysis and wrote the corresponding sections, ATH supervised the neuropsychological test data collection, performed the differences testing analyses and drafted the manuscript, TP was the clinical and diagnostic study leader and commented on the text, JH designed the models for differences testing, JL conceived the research project. Pre-publication history The pre-publication history for this paper can be accessed here: Supplementary Material Additional file 1 Comparison of Cluster 1 (well-performing) and Cluster 2 (impaired). The
original variables (neuropsychological test variables + age) are visualized
using the parallel coordinate plot. The subjects classified as
well-performing are colored green and those classified as impaired are
colored red. Yellow indicates overlap of green and red. There is overlap in
all the original variables.
The above figure was produced using the Crystal Vision software. The
software uses a so-called grand tour technique to systematically (and in a
continous manner) go through all possible rotations in the data space. The
tour can be visually monitored using the parallel coordinate plot. Click here for file Additional file 2 Smaller version of Additional file 1.
Comparison of Cluster 1 (well-performing) and Cluster 2 (impaired). The
original variables (neuropsychological test variables + age) are visualized
using the parallel coordinate plot. The subjects classified as
well-performing are colored green and those classified as impaired are
colored red. Yellow indicates overlap of green and red. There is overlap in
all the original variables.
The above figure was produced using the Crystal Vision software. The
software uses a so-called grand tour technique to systematically (and in a
continous manner) go through all possible rotations in the data space. The
tour can be visually monitored using the parallel coordinate plot. Click here for file Additional file 3 A parallel coordinate plot of the data after a rotation transformation in
the data space. The axes correspond now to linear combinations of the
original variables. The 6th axis from the top shows an interesting
one-dimensional projection of the data. In this direction the two clusters
are well separated.
The above figure was produced using the Crystal Vision software. The
software uses a so-called grand tour technique to systematically (and in a
continous manner) go through all possible rotations in the data space. The
tour can be visually monitored using the parallel coordinate plot. Click here for file Additional file 4 Smaller version of Additional file 3.
A parallel coordinate plot of the data after a rotation transformation in
the data space. The axes correspond now to linear combinations of the
original variables. The 6th axis from the top shows an interesting
one-dimensional projection of the data. In this direction the two clusters
are well separated.
The above figure was produced using the Crystal Vision software. The
software uses a so-called grand tour technique to systematically (and in a
continous manner) go through all possible rotations in the data space. The
tour can be visually monitored using the parallel coordinate plot. Click here for file Additional file 5 A movie which demonstrates the use of the software can be downloaded here.
Smaller resolution CrystalVision.mpg(0.86 Mb). A higer resolution version of
this movie is available at Reference for Crystal Vision software
CrystalVision, copyright (c) 2000 by Crystal Data Technologies (Qiang Luo,
Edward J. Wegman, and Xiaodong Fu), is a Windows 95/98/NT package for Wintel
computers. A demonstration version of CrystalVision is available at Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC512293.xml |
516778 | Prospective open-label study of add-on and monotherapy topiramate in civilians with chronic nonhallucinatory posttraumatic stress disorder | Background In order to confirm therapeutic effects of topiramate on posttraumatic stress disorder (PTSD) observed in a prior study, a new prospective, open-label study was conducted to examine acute responses in chronic, nonhallucinatory PTSD. Methods Thirty-three consecutive newly recruited civilian adult outpatients (mean age 46 years, 85% female) with DSM-IV-diagnosed chronic PTSD, excluding those with concurrent auditory or visual hallucinations, received topiramate either as monotherapy (n = 5) or augmentation (n = 28). The primary measure was a change in the PTSD Checklist-Civilian Version (PCL-C) score from baseline to 4 weeks, with response defined as a ≥ 30% reduction of PTSD symptoms. Results For those taking the PCL-C at both baseline and week 4 (n = 30), total symptoms declined by 49% at week 4 (paired t -test, P < 0.001) with similar subscale reductions for reexperiencing, avoidance/numbing, and hyperarousal symptoms. The response rate at week 4 was 77%. Age, sex, bipolar comorbidity, age at onset of PTSD, duration of symptoms, severity of baseline PCL-C score, and monotherapy versus add-on medication administration did not predict reduction in PTSD symptoms. Median time to full response was 9 days and median dosage was 50 mg/day. Conclusions Promising open-label findings in a new sample converge with findings of a previous study. The use of topiramate for treatment of chronic PTSD, at least in civilians, warrants controlled clinical trials. | Background Posttraumatic stress disorder (PTSD) is a difficult-to-treat condition that over a lifetime affects approximately 10% of the general population [ 1 ]. The condition develops after traumatic events such as combat, terror activities, disaster, or rape, and has 3 main features: (1) reexperiencing the trauma through recollection, dreams, and reliving, (2) avoidance of thoughts, activities, and emotions associated with the trauma, and (3) hyperarousal [ 2 ]. PTSD is usually a chronic disorder, with one third of patients displaying symptoms for ≥ 10 years after experiencing the traumatic event [ 3 , 4 ]. Generally the response to pharmacotherapy has been poor, with many patients completely unresponsive and others only marginally responsive [ 5 ]. Some tricyclic antidepressants, monoamine oxidase inhibitors, and selective serotonin reuptake inhibitors have demonstrated efficacy in double-blind trials [ 3 ]. The complex neurobiology of PTSD involves a number of systems, including dopaminergic, serotonergic, sympathetic, hypothalamic-pituitary-adrenal, and various anatomic regions of the amygdala and other parts of the limbic system [ 3 ]. It has been suggested that after traumatic events limbic nuclei may become kindled or abnormally sensitized [ 6 ], leading to increased susceptibility for psychic and physical arousal and psychiatric disturbance [ 2 ]. Because of the suggested involvement of the kindling phenomenon, several anticonvulsants have been assessed in the treatment of PTSD, including carbamazepine, valproate, lamotrigine, gabapentin, tiagabine, and topiramate [ 3 ]. Topiramate has a broad spectrum of pharmacologic properties, including Na + channel blockade [ 7 - 11 ], inhibition of some high voltage-activated Ca 2+ channels [ 12 ], enhanced γ-aminobutyric acid (GABA) neuroinhibition at novel GABA A receptors [ 13 , 14 ], glutamate inhibition at kainate and α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors [ 15 , 16 ], and promotion of protein phosphorylation of neuronal conductance channels [ 15 ]. These properties, together with inhibitory activity in animal kindling models [ 17 , 18 ], suggest that topiramate may have therapeutic potential in PTSD. Treatment with topiramate has been reported to improve reexperiencing symptoms associated with civilian PTSD [ 6 ]. Results from that study, which included a number of patients who were classified as treatment resistant, suggested that topiramate fully or partially suppressed both intrusions (distressing recollections or nonhallucinatory flashbacks) and nightmares, if present, in 89% of patients with nonhallucinatory PTSD. Although the DSM-IV definition of PTSD lists hallucinations among reexperiencing symptoms [ 19 ], because a few patients displayed varying degrees of impaired reality testing of hallucinations, it was difficult to determine if hallucinations were due to PTSD, to an independent psychotic disorder, or, in some individuals, to both. Although psychotic variants of PTSD have been reported in as many as 40% of veterans with chronic combat-associated PTSD without evidence of a primary psychotic disorder [ 20 , 21 ], it is also possible for patients with psychotic disorders to have comorbid PTSD, leaving questions about whether specific hallucinations should be attributable to the PTSD reexperiencing cluster or to a psychotic disorder. Because the prior study found a more robust effect in the group of PTSD patients without hallucinations [ 6 ], it was decided to focus further on the core group of PTSD patients without ambiguous symptoms confounded by the possible presence of independent psychotic disorders. Based on these earlier results of topiramate in PTSD, a fully prospective open-label study was begun to see whether the previously observed signal of a therapeutic benefit of topiramate for chronic civilian PTSD could be confirmed. None of the patients reported in the earlier study were included in the present study. In order to address methodological limitations of the prior study, the current study is modified in 4 major ways: (1) it excludes patients with hallucinations, (2) it limits the study to 12 weeks, (3) it prospectively employs the PTSD Checklist-Civilian Version (PCL-C) as a validated clinical assessment instrument in all patients, and (4) it determines a clinical response rate as of week 4, using a conventional, predetermined definition for positive response as a ≥ 30% improvement in a standardized clinical rating scale. Methods This sample included all consecutive adults (n = 33) meeting the DSM-IV criteria for chronic civilian PTSD in an outpatient private practice who started open-label topiramate between January 2000 and November 2002 in the course of clinical care. The study excluded patients with concurrent auditory or visual hallucinations associated with either PTSD or a possible psychotic disorder. Topiramate, administered either as monotherapy (n = 5) or added to the patients' existing therapeutic regimen (n = 28) (Table 1 ), was initiated at a starting dosage of 12.5 to 25 mg/day and increased in 25- to 50-mg/day increments every 3 to 4 days as tolerated until a clinical response was achieved. Table 1 Patient characteristics All subjects N = 33 Mean age, years ± SD 46 ± 6.5 Range 29–55 Sex, % Women 85 Men 15 Mean age at onset of PTSD, years ± SD 29 ± 15 Range 2–53 Mean duration of PTSD history, years ± SD 18 ± 15 Range 0–46 Patients with comorbid disorders, N (%) Bipolar disorder 10 (30) Major depressive disorder 21 (64) Substance abuse Current 3 (9) In past 5 (15) SD =standard deviation; PTSD = posttraumatic stress disorder. The study was performed according to the Declaration of Helsinki. Each participant provided verbal informed consent based on disclosure of off-label usage, the availability of alternative medications for PTSD such as sertraline and paroxetine, and expected adverse events and risks. Verbal consent was deemed appropriate and sufficient because of the nonexperimental nature of the treatment. Because this was a descriptive study using aggregate data gathered in the course of treating patients in clinical practice for their own benefit and not an experimental design, there was no protocol that required Institutional Review Board approval. Patients were asked to measure their symptoms using standardized clinical instruments already routinely used as a part of individual care. The study included no features of clinical studies, such as randomization, concealment of treatment, or surrender of patient decision-making rights, as an integral part of conducting the study. The primary outcome was a change in PTSD symptoms calculated from change in PCL-C score from baseline through 4 weeks of treatment [ 22 ]. The PCL-C, a validated self-report instrument, strongly correlates ( k ~0.90) with the Clinician-Administered PTSD Scale (CAPS), a structured clinical interview widely used in PTSD studies [ 23 ]. The PCL-C is scored on a Likert scale of 1 to 5 (1 = not at all; 5 = extremely) for each of 17 items that cover all 3 symptom clusters (B, reexperiencing; C, avoidance/numbing; D, hyperarousal), yielding a score range of 17 to 85. A total score of ≥ 50 meets the threshold for active PTSD. In accordance with conventional usage, a 30% reduction in clinical symptoms of PTSD was defined as a positive clinical response. The secondary measures included were Cluster B symptoms of reexperiencing of trauma, including trauma-related intrusions and nightmares. Secondary measures were limited to Cluster B symptoms because of the practical difficulty patients in clinical practice can have linking a symptom (such as avoidance, numbing, hyperarousal, insomnia, or startling) specifically to a trauma as opposed to some other source, as required by the DSM-IV definitions. Also, Cluster B symptoms appeared to be among the most distressing and salient for many patients, thereby making tracking these distinctive symptoms comparatively easy. Data on patient demographics, concomitant medication, comorbidities, primary trauma, and duration of PTSD were collected at baseline. Assessments were made through clinical interviews over a 12-week period. Cessation of reexperiencing symptoms was considered partial if subjects reported a definite reduction in intensity and frequency of nightmares or intrusive recollections or flashbacks. Full cessation was defined as complete elimination of nightmares and intrusions for a sustained period. Adverse effects were recorded if they resulted in treatment discontinuation. Statistical analyses were performed using Jandel SigmaStat ® Version 2.0 (Jandel Corporation, San Rafael, CA, USA). A single paired t -test of before and after total PCL-C scores was used to analyze the data. Subscale scores for PCL-C Clusters B, C, and D were also evaluated. Results Patient characteristics Baseline patient characteristics and concomitantly administered medications are listed in Tables 1 and 2 , respectively. All monotherapy patients were in the nonbipolar subgroup. Primary traumas for which symptoms were reported are detailed in Table 3 . Of these, 9 were of childhood onset, with 4 involving childhood physical assault, 3 unwanted sex or sexual assault, and 2 sudden unexpected or violent deaths. Table 2 Concomitant medications Medication type Number of patients (N = 33) SSRIs 12 Benzodiazepines 10 Stimulants 6 Atypical neuroleptics 3 Gabapentin 3 Lamotrigine 5 Mirtazapine 3 Venlafaxine 3 Verapamil 1 Other 8 Monotherapy 5 SSRIs = selective serotonin reuptake inhibitors. Table 3 Primary traumas Trauma type Number of patients, N = 33 Unwanted sex 11 Physical assault 6 Sudden violent death 4 Sudden unexpected death of loved one 3 Sexual assault 2 Fire/explosion 1 Weapon assault 1 Combat 1 Life-threatening illness 1 Other 3 Primary measures Reduction in PCL-C symptoms Ninety-one percent of patients (30/33) with baseline PCL-C measurements completed a PCL-C at week 4. In those patients, the total score was 62.6 at baseline and 40.3 at week 4. In addition, the scores for reexperiencing, avoidance, and hyperarousal clusters were 18.2, 25.1, and 19.2 at baseline and 11.2, 17.2, and 12.3 at week 4, respectively. To calculate percentage reduction in PTSD symptoms, to correct for an absence of symptoms being scored as "1" for each item, the change in score between baseline and week 4 was divided by the baseline score minus the minimum score for no symptoms. Therefore, the change in total symptoms was calculated as (baseline – week 4)/(baseline – 17). Total symptoms declined significantly (49%) by week 4 (paired t -test, P < 0.001). In addition, subscale symptoms for reexperiencing (Cluster B), avoidance/numbing (Cluster C), and hyperarousal (Cluster D) decreased by 53%, 43%, and 48%, respectively, from baseline (Figure 1 ). By the end of week 4, 70% of all entrants in the study (n = 33) and 77% of those with both baseline and week 4 PCL-C scores (n = 30) were positive responders, defined as a ≥ 30% improvement in symptoms. Figure 1 Mean percentage symptom reduction at week 4. Symptom reduction in those 30 subjects who completed a PCL-C at baseline and after 4 weeks. *Paired t -test, P < 0.001 versus baseline. PTSD = posttraumatic stress disorder; PCL-C = PTSD Checklist-Civilian Version. Predictors of reduction in PCL-C symptom improvement Comparisons for bipolar versus nonbipolar patients (50% versus 49%, analysis of variance [ANOVA]), duration of symptoms (r = 0.37), age at onset of PTSD (r = -0.36), severity of baseline PCL-C score (r = 0.02), age (r = -0.02), female sex (49% versus 50%, ANOVA), and monotherapy versus add-on medication administration (53% versus 48%, ANOVA) found none significantly associated with reduction in PCL-C symptoms. Secondary measures Results for combined suppression of nightmares and intrusions are shown in Table 4 Table 4 Secondary efficacy measures Combined suppression of nightmares and intrusions Responder status, n (%) Full 26/33 (79%) Partial 3/33 (9%) None 4/33 (12%) Mean time to response, days ± SD, (range) Full response (n = 25) 15 ± 18 (1–83) Partial response (n = 17) 11 ± 13 (2–46) Median time to full response, days 9 Median time to partial response, days 5 Mean dosage at time of response, mg/day Full response 60 ± 47 Partial response 32 ± 15 Median dosage at time of response, mg/day Full response 50 Partial response 25 Modal daily dosage for full response, mg/day 25 % of patients in the 12.5–50-mg/day range 65 Modal dosage at time of partial response, mg/day 25 % of patients in the 12.5–50-mg/day range 100 Improvements in nightmares Full cessation of nightmares 17/18 (94%) Improvements in intrusions Full cessation of intrusions 26/33 (79%) Partial improvement 3/33 (9%) No improvement 4/33 (12%) SD = standard deviation. Of the 33 patients who entered the study, 79% fully ceased having reexperiencing symptoms with a mean time to full cessation of 15 days. The onset of full cessation for all patients occurred at ≤ 200 mg/day and the onset of partial cessation for all patients at ≤ 50 mg/day. Ninety-four percent of patients with nightmares and 79% of patients with intrusions at baseline reported full cessation at the end of week 4. Discontinuations Serious adverse events There were no serious adverse events during this study. Of the 33 patients who entered the study, 7 (21%) discontinued due to adverse events and 5 (15%) for other reasons. The single most common adverse event resulting in discontinuation was intolerable nervous system overstimulation (panic/nervousness/overstimulation/ shakiness), reported by 3 patients. Other adverse events resulting in discontinuation included clumsiness (n = 2), cognitive impairment (n = 1), and severe headache (n = 1). Additional reasons for discontinuation included personal choice (n = 1) and lack of relapse of PTSD symptoms after interrupting medication (n = 4). Of the 12 patients that discontinued, 54% (n = 7) fully ceased having reexperiencing symptoms of nightmares/intrusions and 15% (n = 2) partially ceased having reexperiencing symptoms of nightmares/intrusions. The median time to termination was 25.5 days (range, 4–77). Predictors of discontinuation There were no statistically significant predictors of panic-like adverse events resulting in discontinuation from the study (age, sex, symptom duration, age at onset, bipolarity, substance abuse, baseline severity of illness), although the coadministration of a benzodiazepine nearly reached significance (Fisher exact test, P = 0.05). No monotherapy patients discontinued because of overstimulation/panic-like responses, so the possibility of drug interactions or conditions such as comorbid panic disorder contributing to this reaction should be considered in future investigations. Discussion Results of this prospective open-label study suggest that the use of topiramate for the treatment of chronic PTSD in civilians is indeed promising and warrants controlled clinical trials. Topiramate was able to improve symptoms of reexperiencing, avoidance, and hyperarousal by 49% overall, compared with baseline, and decreased intrusions in 94% of patients and nightmares in 79% of those with this symptom. Although there are numerous other reports describing the use of antiepileptic agents for the treatment of PTSD, most are case reports and case series, and some lack the use of standardized PTSD scales. The first published trials using anticonvulsants date back almost 20 years, when a 5-week trial of carbamazepine in 10 combat veterans produced a 36% reduction in a nonvalidated interviewer-scored instrument [ 24 ]. Studies on adjunctive therapy with valproate reported improved global scores in avoidance and hyperarousal, but not reexperiencing symptoms in veterans over a 14-month period [ 25 ], whereas a subsequent 8-week open-label trial reported a small, 17% reduction in CAPS score in 13 subjects with significant reductions in all 3 cluster criteria [ 26 ]. Two studies in which valproate was used as monotherapy reported varying results, with one showing a 30% reduction in total CAPS scores in 28 veterans [ 27 ], whereas the other in 10 civilian patients with non-combat-related PTSD showed no improvement [ 28 ].The only double-blind placebo-controlled trial evaluating an anticonvulsant in PTSD was conducted with lamotrigine [ 29 ]. Overall, after 10 weeks of treatment, 5 of 10 subjects were characterized as responders, compared with 1 of 4 subjects receiving placebo. However, a last observation carried forward analysis revealed only a 23% decrease in mean pretreatment scores with lamotrigine, compared with 20% in the placebo group. A retrospective clinical series of adjunctive gabapentin therapy in 30 consecutive patients with PTSD assessed improvement of target symptoms such as nightmares, insomnia, and irritability. It was reported that the majority of patients (77%) showed moderate or greater improvement in duration of sleep, and most notably a decrease in the frequency of nightmares [ 30 ]. These studies highlight the need for double-blind placebo-controlled trials using standardized and validated questionnaires. It is encouraging that the results of this current study with topiramate were consistent with the earlier, preliminary data obtained in patients with PTSD. Percentage responders (77% versus 75%), median time to full response (9 d versus 8 d), suppression of intrusions and nightmares (88% versus 89%), and discontinuation rate (39% versus 32%) were all similar. Overall, the open-label experience with topiramate, based on patient statements and self-rating reports, suggests that the drug may have a rapid rate of response with limited risk of clinically significant adverse events and no evidence of tolerance developing over time [ 6 , 31 ]. Moreover, from topiramate studies in epilepsy [Topamax prescribing information] it is known that the drug is not associated with the cardiac, pancreatic, and hematologic toxicity found with valproate or carbamazepine [ 3 ]. In addition, topiramate is not associated with weight gain, a side effect reported with valproate that can predispose patients to diabetes mellitus [ 32 ]. Topiramate also offers potential clinical advantages over serotonin reuptake inhibitors (SSRIs) because it does not appear to destabilize mood in patients with comorbid bipolar disorder and is relatively devoid of common adverse effects of SSRIs, such as sexual dysfunction, weight gain with chronic use, or sedation when used as monotherapy. Limitations of the present study include: (1) the absence of structured assessment data following week 4 to test the maintenance of response over time, (2) use of self-report instruments that, although validated and correlated with results of structured clinical interviews, may be less accurate than structured interviews, and (3) the absence of clinical assessment scores at the time patients discontinued the trial. From a broader perspective, the greatest limitation of this study is characteristic of all open trials: the lack of standard features of clinical trials such as placebo controls, randomization, and blinding of raters. Conclusions Fundamentally, the findings of current and prior topiramate studies are convergent and consistently signal a potentially beneficial therapeutic effect for all 3 criteria clusters of chronic PTSD in civilian adults. In both studies, there is a positive response in a large proportion of patients, at dosages considerably below the usual anticonvulsant dosage levels of 200 to 600 mg/day, and a rapid onset. The effect appears independent of comorbidity with bipolar disorder, age, sex, duration of symptoms, baseline severity of illness, or administration alone or with other psychotropic medications. The purpose of this study was to prospectively test the results of an earlier open-label trial in a new sample; however, the limitations of those open studies will have to be addressed in double-blind, placebo-controlled studies, which are currently in progress. Competing interests The author is a consultant to and a participant in the Speaker's Bureau for Ortho-McNeil Pharmaceutical, Inc. He also holds a licensing agreement with Ortho-McNeil based on issuance of a United States Government patent for topiramate therapy of PTSD. Author's contributions JLB conceived of the study and its design, coordination, statistical analysis, and drafted the manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC516778.xml |
340942 | Engineered Biosynthesis of Regioselectively Modified Aromatic Polyketides Using Bimodular Polyketide Synthases | Bacterial aromatic polyketides such as tetracycline and doxorubicin are a medicinally important class of natural products produced as secondary metabolites by actinomyces bacteria. Their backbones are derived from malonyl-CoA units by polyketide synthases (PKSs). The nascent polyketide chain is synthesized by the minimal PKS, a module consisting of four dissociated enzymes. Although the biosynthesis of most aromatic polyketide backbones is initiated through decarboxylation of a malonyl building block (which results in an acetate group), some polyketides, such as the estrogen receptor antagonist R1128, are derived from nonacetate primers. Understanding the mechanism of nonacetate priming can lead to biosynthesis of novel polyketides that have improved pharmacological properties. Recent biochemical analysis has shown that nonacetate priming is the result of stepwise activity of two dissociated PKS modules with orthogonal molecular recognition features. In these PKSs, an initiation module that synthesizes a starter unit is present in addition to the minimal PKS module. Here we describe a general method for the engineered biosynthesis of regioselectively modified aromatic polyketides. When coexpressed with the R1128 initiation module, the actinorhodin minimal PKS produced novel hexaketides with propionyl and isobutyryl primer units. Analogous octaketides could be synthesized by combining the tetracenomycin minimal PKS with the R1128 initiation module. Tailoring enzymes such as ketoreductases and cyclases were able to process the unnatural polyketides efficiently. Based upon these findings, hybrid PKSs were engineered to synthesize new anthraquinone antibiotics with predictable functional group modifications. Our results demonstrate that (i) bimodular aromatic PKSs present a general mechanism for priming aromatic polyketide backbones with nonacetate precursors; (ii) the minimal PKS controls polyketide chain length by counting the number of atoms incorporated into the backbone rather than the number of elongation cycles; and (iii) in contrast, auxiliary PKS enzymes such as ketoreductases, aromatases, and cyclases recognize specific functional groups in the backbone rather than overall chain length. Among the anthracyclines engineered in this study were compounds with (i) more superior activity than R1128 against the breast cancer cell line MCF-7 and (ii) inhibitory activity against glucose-6-phosphate translocase, an attractive target for the treatment of Type II diabetes. | Introduction Polyketides are a large class of structurally and pharmacologically diverse molecules, including many antibiotics and antitumor drugs ( O'Hagan 1991 ). They are produced as secondary metabolites primarily by bacteria and fungi ( Hopwood 1997 ). Analogous to fatty acid synthases (FASs), polyketide synthases (PKSs) catalyze the biosynthesis of polyketides through repetitive C–C bond-forming reactions between selected acyl-CoA-derived building blocks ( Cane et al. 1998 ). However, in contrast to fatty acid biosynthesis, the carbon chain backbones of polyketides exhibit greater variety with respect to the choice of acyl-CoA building blocks and the degree of reduction of β-ketone functional groups that result after each round of chain elongation. In complex polyketides, such as the macrolide erythromycin, biosynthetic variability arises from independent control of each round of chain elongation by one module of enzymes within a multimodular PKS ( Cane et al. 1998 ). (The term module used in this report refers to a collection of dissociated enzymes. The elongation module consists of enzymes involved in chain extension steps of polyketide biosynthesis, while the initiation module consists of enzymes involved in the nonacetate priming of certain aromatic PKSs.) However, the polyketide backbones of most bacterial aromatic polyketides ( Figure 1 ) are synthesized by a single dissociated enzymatic module comprised of a heterodimeric ketosynthase–chain length factor (KS-CLF) complex that catalyzes chain initiation and iterative elongation, an acyl-carrier protein (ACP) that shuttles malonyl extender units to the active site of KS-CLF as malonyl-S-ACP intermediates, and a malonyl-CoA:ACP acyl transferase (MAT), which catalyzes acyl transfer between malonyl-CoA and the ACP and is shared between the PKS and the housekeeping FAS(s) ( Revill et al. 1995 ; Khosla et al. 1999 ). For example, the minimal PKS from the actinorhodin ( act ) biosynthetic pathway synthesizes an octaketide (C 16 ) backbone from eight malonyl-CoA equivalents, whereas the tetracenomycin ( tcm ) minimal PKS synthesizes a decaketide (C 20 ) backbone from ten equivalents of malonyl-CoA ( Figure 2 ) Although a large number of “unnatural” natural products have been engineered to date by genetic manipulation of bacterial aromatic PKSs ( Khosla and Zawada 1996 ; Rawlings 1999 ; Shen et al. 1999 ), most of this variety has resulted from the combinatorial manipulation of ketoreductases (KRs), aromatases (AROs), and cyclases (CYCs) that ordinarily interact with minimal PKS subunits to channel the exceptionally high reactivity of a poly-β-ketone intermediate ( Figure 2 ) into the observed natural product ( McDaniel et al. 1995 ). The development of generally applicable methods for chemo- and regioselective modification of natural and unnatural bacterial aromatic polyketides is an important goal for the medicinal chemist and, more recently, the biosynthetic engineer. Figure 1 Examples of Aromatic Polyketides Actinorhodin ( S. coelicolor ) and tetracenomycin ( S. glaucescens ) are primed by acetate groups through decarboxylation of malonyl-CoA. Actinorhodin and tetracenomycin are synthesized by the act and tcm PKSs, respectively. Oxytetracycline ( S. rimosus ), frenolicin ( S. roseofulvus) , R1128 ( S. R1128), and doxorubicin ( S. peucetius ) are primed by nonacetate units as shown. Notably, R1128 family of compounds (a–d) are primed with different alkyl units. Oxytetracycline, frenolicin, R1128, and daunorubicin are synthesized by the otc , frn , R1128, and dxr PKSs, respectively. Actinorhodin and tetracenomycin represent much-studied models of aromatic polyketide biosynthesis. Oxytetracycline is a commonly prescribed antibiotic. Frenolicin is a potent antiparasitic agent. R1128 is an estrogen receptor antagonist that shows minimal agonist activity. Doxorubicin is a widely used anticancer drug in treating late-stage tumors. Figure 2 Biosynthesis of Acetate-Primed Polyketides (A) Minimal PKS is necessary and sufficient for the synthesis of a complete polyketide chain. KS-CLF is the condensing enzyme in the minimal PKS, catalyzing each round of condensation between malonyl-ACP and the growing polyketide chain. ACP serves as the carrier for malonyl units, and it is malonylated by the MAT associated with FAS. Chain synthesis initiates with the decarboxylation of malonyl-ACP to acetyl-ACP by the KS-CLF for most aromatic PKSs. The acetyl unit is then transferred to the KS-CLF and primes the enzyme for subsequent condensations. The overall chain length is controlled by the KS-CLF complex. An octaketide synthase (e.g., act PKS) uses a total of eight malonyl equivalents (including the primer), while a decaketide synthase (e.g., tcm PKS) uses a total of ten malonyl equivalents. (B) An octaketide can spontaneously form SEK4 and SEK4b in the absence of tailoring enzymes. Members of the act KR family can regioselectively reduce the octaketide at C-9, which can then spontaneously form mutactin in the absence of AROs and CYCs. When bifunctional ARO/CYC (e.g., act VII) and second-ring CYC (e.g., act IV) are present, the reduced octaketide can be transformed into the anthraquinone DMAC. (C) A decaketide can spontaneously form SEK15 and SEK15b in the absence of tailoring enzymes. KR can regioselectively reduce the C-9 carbonyl. A reduced decaketide can spontaneously form RM20, RM20b, and RM20c. Not shown are the other tailoring enzymes associated with decaketides, which can transform the nascent decaketide into tetracycline or anthracycline structures. The primer unit of a polyketide backbone is an attractive site for introducing unnatural building blocks. For example, genetic and chemobiosynthetic approaches have been devised to replace the natural primer units in the polyketide backbones of erythromycin, avermectin, and rapamycin with a broad range of unnatural functional groups ( Jacobsen et al. 1997 ; Marsden et al. 1998 ; Lowden et al. 2001 ; Long et al. 2002 ). However, most aromatic PKSs initiate polyketide biosynthesis through decarboxylation of malonyl-ACP, resulting in an invariant acetyl primer unit ( Figure 2 A) ( Bisang et al. 1999; Dreier and Khosla 2000 ). Important antitumor antibiotics, such as the anthracycline doxorubicin, are primed with propionate units. Recently, we have investigated the biosynthesis of the estrogen receptor antagonist R1128 ( Hori et al. 1993 ) and the antiparasitic agent frenolicin ( Bibb et al. 1994 ), two aromatic polyketides that are apparently derived from nonacetate primer units (see Figure 1 ). These bimodular PKSs are comprised of a dissociated initiation module consisting of a homodimeric KS (ZhuH, named for R1128 PKS), an acyl transferase (AT) (ZhuC), and an ACP (ZhuG), and an elongation module consisting of a KS-CLF (ZhuA–ZhuB), a second ACP (ZhuN), and the MAT (borrowed from the housekeeping FAS). The proposed biosynthetic mechanism of the R1128 PKS is shown in Figure 3 ( Marti et al. 2000 ). Biochemical studies have revealed that the KS subunits of the initiation and elongation modules have specific protein–protein interactions with ACPs from the same module ( Tang et al. 2003 ), suggesting that it may be possible to functionally coexpress these initiation modules with heterologous minimal PKSs, so as to regioselectively incorporate nonacetate primer units into aromatic polyketides. Of particular interest is the R1128 initiation module, since it is known to have broad substrate specificity ( Meadows and Khosla 2001 ) and the X-ray crystal structure of its KS subunit has been solved ( Pan et al. 2002 ). Here we demonstrate the biosynthetic utility of the R1128 initiation module by synthesizing a variety of anthraquinone antibiotics, some with significant biological activities. In the course of these studies, fundamentally novel and unanticipated properties of bacterial aromatic PKSs have been elucidated. Figure 3 Proposed Priming Mechanisms for R1128 PKS An independent loading module consisting of ZhuG, ZhuH, and ZhuC can generate an alkylacyl-ZhuG intermediate (boxed) from malonyl-CoA and short chain acyl-CoAs such as propionyl-CoA and isobutyryl-CoA. The precursor selectivity is determined by the KSIII analog ZhuH. Ketoreductase, dehydratase, and enoylreductase associated with FAS are presumed to transformed the β-ketoacyl-ZhuG moiety into alkylacyl-ZhuG. The alkylacyl-ZhuG is then able to prime the minimal PKS module (consisting of the ZhuB [KS], ZhuA [CLF], ACP [ZhuN], and MAT) and initiate polyketide synthesis. The mechanism by which the transacylation occurs is not known and is possibly catalyzed by unassigned, but essential, enzyme ZhuC. Homologs of ZhuG, ZhuH, and ZhuC are present in the frn PKS (FrnJ, FrnK, and FrnI, respectively) as well. Results Using the host/vector system first described by McDaniel et al. (1993a ), several combinations of initiation modules, minimal PKSs, and auxiliary PKS subunits were coexpressed and analyzed. Streptomyces coelicolor CH999, which contains a chromosomal deletion of the entire act gene cluster, was used as the host strain, whereas the shuttle vector pRM5 was used as an expression plasmid. The polyketide product profiles of the recombinant strains are summarized in Table 1 . Table 1 Plasmids Constructions and Resulting Polyketide Products Each plasmid is a derivative of pRM5. The constructs are transformed into S. coelicolor strain CH999. Produced are analyzed by LC/MS and NMR a LM: Loading module, consists of ZhuC, ZhuH, and ZhuG b The minimal PKS consists of the indicated KS-CLF, ZhuN, and the endogenous MAT c This construct contains the act KS-CLF, but lacks the minimal ACP ZhuN d NP: no major polyketide product observed e Number in parenthesis indicates yield of the product as percentage of total polyketide recovered Recombination of an Initiation Module and a Heterologous Minimal PKS Guided by our recent observation that the KS–ACP pairs of initiation and elongation PKS modules have orthogonal molecular recognition features, we first attempted to coexpress the R1128 initiation module with the act minimal PKS. The zhuC (AT) , zhuH (KS), and zhuG (ACP) genes from the R1128 gene cluster were coexpressed with the genes encoding the act KS-CLF, ZhuN (ACP), and the act KR (on plasmid pYT46). Control plasmids pYT44, lacking the R1128 initiation module, and pYT45, lacking zhuN , were also constructed and characterized. All three plasmids were introduced into S. coelicolor CH999 by transformation, and polyketides products were analyzed by liquid chromatography/mass spectrometry (LC/MS) and nuclear magnetic resonance (NMR) spectroscopy. S. coelicolor CH999/pYT44 produced mutactin (and its dehydrated derivative, dehydromutactin), the expected products of the act minimal PKS in the presence of act KR ( McDaniel et al. 1994a ), whereas CH999/pYT45 did not produce any polyketide, consistent with the requirement for separate ACPs to support turnover of the initiation and elongation modules in a bimodular aromatic PKS ( Figure 4 A) ( Tang et al. 2003 ). Remarkably, CH999/pYT46 produced two new polyketide products in addition to mutactin; these new polyketides were isolated with a combined yield of 40 mg/l, representing 70% of total polyketides produced by this host. The two compounds had molecular masses of 278 and 292. (The 14 mass unit difference is suggestive of one methylene unit difference between the two compounds.) Isotopic labeling studies indicated that the compounds were derived from six acetate equivalents. NMR analyses ( Table 2 ) revealed that the two compounds, YT46 ( 1 ) and YT46b ( 2 ), had structures shown in Figure 4 . YT46 and YT46b are identical, with the exception of a branched methyl group in YT46b. Each compound contains an α-pyrone moiety, which is commonly observed in aberrantly cyclized aromatic polyketides ( Yu et al. 1998 ). The C-9 carbonyl groups were selectively reduced by the act KR. Supplementing the growth medium with 13 C-labeled sodium propionate revealed that the alkyl moiety of 1 was derived from propionyl-CoA. Similarly, supplementing the growth medium with 1 g/l of L-valine resulted in a 2-fold increase in the level of 2 , suggesting the branched alkyl group observed in 2 was derived from isobutyryl-CoA, a primary catabolite of L-valine ( Zhang et al. 1999 ). These findings are consistent with in vitro characterization of the substrate specificity of ZhuH ( Meadows and Khosla 2001 ). Among substrates tested, ZhuH had an 11-fold higher specificity for both propionyl-CoA and isobutyryl-CoA over the next best substrate, acetyl-CoA ( Meadows and Khosla 2001 ). YT46 analogs generated by incorporation of isovaleryl and butyryl primer units were also detectable by LC/MS, although these compounds were present at lower levels. Dehydrated analogs of 1 and 2 were also observed in CH999/pYT46. (Purified 1 and 2 slowly dehydrate at room temperature.) Figure 4 Engineered Biosynthesis of YT46 (1) and YT46b (2) (A) HPLC trace of extracts from CH999/pYT44-pYT46. A linear gradient of 20%–60% CH 3 CN in water was used. (Left) CH999/pYT44, which only has the minimal PKS module from act PKS and KR, produced the expected mutactin (and the dehydrated dehydromutactin). CH999/pYT45, which contains the R1128 loading module and an incomplete minimal PKS, produced no major polyketides. (Right) Upon coexpressing the R1128 loading module with a functional act minimal PKS (in the presence of KR), two new major polyketides were produced with the indicated masses. When 1 g/l of valine was added to the growth medium, the yield of the compound with mass of 292 doubled (traces not drawn to the same scale). The two compounds are identified as YT46 and YT46b. (B) Engineered biosynthesis of YT46 and YT46b. YT46 (1) is derived from propionate, while YT46b (2) is derived from isobutyrate. Alkylacyl-ZhuG supplied by the loading module is able to prime the act minimal PKS efficiently. Incorporation of the alkylacyl moiety by the KS-CLF led to a decrease in the number of extender units incorporated in the final chain. The octaketide synthase is now only able to synthesize an alkyl hexaketide. The KR is able to regioselectively reduce the alkyl-hexaketide at the expected C-9 position. The reduced hexaketide spontaneously form the novel bicyclic structure observed in 1 and 2. Dehydrated versions of 1 and 2 are also observed (outside of limits shown in [A]). Table 2 Proton and Carbon NMR Data for YT46 ( 1 ) and YT46b ( 2 ) Spectra were obtained at 400 MHz for proton and 100 MHz for carbon and were recorded in acetone- d 6 a Carbons are labeled as shown in Figure 4 B b 1, 2- 13 C-acetate labeling experiments were performed with 2 and the observed carbon–carbon coupling constants are shown in parenthesis The biosynthesis of 1 and 2 by a recombinant bimodular PKS consisting of the act minimal PKS and the R1128 initiation module supported our hypothesis that nonacetate-primed polyketides could be biosynthesized by combinatorial expression of heterologous initiation and elongation PKS modules from bacterial aromatic PKSs. Indeed, the act KS-CLF, which is normally primed exclusively by acetate (generated via decarboxylation of a malonyl unit), has a remarkably strong preference for the diketide product of the R1128 initiation module. However, unexpectedly, the incorporation of a longer chain substrate into the catalytic cycle of the act minimal PKS results in a reduced number of malonyl units utilized during iterative chain extension ( Figure 4 B). Compound 1 is a hexaketide whose backbone consists of 15 C atoms. Thus, upon introduction of an initiation PKS module into the overall catalytic cycle, the octaketide synthase retains its carbon chain length specificity, rather than executing the normal number of extension cycles. In contrast, the act KR retains its selectivity for the C-9 carbonyl in the nascent polyketide backbone, notwithstanding structural differences between a hexaketide and an octaketide. In Vivo Reconstitution of R1128 Biosynthesis Using a Heterologous Bimodular PKS To validate the generality of the above observations, we sought to reconstitute the biosynthesis of R1128 family of compounds in S. coelicolor using a heterologous combination of initiation and elongation PKS modules ( Figure 5 ). It follows from the above analysis that to synthesize an alkyl-primed anthraquinone such as 3 and 4 , the following catalytic components are needed: (i) an initiation module; (ii) a decaketide minimal PKS that can extend the five to six carbon products of the initiation module by seven more extender units to yield a 19–20 carbon polyketide; and (iii) appropriate ARO and CYC subunits. Figure 5 Engineered Biosynthesis of TMAC (7), YT128 (3), and YT128b (4) (A) ZhuI and ZhuJ are CYCs specific for unreduced octaketides. CH999/pYT105, which coexpressed ZhuI and ZhuJ from the R1128 PKS with the act minimal PKS, produced the anthraquinone compound TMAC (7). ZhuI and ZhuJ are thus able to cyclize unreduced octaketides. Previously characterized octaketide CYCs act ARO/CYC and act CYC are specific for reduced octaketides only. ZhuI and ZhuJ are used for reconstituting R1128 biosynthesis. (B) Reconstitution of R1128 biosynthesis using the heterologous combination of tcm minimal PKS, R1128 loading module, and CYCs ZhuI and ZhuJ. The alkylacyl-ZhuG intermediate synthesized by the loading module is able to prime the decaketide synthase from tcm minimal PKS. Owing to backbone size restrictions, the tcm KS-CLF primed with the alkylacyl groups are only able to extend the polyketide by seven additional malonyl units, resulting in an alkyl-octaketide. ZhuI and ZhuJ are able to transform the unreduced octaketide into YT128 (3) and YT128b (4). The decarboxylated versions of 3 and 4, which are R1128b (5) and R1128c (6), respectively, are also observed in the extracts of CH999/pYT128. The R1128 family of antibiotics represents a unique set of anthraquinones that contain an unreduced C-9 carbonyl (present as an enolic C-9 hydroxyl in R1128). Since members of the act ARO/CYC family are unable to cyclize an unreduced octaketide ( McDaniel et al. 1994b ) and since enzymes from the tcm ARO/CYC family have alternative regiospecificity of cyclization ( McDaniel et al. 1995 ), AROs and CYCs were sought from the R1128 biosynthetic pathway. ZhuI and ZhuJ are two putative enzymes present in the R1128 PKS ( Marti et al. 2000 ). ZhuI, which is homologous to the act ARO/CYC, was predicted to be a first ring CYC, while ZhuJ was predicted to be a second ring CYC. To test these hypotheses, plasmids pYT105 and pYT92 (see Table 1 ) were constructed, coexpressing ZhuI and ZhuJ with the act minimal PKS and the tcm minimal PKS, respectively. Analysis of compounds produced by CH999/pYT92 revealed the decaketides SEK15 and SEK15b as the major products, suggesting ZhuI and ZhuJ did not recognize an unreduced decaketide. However, the anthraquinone compound ( Bartel et al. 1990 ), 3,6,8-trihydroxy-1-methylanthraquinone-2-carboxylic acid (TMAC) ( 7 ; also known as laccaic acid D, a well known plant-derived pigment) ( Figure 5 A; Table 2 ; atoms are numbered according to order in polyketide backbone), was isolated from CH999/pYT105 at 10 mg/l, in addition to the known products of act minimal PKS, SEK4 and SEK4b (20 mg/l). Thus, ZhuI and ZhuJ are able to cyclize an unreduced octaketide into the corresponding anthraquinone. The incomplete transformation of nascent octaketides into TMAC may be due to the fact that an acetate-primed octaketide is not a natural substrate of the two CYCs (see below). The identification of ZhuI and ZhuJ as the appropriate CYCs for the synthesis of R1128-like anthraquinones prompted the design of pYT128, which coexpresses tcm KS-CLF, ZhuN, ZhuI, ZhuJ, and the R1128 initiation module. Plasmid pYT92 (which lacks the initiation module) was used as the negative control. In addition to the decaketides SEK15 and SEK15b, two new anthraquinone compounds, YT128 ( 3 ) and YT128b ( 4 ), were isolated at comparable levels (7 mg/l each) from S. coelicolor CH999/pYT128. The two compounds account for 50% of total polyketides produced by this recombinant strain. [ 13 C]Propionate and valine feeding experiments verified the alkyl groups installed at C-16 were indeed derived from either propionyl-CoA or isobutyryl-CoA. NMR and MS analyses confirmed the identities of 3 and 4 as alkyl-primed TMAC analogs ( Table 3 ). The natural products, R1128b ( 5 ) and R1128c ( 6 ) (i.e., decarboxylated derivatives of 3 and 4 ), were present at an approximately 20% level to that of 3 and 4 . Table 3 Proton and Carbon NMR Data for TMAC ( 7) , YT128 ( 3 ), and YT128b ( 4 ) Spectra were obtained at 400 MHz for proton and 100 MHz for carbon and were recorded in methanol- d 4 a Carbons are labeled as shown in Figure 5 The engineered biosynthesis of compounds 3 – 6 further validated our hypothesis that the initiation module of the R1128 PKS could productively interact with elongation modules from any bacterial aromatic PKS. Moreover, by incorporating a primer unit with at least five carbon atoms, the tcm KS-CLF effectively became an octaketide synthase with respect to the rest of the polyketide molecule. This was analogous to the conversion of the act KS-CLF into a hexaketide synthase in the presence of the R1128 initiation module. It also suggested that the R1128 KS-CLF was intrinsically a decaketide synthase. Finally, analogous to the act KR, ZhuI and ZhuJ were programmed to recognize full-length polyketide chains based upon the number of β-carbonyl groups, rather than the carbon chain length of the backbone. Our findings below suggest this is a general property for all CYCs. It should be noted that the CYCs ZhuI and ZhuJ were apparently more efficient in processing the unreduced alkyl-primed octaketide than an acetate-primed octaketide, since no alkyl-primed analogs of SEK4 and SEK4b were observed in extracts of CH999/pYT128. (In the absence of ZhuI and ZhuJ, alkyl-primed versions of both SEK4 and SEK4b are the major polyketides produced [data not shown]). Engineered Biosynthesis of Novel Anthraquinones Using Bimodular PKSs To demonstrate the utility of hybrid bimodular PKSs for the rational design of new analogs of known polyketides, we targeted the engineered biosynthesis of alkyl-primed 3,8-dihydroxy-methylanthraquinone carboxylic acid (DMAC) analogs 8 and 9 ( Figure 6 ). Specifically, we inserted the act KR gene into pYT128, along with replacing zhuI and zhuJ with genes encoding act ARO and act CYC, to arrive at the plasmid pYT127. We rationalized that the act CYCs should be able to recognize the reduced, alkyl-primed octaketide. A control plasmid (pYT90) lacking the R1128 initiation module was also constructed, transformed, and analyzed. Figure 6 Engineered Biosynthesis of DMAC Analogs YT127 (8) and YT127b (9) When the bimodular PKS containing tcm minimal PKS and R1128 loading module are coexpressed with tailoring enzymes KR, act ARO/CYC, and act CYC, the desired compounds 8 and 9 were produced. The decarboxylated versions of 8 and 9 are 10 and 11, respectively. All three tailoring enzymes are able to process the unnatural alkyl-octaketide. In the absence of the initiation module, the tcm minimal PKS outfitted with the act KR produced the expected polyketides RM20, RM20b, and RM20c ( McDaniel et al. 1993a ). The targeted anthraquinone carboxylic acids 8 and 9 were isolated at high titers (15 mg/l each, 70% of total polyketide products) in CH999/pYT127. The identities of 8 and 9 were verified by NMR and MS ( Table 4 ). Decarboxylated analogs of both compounds were also observed; these compounds are alkyl-primed analogs of the natural product aloesaponarin II. These findings confirmed that, analogous to ZhuI and ZhuJ, the act KR, act ARO, and act CYC were able to process the octaketide intermediate possessing unnatural functional groups at C-16. Thus, it appears that the substrate recognition features of all auxiliary PKS subunits have evolved to monitor the number of β-ketone functional groups present in the polyketide chain. Table 4 Proton and Carbon NMR Data for YT127 ( 8 ) and YT127b ( 9 ) Spectra were obtained at 400 MHz for proton and 100 MHz for carbon and were recorded in methanol- d 4 a Carbons are labeled as shown in Figure 6 Cytotoxic Properties of Novel Anthraquinones As described above, the biosynthetic engineering methods reported here have yielded practical routes for the production of several new as well as known anthraquinone compounds. Given the track record of this family of natural products as pharmacologically active molecules, compounds 3 , 4 , 9 , and DMAC were assayed for cytotoxic activities against human mammary adenocarcinoma MCF-7 cells. Apoptosis was observed after 24 h of drug treatment, and IC 50 values were recorded after 5 d of drug addition. The IC 50 values for reduced compounds DMAC and 9 are 26.9 and 21.7 μg/ml, respectively, while the IC 50 values for unreduced anthraquinones 3 and 4 are 3.4 and 1.7 μg/ml, respectively. Thus inserting the hydroxyl group at C-9 results in a 10-fold increase in cytotoxic activity. The new compounds also show modest improvement in cytotoxic activity relative to the natural products 5 (R1128b, IC 50 = 9.5 μg/ml) and 6 (R1128c, IC 50 =6.2 μg/ml) ( Hori et al. 1993 ), suggesting an additive effect of both the C-9 OH and C-2 COOH groups. Inhibition of glucose-6-phosphatase. Recently, the natural product mumbaistatin ( Figure 7 A) was identified as an extremely potent inhibitor (IC 50 = 5 nM) of the glucose-6-phosphate translocase enzyme complex, an attractive target for the treatment of Type II diabetes ( Vertesy et al. 2001 ). The core of mumbaistatin consists of an anthraquinone moiety that is related to several engineered compounds discussed in this report. For example, the carboxylic acid at position C-1 and the reduced C-9 are identical to those present in compounds DMAC, 8 , and 9 . We tested the inhibitory activities of some of these polyketides against glucose-6-phosphate translocase using intact male rat liver microsomes. Figure 7 Inhibition of Glucose-6-Phosphate Translocase Activity by Anthraquinones (A) The chemical structure of mumbaistatin. Mumbaistatin is an extremely potent inhibitor of glucose-6-phosphate translocase (IC 50 = 5 nM). The core of the molecule is a reduced carboxylic acid containing anthraquinone, which is also observed in compounds such as DMAC, YT127, and YT127b. (B) Inhibition of glucose-6-phosphate translocase activity by novel anthraquinones. The inhibition assay is performed as described in the experimental section. The control contains the rat liver microsome and glucose-6-phosphate only. Chlorogenic acid (IC 50 = 0.26 mM) was used as a reference. The G6Pase activity of the microsome in the presence of six anthraquinones (DMAC, 3, 4, 7, 8 , and 9 ) was measured. For each compound, three different concentrations (50, 25, and 12.5 mM) were used to detect dose-dependent inhibition. No inhibition was observed for DMAC or 7 . Dose-dependent inhibition was observed for compounds 3, 4 , 8 , and 9 . The integrity of the microsomes was first verified by comparing the glucose-6-phosphatase (G6Pase) activity in the presence of either glucose-6-phosphate or mannose-6-phosphate. The activity of G6Pase was then measured in the presence of chlorogenic acid (IC 50 = 0.26 mM), DMAC, 3 , 4 , 7 , 8 , or 9 using a colorimetric assay described earlier ( Arion 1989 ) ( Figure 7 B). Dose-dependent inhibition was observed in the 10–50 μM range with the alkylacyl-primed compounds 3 , 4 , 8 , and 9 , but not DMAC or 7 . Our results demonstrate the following: (i) a long substituent at C-16 in an anthraquinone is important for targeting the membrane-bound G6Pase, and (ii) the C-9 position of the anthraquinone can be chemically modified without significantly affecting the enzyme-inhibitor interactions. Discussion The engineering of the primer units of macrolide antibiotics is a well-established strategy for generating new natural product analogs with modified chemical and biological properties ( Jacobsen et al. 1997 ; Marsden et al. 1998 ; Moore and Hertweck 2002 ). In contrast, manipulation of the ordinarily invariant acetate primer unit of bacterial aromatic polyketides has not been recognized as a general methodology in biosynthetic engineering, presumably owing to the apparently high efficiency with which these PKSs decarboxylate malonyl extender units to generate acetate primers. An exception to this principle has been recently demonstrated in the case of the enterocin PKS, which ordinarily incorporates a benzoic acid primer unit, but can also accept a range of aryl acids to generate substituted enterocins ( Kalaitzis et al. 2003 ). In this report, we have described a general method for modifying the primer unit of any aromatic polyketide by engineering hybrid bimodular PKSs. This method can be used to construct hitherto undiscovered polyfunctional aromatic scaffolds, as illustrated by compounds 1 and 2 ; alternatively, regioselective modifications of known polyketides, such as 8 and 9 , can be achieved. Notably, structural analysis of these novel compounds also revealed fundamentally new properties of bacterial aromatic PKSs, as summarized below. The KS-CLF Prefers Nonacetate Priming over Decarboxylative Priming Most bacterial aromatic PKSs catalyze chain initiation by decarboxylating an ACP-bound malonyl extender unit to yield an acetyl primer unit, a reaction that is catalyzed by the KS-CLF. In order to install nonacetate primer units in an aromatic polyketide backbone, one must bypass this decarboxylative priming mechanism. Genetic and enzymological analysis of the R1128 PKS, which utilizes a range of nonacetate primer units, has revealed the existence of two PKS modules. Each module includes a distinct KS and an ACP. Previous studies have shown that these two KS-ACP pairs have orthogonal molecular recognition features, leading to the speculation that the initiation module may be able to productively interact with other bacterial aromatic PKSs to synthesize hybrid polyketides. However, the ability of the R1128 initiation module to kinetically compete with the intrinsic decarboxylative priming mechanism of the heterologous PKS was unexplored. To address this question, we coexpressed the entire R1128 initiation module with either the act or the tcm minimal PKS. The efficient biosynthesis of compounds described in this report shows that, although decarboxylation cannot be completely suppressed, both PKSs have an intrinsic preference for nonacetate primers over decarboxylative chain initiation. It should be noted that although acetate-primed products are observed in conjunction with nonacetate-primed compounds (e.g., CH999/pYT46 cosynthesizes mutactin along with compounds 1 and 2 ), the former class of products may not be derived via decarboxylative priming in strains carrying bimodular PKSs. Instead, they may arise as a result of premature diketide transfer from the initiation module to the elongation module before the β-carbonyl can be reduced. Future isotope labeling studies on such systems should be useful for quantifying the distribution between polyketide chains derived from bimodular PKSs versus those that arise via decarboxylative priming. Our findings are consistent with the fact that the frenolicin PKS from Streptomyces roseofulvus can synthesize both nanaomycin (an acetate-primed aromatic polyketide) and frenolicin (its butyrate-primed analog) ( Tsuzuki et al. 1986 ). They also explain earlier observations that the doxorubicin (a propionate-primed polyketide) and oxytetracycline (a malonamate-primed polyketide) minimal PKSs yield acetate-primed polyketides, when expressed alone ( Fu et al. 1994 ; Rajgarhia et al. 2001 ). Thus, notwithstanding its widespread prevalence, decarboxylative priming by the KS-CLF can be regarded as a default mechanism for chain initiation that occurs when alternative primer units are unavailable. The potential for recombining naturally occurring initiation and elongation PKS modules from bacterial aromatic PKSs is enormous. Other than the R1128 biosynthetic pathway, initiation modules with attractive primer unit specificity can also be found in the doxorubicin, frenolicin, enterocin, and (presumably) oxytetracycline biosynthetic pathways. It should be possible to recombine these synthase units with elongation modules from the act , frn , tcm , and whiE ( Yu and Hopwood 1995 ) PKSs (which synthesize C 16 –C 24 backbones) to yield a range of reactive backbones whose subsequent fates can be controlled by previously analyzed auxiliary PKS subunits and tailoring enzymes. Molecular Recognition Features of the KS-CLF and Auxiliary PKS Subunits By generating a variety of nascent and highly reactive alkylacyl polyketide intermediates in situ, we have been able to probe the properties of the key aromatic PKS components, including the KS-CLF, the KR, and different subclasses of AROs and CYCs. Such studies have provided insight as to whether carbon chain length of the backbone or the repetitive poly-β-ketone functionality are the primary factors influencing the substrate specificity of these proteins. Our results demonstrate that, since chain length specificity of KS-CLF heterodimers is primarily dictated by the backbone size, incorporation of bulky, nonacetate primer units is compensated for by a reduced number of condensation cycles. Thus, hexaketides and octaketides are synthesized by the act and tcm KS-CLF, respectively, when these KSs are primed with pentanoyl (or 4-methylpentanoyl) primer units. In contrast to KS-CLF subunits, the regioselectivity of auxiliary PKS enzymes such as KR, ARO, and CYC is unaffected by the incorporation of nonacetate starter units. For example, the act KR selectively reduced the C-9 keto group in an acetate-primed octaketide (DMAC), an alkyl-hexaketide (see Figure 4 B), and an alkyl-octaketide (see Figure 5 B). This observation confirms an earlier proposal ( McDaniel et al. 1993b ) that such KRs recognize fully synthesized polyketide chains, rather than the β-ketone of a partially elongated intermediate. Similarly, the act ARO and CYC process a reduced (but not unreduced) octaketide, regardless of the primer unit, although they are unable to recognize hexaketides or decaketides. In contrast, the R1128 CYCs act upon an unreduced, but not reduced, backbone, regardless of the primer unit. The Initiation Module of a Bimodular Aromatic PKS Our studies have revealed that each initiation module component, ZhuG (see Figure 4 A), ZhuH (data not shown), and ZhuC (data not shown) are essential for priming with a nonacetate building block. Deletion of any of these three genes from the constructs shown in Table 1 completely abolishes nonacetate priming by the KS-CLF, but leaves the decarboxylative priming mechanism intact. Although the roles of ZhuG and ZhuH in the initiation pathway have been reported ( Meadows and Khosla 2001 ; Tang et al. 2003 ), the role of ZhuC is unclear. ZhuC is homologous to the MAT and was therefore putatively assigned as a second malonyl transferase that malonylates ZhuG. However, subsequently it was shown that MAT can malonylate ZhuG with high efficiency ( k cat , approximately150 s –1 ) ( Tang et al. 2003 ), whereas ZhuC is sluggish in malonylating ACPs (at a rate approximately ten times slower than MAT [data not shown]). In light of these observations, we propose that ZhuC catalyzes transacylation between the diketide-ZhuG and ZhuN, leading to an alkylacyl-ZhuN intermediate that can then be transferred onto the KS-CLF. Future biochemical analysis may be able to verify this property of ZhuC. Engineered Biosynthesis of Diverse Aromatic Polyketides via Bimodular PKSs To further expand the repertoire of primer units that can be introduced into aromatic polyketides using bimodular PKSs, one could (i) alter the substrate specificity of the R1128 initiation module and/or (ii) engineer in vivo metabolic pathways for new types of primer units. KSIII homologs found in initiation modules serve as gatekeepers in primer unit selection and are therefore attractive targets for protein engineering. The X-ray crystal structure of ZhuH has recently been solved and has led to the identification of a binding pocket for the acyl-CoA moiety ( Pan et al. 2002 ). ZhuH adopts a dimeric thiolase fold and selected residues at the interface between the two subunits control size and flexibility of the binding pocket, preventing acyl groups larger than isovaleryl groups from entering the pocket. The corresponding amino acids in FrnI, the homolog to ZhuH in the frn PKS, are occupied by bulkier residues, thus excluding acyl groups larger than acetyl-CoA. Therefore, altering the size and polarity of these gatekeeping residues using rational mutagenesis and directed evolution may enlarge the repertoire of acyl-CoA moieties recognized by KSIII enzymes. Amino acid catabolism in S. coelicolor is the primary source of primer units such as isobutyryl-CoA (valine) and isovaleryl-CoA (leucine). These pathways involve transamination (catalyzed by branched-chain amino acid transaminases) to convert the amino acid into the corresponding α-ketoacid, followed by decarboxylation (catalyzed by acyl-CoA dehydrogenase [AcdH]) to yield the corresponding acyl-CoA ( Zhang et al. 1999 ). These catabolic pathways in actinomyces are presumably tolerant of unnatural amino/α-keto acids, as illustrated by the incorporation of a large variety of primer units into the macrolide avermectin through precursor feeding ( Dutton et al. 1991 ). It may therefore be possible to expand the repertoire of primer units in S. coelicolor by feeding the recombinant strains, constructed as above, with unnatural amino acids such as allylglycine, norvaline, norleucine, and fluorinated derivatives thereof. In addition, heterologous expression of enzymes involved in the biosynthesis of novel acyl-CoA moieties, such as cyclohexynoyl-CoA ( Cropp et al. 2000 ) and benzoyl-CoA ( Xiang and Moore 2003 ), can be efficient sources of loading module substrates. Successful elaboration of the corresponding acyl-CoA primers into full-length polyketides, through both KSIII-based protein engineering and Streptomyces metabolic engineering, can yield additional polyketide variants. Our observation that anthracycline derivatives generated via such engineering approaches can have improved properties over their natural product counterparts provides further motivation to expand the biosynthetic potential of bimodular PKSs. Materials and Methods Bacterial strains and general methods for DNA manipulation. S. coelicolor strain CH999 was used as the host for transformation by shuttle vectors. Protoplast preparation and PEG-assisted transformation were performed as described by Hopwood et al. (1985 ). All cloning steps were performed in Escherichia coli strain XL-1 Blue. PCR was performed using the pfuTurbo polymerase (Strategene, La Jolla, California, United States). PCR products were first cloned into pCRBlunt vector (Invitrogen, Carlsbad, California, United States), followed by DNA sequencing (Stanford PAN Facility, Stanford, California, United States). Unmethylated DNA was obtained by using the methylase-deficient strain GM2163 (New England Biolabs, Beverly, Massachusetts, United States). Construction of plasmids. The following primers were used to amplify the individual genes: zhuC : 5′-CC TCTAGA TGTACTCGGGTCGAGGAGACCTCCG-3′, 5′-GG ACTAGT GCCACGTTCACCGTTCCGCCGCG-3′; zhuN : 5′-CATGCGACCCG TCTAGA GAAGGAGATTCCG-3′, 5′-CGCGGTTCTGC ACTAGT CAGGCCGCGGCC-3′; zhuG : 5′-CCTG TCTAG AGGGAGGACGAACCC-3′, 5′-TG CTGCAG TCAGCCCGCGGTCTCG-3′; zhuH : 5′-GA CTGCAG CAGAACCGCGAAAGGTGG-3′, 5′-AGTAGTAC GTTTAAAC TCAAGCCGGAGTGGACGGC-3′; act VII /act IV: 5′-GCC GTTTAAAC GCTGGCGCCAAGCTTCTC-3′, 5′-CCGGAGACGCGTCACGGCCGAAGC-3′; zhuI/zhuJ : 5′-GCC GTTTAAAC CGAGGAGCACCCTCATGCGTC-3′, 5′-GGACTAGTCCTCCTCTTCCTGCTCG-3′. The introduced restriction sites are shown in italics. Genes encoding zhuC , zhuH , zhuN , zhuG , and zhuI/zhuJ were amplified from pHU235 ( Marti et al. 2000 ), and genes encoding act VII /act IV were amplified from pRM5 ( McDaniel et al. 1993a ). zhuC , zhuN , and zhuG were cloned as a single 2.1 kb XbaI–PstI cassette; zhuH was cloned as a 1.3 kb PstI–PmeI cassette; act VII /act IV (2.5 kb) and zhuI/zhuJ (1.3 kb) were each cloned as a PmeI–EcoRI cassette. Different combinations of cassettes, as shown in Table 1 , were introduced into either pRM5 (KR- act KS/CLF), pSEK24 ( act KS/CLF), pSEK33 ( tcm KS-CLF), or pRM20 (KR- tcm KS/CLF) to yield pYT46, pYT105, pYT128, and pYT127, respectively. Culture conditions, extraction, and small-scale analysis. The strains were grown on R5 plates containing 50 mg/l thiostrepton at 30°C for 7–10 d. Acyl-CoA precursors such as sodium propionate and valine were added at 1 g/l when needed. For LC/MS and analytic HPLC analysis, one plate was sufficient. The plate was chopped into fine pieces and extracted with 30 ml of ethyl acetate (EA)/methanol/acetic acid (89:10:1). The solvent was removed in vacuo and the residue was redissolved in 1 ml of methanol. The polyketide products were separated and detected by analytical reversed-phase HPLC using a diode array detector at 280 and 410 nm using an Alltech (Vienna, Virginia, United States) Econosphere C18 column (50 mm × 4.6 mm); linear gradient: 20% acetonitrile (ACN) in water (0.1% TFA) to 60% ACN in water (0.1% TFA) over 30 min; 1 ml/min. HPLC retention times (t R , minutes) were as follows: 1 : 13.8; 2 : 15.4; 3 : 20.4; 4 : 21.8; 5 : 24.9; 6 : 26.4; 7 : 17.1; 8 : 22.8; 9 : 24.4; 10 : 27.7; 11 : 29.4. LC/MS was performed at the Vincent Coates Foundation Mass Spectrometry Laboratory at Stanford University using a ThermoFinnigan (San Jose, California, United States) quadrupole ion trap LC/MS system and electrospray ionization (both positive and negative ionization). Large-scale production and isolation. Sufficient number of R5 plates (20–60 plates, depending on the yield of the product) streaked with the desired CH999 strains were grown at 30°C for 1 wk. The plates were chopped into fine pieces and extracted with a minimum of 1 l of EA/methanol/acetic acid (89:10:1). The organic solvents were removed and the residuals were dissolved in 5 ml of methanol. The solution was filtered and injected into a preparative reversed-phase HPLC column (250 × 22.5 mm C-18 column; Alltech Econosil). A 20%–60% ACN in water (0.1 % TFA) gradient (50 min, 5 ml/min) was used to separate the polyketide products. Fractions containing the desired polyketides were combined and concentrated in vacuo. The residuals were redissolved in acetone and applied to a preparative TLC plate (20 cm × 20 cm, 0.25 mm E. Merck [Readington Township, New Jersey, United States] silica gel plates [60F-254]). TLC plates spotted with 1 ( R f = 0.34) and 2 ( R f = 0.41) were developed with EA/methanol/acetic acid (97:2:1), while those spotted with 3 ( R f = 0.29), 4 ( R f = 0.37), 7 ( R f = 0.21), 8 ( R f = 0.34), and 9 ( R f = 0.43) were developed with EA/hexane/acetic acid (90:10:1). The desired bands were excised from the TLC plates and stirred in EA/methanol (10 ml, 9:1) for 2 h. The compounds were eluted from silica using the same solvent and dried in vacuo. NMR and MS characterization of novel compounds. NMR spectra were recorded on Varian (Salt Lake City, Utah, United States) Inova 500 or Mercury 400 instruments and calibrated using residual undeuterated solvent as an internal reference. 1 H and 13 C NMR spectra data are shown in Tables 2–4 . HRFABMS were collected under negative ionization mode as follows: HRFABMS m/z: 1 : 277.1082 (calcd for C 15 H 17 O 5 : 277.1076); 2 : 291.1247 (calcd for C 16 H 19 O 5 : 291.1232); 3 : 355.0823 (calcd for C 19 H 15 O 7 : 355.0818); 4 : 369.0981 (calcd for C 20 H 17 O 7 : 369.0974); 7 : 313.0354 (calcd for C 16 H 9 O 7 : 313.0348); 8 : 339.0871 (calcd for C 19 H 15 O 6 : 339.0869); 9 : 353.1036 (calcd for C 20 H 17 O 6 : 353.1025). Cytotoxicity studies. The studies were performed as described by Hori et al (1993). The cells were maintained at 37°C and growth was measured with the colorimetric MTT assay after each day. The IC 50 values were measured after 5 d. G6Pase activity assay. Male rat liver microsome (Sprague Dawley) was purchased from BD Gentest TM (Becton Dickinson, Franklin Lakes, New Jersey, United States). Aliquots (100 μl, 2 mg/ml) in 0.25 M sucrose were stored at –80°C. Compound stock solutions were prepared in 95% ethanol and diluted with DMSO. Glucose-6-phosphate and mannose-6-phosphate were purchased from Sigma (St. Louis, Missouri, United States). The integrity of microsomes and G6Pase activity was measured based on the colorimetric reaction of inorganic phosphate as previously reported ( Arion 1989 ). The enzyme reaction was initiated by adding 3 μl of microsome to the reaction mixture, which contained 51 μl of assay buffer (50 mM HEPES, 100 mM KCl, 2.5 mM EDTA, 2.5 mM MgCl 2 , and 1 mM DTT at pH 7.2), 3 μl of glucose-6-phosphate (final concentration, 1 mM), and 3 μl of inhibitor sample in DMSO. The reaction mixture was incubated at room temperature, and 13 μl of reaction mixture was taken every 10 min and quenched with 117 μl of working solution (6:2:1 mixture of 0.42% ammonium molybdate tetrahydrate in 1N H 2 SO 4 , 10% SDS in water, and 10% ascorbic acid in water). The blue reduced phosphomolybdate complex is formed after incubation at 50°C for 20 min. The absorbance was measured at 820 nm. Supporting Information Accession Numbers The SwissProt ( www.ebi.ac.uk/swissprot/ ) accession numbers for the proteins and genes discussed in this paper are act ARO/CYC (Q02055), act KS-CLF (Q02059 and Q02062), ZhuA (Q9F6E1), ZhuB (Q9F6E0), ZhuC (Q9F6D6), ZhuG (Q9F6D5), ZhuH (Q9F6D4), ZhuI (Q9F6D3), ZhuJ (Q9F6D2), and ZhuN (Q9F6C8). | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC340942.xml |
516036 | Microarray analysis reveals genetic pathways modulated by tipifarnib in acute myeloid leukemia | Background Farnesyl protein transferase inhibitors (FTIs) were originally developed to inhibit oncogenic ras , however it is now clear that there are several other potential targets for this drug class. The FTI tipifarnib (ZARNESTRA™, R115777) has recently demonstrated clinical responses in adults with refractory and relapsed acute leukemias. This study was conducted to identify genetic markers and pathways that are regulated by tipifarnib in acute myeloid leukemia (AML). Methods Tipifarnib-mediated gene expression changes in 3 AML cell lines and bone marrow samples from two patients with AML were analyzed on a cDNA microarray containing approximately 7000 human genes. Pathways associated with these expression changes were identified using the Ingenuity Pathway Analysis tool. Results The expression analysis identified a common set of genes that were regulated by tipifarnib in three leukemic cell lines and in leukemic blast cells isolated from two patients who had been treated with tipifarnib. Association of modulated genes with biological functional groups identified several pathways affected by tipifarnib including cell signaling, cytoskeletal organization, immunity, and apoptosis. Gene expression changes were verified in a subset of genes using real time RT-PCR. Additionally, regulation of apoptotic genes was found to correlate with increased Annexin V staining in the THP-1 cell line but not in the HL-60 cell line. Conclusions The genetic networks derived from these studies illuminate some of the biological pathways affected by FTI treatment while providing a proof of principle for identifying candidate genes that might be used as surrogate biomarkers of drug activity. | Background The investigative agent tipifarnib is a member of a new class of drugs that were designed to function as a non-peptidomimetic competitive farnesyltransferase inhibitor (FTI). The principal behind this drug class is that protein farnesylation is required for many cell-signaling processes and that dysregulation of cell signaling is thought to be instrumental in driving cell proliferation in several malignancies. The hypothesis that gave rise to this exciting class of drugs is that the inhibition of this enzyme would reduce the uncontrolled cell signaling and provide some control over cell division and malignant cell proliferation. In hematological cancers, tipifarnib has shown significant inhibition of the proliferation of a variety of human tumor cell lines both in vitro and in vivo [ 1 - 3 ]. A recent phase I clinical trial of tipifarnib demonstrated a 32% response rate in patients with refractory or relapsed acute myeloid leukemia [ 4 ]. Furthermore, tipifarnib activity has also been seen in early clinical trials for patients with myelodysplastic syndrome (MDS) [ 5 , 6 ], multiple myeloma (MM) [ 7 ], and chronic myeloid leukemia (CML) [ 8 ]. Mechanism of action (MOA) and biomarker studies with tipifarnib have focused on the oncogenic Ras protein. However, it has since been shown that inhibition of Ras farnesylation does not account for all of the compound's actions. For example, FTIs do not require the presence of mutant Ras protein to produce anti-tumor effects [ 4 ]. Several other proteins have been implicated as downstream targets that mediate the anti-tumorigenic effects of FTIs. The regulation of RhoB, a small GTPase that acts down-stream of Ras and is involved in many cellular processes including cytoskeletal regulation and apoptosis, has been proposed as a mechanism of FTI-mediated anti-tumorogenesis [ 9 ]. Additional proteins involved in cytoskeletal organization are also known to be farnesylated including the centromere proteins, CENP-E and CENP-F, protein tyrosine phosphatase, and lamins A and B. Thus, one possible mode of action of FTI's may be due to their inhibiting effects on cellular reorganization and mitosis. In addition to possibly inhibiting cellular reorganization and mitotic pathways, it is also known that FTIs indirectly modulate several important signaling molecules including TGFβRII [ 10 ], MAPK/ERK [ 11 ], PI3K/AKT2 [ 12 ], Fas (CD95) and VEGF [ 13 ]. The regulation of these effectors can lead to the modulation of signaling pathways involving cell growth and proliferation, and apoptosis. Thus, FTIs may have complex inhibitory effects on a number of cellular events. Where there are multiple candidate pharmacologic biomarkers as is the case with tipifarnib, a comprehensive, parallel study of all candidates is required. Here we describe the application of DNA microarray technology to the measurement of the steady-state mRNA level of thousands of genes simultaneously. This comprehensive experimental approach allows for the simultaneous analysis of candidate biomarkers as well as the generation of novel hypothesis on MOA and previously uncharacterized biomarkers. Biomarkers that enable the monitoring of drug response have the potential to facilitate clinical evaluation of the compound's safety and efficacy in humans. In the present paper we describe the use of global gene expression monitoring to identify genes and gene pathways that are modulated in acute myeloid leukemia (AML) following treatment with tipifarnib. Several genes involved in FTI biology were identified as being modulated following treatment with tipifarnib in addition to pathways involved with cytoskeletal organization, cell signaling, immunity, and apoptosis. This genome-wide approach of gene expression analysis has provided insight into genes that can be used as surrogate biomarkers for FTI drug activity as well as identifying putative pathways that are involved in the drug's anti-leukemic mechanism of action. This is the first successful report of the application of genomics to this novel class of drugs. Methods Cell culture The AML cell lines AML-193, HL-60, THP-1, and U-937 were obtained from the American Type Culture Collection (Manassas, VA). Cells were grown in RPMI supplemented with 20% FBS. AML-193 was also supplemented with GM-CSF (10 ng/ml; PeproTech Inc., Rocky Hill, NJ), insulin (0.005 mg/ml; Sigma-Aldrich, St. Louis, MO), and transferrin (0.005 mg/ml; Sigma-Aldrich, St. Louis, MO). Cell numbers were counted in a hemocytometer and cell viability was determined by trypan blue dye exclusion assay. Tipifarnib was dissolved in 0.1% DMSO. The IC 50 was defined as the dose at which the number of viable cells in the treated sample was 50% of that in the control. This was determined after 7 days of drug treatment. Cytotoxicity assays were performed in duplicate. Control cultures were grown in medium containing vehicle (0.1% DMSO) only. Cells were analyzed for apoptosis by treating with vehicle or tipifarnib (100 nM and 1 μM) over a 5-day time course. Cells were stained with Annexin V and propidium iodide daily according to the manufacturers protocol (Roche Applied Science, Indianapolis, IN) and analyzed by FACS. Bone marrow processing Bone marrow samples were collected from consenting patients both before and during treatment with tipifarnib [ 4 ], diluted with PBS and centrifuged with Ficoll-diatrizoate (1.077 g/ml). White blood cells were washed twice with PBS, resuspended in FBS with 10% DMSO and immediately frozen at -80°C. Some characteristics of the two patient samples used in the present study are shown in Table 1 . Ras mutational status Analysis of activating mutations in N- ras , K- ras , and H- ras codons was determined by PCR and RFLP analysis as previously described [ 1 ]. Microarray analysis Total RNA was isolated using the Qiagen RNeasy kit (Qiagen, Valencia, CA) and treated with DNase1 (DNase1 kit, Qiagen, Valencia, CA) to remove any residual genomic DNA. Probe preparation was performed as previously described [ 14 ]. Linear amplification was performed on total RNA to obtain at least 15 μg of amplified RNA. Cell line mRNA and patient sample mRNA underwent one and two rounds of linear amplification respectively. Microarrays were generated and probes hybridized as described [ 15 ]. Samples were hybridized to arrays that contained 7452 cDNAs from the IMAGE consortium (Integrated Molecular Analysis of Genome and their Expression: ResGen™, Invitrogen Life Technologies, Carlsbad, CA) and Incyte libraries (Incyte, Palo Alto, CA). The intensity level of each microarray was scaled so that the 75 th percentile of the expression levels was equal across micro-arrays. To control for chip errors, replicate clones on each chip that displayed a coefficient of variance (CV) greater than 50% of the mean were excluded from the analysis. Since background intensity was a maximum of 30 relative fluorescent units (RFU) for all experiments, a threshold of 30 RFU was assigned to all clones exhibiting an expression level lower than this. The microarray data were then normalized by quantile normalization and logarithmically transformed before further analysis. Statistical analysis Analysis of variance (ANOVA) and t-tests were used to investigate the effect of drug treatment and time and their interactions for each gene. Multiple hypotheses testing was controlled by applying the false discovery rate (FDR) algorithm [ 16 ]. All statistical analyses were performed in S-Plus 6.1 (Insightful Corporation). Ratio matrices were generated based on pair-wise analysis of treated versus control samples. Hierarchical clustering was performed using a correlation metric and complete linkage (OmniViz Pro™, OmniViz, Maynard, MA). Pathway analysis A total of 1198 genes that had a false discovery rate (FDR) < 0.1 (p < 0.05) in at least one cell line were used for the pathway analysis. Gene refseq accession numbers were imported into the Ingenuity Pathway Analysis software (Ingenuity Systems). 898 of these genes were mapped to the Ingenuity database. Seventy-two of these genes were also affected in patient samples (p < 0.05, FDR < 0.3) and were, therefore considered to be significantly regulated by tipifarnib. The identified genes were mapped to genetic networks available in the Ingenuity database and were then ranked by score. The score is the probability that a collection of genes equal to or greater than the number in a network could be achieved by chance alone. A score of 3 indicates that there is a 1/1000 chance that the focus genes are in a network due to random chance. Therefore, scores of 3 or higher have a 99.9% confidence of not being generated by random chance alone. This score was used as the cut-off for identifying gene networks significantly affected by tipifarnib. Real Time RT-PCR The genes and primers used for RT-PCR are listed in Table 2 . Due to the limited amount total RNA from the patient samples, RNA that had been through one round of linear amplification was used. The Roche Molecular LightCycler (Roche Applied Science, Indianapolis, IN) with Syber Green I system detection was used for real time PCR. PCR thermocycling consisted of denaturation at 95°C for 45 seconds, followed by 30 cycles at 62°C for 10 seconds, and 72°C for 12 seconds. Samples were run in triplicate with both test primer sets and the control gene eukaryotic elongation factor 1 alpha (EEF1A1). This gene was used to control for differences in the amount of target material since initial microarray experiments found that expression of the EEF1A1 gene did not vary significantly between drug-treated and control cells. A standard curve was also run in each PCR reaction. Fold changes were calculated by normalizing the test crossing thresholds (Ct) with the EEF1A1 amplified control Ct. Results and Discussion Response of AML-like cell lines to tipifarnib Tipifarnib inhibited the growth of 4 human AML cell lines in a dose-dependent manner. The IC 50 of these cell lines when treated with tipifarnib ranged from 19 to 134 nM (Table 3 ). The mutation status of the ras oncogenes in the AML cell lines are also shown. These data indicate that the four AML-like cell lines are sensitive to tipifarnib treatment at concentrations well below the micromolar concentrations that is achievable in the bone marrow of leukemia patients [ 4 ]. However, there was no correlation between the type of ras mutation and sensitivity to the drug. These data are consistent with the activity of tipifarnib in vivo and allowed for further characterization of gene expression changes in these cells after treatment with pharmacologically relevant drug concentrations. Identification of genes differentially expressed in tipifarnib-treated AML cells We next asked what genes are modulated following treatment of AML cells with tipifarnib and if there are differences between the affected gene networks in cell lines compared to primary cells from patients. To this end we first selected the three most sensitive cell lines and treated them with tipifarnib or vehicle alone over a 6-day time course. A standard concentration of 100 nM tipifarnib was chosen to ensure exposure within the pharmacologically active range of the compound (Fig 1 ). Samples for RNA analysis were harvested daily from duplicate cell cultures. Message RNA was isolated, amplified and hybridized to the cDNA microarrays containing approximately 7000 genes. Based on scatter plot analysis the microarray data was found to be highly reproducible between duplicate samples (Fig 2 ). A one-way ANOVA was employed to identify genes that were significantly changed over the 6-day time course compared to time-matched controls. A total of 1198 genes were significantly regulated (p < 0.05 with a false discovery rate of less than 10%) in at least one of the cell lines over the time course (Supplementary Figure A [see Additional file 1 ]). We also had access to bone marrow samples from two newly diagnosed AML patients enrolled in a phase I trial for tipifarnib [ 4 ]. The gene expression profiles in pre-treated leukemia cells were compared to those during drug treatment at days 8, 15 and 22. 1016 genes were significantly changed (p < 0.05, FDR < 0.3) during farnesyltransferase inhibition in vivo (Supplementary Figure B [see Additional file 1 ]). A total of 180 genes were common between the cell line and patient data sets, 141 of these had known functions (see Additional file 2 ). Real time RT-PCR showed good agreement with the microarray data (r 2 = 0.87; Fig 3 ). There are several known targets of FTIs including ras, RhoB, centromere proteins, lamins, PI3K/AKT, and TGFβRII [ 3 , 10 ]. While the majority of these genes were present on our expression array (except the lamins) we only found k-ras to be significantly regulated. However, while not significant, up-regulation of TGFβRII was confirmed by RT-PCR (Fig 3 ). The absence of strong regulation of TGFβRII in the current data set may be due to the different FTI and/or the different culture conditions that were employed compared to previous reports [ 10 ]. Interestingly, k-ras was significantly down-regulated in our system. While k-ras is a target of FTIs it has been shown to undergo alternative geranylgeranylation when farnesylation is inhibited and may therefore not be an important anti-tumorgenic target post-translationally; however, it maybe a relevant target at the transcriptional level [ 17 ]. Repression of k-ras transcription has also been shown recently in a mouse model designed to identify genes that are related to the transformation-selective apoptotic program triggered by FTIs [ 18 ]. K-ras may therefore warrant further investigation as a candidate transcriptional target of FTIs. Identification of genetic networks affected by tipifarnib To further refine the list of FTI-affected genes we next investigated which of these genes are known to interact biologically. To this end we carried out pathway analysis on the above 180 genes using the Ingenuity Pathway Analysis (IPA) tool. Seventy-nine (72 unique) of these 180 genes mapped to genetic networks as defined by the IPA tool. These networks describe functional relationships between gene products based on known interactions in the literature. The tool then associates these networks with known biological pathways. Five networks were found to be highly significant in that they had more of the identified genes present than would be expected by chance (Table 3 ). These networks were associated with the cell cycle, apoptosis, proliferation, chemotaxis, and immunity pathways. The study by Kamasani et al also found cell cycle pathways were repressed and immunity and cell adhesion pathways were activated by FTI treatment [ 18 ]. The 79 genes were then analyzed by two-way hierarchical clustering to compare the expression profiles of the AML samples (Fig 4 ). A number of observations could be made using this visual approach. First, although there were some outliers, the majority of duplicate samples clustered close together again demonstrating the reproducibility of the results. Similarly, a number of replicate clones of the same gene clustered next to each other thereby improving the confidence of the microarray data. As expected, samples from the same cell line or patient clustered together. However, samples from late in the time courses have very different expression profiles possibly reflecting greater differences in the transcriptional activity between control and treated cells at this late stage of drug treatment. Interestingly, the cluster analysis showed that the HL-60 profile was most similar to the patient samples indicating it has a more similar response to tipifarnib compared to the patient cells than THP-1 and U-937. This similarity cannot be associated with FAB sub-type since HL-60 was isolated from a patient with M2 AML and the patients examined in this study were M4 and M5 sub-types. Therefore, it is suggested that the different expression profiles seen are due to other genetic differences that impact the specific down-stream effects of FTI inhibition. This may be important when considering appropriate models for FTI investigations. While the cell lines portrayed higher heterogeneity in expression changes compared with the patient samples, the hierarchical clustering did reveal a common set of up- and down-regulated genes. A set of 23 genes was found to be down-regulated in the cell line and patient samples (Fig. 4 ). The major network associated with these genes contained several involved in proliferation including CSK, FGFR3, KRAS2, PPARG, RET, and USF1. Alternatively, 29 genes were commonly up-regulated and network analysis of these revealed activation of apoptotic- and immune-related genes, including CASP6, CD48, FGR, IGF2R, PECAM1, and TNFRSF5. It will be of interest to investigate these genes further to see if they are transcriptional targets of FTIs and if their regulation is additive or synergistic to FTI efficacy. Due to the stringency of our gene selection process it is likely that many genes that are indeed regulated by FTIs, were not identified. For instance, as noted above, of the targets known to be affected by FTIs we identified only k-ras at the transcriptional level. However, the use of pathway analysis tools allows for the identification of networks of genes that are known to interact with each other. This procedure therefore provides additional confidence in the selected genes as well as clues to other genes that may also be regulated but not identified as being significant by the microarray analysis. For example, the network of up-regulated genes (Fig 5A ) includes the lamin B gene, which is indeed a direct target of FTIs. Also, the PIK3R2 gene, which regulates AKT and is a known target of FTIs [ 3 ], can be found in the down-regulated network of genes (Fig. 5B ). This illustrates that the pathway analyses correctly identifies genes that have previously been demonstrated to be either direct or indirect targets of farnesyltransferase inhibition and provides a greater context for screening candidate genes modulated by FTIs. Investigation of apoptosis Since a number of apoptotic genes were identified as being affected by tipifarnib we performed experiments in THP-1 and HL-60 cell lines to verify if they were indeed undergoing apoptosis. Previous reports have shown that two other FTIs can induce apoptosis in myeloid leukemia cell lines [ 11 ] and that tipifarnib causes apoptosis in other malignancies including multiple myeloma [ 19 ], and melanoma [ 1 ]. Annexin V staining demonstrated a significant increase in FTI-mediated apoptosis in THP-1 for both 100 nM (p = 0.027) and 1 uM (p = 0.032) concentrations of tipifarnib (Fig 6 ). A maximum of 23% apoptotic cells were demonstrated at day 5 (Fig. 6 ). No difference in the level of apoptosis was seen between 100 nM and 1 μM of tipifarnib. While apoptosis was activated in the HL-60 cell line this was found to be non-specific since control cells also exhibited this phenomenon during cell culture (data not shown). The lack of FTI-specific apoptosis in HL-60 is consistent with a recent report that also failed to demonstrate tipifarnib-mediated apoptosis in primary AML blasts [ 20 ]. However, in that report apoptosis was measured only two days after treatment where here we found a marked increase in apoptosis at days 3–5. Therefore, our data indicate that tipifarnib can cause apoptosis in AML but may not be detectable at early time points or in AML with certain genetic backgrounds. Conclusions Tipifarnib is one of three FTIs that are currently in clinical trials for treating a variety of cancers [ 21 ] and it is showing promise in hematological malignancies [ 3 - 8 ]. While FTIs were originally designed to inhibit the function of the ras oncogene it has been recently demonstrated that there is no correlation between patient response and ras mutational status [ 4 ]. Additionally, it is clear that other targets of FTIs exist that provide equally important anti-cancer properties. We have reported the use of microarray analysis of both primary human AML cells and AML cell lines following treatment with tipifarnib in order to identify genes and gene pathways that are modulated by this FTI. In particular, genes involved in signaling pathways, down-stream cytoskeletal pathways, and apoptotic events were described. Pharmacodynamic markers that are currently used in the clinic, such as lamin A and HDJ2 [ 22 ], are direct markers of farnesyltransferase inhibition while the majority of genes identified in this work are likely downstream transcriptional targets. Both of these current candidate markers were not present on our microarrays so we did not report on their expression changes. Further analysis will be required to elucidate whether the expression changes seen in our work are due to direct or indirect effects of FTIs. Also, while the currently used clinical biomarkers do not correlate with patient response to FTIs the genes identified here may be candidates for patient stratification [ 3 ]. We are therefore in the process of examining bone marrow specimens from larger phase 2 clinical trials with the aim of validating the panel of pharmacodynamic gene expression markers we have identified here. Such pharmacogenomic analysis will be very important in further elucidating the action of FTIs while providing a platform for identifying patients who could potentially respond to tipifarnib therapy. Competing interests MR, RTB, DA, and YW are employees of Veridex LLC., a Johnson and Johnson company. JEK and JEL have received fees from Johnson & Johnson Pharmaceutical Research and Development in the last 5 years. Author's contributions MR wrote the manuscript and performed experiments. RTB performed Ras analysis. DA helped conceive the experiments. JEK and JEL were principal investigators in the clinical trials from which samples were received. YW helped conceive the experiments. All authors read and approved the final manuscript. List of abbreviations AML: acute myeloid leukemia; FDR: false discovery rate; FTI: farnesyltransferase inhibitor; MDS: myelodysplastic syndrome; MOA: mechanism of action; MM: multiple myeloma; CML: chronic myeloid leukemia Pre-publication history The pre-publication history for this paper can be accessed here: Supplementary Material Additional File 1 A. Two-way hierarchical clustering of 1198 genes regulated in three AML cell lines after tipifarnib treatment. A fold-change ratio was calculated using the treated sample and its matched untreated sample. B. Gene expression changes in patient AML cells. Two-way hierarchical clustering of 1016 genes regulated in two AML patients over a three-week time course. Click here for file Additional File 2 A list of 141 genes with known function that were regulated in both the cell line and patient data sets. Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC516036.xml |
549526 | Epidemiological, clinical and biological features of malaria among children in Niamey, Niger | Background Malaria takes a heavy toll in Niger, one of the world's poorest countries. Previous evaluations conducted in the context of the strategy for the Integrated Management of Childhood Illness, showed that 84% of severe malaria cases and 64 % of ordinary cases are not correctly managed. The aim of this survey was to describe epidemiological, clinical and biological features of malaria among <5 year-old children in the paediatric department of the National Hospital of Niamey, Niger's main referral hospital. Methods The study was performed in 2003 during the rainy season from July 25 th to October 25 th . Microscopic diagnosis of malaria, complete blood cell counts and measurement of glycaemia were performed in compliance with the routine procedure of the laboratory. Epidemiological data was collected through interviews with mothers. Results 256 children aged 3–60 months were included in the study. Anthropometrics and epidemiological data were typical of a very underprivileged population: 58% of the children were suffering from malnutrition and all were from poor families. Diagnosis of malaria was confirmed by microscopy in 52% of the cases. Clinical symptoms upon admission were non-specific, but there was a significant combination between a positive thick blood smear and neurological symptoms, and between a positive thick blood smear and splenomegaly. Thrombopaenia was also statistically more frequent among confirmed cases of malaria. The prevalence of severe malaria was 86%, including cases of severe anaemia among < 2 year-old children and neurological forms after 2 years of age. Overall mortality was 20% among confirmed cases and 21% among severe cases. Conclusions The study confirmed that malaria was a major burden for the National Hospital of Niamey. Children hospitalized for malaria had an underprivileged background. Two distinctive features were the prevalence of severe malaria and a high mortality rate. Medical and non-medical underlying factors which may explain such a situation are discussed. | Background Malaria accounts for 1 in 5 of all children deaths in Africa. Late presentation, misdiagnosis, inadequate management, unavailability or stock-outs of effective drugs are factors that influence high case-fatality rates even among hospital in-patients [ 1 ]. One of the targets of the Roll Back Malaria (RBM) initiative [ 2 ] is to establish systems to guarantee that adequate stocks of drugs and clinical consumables are available to health facilities, and that health facility staff are trained and supervised for the rapid identification, resuscitation and subsequent clinical care of children with severe malaria. But many African countries are still a long way from meeting this objective, and it is admitted now that malaria disproportionally affects the poorest populations [ 3 ]. Geographical and socio-economic situation Niger is situated in the eastern part of West Africa in the Sahelo-Saharan zone. It is a landlocked country surrounded by Algeria, Libya, Chad, Nigeria, Benin, Burkina-Faso and Mali. The country's climate is a rainy season from June to October and a dry season during the rest of the year. The hottest months are April and May. According to the Human Development Index (UNDP 2003), Niger ranks 174 th of 175 countries with an annual GDP of about $160 per capita (2001). The demographic situation is characterized by a population of 11,544,000 inhabitants (2002) which is expected to double by 2025 with an annual growth rate of 3.5% (1999–2002). The infant and child mortality rate is 255‰ and 63% of the population live below the poverty line ($1 per day per person) [ 4 - 6 ]. The Bamako Initiative of 1987 was adopted by Niger but in a context of extreme poverty, the generalization of the cost recovery may have an impact on the access to health facilities, which should be assessed. Malaria in Niger Malaria is endemic in Niger with 97% of the population exposed to the risk of malaria [ 1 ], but some areas of the country are much more affected with the disease than others. Varying environmental conditions mean that the level of transmission will vary from one area to another: the northern part has an unstable situation while the South of the country (including Niamey, the capital), which is exposed to the annual monsoon rainfall following a long dry season [ 7 , 8 ] has an intermediate, stable situation. In the Sudan-Sahelian savanna, close to the River Niger, immunity and partial protection are acquired in children over 5 years of age [ 9 ]. Plasmodium falciparum is the predominant species partially associated with Plasmodium malariae and less frequently with Plasmodium ovale , according to geographical variations [ 7 , 8 ]. Among the anopheles of epidemiological importance, only Anopheles gambiae and Anopheles arabiensis may play a substantial role in malaria transmission but Anopheles funestus which was no longer found after 1970 seems to have reappeared in several sudano-sahelian zones of Niger [ 10 , 11 ]. Resistance to chloroquine was first described in 1991 [ 12 ] but few surveys have been conducted in the country so far [ 13 , 14 ]. Chloroquine remains as the first line treatment but such antimalarial treatment policy is now being challenged (2004). Malaria remains a main cause of morbidity in Niger with an average of 850,000 cases per annum, corresponding to an annual incidence of 80‰ inhabitants, and 33.97% of reported deaths are due to malaria [ 4 ]. These estimates, however, lack accuracy as most of the reported cases are only "presumed cases" and many deaths do not occur at the hospital. It has also been estimated that only 50% of children under 5 with reported fever in the two previous weeks received chloroquine or any anti-malarial treatments, and less than 5% of households have insecticide-treated nets [ 1 ]. Rationale In Niger, as in many sub-Saharan countries, malaria takes a heavy toll, especially among young children in poor households. An evaluation – conducted in 2003 by the University Research Corporation [ 15 ] in the context of the strategy for the Integrated Management of Childhood Illness (IMCI) adopted by Niger in 1996 – showed that 84% of severe cases and 64% of ordinary cases are not correctly managed in national, regional and district hospitals. A high proportion of deaths in the National Hospital of Niamey (NHN) is related to malaria. The poor quality of medical care makes it difficult to assess the real burden of malaria in terms of mortality and morbidity. Misdiagnosis is likely to be quite common, either in excess or default. The objective of the present prospective study was to describe the epidemiological, biological and clinical features of cases of malaria in children in the paediatric department of the NHN in order to recommend more relevant decision-making and interventions. Methods Study area The study was carried out in the NHN during the rainy season of 2003 from July 25 th to October 25 th , which is the peak transmission period. The NHN is the country's main referral hospital and has 852 beds. The paediatric department has two sections: section A (60 beds) for children from 0 to 24 months, and section B (60 beds) for children from 25 months to 15 years of age. A majority of patients come from the urban community of Niamey but some may be referred from all parts of the country. Patients and methods Children from 3 months to 5 years of age (60 months) admitted to Paediatric department A or B for presumed malaria were included in the study and treated following procedures normally used in the department: children went through the emergency department where a clinical examination was performed by doctors or by medical students in their final year, venous blood was drawn for biological tests (for the study, laboratory tests were free of charge), confirmed cases of malaria were treated intravenously with quinine twice a day for 2–3 days, followed by oral treatment (WHO protocol adapted for Niger by the National Malaria Control Program). Upon admission, or within 24 hours, a physician conducted interviews with mothers in order to collect epidemiological data. Information such as age of the parents, number of siblings, the child's rank in relation to siblings, care-seeking previously to hospitalization, delay between the onset of disease and admission to the NHN, and treatment before hospitalization, was collected. The socio-economic background of households was evaluated through the Niamey neighbourhood where children originated from, the educational level of the mother (whether illiterate or not), the occupation of the mother and father, and the existence or non existence of a regular salary. Biological tests were performed along the routine procedures of the NHN's laboratory. Thick and thin blood smears were prepared and stained following standard procedures: Giemsa 10% for 20 minutes (Giemsa R, RAL, CML, Nemours, France) for thick smear, and rapid staining (RAL 555, CML) for thin smear. A thick smear was declared negative only after examination of 200 fields (obj × 100). All smears were checked the following day by a different technician. Parasitaemia was measured on thin smear and expressed as a percentage of parasitised red blood cells. Haematological tests – haemoglobin concentration, hematocrit and complete blood cell count – were performed automatically using a KX21 (Sysmex). The blood glucose rate was performed with a COBAS Mira (Roche) using the glucose oxidase technique (Biomérieux, Marcy l'Etoile, France). According to the 2000 WHO criteria [ 16 ], severe malaria is defined as the presence of P. falciparum on thick smear and at least one of the following clinical or biological criteria : coma (Blantyre coma scale ≤ 2), impaired consciousness (Blantyre >2 and < 5), repeated convulsions (≥ 2/24 hours), prostration, respiratory distress, jaundice, metabolic acidosis (bicarbonates < 15 mmol/L), severe anaemia (Hb < 5 g/dL or Ht < 15%), hyperparasitaemia (parasitaemia > 4% in non-immune patients), macroscopic haemoglobinuria, renal failure, collapse (TAS < 60 mmHg before 5 years of age), abnormal bleeding or pulmonary oedema (X-ray criterion). The conditions in the NHN in 2003 made it impossible to asses acidosis, renal failure (based on creatinine) and pulmonary oedema. Only the following criteria could be considered: coma, impaired consciousness, convulsions, respiratory distress, jaundice, abnormal bleeding, severe anaemia, hyperparasitaemia, hypoglycaemia. Children were weighed. The index weight-for-age (weight/age ratio) was calculated and analysed with the Epi-Info software, based on the reference population defined by the US National Center for Health Statistics.(NCHS). Malnutrition was defined as a weight/age ratio more than 2 SD below the NCHS's reference population [ 17 ]. Malnutrition was considered moderate between -2 and -3 SD and severe below 3 SD. Statistical methods Version 6 of the program Epi-Info was used for statistical analysis. ANOVA and Kruskal Wallis tests were used for mean comparisons whenever appropriate. Proportions were compared using chi-square tests. Bivariate odds ratio and their 95% confidence intervals were calculated to measure the combination between fatal outcome and different variables. For all tests, a p-value below 0.05 was considered significant. Ethical issues The routine management of the children was not changed. Verbal consent was obtained from the parents after informing them in their native language. The parents were allowed to remove their child from the study at any time during follow-up. All data was entered anonymously into a database and identification numbers were coded. No ethnic data was registered. The protocol was submitted to and approved by the Ministry of Public Health and Endemic Diseases Control's National Malaria Control Program. Results Population 256 children were recorded over the 3 month survey: 138 males, 118 females (sex ratio = 1.17). The average age was 20.2 months: 3–12 months, n = 103 (40%), 13–24 months, n = 103 (40%) and 25–60 months, n = 50 (20%). The average age of parents was 27 years for the mothers, 39 years for the fathers. The number of living siblings averaged 2.8, and 35% of the children were first born. All children came from the urban community of Niamey, and 241/256 (94 %) from poor neighbourhoods: unsanitary neighbourhoods, near the river or near permanent ponds, some of them without electricity or tap water. The educational level of the mothers was low: 246/256 (96%) were illiterate, and 248/256 (97%) were home workers. Only 58/256 fathers (23%) had a monthly fixed income (24 civil servants and 34 salaried workers). The others had unsteady jobs with no stable income. 77 children out of 217 documented files for this item (35%) were considered malnourished and 42% dehydrated as a result of clinical examination. The percentage of malnutrition, however, was 58% based on the weight/age ratio, of which 30% were moderate and 28% severe cases of malnutrition. The youngest children (under 2 years old), particularly the 13–24 months age group, which corresponds to the weaning period, were most affected (Table 1 ). Table 1 Nutritional condition of children upon admission. 3–12 13–24 24–60 number n = 103 % number n = 103 % number n = 50 % Dehydration severe 8 (7.8) 3 (2.9) 0 (0.0) mild 42 (40.8) 32 (31.1) 16 (32.0) absence 46 (44.7) 60 (58.3) 32 (64.0) ND 7 (6.8) 8 (7.8) 2 (4.0) Clinical malnutrition a severe 4 (3.9) 8 (7.8) 0 (0.0) mild 29 (28.1) 27 (26.2) 9 (18.0) absence 53 (51.5) 50 (48.5) 37 (74.0) ND 17 (16.5) 18 (17.5) 4 (8.0) Weight / age ratio b severe 26 (25.2) 33 (32.0) 4 (8.0) mild 30 (29.1) 32 (31.1) 6 (12.0) absence 39 (37.9) 30 (29.1) 16 (32.0) ND 8 (7.8) 8 (7.8) 24 (48.0) a : clinically assessed malnutrition b : malnutrition assessed through weight/age ratio ND: not done Clinical and biological diagnosis of malaria Upon admission, mothers reported fever (91% of cases), digestive disorder (51%) and convulsions (19%). These symptoms were often combined (Figure 1 ). The most common combination was fever + digestive disorders (43% of the admissions). Convulsions are always reported in a context of fever. The average time (as reported by parents) between the onset of disease and admission at the NHN was 5.6 days (in the 176 documented files) ranging from 1 (arrival at the hospital the same day) to 35 days. 54% of the children were brought to hospital within 3 days after the early symptoms, and 16% after 7 days.175 children (86%) of the 203 documented cases had already received a treatment before they were admitted to the NHN (home treatment or previous care in another health structure). We could not get accurate information on the kind of treatment they received, whether antimalaria, antibiotic, antipyretic or any other treatment. Parents did not know and most of the time had no written prescription, although 53% had a health book. Only the route of administration was precisely established: oral (49%), intramuscular (41%), intravenous (7%) or intrarectal (3%). Figure 1 Schematic representation of reasons for admission. Number of cases showing each symptom against the total number of children admitted. 132/256 children admitted for presumed malaria (52%) had a positive thick blood smear. The percentage of treated children was the same among children with a negative thick blood smear as among those with a positive one (87% versus 86%). Both groups had quite a similar corrected axillary temperature (38.6° with positive thick smear versus 38.4°). The clinical presentation of children admitted to hospital with a respective prevalence of the various observed symptoms is summarised in Table 2 . Neurological symptoms – coma (Blantyre ≤ 2), impaired consciousness (Blantyre 3–4) or convulsions – were significantly more frequent when the thick smear was positive. Other unspecific symptoms (digestive disorders, dyspnea, hepatomegaly) were also more frequent with a positive blood smear although the difference was not significant. Only 3 cases of respiratory distress were observed, all three among children with a positive blood smear. The presence of splenomegaly was more frequent in children with a positive blood smear (26% versus 13%, p = 0.017). Table 2 Prevalence of clinical symptoms and biological markers upon admission Positive blood smear Negative blood smear p value number/total (%) number/total (%) CLINICAL SYMPTOMS coma a 22/77 (28,6) 7/64 (10,9) p = 0,022 impaired consciousness b 15/77 (19,5) 11/64 (17,2) convulsions c 71/119 (59,7) 36/103 (35,0) < 0,001 digestive disorders 58/93 (62,4) 63/104 (60,6) NS dyspnea 19/98 (19,4) 16/91 (17,6) NS respiratory distress 3/96 (3,1) 0/96 (0,0) NS hepatomegaly 36/107 (33,6) 28/95 (29,5) NS splenomegaly 28/106 (26,4) 12/93 (12,9) 0,017 BIOLOGY anaemia d 125/132 (94,7) 114/124 (91,9) NS severe anaemia e 55/132 (41,7) 39/124 (31,5) NS hypoglycaemia f 30/132 (22,7) 21/124 (16,9) NS thrombopaenia g 93/132 (70,5) 43/124 (34,7) <0,001 a : Blantyre score ≤ 2 b : Blantyre score = 3–4 c : ≥ 2 / 24 hours d : haemoglobin < 11 g/dL e : haemoglobin < 5 g/dL f : glyceamia < 2.2 mmol/L g : platelets < 150 000 / mm3 The haemoglobin rate in the overall population was 6.5 g/dL (6.17 with positive blood smear, 6.94 with negative blood smear, p = 0.038). 94 % of children had anaemia (haemoglobin < 11 g/dL). This massive proportion of anaemia was observed in both groups: 95% versus 92% (NS) for "simple anaemia" (haemoglobin < 11 g/dL), and 42% versus 31% (NS) for severe anaemia (haemoglobin < 5 g/dL) (Table 2 ). Similarly, hypoglycaemia was more frequent with a positive blood smear but the difference was not significant (23% versus 17%, NS) (Table 2 ). The average of platelets was 141,000 in children with a positive blood smear, and 292,000 for those with a negative blood smear (p < 0.001). Based on a definition of thrombopaenia as a number of platelets below 150000/mm 3 , we found that thrombopaenia was statistically combined with a positive blood smear (70% versus 35%, p < 0.001) (Table 2 ). A similar result (p < 0.001) was obtained with the cut off = 100 000/mm 3 . Severe malaria attacks 114/132 children with a positive blood smear (86%) met the criteria of severe malaria according to the clinical and biological criteria explored in this study. The most frequent criteria were convulsions (71/114 = 62%) and severe anaemia (55/114 = 48%) (Table 3 ). Some children were affected with combined criteria: 24/114 (21%) showed impaired consciousness and convulsions, 28/114 (25%) showed a neurological form (coma, and /or impaired consciousness and/or convulsions) and severe anaemia. No collapse, jaundice or spontaneous bleeding were recorded. Total mortality in the studied population was 17%, with a higher -although not significant (20% versus 13%) – rate among individuals with a positive blood smear. Mortality among children with severe malaria was 21%. Table 3 shows specific case fatality rate according to observed criteria of severe malaria. Only hypoglycaemia and coma were related to a higher mortality. Clinical presentation of severe malaria was analysed in terms of age. Figure 2 shows that severe anaemia is more frequent among children of less than 24 months (56% versus 31 %) and, conversely, the neurological forms are more frequent after 24 months of age. The prevalence of hypoglycaemia and high parasitaemia were similar in the two age groups. Table 3 Prevalence of severity criteria among severe malaria cases with respective case fatality rate and relative risk of dying prevalence case fatality rate Odd Ratio p value number n = 114 % number/total % 95 % IC coma a 22 (19.3) 12/21 (57.1) 9.33 (2.85 – 31.58) < 0.001 impaired consciousness b 15 (13.2) 1/14 (7.1) 0.26 (0.01 – 2.07) NS convulsions c 71 (62.3) 18/71 (25.3) 4.42 (0.88 – 29.83) 0.04 respiratory distress 3 (2.6) 3/3 0.002 severe anaemia d 55 (48.2) 11/50 (22.0) 1.10 (0.40 – 3.05) NS hypoglycaemia e 30 (26.3) 11/28 (39.3) 3.72 (1.26 – 11.05) 0.006 high parasitaemia f 39 (34.2) 7/37 (18.9) 0.82 (0.27 – 2.42) NS a : Blantyre score ≤ 2 b : Blantyre score = 3–4 c : ≥ 2 / 24 hours d : haemoglobin < 5 g/dL e : glycaemia < 2.2 mmol/L f : parasitaemia > 4% Figure 2 Prevalence of severe malaria criteria by age group. Prevalence (%) of different WHO criteria among severe cases of malaria in children by age: 3–24 months (black) and 25–60 months (white) Discussion Description of the population Hospitalized children in NHN were from underprivileged families: destitute neighbourhoods and disadvantaged socio-economic groups. This was confirmed by their poor nutritional conditions. 58% of the children had a weight/age ratio below 2 SD. This percentage is higher than national results previously published: the percentage of children with a weight/age ratio below 2 SD was 36% in 1992, and 50% in 1998 [ 18 ]. The delay before admission (5.6 days – ranging 1 to 35) is high but quite similar to other studies conducted in Africa: 6 days (ranging 1 – 30) in Dakar [ 19 ], 3.1 days (ranging 0–61) in Ouagadougou among urban patients [ 20 ]. One of the consequences of such a delay is that patients arrive at an advanced stage of disease. But in most cases, the NHN was not the first form of care sought. Interviews with the mothers highlighted the fact that traditional beliefs are still deeply rooted in families, and they interfere with the care-seeking pattern. For example malaria, "hemar ize" (= "the product of harvest season" in Zarma language) is still considered to be caused by "the smell of new plants during the rain season", despite the fact that there is some understanding of the part played by the mosquito as a result of health education. Convulsions are often linked to popular nosological entities, either "humburukumey" ("fear" in Zarma) whose origin is attributed to djins or witches, or "kyura" ("bird" in Zarma), i.e. "a bird having flown over the pregnant woman and dropped some graveyard soil onto her" (JP Olivier de Sardan, personal communication). This idea of a bird being responsible for children's convulsions is common to most of African societies and is beyond the medical context of malaria [ 21 ]. The occurrence of these convulsions, despite the parent's worry, does not necessarily encourage them to apply for a "modern" medical structure, but to refer more frequently to a "traditional" type of medicine or even to religious or superstitious practices. Thus, the neurological signs of malaria are identified and cured without mentioning malaria. The consequences can be serious, leading to delays in diagnosis and use of inefficient or even harmful traditional treatments. When finally reaching the hospital, because of the requirement for cost recovery, parents cannot afford all the medical examinations or treatments which should be undertaken. Malaria at NHN Diagnosis was confirmed by thick blood film in 52% of cases which shows that clinical diagnosis is not accurate and needs to be confirmed by microscopic examination. This range of 50% of confirmed malaria attacks has already been observed 20 years ago in quite a similar health situation [ 22 ]. The clinical presentation was unspecific, but a significant combination of positive blood smear with neurological symptoms, and with splenomegaly, was observed. From a biological point of view, only thrombopaenia was significantly found to be combined with a positive blood smear. This importance of splenomegaly [ 23 , 24 ] and thrombopaenia [ 25 , 26 ] in the diagnosis of malaria has already been described and discussed in previous studies. With 94% of anaemic children, the "anaemia" symptom is not typical of malaria in this population. Most children must have had one or several previous attacks of malaria during the transmission season even if their test is negative on arrival at the hospital. It would be interesting to assess the impact of treatments carried out before hospitalization (frequent, but accurate reporting is problematic as mentioned above) on the biological diagnosis. Non-specific clinical diagnosis followed by a presumptive treatment carried out without any biological confirmation, with an almost total absence of written follow-up in the management of children, are elements showing how difficult it is to diagnose malaria attacks in endemic areas. Both theoretical and practical parameters interfering in the diagnosis of malaria have been reviewed by Rogier [ 27 ]. Severity of malaria at NHN A particularly high percentage of severe malaria cases (86 %) was noted in the study population. Five children aged 6 months or under (one of them aged 3 months) were suffering from severe malaria, confirming the existence of severe malaria among very young children as already pointed out in other studies [ 28 , 29 ]. The new WHO criteria increase the percentage of severe cases [ 30 ]. This percentage may have been underestimated in this survey as some criteria could not be included for practical reasons. One of the problems was the difficulty of taking prostration into account: even if the WHO definition is clear (inability to sit upright for a child normally able to do so, or to drink in the case of children too young to sit) [ 16 ] this criterion could not be retained because of subjective interpretation. The absence of systematic chest X-ray and systematic creatinine measurements is unlikely to modify the results significantly as other surveys showed that pulmonary oedema and renal failure are infrequent criteria of severity among children [ 19 , 20 ]. The impossibility of performing blood gas analysis was more problematic since the importance of acidosis has been pointed out in several studies [ 16 , 31 , 32 ]. Consequently two major clinical presentations were observed: neurological form (coma, impaired consciousness or convulsions) and severe anaemia. Different breakdowns by age of these clinical presentations with a strong predominance of severe anaemia among 0–2 year-old children were reported. Several studies have already described links between clinical presentation, age of patients, and transmission level [ 19 , 20 , 33 ]. In Niamey, the severity of anaemia is probably enhanced by the very precarious nutritional condition of the infant population. Other severity criteria in the study population were hypoglycaemia and hyperparasitaemia. The small number of cases of respiratory distress (3/114 severe cases) is surprising compared to other studies [ 19 , 34 ], but the 3 children showing a respiratory distress in Niamey died. One may surmise that some health workers are not always sufficiently trained in assessing the severity of patients. The dramatic lack of therapeutic facilities in this hospital, and more specifically the insufficiencies of the intensive care service (5 intensive care beds for a 800-bed hospital, none of which is designated for paediatric resuscitation), does not encourage adequate concern in medical workers to make an accurate assessment based on severity criteria. Mortality The overall case fatality rate of malaria was 20% among confirmed cases and 21% among severe cases. These values are higher than those observed in other African countries also based on 2000 WHO criteria: 11.9% in Ethiopia [ 35 ], 12% in Senegal [ 30 ], and 14.58% in Madagascar [ 28 ]. The first line treatment in Niger remains chloroquine but in the hospital, according to WHO recommendations, children were treated intravenously with quinine twice a day for 2–3 days, followed by oral treatment (WHO protocol adapted for Niger by the National Malaria Control Program). However, such a level of hospital mortality among children who have already received treatment before admission could indicate an increase in the level of chloroquine resistance. Data on resistance is still insufficient in Niger but studies are presently being carried out. Detection of the pfcrt (T76) mutation which confers chloroquine resistance [ 36 ] was performed on 30 isolates among the 114 severe cases. The pfcrt mutation was present in 10/30 samples (ML Ibrahim, personal communication). These preliminary results need to be completed but such a level of genotypic drug resistance clearly seems insufficient to explain the high mortality rate observed at the NHN. Other implied factors have to be identified. One of them is anaemia. As the prevalence of severe anaemia is very high, the capacity of the blood bank to provide blood in a short time is an important factor in reducing mortality. Transfusion facilities are woefully inadequate in the NHN. Conclusions This survey conducted at the National Hospital of Niamey during the 2003 rainy season was an opportunity to examine the features of childhood malaria in a referral hospital of one of the world's lowest income countries. One of the dominant features in the NHN is that malaria affects an underprivileged population already affected by malnutrition. Only 52% of presumed cases are parasitologically confirmed. The rate of severe cases is high (86%) with two main clinical presentations: severe anaemia in less than 2-year old children and neurological form between 2 and 5-year old children. The mortality rate (21% among severe cases) is higher than established by most previously published data. The study highlights several underlying factors that contribute to such an alarming situation: (1) Patient factors: poor socio-economic background leading to multiple difficulties in getting access to healthcare (e.g. transport fees, hospital admission fees, biological tests, drug treatments...), and also the persistence of traditional beliefs interfering with care seeking behaviour. (2) Provider factors: the recovery of medical care costs, although essential for the financial survival of the hospital, leads to multiple difficulties for the patient as seen above. Healthcare quality standards are also too low at all levels of the health organization, not just because of technical deficiencies (e.g. deficiency in intensive care units or in transfusion supplies) but also because of the inadequate training and attitudes of health-care workers. All these factors need to be taken into account in order to find ways of improving the management of malaria in children. A recent awareness initiative is under way at the NHN, and several initiatives are being taken towards a gradual involvement of the hospital in a quality-oriented policy. But such local interventions will not reach their goal if extra efforts are not made at all levels to fight against disadvantages in the access to health facilities. Authors' contributions FGA was an initiator of the study and took part in its development, supervised laboratory staff, took part in the analysis of the data, and wrote this article. EA was an initiator of the study, took part in its development, coordinated the clinical study, took part in the analysis of the data and in the writing of the paper. VL and MG took part in the development of the study, collected clinical and epidemiological data, and analysed the data. MLI conducted molecular tests in the Institut Pasteur de Madagascar. HK and HB supervised medical and nursing staff in the paediatric department. All authors have read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549526.xml |
549532 | Hesperidin, a citrus bioflavonoid, decreases the oxidative stress produced by carbon tetrachloride in rat liver and kidney | Background CCl 4 is a well-established hepatotoxin inducing liver injury by producing free radicals. Exposure to CCl 4 also induces acute and chronic renal injuries. The present study was designed to establish the protective effect of hesperidin (HDN), a citrus bioflavonoid, on CCl 4 -induced oxidative stress and resultant dysfunction of rat liver and kidney. Methods Animals were pretreated with HDN (100 and 200 mg/kg orally) for one week and then challenged with CCl 4 (2 ml/kg/s.c.) in olive oil. Rats were sacrificed by carotid bleeding under ether anesthesia. Liver enzymes, urea and creatinine were estimated in serum. Oxidative stress in liver and kidney tissue was estimated using Thiobarbituric acid reactive substances (TBARS), glutathione (GSH) content, superoxide dismutase(SOD), and Catalase (CAT) Results CCl 4 caused a marked rise in serum levels of ALT and AST (P < 0.05). TBARS levels were significantly increased whereas GSH, SOD and CAT levels decreased in the liver and kidney homogenates of CCl 4 treated rats. HDN (200 mg/kg) successfully attenuated these effects of CCl 4 Conclusion In conclusion, our study demonstrated a protective effect of HDN in CCl 4 induced oxidative stress in rat liver and kidney. This protective effect of HDN can be correlated to its direct antioxidant effect. | Background Drug exposure, ionizing radiations and environmental pro-oxidant pollutants induce free radical formation. Lipid peroxidation initiated by free radicals is considered to be deleterious for cell membranes and has been implicated in a number of pathological situations. Carbontetrachloride (CCl 4 ), an industrial solvent, is a well-established hepatotoxin [ 1 - 3 ]. Various Studies demonstrated that liver is not the only target organ of CCl 4 and it causes free radical generation in other tissues also such as kidneys, heart, lung, testis, brain and blood [ 4 - 6 ]. It has also been reported that exposure to CCl 4 induces acute and chronic renal injuries [ 7 , 8 ]. Case control studies and various documented case reports increasingly establish that hydrocarbon solvents produce renal diseases in humans [ 9 ]. Extensive evidence demonstrates that as a result of the metabolic activation of CCl 4 , • CCl 4 and • Cl, are formed which initiate lipid peroxidation process. Vitamin E protected CCl 4 -induced liver injury indicating the role of oxidative stress in this model [ 10 ]. Studies also show that certain natural extracts containing antioxidants protect against the CCl 4 -induced increased lipid peroxide levels and impairment in hepatic GSH status [ 11 ]. Hesperidin is a flavanone glycoside abundantly found in sweet orange and lemon and is an inexpensive by-product of citrus cultivation [ 12 ]. Hesperidin is effectively used as a supplemental agent in the treatment protocols of complementary settings. Its deficiency has been linked to abnormal capillary leakiness as well as pain in the extremities causing aches, weakness and night leg cramps. Supplemental hesperidin also helps in reducing oedema or excess swelling in the legs due to fluid accumulation. A number of researchers have examined the antioxidant activity and radical scavenging properties of hesperidin using a variety of assay systems [ 13 - 16 ]. Thus the present study was designed to investigate the effect of HDN on CCl 4 -induced oxidative stress and resultant dysfunction of rat liver and kidney. Results Effect on liver enzymes CCl 4 caused a marked rise in serum levels of ALT (control = 45 IU/L) and AST (control = 135 IU/L) demonstrating a marked liver damage. Treatment with HDN decreases the elevated levels of ALT and AST in serum (P < 0.05)(Table 1 ). Both the doses of HDN also attenuated the CCl 4 -induced elevated levels of total bilirubin (control = 0.184 mg/dl). Table 1 Effect of different doses of Hesperidin on CCl 4 induced rise in AST, ALT and total bilirubin. ALT (IU/L) AST(IU/L) Bilirubin total (mg/dl) Control 100 100 100 CCl 4 255.55 ± 23.66 a 444.44 ± 22.56 a 205.21 ± 11.66 a HDN (200) 112.81 ± 13.66 126.66 ± 28.66 108.15 ± 9.66 CCl 4 +HDN (100) 185.18 ± 17.56 a,b 333.33 ± 29.66 a,b 176.63 ± 15.29 a,b CCl 4 +HDN (200) 144.44 ± 15.22 a,b,c 244.44 ± 22.66 a,b,c 149.45 ± 12.45 a,b,c Values are expressed as percent response compared to control rats. a = Statistical significant at P < 0.05 as compare to control, b = Statistical significant at P < 0.05 as compare to CCl 4 , c = Statistical significant at P < 0.05 as compare to CCl 4 + HDN(100) Effect on hepatic and renal TBARS levels CCl 4 challenge caused a marked lipid peroxidation in both liver (control = 1.5 micromoles/mg protein) and kidney (control = 33.86 nmoles/mg protein). Both the doses of HDN decreased the level of lipid peroxidation in liver, but in the kidney, no effect on lipid peroxidation was seen with 100-mg/kg dose, and only the higher dose of HDN (200 mg/kg) could attenuate the increased level of lipid peroxidation (P < 0.05) (Fig 1 ). 7-day oral feeding of HDN per se (200 mg/kg) did not result in a significant alteration of either hepatic or renal TBARS levels. Figure 1 Effect of different doses of Hesperidin on CCl 4 induced lipid peroxidation in rat liver and kidney. Values are expressed as percent response compared to control rats. a = Statistical significant at P < 0.05 as compare to control, b = Statistical significant at P < 0.05 as compare to CCl 4 , c = Statistical significant at P < 0.05 as compare to CCl 4 + HDN(100) Effect on the glutathione levels in CCl 4 treated rats CCl 4 administration markedly decreased the levels of reduced glutathione in both the liver (control = 35.99 micromoles/mg protein) and kidneys (control = 27.99 micromoles/mg protein) demonstrating oxidative stress. HDN (200 mg/kg) per se did not produce any change in the levels of reduced glutathione either in liver or kidney. HDN (100 mg/kg) showed no effect on the levels of reduced glutathione either in liver or kidney in CCl 4 treated rats whereas HDN (200 mg/kg) significantly ameliorated CCl 4 -induced depletion of GSH in both liver and kidney (P < 0.05)(Fig- 2 ). HDN per se (200 mg/kg) did not result in a significant alteration of either hepatic or renal GSH levels. Figure 2 Effect of different doses of Hesperidin on CCl 4 induced depletion in GSH levels in rat liver and kidney. Values are expressed as percent response compared to control rats. a = Statistical significant at P < 0.05 as compare to control, b = Statistical significant at P < 0.05 as compare to CCl 4 , c = Statistical significant at P < 0.05 as compare to CCl 4 + HDN(100) Effect on the antioxidant enzymes in liver and kidneys in CCl 4 treated Rats CCl 4 challenge significantly decreased the levels of SOD and catalase in both liver (SOD: control = 25.66 U/mg protein; Catalase: control = 0.32 K/min) and kidneys (SOD: control = 99.22 U/mg protein; Catalase: control = 0.32 K/min). HDN per se (200 mg/kg) had no effect on these enzymes either in liver or in kidneys. HDN (100 mg/kg) failed to improve the levels of SOD or catalase either in liver or kidneys of CCl 4 administered rats but HDN (200 mg/kg) significantly increased the levels of both enzymes in liver and kidneys of CCl 4 treated rats (P < 0.05)(Fig- 3 and 4 ). 7-day oral feeding of HDN per se (200 mg/kg) did not result in a significant alteration of any of these antioxidant enzymes either in liver or kidney. Figure 3 Effect of different doses of Hesperidin on CCl 4 induced depletion in SOD levels in rat liver and kidney. Values are expressed as percent response compared to control rats. a = Statistical significant at P < 0.05 as compare to control, b = Statistical significant at P < 0.05 as compare to CCl 4 Figure 4 Effect of different doses of Hesperidin on CCl 4 induced depletion in Catalase levels in rat liver and kidney. Values are expressed as percent response compared to control rats. a = Statistical significant at P < 0.05 as compare to control, b = Statistical significant at P < 0.05 as compare to CCl 4 Discussion CCl 4 -induced lipid peroxidation is highly dependent on its bioactivation to the trichoromethyl radical and trichloromethyl peroxy radical [ 17 - 19 ]. It is well known that CCl 4 is activated by the cytochrome P450 system. The initial metabolite is the trichloromethyl free radical, which is believed to initate the biochemical events that ultimately culminate in liver cell necrosis[ 20 , 21 ]. The trichloromethyl radical can form covalent adducts with lipids and proteins, interact with O 2 to form a tricholoromethyl peroxy radical or abstract hydrogen atoms to form chloroform [ 22 ]. Other products include conjugated dienes, lipid hydroperoxides, malonaldehyde-like substances, and other short-chain hydrocarbons [ 23 - 25 ]. In response to hepatocellular injury initiated by the biotransformation of CCl 4 to reactive radicals, "activated" Kupffer cells in liver respond by releasing increased amounts of active oxygen species and other bioactive agents [ 26 ]. Protective effects of various natural products in CCl 4 hepatotoxicity have been reported [ 27 ]. Studies done with Ginseng showed that the antioxidant property of ginsenosides contributes to protection against CCl 4 induced hepatotoxicity in rats [ 28 ]. In the present study, CCl 4 induced a severe hepatic damage as represented by markedly elevated levels of ALT, AST and bilirubin coupled with a marked hepatic oxidative stress. CCl 4 -induced generation of peroxy radicals and O 2 -• leads to inactivation of catalase and SOD. We too observed that CCl 4 challenge significantly decreased the levels of SOD and catalase in liver and kidney. Recently, Szymonik-Lesiuk et al [ 2 ] have shown that CCl 4 intoxication can lead to alteration in gene expression and depletion of SOD and catalase levels in kidney and heart. Oxidative stress causes depletion of intracellular GSH, leading to serious consequences. HDN administration ameliorated the increased level of lipid peroxidation after CCl 4 treatment. Interestingly, only the higher dose of HDN (200 mg/kg) was able to show improvement in the levels of endogenous antioxidant enzymes (SOD and catalase) and GSH in liver. Improvement of hepatic GSH levels in HDN-treated rats in comparison to CCl 4 intoxicated rats demonstrates the antioxidant effect of HDN. We failed to observe any effect of CCl 4 on renal function. Neither BUN nor serum creatinine levels increased after CCl 4 administration (data not shown). Studies by Zimmerman et al [ 29 ] also did not report any rise in BUN levels even after chronic treatment of CCl 4 in nephrectomized rats. They found an increased frequency of glomerulosclerosis and tubulointerstitial alterations in rats with reduced renal mass on CCl 4 administration thereby indicating nephrotoxicity on long-term CCl 4 administration in rats. These findings raise the possibility that renal disease in man is related to hydrocarbon solvent exposure and may also be potentiated by concomitant renal disease or impaired renal function. Ogawa et al [ 30 ] also reported that chronic renal injuries and BUN elevations developed in Balb/c mice only after 12 weeks of CCl 4 intoxication. On the contrary, we estimated the renal function just after 48 hrs of CCl 4 challenge. Thus this brief period might not be sufficient to demonstrate any rise in serum BUN and creatinine levels. Though renal function did not alter after 48 hrs of CCl 4 administration but even this short period of exposure led to a significant oxidative stress in kidneys. Fadhel and coworkers [ 31 ] had also reported increased levels of renal TBARS in rats after CCl 4 exposure which could be improved by black tea extract. Similar observations were also reported with certain Indian ayurvedic Indian preparations [ 32 ]. HDN treatment has been previously demonstrated to improve GSH levels in liver and kidneys of diabetic rats and a decrease in levels of 8-hydroxydeoxyguanosine (8-OHdG), a marker of DNA fragmentation, in the urine of diabetic rats [ 33 ]. HDN in combination with Diosmin has also been shown to inhibit the reactive oxygen radicals production in Zymosan-stimulated human polymorphonuclear neutrophils [ 34 ]. Thus HDN has been shown to reduce oxidative stress in various in-vivo and in-vitro studies. Conclusions In conclusion, our study demonstrated that CCl 4 induces a marked oxidative stress in rat liver and kidney, which is amenable to attenuation by HDN. This protective effect of HDN can be correlated directly to its antioxidant property. Methods Animals Male wistar rats (150 g–200 g), bred in the central animal house of Panjab University (Chandigarh, India) were used. The animals were housed under standard conditions of light and dark cycle with free access to food (Hindustan Lever Products, Kolkata, India) and water. The experimental protocols were approved by the Institutional Ethical Committee of Panjab University, Chandigarh. Drugs Chemicals employed in these studies were reagent grade. Carbon tetrachloride (E Merck, India) was administered subcutaneously in olive oil Hesperidin (Sigma chemical USA) was suspended in 0.5% sodium carboxy methyl cellulose (CMC) and administered orally. Experimental groups and protocol Animals were divided into following groups, each containing 6–8 animals: Control : These animals received a vehicle for HDN (i.e. CMC) by oral route for eight days and on 8 th day, they were administered the subcutaneous injection of olive oil. CCl 4 group : These animals received vehicle for 10 days and were challenged with CCl 4 2 ml/kg/s.c. (40% v/v in olive oil) on 8 th day. In the preliminary studies done in our lab, we observed a very high mortality rate (50–60%) when CCl 4 was administered interaperitoneally. Thus we adopted the subcutaneous route of CCl 4 administration as reported in the literature [ 35 ]. With this route and dose of CCl 4 , the mortality rate reduced to about 20%(1-2/8 animal) HD N group : These rats received only HDN 200 mg/kg/p.o. daily for 10 days CCl 4 + HDN (100) : Rats received HDN continuously for 8 days. On eight day just after HDN treatment they received CCl 4 2 ml/kg/s.c in olive oil. HDN was further continued for 2 more days. CCl 4 + HDN (200) : This group is similar to the above one except that the dose of HDN administered was 200 mg/kg/p.o. On the 10 th day, animals were sacrificed 2 hr, after the last dose of HDN and blood was collected, by carotid bleeding, in centrifuge tubes. Serum was separated and was used freshly for the assessment of renal and liver function tests. Both the kidneys and the liver were quickly harvested and immediately stored at -20°C till further biochemical estimations. Assessment of renal functions Before sacrifice, rats were kept individually in metabolic cages for 24 h to collect urine for estimation of renal function. Serum samples were assayed for blood urea nitrogen (BUN), urea clearance, serum creatinine & creatinine clearance by using standard diagnostic kits (Span Diagnostics, Gujarat, India). Assessment of liver function Serum alanine aminotransferase (ALT) and serum aspartate aminotransferase (AST) were estimated by International Federation of Clinical Chemistry [ 36 ] (ERBA test kits). Serum bilirubin was estimated by Diazo method [ 37 ] (ERBA test kits). Assessment of oxidative stress Post mitochondrial supernatant preparation (PMS) Kidneys and liver were, perfused with ice cold saline (0.9% sodium chloride) and homogenized in chilled potassium chloride (1.17%) using a homogenizer. The homogenates were centrifuged at 800 g for 5 minutes at 4°C to separate the nuclear debris. The supernatant so obtained was centrifuged at 10,500 g for 20 minutes at 4°C to get the post mitochondrial supernatant which was used to assay catalase and superoxide dismutase (SOD) activity. Estimation of lipid peroxidation The malondialdehyde (MDA) content, a measure of lipid peroxidation, was assayed in the form of thiobarbituric acid reacting substances (TBARS) by method of Okhawa et al. [ 38 ] Briefly, the reaction mixture consisted of 0.2 ml of 8.1% sodium lauryl sulphate, 1.5 ml of 20% acetic acid solution adjusted to pH 3.5 with sodium hydroxide and 1.5 ml of 0.8% aqueous solution of thiobarbituric acid was added to 0.2 ml of 10%(w/v) of PMS. The mixture was brought up to 4.0 ml with distilled water and heated at 95°C for 60 minutes. After cooling with tap water, 1.0 ml distilled water and 5.0 ml of the mixture of n-butanol & pyridine (15:1 v/v) was added and centrifuged. The organic layer was taken out and its absorbance was measured at 532 nm. TBARS were quantified using an extinction coefficient of 1.56 × 10 5 M -1 /cm -1 and expressed as nmol of TBARS per mg protein. Tissue protein was estimated using Biuret method of protein assay and the TBARS content expressed as nanomoles per milligram of protein. Estimation of reduced glutathione Reduced glutathione (GSH) in the kidneys and liver was assayed by the method of Jollow et al [ 39 ]. Briefly, 1.0 ml of PMS (10%) was precipitated with 1.0 ml of sulphosalicylic acid (4%). The samples were kept at 4°C for at least 1 hour and then subjected to centrifugation at 1200 g for 15 minutes at 4°C. The assay mixture contained 0.1 ml filtered aliquot and 2.7 ml phosphate buffer (0.1 M, pH 7.4) in a total volume of 3.0 ml. The yellow colour developed was read immediately at 412 nm on a spectrophotometer. Estimation of SOD SOD activity was assayed by the method of Kono et al.[ 40 ] The assay system consisted of EDTA 0.1 mM, sodium carbonate 50 mM and 96 mM of nitro blue tetrazolium (NBT). In the cuvette, 2 ml of above mixture, 0.05 ml hydroxylamine and 0.05 ml of PMS were taken and the auto-oxidation of hydroxylamine was observed by measuring the absorbance at 560 nm. Estimation of catalase Catalase activity was assayed by the method of Claiborne et al [ 41 ]. Briefly, the assay mixture consisted of 1.95 ml phosphate buffer (0.05 M, pH 7.0), 1.0 ml hydrogen peroxide (0.019 M) and 0.05 ml PMS (10%) in a final volume of 3.0 ml. Changes in absorbance were recorded at 240 nm. Catalase activity was calculated in terms of k minutes -1 . Statistical analysis Results were expressed as mean ± SEM. The intergroup variation was measured by one way analysis of variance (ANOVA) followed by Fischer's LSD test. Statistical significance was considered at p < 0.05. The statistical analysis was done using the Jandel Sigma Stat Statistical Software version 2.0. Authors' contributions Naveen Tirkey, Sangeeta Pilkhwal and Anurag did all the biochemical estimations in kidney and liver. Kanwaljit Chopra did the data interpretation after statistical analysis and contributed in manuscript preparation. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549532.xml |
547912 | Histopathological changes in the human larynx following expanded polytetrafluroethylene (Gore-Tex®) implantation | Background Expanded polytetrafluroethelyne (e PTFE, Gore-Tex ® ) has been advocated as an implant material for medialization of the vocal fold. Animal studies involving rabbits and a porcine model have demonstrated host tolerance of the implant. There have been no reports describing the histological changes in a human laryngectomy specimen with a Gore-Tex implant. Case presentation The histological findings in a laryngectomy specimen of a patient previously implanted with e PTFE for medialization of a paralyzed vocal fold following excision of a vagal neurofibroma were studied. Histopathology revealed a mild foreign-body giant cell granulomatous reaction with some associated fibrosis. The granulomatous response was limited to the periphery of the Gore-Tex and although it closely followed the profile of the material it did not encroach into or significantly break up the material. There was no significant neutrophilic or lymphocytic inflammation. Conclusions Our findings are consistent with the animal models confirming that Gore-Tex implantation does not result in a significant granulomatous reaction in the human larynx over a 13-month period. Moreover, there is no evidence of resorption or infection. Further, the lack of lymphocytes in association with the granulomas indicates that there is no significant immunological hypersensitivity. Histologically, the slight permeation by connective tissue is similar to that seen in Gore-Tex vascular and cardiac implants. The degree of the slight giant cell response appears to be dependent on the profile of the material; a sharp edge incited more of a response than a flat surface. | Background Unilateral vocal fold paralysis is symptomatic when it results in failure of the mobile vocal fold to approximate the paralyzed vocal fold during adduction. Despite the lack of movement, the paralyzed vocal fold will often contact the contra lateral mobile vocal fold permitting adequate glottic closure [ 1 ]. Medialization laryngoplasty is a common procedure used to restore glottic competence. This procedure was popularized by Ishiki and initially used in patients with vocal fold paralysis. In recent years, the indications for this procedure have expanded to include most forms of glottic incompetence, including the use of bilateral medialization in mobile vocal folds for vocal fold bowing and atrophy [ 2 ]. Although medialization is now widely accepted, the choice of implant material is still a subject of controversy. The ideal vocal fold injection material would be readily available, have excellent biointegration with no or minimal immunologic response, has an excellent biomechanical in vivo match to the injection site tissues, and is deliverable through a fine-gauge needle. Such a material does not presently exist. Teflon (Polytetrafluoroethylene) was the first modern material used for vocal fold injection. Long-term results, unfortunately, have shown an unacceptably high rate of granuloma formation associated with Teflon injection. This results in a gradual degradation of the injected material by foreign body giant cells and fibroblastic proliferation in the location of the implant [ 3 ]. Autologous tissues such as fat or fascia are well described. These tissues have the benefit of medializing the vocal fold with minimal tissue reaction [ 4 , 5 ]. Collagen injection has been used with some success, its major drawback is the need for serial injections in some cases [ 2 ]. Recently, an expanded polytetrafluroethelyne implant (e PTFE, Gore-Tex ® ) has been advocated as an implant material for medialization of the vocal fold [ 6 ]. This material has been widely used in cardiac- and vascular surgery and soft tissue augmentation with minimal complications [ 2 ]. Cashman et al published a histological study of changes in the rabbit larynx in response to Gore-Tex implantation. They found evidence of a foreign body giant cell reaction and a thin rim of fibrous tissue surrounding the implant. Although the fibrous capsule invaginated between the folds of the implant, it did not appear to grow into the implant itself. The implant was secure in the soft tissue with no migration or spontaneous extrusion noted [ 2 ]. Case presentation A 64-year-old female patient underwent transcervical excision of a right vagal neurofibroma. This was complicated by right vocal fold paralysis with dysphonia and aspiration. Six months later, the patient underwent fat injection for medialization of the right vocal fold. After an initial improvement, the patient suffered a recurrence of the aspiration, chest infections and dysphonia. The patient subsequently underwent right medialization using Gore-Tex implant (Gore Thyroplasty Device, Medtronics, Xomed ® ) with cricothyroid approximation. Videofluroscopy and Speech assessment confirmed an improvement in aspiration and speech. Six months later, her symptoms recurred with subsequent aspiration pneumonia. A decision was made after lengthy counseling to proceed with a laryngectomy 13 months following the implantation of Gore-Tex. Figure 1 shows the Gore-Tex in the laryngectomy specimen. Histopathology Light microscopy revealed a mild foreign-body giant cell granulomatous reaction with some associated fibrosis (Figures 2 and 3 ). The granulomatous response was limited to the periphery of the Gore-Tex and although it closely followed the profile of the material it did not encroach into or significantly break up the material. However, some fibrosis penetrated a short way into the Gore-Tex in a few areas. The giant cells were present mainly in a multinucleated form. Figure 4 shows the Gore-Tex strip under crossed Polaroids showing birefringence of the implanted material. Birefringent fragments were only very occasionally seen within isolated giant cells. The giant cell response was mainly seen in areas where the end of a Gore-Tex strip formed a sharp corner. In one area a very small focus of metaplastic bone formation was seen. There was no significant lymphocytic, neutrophilic inflammation or vascular ingrowth. Discussion Gore-Tex has been used as an effective implant for medialization laryngoplasty in the management of paralytic dysphonia and to a lesser extent aspiration [ 6 , 7 ]. Animal studies involving rabbits and a porcine model have demonstrated host tolerance of the implant, with no evidence of granuloma formation, resorption, extrusion or migration of the implant [ 2 , 8 ]. However, adverse responses to Gore-Tex in humans requiring implant removal from the lips, face, nose and larynx have been reported [ 9 - 12 ]. To date, there have been no reports describing the histological changes in a human laryngectomy specimen with a Gore-Tex implant. Conclusions Our findings are consistent with the animal models confirming that Gore-Tex implantation does not result in a significant granulomatous reaction in the human larynx over a 13-month period. Moreover, there is no evidence of resorption or infection. Further, the lack of lymphocytes in association with the granulomas indicates that there is no significant immunological hypersensitivity. Histologically, the slight permeation by connective tissue is similar to that seen in Gore-Tex vascular and cardiac implants. The degree of the slight giant cell response appears to be dependent on the profile of the material; a sharp edge incited more of a response than a flat surface. Competing interests The author(s) declare that they have no competing interests. Authors' contributions All authors equally contributed towards the background research and writing of the paper. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC547912.xml |
516022 | Immunopathogenesis of brain abscess | Brain abscess represents a significant medical problem despite recent advances made in detection and therapy. Due to the emergence of multi-drug resistant strains and the ubiquitous nature of bacteria, the occurrence of brain abscess is likely to persist. Our laboratory has developed a mouse experimental brain abscess model allowing for the identification of key mediators in the CNS anti-bacterial immune response through the use of cytokine and chemokine knockout mice. Studies of primary microglia and astrocytes from neonatal mice have revealed that S. aureus , one of the main etiologic agents of brain abscess in humans, is a potent stimulus for proinflammatory mediator production. Recent evidence from our laboratory indicates that Toll-like receptor 2 plays a pivotal role in the recognition of S. aureus and its cell wall product peptidoglycan by glia, although other receptors also participate in the recognition event. This review will summarize the consequences of S. aureus on CNS glial activation and the resultant neuroinflammatory response in the experimental brain abscess model. | Pathogenesis of brain abscess Brain abscesses develop in response to a parenchymal infection with pyogenic bacteria, beginning as a localized area of cerebritis and evolving into a suppurative lesion surrounded by a well-vascularized fibrotic capsule. The leading etiologic agents of brain abscess are the streptococcal strains and S. aureus , although a myriad of other organisms have also been reported [ 1 , 2 ]. Brain abscess represents a significant medical problem, accounting for one in every 10,000 hospital admissions in the United States, and remains a serious situation despite recent advances made in detection and therapy [ 2 ]. In addition, the emergence of multi-drug resistant strains of bacteria has become a confounding factor. Following infection, the potential sequelae of brain abscess include the replacement of the abscessed area with a fibrotic scar, loss of brain tissue by surgical excision, or abscess rupture and death. Indeed, if not detected early, an abscess has the potential to rupture into the ventricular space, a serious complication with an 80% mortality rate [ 1 ]. The most common sources of brain abscess are direct or indirect cranial infection arising from the paranasal sinuses, middle ear, and teeth. Other routes include seeding of the brain from distant sites of infection in the body (i.e. endocarditis) or penetrating trauma to the head. Following brain abscess resolution patients may experience long-term complications including seizures, loss of mental acuity, and focal neurological defects that are lesion site-dependent. At the histological level, brain abscess is typified by a sequential series of pathological changes that have been elucidated using the experimental rodent models described in detail below [ 3 - 7 ]. Staging of brain abscess in humans has been based on findings obtained during CT or MRI scans. The early stage or early cerebritis occurs from days 1–3 and is typified by neutrophil accumulation, tissue necrosis, and edema. Microglial and astrocyte activation is also evident at this stage and persists throughout abscess development. The intermediate, or late cerebritis stage, occurs from days 4–9 and is associated with a predominant macrophage and lymphocyte infiltrate. The final or capsule stage occurs from days 10 onward and is associated with the formation of a well-vascularized abscess wall, in effect sequestering the lesion and protecting the surrounding normal brain parenchyma from additional damage. In addition to limiting the extent of infection, the immune response that is an essential part of abscess formation also destroys surrounding normal brain tissue. This is supported by findings in experimental models where lesion sites are greatly exaggerated compared to the localized nature of bacterial growth, reminiscent of an over-active immune response [ 5 , 8 , 9 ]. This phenomenon is also observed in human brain abscess, where lesions can encompass a large portion of brain tissue, often spreading well beyond the initial focus of infection. Therefore, controlling the intensity and/or duration of the anti-bacterial immune response in the brain may allow for effective elimination of bacteria while minimizing damage to surrounding brain tissue. The mechanisms elucidated to date in the immunopathogenesis of brain abscess are depicted in Figure 1 . Figure 1 Immunopathogenesis of brain abscess. Pyogenic bacteria such as S. aureus induce a localized suppurative lesion typified by direct damage to CNS parenchyma and subsequent tissue necrosis. Bacterial recognition by Toll-like receptor 2 (TLR2; Y ) leads to the activation of resident astrocytes and the elaboration of numerous proinflammatory cytokines and chemokines. Microglia produce a similar array of proinflammatory mediators following bacterial stimulation; however, the receptor(s) responsible for S. aureus recognition and subsequent cell activation remain to be identified. Both microglia and astrocytes utilize TLR2 to recognize peptidoglycan (PGN) from the bacterial cell wall. Proinflammatory cytokine release leads to blood-brain barrier (BBB) compromise and the entry of macromolecules such as albumin and IgG into the CNS parenchyma. In addition, cytokines induce the expression of adhesion molecules (ICAM, intercellular adhesion molecule; VCAM, vascular cell adhesion molecule) which facilitate the extravasation of peripheral immune cells such as neutrophils, macrophages, and T cells into the evolving abscess. Newly recruited peripheral immune cells can be activated by both bacteria and cytokines released by activated glia, effectively perpetuating the anti-bacterial immune response that is thought to contribute, in part, to disease pathogenesis. S. aureus -induced experimental brain abscess model Although case reports of brain abscess in humans are relatively numerous, studies describing the nature of the ensuing CNS and peripheral immune responses are rare. Therefore, our laboratory has developed a mouse experimental brain abscess model to elucidate the importance of host immune factors in disease pathogenesis [ 5 , 7 - 9 ]. Our mouse model was modified based on a previously published model in the rat [ 3 ] and utilizes S. aureus , one of the main etiologic agents of brain abscess in humans. The mouse brain abscess model accurately reflects the course of disease progression in humans, providing an excellent model system to study immunological pathways influencing abscess pathogenesis and the effects of therapeutic agents on disease outcome. We have successfully utilized this model to characterize inflammatory mediators induced in the brain immediately following S. aureus exposure [ 5 ] as well as identification of bacterial virulence factors critical for pathogenesis in vivo [ 8 ]. For example, we have demonstrated that S. aureus leads to the immediate and sustained expression of numerous proinflammatory cytokines and chemokines in the brain including tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), IL-1α,β, macrophage inflammatory protein-2 (MIP-2/CXCL2), monocyte chemoattractant protein-1 (MCP-1/CCL2), MIP-1α/CCL3, MIP-1β/CCL4, and regulated upon activation T cell expressed and secreted (RANTES/CCL5) [ 5 , 7 - 9 ]. As mentioned earlier, lesion sites in both our experimental model and in human brain abscess are greatly exaggerated compared to the localized nature of bacterial growth, reminiscent of an over-active immune response. To account for the enlarged region of affected tissue involvement associated with brain abscesses compared to the relatively focal nature of the initial insult, we have proposed that proinflammatory mediator production following S. aureus infection persists, effectively augmenting damage to surrounding normal brain parenchyma [ 10 ]. Specifically, the continued release of proinflammatory mediators by activated glia and infiltrating peripheral immune cells may act through a positive feedback loop to potentiate the subsequent recruitment and activation of newly recruited inflammatory cells and glia. This would effectively perpetuate the anti-bacterial inflammatory response via a vicious pathological circle culminating in extensive collateral damage to normal brain tissue. Recent studies support persistent immune activation associated with experimental brain abscesses with elevated levels of IL-1β, TNF-α, and MIP-2 detected from 14 to 21 days following S. aureus exposure [ 9 ]. Concomitant with prolonged proinflammatory mediator expression, S. aureus infection was found to induce a chronic disruption of the blood-brain barrier, which correlated with the continued presence of peripheral immune cell infiltrates and glial activation [ 9 ]. Collectively, these findings suggest that intervention with anti-inflammatory compounds subsequent to sufficient bacterial neutralization may be an effective strategy to minimize damage to surrounding brain parenchyma during the course of brain abscess development, leading to improvements in cognition and neurological outcomes. Besides the potential detrimental roles cytokines may exert on surrounding normal brain parenchyma during the later stages of brain abscess, numerous proinflammatory cytokines such as IL-1β, TNF-α, and IL-6 may have beneficial effects on the establishment of host anti-bacterial immune responses. These cytokines exert numerous functions within CNS tissues including modulation of blood-brain barrier integrity, induction of adhesion molecule expression on cerebral microvascular endothelial cells, and subsequent activation of resident glia and infiltrating peripheral immune cells [ 11 - 17 ]. We recently examined the relative importance of IL-1, TNF-α and IL-6 in experimental brain abscess using cytokine knockout (KO) mice [ 7 ]. The IL-1 KO animals used for these studies were deficient in both IL-1α and IL-1β; therefore, potential caveats arising from redundancy in the activities of these two proteins were avoided. Despite the fact that these cytokines share many overlapping functional activities, IL-1 and TNF-α appear to play an important role in dictating the ensuing anti-bacterial response in brain abscess. This was evident by the finding that bacterial burdens were significantly higher in both IL-1 and TNF-α KOs compared to wild type mice which correlated with enhanced mortality rates in both KO strains [ 7 ]. In contrast, IL-6 was not found to be a major contributor to the host anti-bacterial immune response. These studies established important roles for IL-1 and TNF-α during the acute phase of experimental brain abscess development, indicating that these cytokines individually dictate essential functions for the establishment of an effective anti-bacterial response in the CNS parenchyma. Neutrophils are potent bactericidal effector cells and represent the major peripheral cell infiltrate associated with developing brain abscesses [ 5 , 9 ]. Neutrophils exert their bactericidal activity through the production of reactive oxygen and nitrogen intermediates and hydrolytic enzymes that directly destroy bacteria. In addition, neutrophils serve as a source of proinflammatory cytokines, such as TNF-α that serve to amplify the host anti-bacterial immune response [ 18 , 19 ]. However, the continuous release of these products by newly recruited and activated neutrophils can also contribute to tissue damage. Therefore, depending on the context of inflammation, neutrophils can have either beneficial or detrimental effects on the course of infectious diseases. We have recently revealed the functional importance of neutrophils in brain abscess development using antibody-mediated neutrophil depletion and CXCR2 KO mice where neutrophils lack the high-affinity receptor for the neutrophil chemoattractants MIP-2/CXCL2 and KC/CXCL2 [ 5 ]. Interestingly, in spite of elevated levels of the CXCR2 ligands MIP-2 and KC, neutrophil extravasation was impaired in CXCR2 KO mice, with cells remaining sequestered within small vessels in developing brain abscesses. Impaired neutrophil influx into evolving brain abscesses in both CXCR2 KO and neutrophil-depleted mice led to exacerbated disease typified by elevated bacterial burdens compared to wild type animals [ 5 ]. These studies demonstrate that CXCR2 ligands are the major chemotactic signals required for neutrophil influx into brain abscesses and that their activity cannot be substituted by alternative chemotactic factors such as complement split products (i.e. C3a, C5a), prostaglandins, leukotrienes, or other chemokines. Similar to our findings, the importance of neutrophils in S. aureus -induced acute cerebritis was demonstrated by Lo et al. where transient neutrophil depletion resulted in enhanced pathology [ 20 ]. In addition to MIP-2 and KC, numerous other chemokines are also detected within evolving brain abscesses including MIP-1α, MIP-1β, MCP-1, and TCA-3/CCL1 [ 5 , 8 ]. The potential roles these chemokines play in the pathogenesis of brain abscess development remain to be defined. However, they could be envisioned to influence the accumulation of monocytes and lymphocytes into the brain and possibly the establishment of adaptive immune responses. Indeed, we and others have demonstrated the influx [ 21 ](Kielian, unpublished observations) and generation of S. aureus -specific lymphocytes [ 9 ] in experimental brain abscess. Staphylococci produce a wide array of virulence determinants that play a role in disease pathogenesis [ 22 , 23 ]. These can be broadly subdivided into surface and extracellular secreted proteins. Surface proteins include structural components of the bacterial cell wall such as lipoteichoic acid and peptidoglycan. Secreted proteins are generally expressed during the exponential phase of bacterial growth and include such proteins as α-toxin, lipase, and enterotoxin. We recently reported that virulence factor production by S. aureus is essential for the establishment of brain abscess in the experimental mouse model [ 8 ]. Specifically, a requirement for ongoing bacterial replication and/or virulence factor production was supported by the finding that heat-inactivated bacteria were not sufficient to induce proinflammatory cytokine/chemokine expression or abscess formation in the brain. Using a series of S. aureus mutants with various defects in virulence factor expression, we identified α-toxin as a critical virulence factor determinant in the experimental brain abscess model. Replication of a S. aureus α-toxin mutant was significantly attenuated in the brain, which correlated with a reduction in proinflammatory mediator expression and the failure to establish a well-defined abscess [ 8 ]. We proposed that in wild type bacteria, α-toxin, which leads to pore formation in mammalian cell membranes and subsequent osmotic lysis, serves as an effective mechanism to eliminate CNS resident immunocompetent cells (i.e. microglia and astrocytes) as well as professional phagocytes that infiltrate brain abscesses and exert potent anti-bacterial activity (i.e. neutrophils and macrophages). This would effectively impair the efficacy of the ensuing anti-bacterial immune response, allowing bacterial burdens to expand unchecked during the acute phase of disease. In contrast, in the absence of α-toxin secretion, resident glia and infiltrating leukocytes would be capable of rapidly neutralizing bacteria, effectively facilitating the resolution of infection in a timely manner and thus preventing the establishment of a well-formed abscess. However, it is likely that additional virulence factors participate in S. aureus infection in the brain since the α-toxin mutant was not completely avirulent. Potential candidates include V8 protease, staphylococcal enterotoxin B, and protein A, the latter of which has been shown to bind to TNF receptor I in the host [ 24 ]. Recently, the S. aureus -induced experimental brain abscess model has been utilized by Stenzel et al. to demonstrate an important role for astrocytes in dictating the extent of brain abscess pathology [ 21 ]. Using glial fibrillary acidic protein (GFAP) KO mice, this group showed that brain abscess pathogenesis was exacerbated in KO animals where lesions were larger and typified by ill-defined borders, severe brain edema, and enhanced levels of vasculitis compared to wild type mice. In addition, GFAP KO mice exhibited a diffuse leukocyte infiltrate that extended into the uninfected contralateral hemisphere. Exacerbation of brain abscess severity in GFAP KO mice was attributed to the absence of a bordering function by astrocytes to contain the infection since strong GFAP immunoreactivity was observed along the abscess margins in wild type animals. It is intriguing that the absence of GFAP influences brain abscess evolution in such a dramatic manner, as astrocytes are still present and functional in these mice. It is possible that GFAP expression in activated astrocytes induces structural changes that influence the local cytoarchitecture leading to bacterial dissemination in brain abscess. Collectively, the studies to date performed in the mouse experimental brain abscess model have begun to elucidate critical mediators in the pathogenesis of disease and host cytokines that play a pivotal role in the generation of the CNS anti-bacterial immune response. However, there are numerous issues that remain to be resolved regarding the role of inflammatory mediators in the evolution of brain abscess. For example, the potential importance of other proinflammatory cytokines and chemokines detected in brain abscess remain to be defined. In addition, factor(s) that participate in the initiation of the anti-bacterial adaptive immune response remain to be elucidated. Evidence to support the establishment of an adaptive immune response is provided by our recent findings that S. aureus -specific lymphocytes are formed during the later stages of experimental brain abscess development [ 9 ]. It is not known whether the immune response generated during a previous brain abscess episode is capable of providing protection against a second CNS challenge. Another question relates to the potential dual role of various proinflammatory mediators during the course of brain abscess pathogenesis. As mentioned above, a dual role for IL-1 and TNF-α has been suggested by our findings that these cytokines are critical for establishing an effective host anti-bacterial immune response during the acute stage of brain abscess development. However, IL-1 and TNF-α expression persists within brain abscesses for at least 14 to 21 days following infection, suggesting an over-active immune response that is not down-regulated in a timely manner. We are currently using knockout mice to investigate the potential dual role these cytokines may exert during the evolution of brain abscess. Addressing these issues may facilitate the design of effective therapeutic regimens for brain abscess that would be capable of pathogen elimination without the accompanying destruction of surrounding brain parenchyma that normally occurs in disease. Responses of microglia to the brain abscess pathogen S. aureus Relevant to our experimental brain abscess model, recent studies from our laboratory have established that both intact S. aureus and its cell wall product peptidoglycan (PGN) serve as potent stimuli for proinflammatory mediator production in primary microglia [ 5 , 10 , 25 ]. Specifically, exposure to both stimuli led to a dose- and time-dependent induction of the proinflammatory cytokines IL-1β, TNF-α, IL-12 p40, and several chemokines including MIP-2, MCP-1, MIP-1α, and MIP-1β. The importance of microglia in the early host response to infection in brain abscess is suggested by the fact that proinflammatory mediator production is detected within 1 to 3 hours following the initial S. aureus infection, well before the significant accumulation of peripheral immune cell infiltrates [ 4 ]. Another study has also demonstrated that S. aureus induces IL-1β expression in neonatal rat microglia [ 26 ]. Microglia represent one of the main antigen presenting cells in the CNS [ 11 , 27 ]. To achieve efficient activation of antigen-specific T cells, microglia must express sufficient levels of major histocompatability complex (MHC) class II (signal I) and co-stimulatory molecules such as CD40, CD80, and CD86 (signal II). Recognition of signal I without the concomitant engagement of signal II results in T cell non-responsiveness or anergy. Our group found that both heat-inactivated S. aureus and PGN are capable of inducing microglial MHC class II [ 10 , 25 ], CD40, CD80, and CD86 receptor expression, similar to what has been described for microglia in response to the gram-negative bacterial product lipopolysaccharide (LPS) and interferon-γ (IFN-γ) [ 27 - 31 ]. The ability of S. aureus to augment the expression of receptors that are important for antigen presentation suggests that the ability of microglia to present bacterial peptides to antigen-specific T cells may be greatly enhanced following an initial exposure to S. aureus . The effects of S. aureus and PGN on microglial CD40, CD80, CD86, and MHC class II expression may either be a direct consequence of bacterial stimulation or indirect via the autocrine action of cytokines produced by activated microglia. Microglial activation is a hallmark of brain abscess [ 4 , 5 , 9 ]. They respond robustly to both S. aureus and PGN with significant proinflammatory mediator expression, and many of these same mediators are persistently elevated in brain abscess. Drawing on this relationship, we have proposed that chronic microglial activation may contribute, in part, to the excessive tissue damage characteristic of brain abscess. Therefore, attenuating chronic microglial activation subsequent to effective bacterial elimination in the brain may result in attenuation of the structural and functional damage associated with brain abscess. We have recently examined the efficacy of the cyclopentenone prostaglandin 15d-PGJ 2 to modulate microglial responses to S. aureus [ 10 ]. 15d-PGJ 2 was found to be a selective and potent inhibitor of S. aureus -dependent microglial activation through its ability to significantly attenuate the expression of numerous proinflammatory cytokines and chemokines of the CC family including IL-1β, TNF-α, IL-12 p40, MCP-1, and MIP-1β. In addition, 15d-PGJ 2 also selectively inhibited the S. aureus -dependent increase in microglial TLR2, CD14, MHC class II, and CD40 expression whereas it had no effect on the co-stimulatory molecules CD80 and CD86. The ability of 15d-PGJ 2 to modulate the expression of these receptors may serve as a means to regulate microglial and T cell activation during gram-positive bacterial infections in the CNS. Preventing microglial activation by 15d-PGJ 2 or related compounds may help to resolve inflammation earlier, resulting in reductions in brain abscess size and associated damage to surrounding normal brain parenchyma. Receptors utilized by microglia for bacterial recognition As detailed above, our laboratory has established that microglia are capable of recognizing S. aureus and respond with robust production of numerous proinflammatory mediators. However, to date, the receptor repertoire responsible for bacterial recognition remains to be defined. In macrophages, numerous receptors have been implicated in bacterial phagocytosis and subsequent activation leading to proinflammatory mediator release including Toll-like receptors (TLR), scavenger receptors, and mannose receptors. The fact that microglia and macrophages share many functional and phenotypical characteristics supports the contention that these receptors may play an important role in microglial responses to bacteria. Toll-like receptors are a family of surface receptors expressed on cells of the innate immune system that allow for the recognition of conserved structural motifs on a wide array of pathogens (referred to as pathogen-associated molecular patterns) [ 32 , 33 ]. To date, eleven TLR have been identified, with TLR2 playing a pivotal role in recognizing structural components of various gram-positive bacteria, fungi, and protozoa [ 34 ]. Several groups have reported TLR2 expression in microglia, with receptor expression augmented following inflammatory activation [ 25 , 35 - 38 ]. Relevant to brain abscess, we have demonstrated that both S. aureus and PGN lead to significant increases in TLR2 mRNA and protein expression, which may enhance microglial sensitivity to bacteria during the course of experimental brain abscess development [ 25 ]. Recent studies from our laboratory using primary microglia from TLR2 KO mice have revealed that TLR2 plays a pivotal role in recognition of PGN but not intact S. aureus (Kielian, manuscript in preparation). These findings indicate that an alternative receptor(s) is involved in mediating responses to intact bacteria. Candidates include the mannose receptor and members of the scavenger receptor family. Scavenger receptors encompass a broad range of molecules involved in receptor-mediated phagocytosis of select polyanionic acids such as lipoteichoic acid of S. aureus [ 39 ]. Although adult microglia do not express scavenger receptors in the normal CNS, their expression is induced following inflammation or injury [ 40 ]. In the context of brain abscess, a potential tripartite role for microglial scavenger receptors can be envisioned that would include regulating cell adhesion and retention within the inflammatory milieu, facilitating bacterial phagocytosis, and promoting the removal of apoptotic cell debris associated with the evolving abscess [ 41 ]. Preliminary data suggest that S. aureus and PGN differentially modulate the expression of several distinct scavenger receptors that may influence the nature and extent of phagocytosis (Kielian, unpublished observations). Scavenger receptors have been implicated in β-amyloid phagocytosis by microglia in the context of Alzheimer's disease, in part, by the finding that microglia associated with senile plaques express a high degree of scavenger receptor immunoreactivity [ 42 , 43 ]. In addition, scavenger receptors have been implicated in β-amyloid uptake by microglia [ 44 - 47 ]. The functional importance of scavenger receptors in S. aureus phagocytosis by microglia remains to be established. Microglia have been shown to express functional mannose receptors that are responsible for the binding and phagocytosis of mannosylated and fucosylated ligands of bacteria [ 48 , 49 ]. Interestingly, proinflammatory cytokines such as IFN-γ and LPS have been shown to downregulate mannose receptor expression on microglia [ 48 , 49 ]. Using microarray analysis, we also recently demonstrated that mannose receptor levels were significantly attenuated in microglia following S. aureus exposure, suggesting that the regulation of mannose receptor expression is conserved among diverse stimuli [ 25 ]. Following the subsequent internalization of molecules via the mannose receptor by antigen presenting cells, an immune response can be generated in either a MHC class I, class II, or CD1-restricted manner [ 50 - 52 ]. In addition, some studies have indicated a functional coupling of the mannose receptor to microbiocidal activities, strongly suggesting a cytotoxic activity linked to mannose receptor-ligand interactions [ 53 ]. The functional importance of mannose receptors in the initial recognition and phagocytic events in microglia following S. aureus exposure remain to be defined. In addition to the receptors described above, there are additional candidates that may serve as receptors for S. aureus phagocytosis in microglia including complement receptor 3 (also known as CD11b/CD18) and CD14, the latter of which we have shown to be expressed on microglia and significantly upregulated following activation with either S. aureus or PGN [ 10 , 25 ]. Responses of astrocytes to the brain abscess pathogen S. aureus Astrocytes play a pivotal role in the type and extent of CNS inflammatory responses. These cells likely play an important role in the initial recruitment and activation of peripheral immune cells into the CNS during neuroinflammation through the production of several cytokines and chemokines, such as IL-1, IL-6, IL-10, TNF-α, IFN-α/β, granulocyte-macrophage colony-stimulating factor (GM-CSF), macrophage-CSF (M-CSF), granulocyte-CSF (G-CSF), transforming growth factor-beta (TGF-β), RANTES, MCP-1, and IFN-γ-inducible protein-10 (IP-10/CXCL10) [ 12 , 54 ]. Various studies have documented the ability of LPS to induce nitric oxide (NO), cytokine, and chemokine production in astrocytes [ 55 , 56 ]. In contrast, the characterization of products produced by astrocytes following exposure to gram-positive bacteria had remained largely undefined until recently. Studies from our group have revealed that primary astrocytes are capable of recognizing both intact S. aureus and PGN and that they respond with vigorous proinflammatory cytokine and chemokine production [ 57 ]. Among the factors produced by S. aureus -activated astrocytes are NO, TNF-α, IL-1β, MIP-2, MCP-1, MIP-1α, and MIP-1β. These proinflammatory chemokines may serve as signals for neutrophil (MIP-2), monocyte and lymphocyte (MCP-1, MIP-1β) recruitment in vivo , whereas IL-1β and TNF-α likely alter blood-brain barrier permeability and induce the expression of critical adhesion molecules on CNS vascular endothelium required for immune cell extravasation into brain abscesses. Receptors utilized by astrocytes for bacterial recognition Astrocytes have recently been shown to express TLR2 [ 38 , 58 ], and although these cells are capable of responding to the well-characterized TLR2 ligand PGN [ 58 ], the functional significance of this receptor was not directly demonstrated until recently. Using primary astrocytes from TLR2 KO and wild type mice, our laboratory was the first to report that TLR2 plays a pivotal role in the recognition of S. aureus and PGN and in subsequent cytokine and chemokine expression by astrocytes [ 57 ]. Interestingly, the production of these cytokines and chemokines was only partially attenuated in TLR2 KO astrocytes, suggesting that alternative receptors are also involved in bacterial recognition. There are numerous candidates for alternative receptors in astrocytes for gram-positive pathogens like S. aureus . For example, TLR2 has been shown to form functional heterodimers with TLR1 and/or TLR6 [ 59 , 60 ], thereby increasing its range of antigen detection. It has recently been suggested that CD14 serves as a co-receptor for TLR2 [ 61 ] and enhances the recognition efficiency of many TLR2-specific ligands including PGN and lipoteichoic acid [ 62 - 64 ]. Recently, several studies have reported data that support the involvement of additional, as of yet uncharacterized pattern recognition receptors in bacterial recognition [ 61 , 65 ]. Alternatively, activation through mannose and scavenger receptors that play an important role in the phagocytic uptake of bacteria and have been reported to be expressed by astrocytes [ 66 - 68 ] may be responsible for the residual proinflammatory mediator expression in TLR2 KO astrocytes. However, to date, the functional importance of these alternative receptors in mediating astrocyte activation in response to S. aureus and PGN is currently not known. Although astrocytes have been shown to possess phagocytic activity in response to β-amyloid [ 69 ], apoptotic cells [ 70 ], and yeast [ 71 , 72 ], the phagocytic potential of astrocytes is still a subject of controversy. Data from our laboratory indicates that primary astrocytes are capable of phagocytosing S. aureus [ 57 ]. An active phagocytic process is supported by the finding that astrocytes rapidly internalize heat-killed S. aureus , indicating that bacterial uptake occurs via a phagocytic pathway and is not simply the result of productive infection by live organisms. Interestingly, TLR2 is not a major receptor for bacterial phagocytosis in astrocytes since both TLR2 KO and wild type astrocytes were equally capable of phagocytosing intact S. aureus organisms in vitro [ 57 ]. The receptor(s) responsible for mediating bacterial uptake in astrocytes are not known but could include the mannose and/or scavenger receptors described above. Studies to identify receptors responsible for S. aureus phagocytosis by astrocytes and the optimal conditions required for bacterial uptake are currently ongoing in our laboratory. Issues such as whether bacterial internalization is serum-dependent or requires other bacterial binding proteins must also be addressed. Conclusions and perspectives The incidence of brain abscess is expected to persist in the human population due to the ubiquitous nature of bacteria coupled with the recent emergence of antibiotic-resistant bacterial strains. Therefore, understanding the roles of both host anti-bacterial immune responses along with bacterial virulence factors may lead to the establishment of novel therapeutic treatments for brain abscess. The mouse S. aureus experimental brain abscess model provides an excellent tool for deciphering the importance of various mediators in disease pathogenesis. Especially appealing is the ability to examine the role of specific factors using transgenic and knockout mice because, in our experience, all of the mouse strains examined with this model have qualitatively similar inflammatory profiles following bacterial challenge. In addition, the consequences of S. aureus infection do not appear to be influenced by gender, as the responses of female and male mice are similar- another advantage when performing studies with knockout or transgenic mice where animal numbers are often limiting. The responses of microglia and astrocytes to S. aureus have been elucidated in terms of proinflammatory mediator expression and in general, have been found to be qualitatively similar to those observed following LPS exposure. Although studies with primary microglia and astrocytes from TLR2 KO mice reveal an important role for this receptor in mediating S. aureus -dependent activation, it is clear that additional receptors are also involved in glial responses to this bacterium. This functional redundancy is not surprising because these pathogens have the potential for devastating consequences in a tissue that has limited regenerative capacity such as the CNS. The implications of glial cell activation in the context of brain abscess are likely several-fold. First, parenchymal microglia and astrocytes may be involved in the initial recruitment of professional bactericidal phagocytes into the CNS through their elaboration of chemokines and proinflammatory cytokines. Second, microglia exhibit S. aureus bactericidal activity in vitro , suggesting that they may also participate in the initial containment of bacterial replication in the CNS. However, their bactericidal activity in vitro is not comparable to that of neutrophils or macrophages, suggesting that this activity may not be a major effector mechanism for microglia during acute infection. Third, activated microglia have the potential to influence the type and extent of anti-bacterial adaptive immune responses through their upregulation of MHC class II and co-stimulatory molecule expression. Finally, if glial activation persists in the context of ongoing inflammation, the continued release of proinflammatory mediators could damage surrounding normal brain parenchyma. Indeed, inappropriate glial activation has been implicated in several CNS diseases including multiple sclerosis and its animal model experimental autoimmune encephalomyelitis as well as Alzheimer's disease. The continued use of transgenic and knockout mice for in vivo studies will facilitate our understanding of immune mechanisms contributing to brain abscess pathogenesis. List of abbreviations BBB blood-brain barrier CCL CC chemokine ligand CD cluster of differentiation CSF cerebral spinal fluid CXCL CXC chemokine ligand CXCR CXC chemokine receptor GFAP glial fibrillary acidic protein GM-CSF granulocyte-macrophage colony-stimulating factor IFN interferon IL interleukin IP-10 interferon-inducible protein-10 KO knockout LPS lipopolysaccharide M-CSF macrophage colony-stimulating factor MCP monocyte chemoattractant protein MHC major histocompatability complex MIP macrophage inflammatory protein NO nitric oxide PGN peptidoglycan RANTES regulated upon activation T cell expressed and secreted TGF transforming growth factor TNF tumor necrosis factor Competing interests None declared. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC516022.xml |
522822 | Effect of time of administration on cholesterol-lowering by psyllium: a randomized cross-over study in normocholesterolemic or slightly hypercholesterolemic subjects | Background Reports of the use of psyllium, largely in hypercholesterolemic men, have suggested that it lowers serum cholesterol as a result of the binding of bile acids in the intestinal lumen. Widespread advertisements have claimed an association between the use of soluble fibre from psyllium seed husk and a reduced risk of coronary heart disease. Given the purported mechanism of cholesterol-lowering by psyllium, we hypothesized that there would be a greater effect when psyllium is taken with breakfast than when taken at bedtime. Secondarily, we expected to confirm a cholesterol-lowering effect of psyllium in subjects with "average" cholesterol levels. Methods Sixteen men and 47 women ranging in age from 18 to 77 years [mean 53 +/- 13] with LDL cholesterol levels that were normal or slightly elevated but acceptable for subjects at low risk of coronary artery disease were recruited from general gastroenterology and low risk lipid clinics. Following a one month dietary stabilization period, they received an average daily dose of 12.7 g of psyllium hydrophilic mucilloid, in randomized order, for 8 weeks in the morning and 8 weeks in the evening. Change from baseline was determined for serum total cholesterol, LDL, HDL and triglycerides. Results Total cholesterol for the "AM first" group at baseline, 8 and 16 weeks was 5.76, 5.77 and 5.80 mmol/L and for the "PM first" group the corresponding values were 5.47, 5.61 and 5.57 mmol/L. No effect on any lipid parameter was demonstrated for the group as a whole or in any sub-group analysis. Conclusion The timing of psyllium administration had no effect on cholesterol-lowering and, in fact, no cholesterol-lowering was observed. Conclusions regarding the effectiveness of psyllium for the prevention of heart disease in the population at large may be premature. | Background A cholesterol lowering effect has been reported for a variety of soluble dietary fibres [ 1 - 5 ]. In February, 1998, the U.S. Food and Drug Administration authorized the use, on food labels and food labelling, of health claims on the association between soluble fibre from psyllium seed husk and a reduced risk of coronary heart disease [ 6 ]. Among the suggested mechanisms by which soluble fibre lowers cholesterol is the binding of bile acids in the intestinal lumen resulting in decreased absorption and increased faecal excretion of them [ 7 - 15 ]. The ensuing bile acid depletion increases hepatic demand for the de novo synthesis of bile acids from cholesterol. This requirement is met, in part, by increased hepatic LDL receptor activity, which in turn reduces circulating LDL. The accumulation and concentration of bile in the gallbladder is a continuous process. In rats which lack a gallbladder, the biliary excretion rate of bile salts is maximal at night [ 16 ] and bile is stored in the gallbladder during an overnight fast. To the extent that the cholesterol-lowering effect of psyllium requires an interaction with bile, the magnitude of its cholesterol-lowering effect should vary with the quantity of bile in contact with a given amount of psyllium. The gallbladder empties with a meal and the total quantity of bile salt presented to the small bowel should be highest when this emptying occurs at breakfast, following a fast and the associated overnight accumulation of secreted bile. Conversely, psyllium presented to the gut following a short fast and without the full stimulus to gallbladder emptying of an associated meal should result in an encounter with a smaller amount of bile. We accordingly set out to test the hypothesis that administering psyllium with breakfast would have a significantly greater cholesterol-lowering effect than would taking a similar dose at bedtime. Since the predominance of the literature with regard to the cholesterol-lowering effect of psyllium is in individuals with hypercholesterolemia, a secondary goal of our study was to confirm the cholesterol-lowering effect of psyllium in subjects with "average" cholesterol levels. Methods Patients identified in gastroenterology practices as requiring long-term treatment with psyllium, typically for chronic constipation or the irritable bowel syndrome, were invited to participate in the study. In addition, individuals who had received dietary counselling in a lipid clinic regarding cholesterol-lowering and who subsequently had cholesterol levels deemed not to require further intervention because they met targets set out in clinical practice guidelines were invited by clinic staff to participate. Subjects were deemed ineligible if they were under age 18 years, were under active treatment for hyperlipidemia, had total cholesterol greater than 7.00 mmol/L, required alterations in dosage of medications which might have an effect on lipid levels, had had a gastrectomy, had any disease which is associated with hyperlipidemia, were receiving a bile acid binding resin, or if they did not eat breakfast regularly. The study was approved by the Conjoint Health Research Ethics Board of the University of Calgary, Faculty of Medicine. Subjects were given a description of the study indicating our interest in comparing the relative efficacy of hs versus am dosing with psyllium without indicating the specific hypothesis, and were asked to sign a consent form. Gastroenterology patients were given a high fibre diet sheet as part of their therapeutic regimen and were asked to take this on a continuing basis, beginning one month prior to the initiation of the study. The diet sheet emphasized dietary sources containing predominantly insoluble fibre. Lipid clinic patients had all received in-depth counselling regarding dietary measures for hypercholesterolemia, including a high fibre regimen, and had implemented their dietary changes at least one month before beginning the study. An unsweetened psyllium preparation, "Novo-Mucilax" [NovoPharm], providing three grams of hydrophilic mucilloid per 6.2 gram powder, and a scoop known to provide at least ten grams of psyllium were provided. Containers were numbered and weighed at the conclusion of each test interval. Subjects were randomized to initially take a scoop full of psyllium either with breakfast or at bedtime. Using a crossover design, psyllium was taken in the morning or evening for eight weeks and at the alternate time for the subsequent eight weeks. Determinations of serum total cholesterol, LDL, HDL and triglycerides were made before beginning psyllium, at eight weeks and at sixteen weeks after commencing its use. A trained dietician obtained a dietary history and patients were weighed at the beginning and at the conclusion of the study. All lipid determinations were undertaken following a 14 hour fast and analyses were done in a central laboratory. After initial data inspection based on boxplots and summary measures, cholesterol values at 8 and 16 weeks were examined using analysis of variance, taking into account treatment, period, and between and within subject effects in accordance with the cross-over design. The pattern of change from baseline to 16 weeks was evaluated using paired t-tests. Results Of 86 subjects beginning the study, 33 of those referred from the gastroenterology clinics and 30 of those referred from the lipid clinic completed it. Of those withdrawing, eight did so because they could not tolerate the psyllium or it was felt to interfere with prescribed medication, 2 had elevated lipids, 4 did not complete all the required blood work, 5 were unable to comply with the protocol because of work or lifestyle changes and 4 developed intercurrent diseases which precluded completing the protocol. The age range of subjects was 18 to 77 years, mean 53 +/- 13 years, including 16 men and 47 women. The mean dose of psyllium taken was 12.7 +/ - 2.3 g for morning dosing and 12.7 +/- 2.2 g when taken in the evenings. Values for total, LDL, HDL cholesterol and triglycerides at baseline, eight and sixteen weeks for various subgroupings are tabulated in tables 1 , 2 , 3 and 4 , respectively. Data for changes from baseline for total, LDL and HDL cholesterol and triglycerides are listed in table 5 . Changes from baseline were not significant for any parameter. Furthermore, the confidence intervals given in Table 5 show that this data essentially excludes a clinically significant psyllium effect. Table 1 Total Cholesterol in mmol/L [SD] [n] Baseline 8 wks 16 wks AM first 5.76 [0.94] 5.77 [1.01] 5.80 [0.95] PM first 5.47 [0.98] 5.61 [1.10] 5.57 [1.06] Females [47] 5.72 [0.96] 5.79 [1.07] 5.74 [0.99] Males [16] 5.29 [0.94] 5.39 [0.96] 5.48 [1.04] GI subjects [33] 5.00 [0.83] 5.06 [0.91] 5.07 [0.84] Lipid clinic Subjects [30] 6.28 [0.59] 6.39 [0.69] 6.35 [0.70] All subjects [63] 5.61 [0.96] 5.69 [1.05] 5.68 [1.00] "AM first" are subjects taking psyllium in the morning for the first 8 weeks and in the evening for the second 8 weeks. Vice versa for "PM first". Table 2 LDL CHOLESTEROL in mmol/L [SD] Baseline 8 wks 16 wks AM first 3.55 [1.01] 3.51 [1.04] 3.52 [0.94] PM first 3.39 [0.90] 3.43 [1.03] 3.45 [0.92] Females 3.55 [0.92] 3.54 [1.04] 3.52 [0.92] Males 3.22 [1.03] 3.24 [0.99] 3.39 [0.97] GI subjects 2.89 [0.81] 2.85 [0.85] 2.91 [0.72] Lipid clinic subjects 4.09 [0.66] 4.15 [0.72] 4.11 [0.69] All subjects 3.47 [0.95] 3.47 [1.03] 3.48 [0.92] Table 3 HDL CHOLESTEROL in mmol/L [SD] Baseline 8 wks 16 wks AM first 1.38 [0.29] 1.42 [0.37] 1.42 [0.33] PM first 1.38 [0.40] 1.39 [0.36] 1.36 [0.33] Females 1.41 [0.30] 1.44 [0.30] 1.42 [0.29] Males 1.28 [0.45] 1.29 [0.50] 1.30 [0.43] GI subjects 1.36 [0.39] 1.41 [0.43] 1.37 [0.38] Lipid clinic subjects 1.40 [0.30] 1.40 [0.28] 1.41 [0.27] All subjects 1.38 [0.35] 1.41 [0.36] 1.39 [0.33] Table 4 TRIGLYCERIDES in mmol/L [SD] Baseline 8 wks 16 wks AM first 1.81 [0.70] 1.84 [0.86] 1.86 [0.91] PM first 1.53 [0.52] 1.69 [0.70] 1.65 [0.61] Females 1.66 [0.61] 1.72 [0.70] 1.76 [0.71 Males 1.70 [0.68] 1.89 [0.99] 1.72 [0.94] GI subjects 1.63 [0.54] 1.71 [0.82] 1.70 [0.72] Lipid clinic subjects 1.71 [0.71] 1.82 [0.74] 1.80 [0.83] All subjects 1.67 [0.35] 1.76 [0.78] 1.75 [0.77] Table 5 Mean changes from baseline at 8 and 16 weeks [mmol/L] and 95% confidence intervals [all patients, n = 63] 8 wks 16 wks TOTAL 0.080 p = 0.26 0.068 p = 0.28 CHOLESTEROL [-0.061,0.22] [-0.056,0.19] LDL 0.002 p = 0.98 0.018 p = 0.75 CHOLESTEROL [-0.13,0.14] [-0.098,0.13] HDL 0.028 p = 0.28 0.012 p = 0.61 CHOLESTEROL [-0.023,0.08] [-0.035,0.059] TRIGLYCERIDES 0.094 p = 0.22 0.082 p = 0.22 [-0.059,0.25] [-0.050,0.21] The mean caloric intake or the intakes of fat or fibre did not change significantly during the study [table 6 ]. There was a small but significant increase in the weights of subjects, which was felt to reflect the high proportion of subjects beginning the study in the fall and the associated reduction of physical activity during the subsequent winter months. There was no apparent relationship between change in cholesterol and change in weight [correlation = .1, p = .15]. Table 6 Weight and Dietary Intake [SD] at Baseline and at 16 Weeks Baseline 16 wks WEIGHT [Kg] 74.6 [16.5] 75.4 [16.9] p = 0.0039 ENERGY [kcal] 1649 [409] 1689 [419] p = 0.29 FAT [g] 49.1 [20.3] 49.8 [20.2] p = 0.66 FIBER [g] 19.7 [7.96] 19.7 [7.28] p = 0.99 Discussion We failed to prove our hypothesis that administration of psyllium in the morning would have a greater cholesterol-lowering effect than it would in the evening. Not only was there no observable difference in lipid levels between the crossover periods but the daily ingestion of a greater daily dose than the 10.2 g of psyllium for which the FDA allows health claims to be made [ 6 ] had no effect on lipid levels in our study group. No change in any lipid parameter, including total and LDL cholesterol was observed. No difference was found when subgroup analysis was undertaken for the sex of the patients, the time of day they took their psyllium, or whether they were recruited from the gastroenterology clinic or the lipid clinic. We used a crossover design since this was the most appropriate one for the primary question being addressed; accordingly, our study did not include a control group. However, the nature of the study should not have provided any motivation for study subjects to adopt any new lifestyle or dietary changes beyond those implemented well before the introduction of psyllium. Observational data has been shown to provide valid information, which is consistent with that observed in randomized, controlled trials [ 17 , 18 ]. The nature of the intervention was in keeping with those undertaken in day-to-day clinical practice and the protocol used should, therefore, have high "clinical relevance". Failure of lipid-lowering by psyllium has also been demonstrated in twenty hypercholesterolemic children [ 19 ], in twenty-four hyperlipidemic adults [ 20 ] and in a large observational study of elderly patients taking psyllium [ 21 ]. A report of lipid-lowering therapies in hypercholesterolemic veterans showed only a 2% reduction in LDL cholesterol and a small increase in LDL/HDL ratio in patients taking psyllium, but does not provide a measure of statistical significance [ 22 ]. One study revealed no difference in total cholesterol-lowering compared to placebo, but a reduction of LDL cholesterol resulted from psyllium treatment [ 23 ]. Another demonstrated no difference in total cholesterol-lowering compared to placebo, a reduction of LDL in 11 "responders" and no change in 9 "nonresponders" [ 24 ]. A reduction of HDL cholesterol has been noted in some studies [ 25 - 27 ] and was associated with changes in LDL/HDL ratios similar to placebo treatment [ 25 , 26 ]. Published studies include few normocholesterolemic subjects. Cholesterol reduction was observed in 7 normal men [ 28 ] and in 5 of 9 subjects [ 29 ], in both studies after 3 weeks of treatment. A reduction of cholesterol levels was also observed in 12 elderly patients given psyllium for 4 months [ 30 ], while 5 normocholesterolemic subjects in another study showed no reduction after 2 to 7 months of treatment [ 31 ]. A meta-analysis of 17 studies of patients with hypercholesterolemia has suggested a small but significant cholesterol-lowering effect of psyllium [ 2 ]. All of these investigations were associated in one way or another with the product manufacturer. Additional studies have also indicated some cholesterol-lowering by psyllium in hypercholesterolemic individuals [ 32 - 37 ] or in diabetics [ 38 - 40 ]; however, much of this work is uncontrolled and some protocols have specifically excluded premenopausal women [ 33 , 38 ]. The association of cholesterol-lowering effects with psyllium may be weakened in some studies by the use of a supplement containing additional forms of soluble fibre [ 42 ] or by apparent differences in intake of calories [ 43 - 46 ], soluble fibre [ 25 ] or cholesterol [ 47 ] in control and treatment groups or periods. Several reports include only small numbers of patients and/or are of short duration. There is a strong predominance of male subjects in these publications and some protocols incorporate additional treatment interventions [ 20 , 48 ]. Several factors may contribute to the difference between our observations and those of others. A meta-analysis has demonstrated that the initial level of cholesterol was highly predictive of the subsequent reduction of cholesterol by oat bran [ 49 ]. A greater effect of psyllium in men compared to women has been suggested [ 23 , 46 ] and a diet high in soluble fibre produced less cholesterol-lowering in post menopausal women than in men [ 10 ]. Soluble fibre has a lesser effect on lipid metabolism in female than in male guinea pigs [ 50 ] and there is a sex-based difference in mechanism of action in this animal [ 51 ]. Oat bran fails to lower cholesterol in young women, in contrast to men and older women [ 52 ]. The dominance of women in our study, the "normal", or only slightly abnormal cholesterol states of our subjects and the relatively young ages of some of them may, accordingly, account for some of the variance of our observations with some of those previously reported. The small increase in the weight of subjects is believed to be have resulted from reduced physical activity. In a meta-analysis of the effect of weight reduction on lipids, predominantly through dietary change, a reduction in total cholesterol of 0.05 mmol/L and of 0.02 mmol/L in LDL cholesterol per kilogram of weight lost was identified [ 53 ]. Dietary intakes were stable throughout our study and the average weight gain of less than one kilogram is very unlikely to have raised cholesterol levels to a degree sufficient to offset a significant cholesterol lowering effect of psyllium. A small cholesterol-lowering effect of psyllium appears to occur in hypercholesterolemic individuals, at least in men and possibly postmenopausal women. The notion of a benefit accruing to the general population requires additional study. The promotion of foods containing psyllium as reducing the risk of heart disease for the population at large [ 6 ] may be premature. Additional study is required and this should be undertaken in a manner that is free from concern regarding the possibility of publication bias which Brown L, Rosner B, Willett WW and Sacks FM have raised [ 2 ]. Conclusion The timing of psyllium administration had no effect on cholesterol-lowering and, in fact, no cholesterol-lowering was observed. Conclusions regarding the effectiveness of psyllium for the prevention of heart disease in the population at large may be premature. Competing interests The authors declare that they have no competing interests. Authors' contributions GVR carried out the study design, data review and writing. EAS carried out the study design and data review. RB carried out the study design, data review and statistical analysis. ALE carried out the study design and data review. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC522822.xml |
514609 | Ancestry reported by white adults with cutaneous melanoma and control subjects in central Alabama | Background We sought to evaluate the hypothesis that the high incidence of cutaneous melanoma in white persons in central Alabama is associated with a predominance of Irish and Scots descent. Methods Frequencies of country of ancestry reports were tabulated. The reports were also converted to scores that reflect proportional countries of ancestry in individuals. Using the scores, we computed aggregate country of ancestry indices as estimates of group ancestry composition. HLA-DRB1*04 allele frequencies and relationships to countries of ancestry were compared in probands and controls. Results were compared to those of European populations with HLA-DRB1*04 frequencies. Results Ninety evaluable adult white cutaneous melanoma probands and 324 adult white controls reported countries of ancestry of their grandparents. The respective frequencies of Ireland, and Scotland and "British Isles" reported countries of ancestry were significantly greater in probands than in controls. The respective frequencies of Wales, France, Italy and Poland were significantly greater in controls. 16.7% of melanoma probands and 23.8% of controls reported "Native American" ancestry; the corresponding "Native American" country of ancestry index was not significantly different in probands and controls. The frequency of HLA-DRB1*04 was significantly greater in probands, but was not significantly associated with individual or aggregate countries of ancestry. The frequency of DRB1*04 observed in Alabama was compared to DRB1*04 frequencies reported from England, Wales, Ireland, Orkney Island, France, Germany, and Australia. Conclusion White adults with cutaneous melanoma in central Alabama have a predominance of Irish, Scots, and "British Isles" ancestry and HLA-DRB1*04 that likely contributes to their high incidence of cutaneous melanoma. | Background The incidence of cutaneous melanoma in Alabama is high (16.9 and 8.6 per 100,000 men and women, respectively, in 1998) [ 1 ]. The average annual incidence rate in Alabama during 1996 was 15.5 per 100,000 for men and 8.8 per 100,000 for women, indicating that the incidence in men is increasing. This is consistent with reports that cutaneous melanoma is one of few cancers the incidence and mortality of which is increasing in white Americans [ 2 ]. Lifestyle, sun exposure, fair skin, light eye color, poor ability to tan, Northern European or Celtic ethnicity, family history of melanoma, benign nevi, various major histocompatibility complex alleles, and mutations in the p16 gene have been reported as risk factors for cutaneous melanoma [ 3 - 16 ]. HLA-DRB1*04 is significantly associated with cutaneous melanoma in white persons in Alabama [ 9 ] and Australia [ 11 , 12 ]. In addition, HLA-DRB1*04 was reported positively associated with cutaneous melanoma in white patients who had a high index of Celtic ancestry [ 14 ]. It has been hypothesized that the high incidence of melanoma in Alabama, a Sunbelt state with a mild climate, is due to the settlement of Alabama by a predominance of persons of Irish and Scots descent [ 8 - 10 ]. In the present study, we sought to evaluate this hypothesis by using questionnaires to obtain information about the countries of ancestry of the grandparents of white adults in central Alabama with cutaneous melanoma and of white control subjects from the central Alabama general population. Using data from the questionnaires, we evaluated the frequency of country of ancestry reports in cutaneous melanoma and control participants and computed country of ancestry indices to permit quantification and comparison of group ancestry data, as previously described [ 17 ]. We also compared HLA-DRB1*04 frequencies in participants with melanoma who reported "Celtic" and "non-Celtic" ancestry. The rationale for using country of ancestry information and HLA-DRB1*04 to identify persons who have increased risk to develop cutaneous melanoma is discussed. Methods General criteria for selection of study subjects The performance of this work was approved by the Institutional Review Boards of the University of Alabama at Birmingham and Brookwood Medical Center. All subjects were adults (≥ 18 years of age) who were central Alabama residents; each identified himself/herself as white. Persons with melanoma were diagnosed in routine medical care in the interval 1981 – 2002, but were otherwise unselected. We excluded persons of sub-Saharan African or African American descent, because most of these persons have non-European ancestry and their incidence of cutaneous melanoma is much lower than in white persons in central Alabama [ 1 , 2 ]. Melanoma probands Cutaneous melanoma was diagnosed as previously described [ 9 ]. HLA-DRB1 typing was performed by either the microdroplet lymphocytotoxicity test using B-lymphocytes isolated by the nylon wool column procedure, or by low-resolution SSP using genomic DNA obtained from peripheral blood buffy coat, as previously described [ 9 , 18 ]. Further, we included only the first persons diagnosed to have cutaneous melanoma in respective families; they were designated as probands. Control subjects Adult control subjects residing in central Alabama were recruited in two groups. The first group consisted of 83 spouses of persons with cutaneous melanoma and randomly recruited subjects. The second group consisted of 260 unselected volunteers who completed the present questionnaire (described below); they were recruited from hospital workers, employees of two universities, spouses of patients who attended a hematology/medical oncology outpatient clinic, and members of the general public encountered in two retail shopping malls. In both groups, we excluded persons who were known to be relatives of other study participants. We did not evaluate medical histories or perform physical examinations in control subjects. These data were pooled to yield a group of 343 unrelated controls, of whom 24 were eliminated because they did not know the country of ancestry of any of their grandparents (as indicated below). This left a group of 319 control subjects whose data were deemed evaluable for the present study. HLA typing was performed on some control subjects as indicated below. Questionnaire and interview design A one-page questionnaire was designed to permit each study participant to indicate the countries of ancestry of his/her paternal and maternal grandparents. This method is identical to that previously validated in a study of the ancestry of hemochromatosis probands [ 17 ]; in part, this method was modified from previously reported methods by including only the country of birth of grandparents [ 6 , 14 , 19 ]. We defined aggregate country categories as the composite of reports from respective countries [ 17 ]. Reports based solely on association of family names with specific countries were tabulated as "Don't Know." Reports from participants for whom each of four grandparents were categorized as "Don't Know" were defined as inevaluable and were excluded from final analysis [ 17 ]. We did not evaluate relationships of country of ancestry to gender of the grandparents or to the paternal or maternal side of the participant's family [ 17 ]. Frequencies of countries of ancestry reports We tabulated the number of participants who reported specific countries of ancestry or aggregate country categories (as defined above) for one or more grandparents. This method is identical to that previously described [ 17 ]; in part, this method was modified from previously reported methods by including only the country of birth of grandparents [ 14 , 19 ]. Country of ancestry indices The questionnaire and interview reports from each participant were evaluated to yield individual country of ancestry scores that reflect proportional national ancestry, as previously described [ 17 ]. We also computed aggregate country of ancestry indices for cutaneous melanoma probands and control subjects as estimates of group ancestry composition. These indices were expressed as the quotient of total individual scores for respective countries and the total number of cutaneous melanoma probands or control subjects, as appropriate [ 17 ]. Index of "Celtic" ancestry An index of Celtic ancestry was computed in a manner similar to that used for country of ancestry indices. In the present analysis, persons who reported one or more grandparents whose country of origin was Ireland or Scotland were defined as "Celts." Other participants were defined as "non-Celts." We did not use grandparental surnames or maiden names to quantify the degree of Celtic ancestry as did previous investigators [ 14 , 19 ]. Review of HLA-DRB1*04 phenotype frequencies in europe and australia A manual and computerized literature search was performed to identify reports of HLA-DRB1*04 phenotype frequencies in white persons with or without melanoma in countries or areas of Europe and Australia that are known to have had Celt settlements The HLA-DRB1*04 phenotype frequency among white control subjects in central Alabama is approximately 0.2148, based on the combined data from two previous reports [ 9 , 20 ]. Thus, we tabulated only those reports from countries or regions in which the HLA-DRB1*04 phenotype frequency was > 0.2148. We excluded studies which reported HLA-DRB1*04 frequency estimates on control cohorts of fewer than 100 subjects. Statistical considerations The dataset consisted of observations on 90 cutaneous melanoma probands and two groups of controls (83 interview subjects and 261 questionnaire responders, respectively). A computer spreadsheet (Excel 2000, Microsoft Corp., Redmond, WA) and a statistical program (GB-Stat, v. 8.0 2000, Dynamic Microsystems, Inc., Silver Spring, MD) were used to perform the present analyses. In a preliminary evaluation, we determined that the proportions of men and women, mean ages, country of ancestry reports, and frequencies of country of ancestry reports were not significantly different in the two control groups. Therefore, we pooled data from the two groups for comparison with those of melanoma probands. Frequencies of men and women, clinical abnormalities, and countries of ancestry were counted. General descriptive data are presented as percentages or mean ± 1 S.D. Comparisons between groups were tested for statistical differences using chi-square analysis, Fisher exact test or two-tailed Student t test, as appropriate. However, Student t-test can not be used to compare country of ancestry data groups in which all values are 0 (no variability). Accordingly, we arbitrarily assigned a country of ancestry datum of 0.25 for one person in each proband country group for which there were no actual country of ancestry reports. Student t-test was then performed using this modified data group; estimated p values from these tests are displayed in parentheses. A p value < 0.05 was defined as significant. Odds ratios (OR) were calculated as described by Woolf [ 21 ]. Results Characteristics of cutaneous melanoma probands There were reports from 90 evaluable probands (46 men, 44 women). Their mean age at the time of participation in the present study was 49 ± 16 years (range 19 – 80 years). Characteristics of control subjects There were reports from 319 evaluable control subjects (128 men, 191 women). The percentages of men and women in the control group were similar to those in the cutaneous melanoma probands (p = 0.0621, chi square analysis). Their mean age at the time of participation in the present study was 43 ± 16 years (range 18 – 82 years). The mean age of the cutaneous melanoma probands was significantly greater than that of the control subjects (49 ± 16 years vs. 43 ± 16 years, respectively; p = 0.0016). General analysis of questionnaire and interview reports No person declined to participate in the study, and there were no incomplete, equivocal, or unintelligible questionnaire or interview reports. Some participants reported that they were unaware of their ancestry due to adoption, family estrangement, or disinterest in genealogy. Ninety of 110 cutaneous melanoma probands (81.8%) and 319 of 347 control subjects (91.3%) provided reports for at least one of four grandparents; these differences were significant (p = 0.0026). Data from these participants were included for further analysis. Thus, there were reports from 90 evaluable cutaneous melanoma probands and 319 evaluable control subjects. Among evaluable cutaneous melanoma probands, the mean number of countries reported was 2.2 ± 1.1 (range 1 – 5 countries). The mean number of countries reported by evaluable control subjects was similar (2.4 ± 1.1 countries (range 1 – 7 countries); p = 0.2246). Frequencies of country of ancestry reports These data are displayed in Table 1 . Most participants reported European countries of ancestry. Frequencies of Ireland and Scotland country reports of ancestry obtained from cutaneous melanoma probands were significantly greater than those from control subjects (p = 0.0042, OR = 2.0 and p = 0.0015, OR = 2.2, respectively). The "Celtic" country reports of ancestry was also significantly higher in melanoma probands than in control subjects (Table 1 ). The frequency of "British Isles" and "Europe Not British Isles" ancestry reports tabulated in cutaneous melanoma probands were not significantly different than those in controls subjects. The respective frequencies of Wales, France, Italy, and Poland ancestry reported by cutaneous melanoma probands were significantly lower than those in control subjects (Table 1 ). 16.7% of country of ancestry reports from cutaneous melanoma probands and 23.8% in control subjects indicated "Native American" ancestry; these percentages were not significantly different (Table 1 ). The percentage of cutaneous melanoma probands who reported "Don't Know" for the countries of ancestry of one, two, or three grandparents was similar to that in control subjects (30.0% vs. 30.1%, respectively) (Table 1 ). Country of ancestry indices These data are displayed in Table 2 . The respective Ireland and Scotland indices in cutaneous melanoma probands were significantly greater than those in control subjects. The "Celtic" ancestry index was significantly higher in melanoma probands than in control subjects. The aggregate "British Isles" index was also significantly greater in melanoma probands (Table 2 ). The Wales, France, Spain, Austria, Italy, Poland, Russia, Sweden and "Europe Not British Isles" indices in probands were significantly lower than those in control subjects (Table 2 ). There were no other significant differences. HLA-DRB1*04 frequencies HLA-DRB1 phenotypes were available for 63 of the 90 present cutaneous melanoma probands. We divided the probands for which we had both country of ancestry data and HLA-DRB1 phenotypes into "Celts" (n = 19) and "non-Celts" (n = 44), as defined above. There was no significant difference in the frequency of HLA-DRB1*04 in the two groups (0.5263 vs. 0.3636, respectively; p = 0.3551, OR = 1.9). Because this finding was unexpected and differed from that of previous reports [ 14 ] we then computed the frequencies of HLA-DRB1*04 in Alabama white cutaneous melanoma probands and in control subjects from two previous Alabama reports [ 9 , 20 ]. Thus, the frequency of HLA-DRB1*04 in 123 probands (38.2%) was significantly different from that in 340 controls (21.5%) (p = 0.0005, OR = 2.3). DRB1*04 phenotype frequencies > 0.2148 in Europe and Australia We identified HLA-DRB1*04 phenotype frequency estimates of > 0.2148 from England, Ireland, Scotland, Wales, Orkney Islands, Brest, Germany, and Australia, areas populated predominantly by persons of Celtic ancestry [ 9 , 11 , 13 , 20 , 22 ]. These data and the results of studies in which HLA-DRB1*04 phenotype frequencies were assessed in melanoma patients are displayed in Table 3 . The frequencies of HLA-DRB1*04 in these countries or regions are significantly higher than that in Alabama subjects (p ≤ 0.005) (Table 3 ). Discussion The present results indicate that England, Ireland, Scotland and the aggregate "British Isles" are the countries of ancestry reported most frequently by cutaneous melanoma probands and control subjects in central Alabama. Germany is another country of ancestry often reported by the present study participants. These results are consistent with historical accounts of early migrations of persons of English, Irish and Scots descent into central Alabama [ 23 - 26 ], with the national associations of surnames recorded in Alabama Census returns for 1820 and 1830 [ 27 ], and with the present composition of the southern United States [ 25 ]. In U.S. Census 2000, country of ancestry information (maximum of two countries) was reported on a "long form" provided to one in six census participants, and was tabulated as numbers of country-specific reports [ 28 ]. Thus, the data of U.S. Census 2000 cannot be compared statistically with the results of the present study, but the percentages of European countries of ancestry of white Alabama residents compiled in both studies reveal similar trends. In the present study, the largest subgroups of reports in the "North, Central, and South American Countries" category are those of "U.S." or "American" ancestry. Some participants did not report or know the country of ancestry of their grandparents. These findings are also consistent with trends in the U.S. Census 1990 and U.S. Census 2000, in which the percentages of white Americans who report "American" ancestry are increasing, and the percentages of those who report various European countries of ancestry are decreasing [ 28 , 29 ]. "Native American" ancestry, especially Cherokee or Creek heritage, was reported by many of the present study participants, and this is consistent with accounts of early Alabama history [ 24 , 30 - 32 ] and with U.S. Census data on Alabama since 1820 [ 27 - 29 ]. However, the corresponding aggregate "Native American" frequency of country ancestry reports and country of ancestry indices were not significantly different in melanoma probands and control subjects. This supports the postulate that native American ancestry does not contribute significantly to the increased frequency of cutaneous melanoma in central Alabama. The percentages of men and women were not significantly different in cutaneous melanoma probands and control subjects. Analyses of the two study groups indicate that age is not significantly correlated with country of ancestry indices. The mean age of cutaneous melanoma probands was significantly greater than that of control subjects. However, the mean number of countries reported by cutaneous melanoma probands and control subjects did not differ significantly. Thus it appears that diagnosis of cutaneous melanoma or greater age is not associated with greater interest or knowledge in personal ancestry, although this is unproven. 30.0% of cutaneous melanoma probands and 30.1% of control subjects did not know the country of ancestry of any of their four grandparents and were thus declared inevaluable. Many others did not know the ancestry of some of their grandparents. Some participants reported that they were unaware of their ancestry due to adoption, family estrangement, or disinterest in genealogy. Other participants could have been incorrect in their reporting. The percentages of evaluable cutaneous melanoma probands and control subjects who reported grandparents in the "Don't Know" category were similar. Altogether, it is unlikely that exclusion of subjects who did not know the country of ancestry of each of their grandparents would significantly change the outcomes of the present study. The overall trends in frequency of country reporting and country of ancestry indices in "British Isles" and "Europe Not British Isles" categories in the present study were similar. This suggests that uncertainty of participants about the exact degree of country of ancestry of some of their grandparents was probably not a significant contributor to the major conclusions of the present study. It is possible that control subjects who were included also had cutaneous melanoma. Nonetheless, it is unlikely that identification of a presumably small number of undiscovered control subjects with cutaneous melanoma would significantly change the conclusions of the present study. Previous studies reported that Celtic ancestry is a risk factor for cutaneous melanoma, particularly when people of Celtic ancestry inhabit areas of high flux of ultraviolet radiation [ 4 , 6 ]. Persons of Celtic ancestry usually have fair skin, eyes and hair of light color, poor ability to tan, and tendency to burn easily after sun exposure [ 4 , 6 ]. Areas of the British Isles where persons have a high degree of Celtic ancestry include Wales, Cornwall, Scotland and Ireland [ 6 ]. We did not specifically ask about nor did we receive reports of Cornwall as a region of origin in the present study. Similarly, Wales was reported as a country of origin by none of the present probands and few of the control subjects. However, the frequency of country reports and ancestry indices for Ireland, Scotland, aggregate "British Isles," and combined Ireland and Scotland ("Celts") in the present melanoma probands was significantly greater than that in control subjects, consistent with the hypothesis that Celtic ancestry is a risk factor for cutaneous melanoma. The present observations are consistent with those of another study in Alabama in which white persons with hemochromatosis, a disease thought to be of Celtic origin [ 33 , 34 ], reported a significantly greater Scotland and "British Isles" ancestry than controls [ 17 ]. The countries of ancestry reported was also consistent with the relative high frequency of C282Y, the major allele associated with hemochromatosis, in Alabama hemochromatosis probands compared to controls [ 17 ]. Using an index of Celtic ancestry that included grandparental surnames, maiden names, and country of birth to evaluate persons in Wales, other investigators observed that "high-scoring Celts" were significantly more likely to have Fitzpatrick skin type I or II (poor ability to tan, and tendency to burn easily after sun exposure) [ 35 ] than non-Celtic subjects [ 14 , 19 ]. Moreover, the frequency of HLA-DRB1*04 was significantly greater in "high-scoring Celts." These authors concluded that the increased risk of cutaneous melanoma and other types of skin cancer in persons of Celtic ancestry in Wales is due not only to paler skin, but also to HLA-DRB1*04 and associated immunologic factors. Observations from previous studies demonstrate that the frequency of positivity for the HLA-DRB1*04 phenotype is significantly greater in Alabama cutaneous melanoma probands than in control subjects [ 8 , 9 , 20 ]. This is consistent with reports from Australia and Wales [ 11 , 12 , 14 ]. Further, previous reports from Alabama suggest that the subgroup of individuals who possess HLA-DRB1*04 are at increased risk for cutaneous melanoma, independent of eye color, hair color, amount of melanin in the skin, or ethnic origin [ 8 ]. However, the failure to reach statistical significance in the comparison of HLA-DRB1*04 phenotypes in the present 19 "Celts" and the 44 "non-Celts" is likely due to the small number of probands who reported Celtic ancestry and for whom HLA typing data were available. Investigators in Texas and England reported that the HLA genotype DQB1*0301 influences either susceptibility to and/or severity of melanoma [ 15 , 36 - 38 ]. This conclusion is consistent with previous studies in which HLA-DRB1*04 was associated with melanoma, because HLA-DQB1*0301 is in linkage disequilibrium with HLA-DRB1*04 [ 37 ]. The highest frequencies of HLA-DRB1*04 in Europe occur in England, Wales, Ireland, the Orkney Islands, and the Brest area of France [ 22 ]. These geographic results could explain the significantly increased frequency of HLA-DRB1*04 in cutaneous melanoma probands from central Alabama, because the present melanoma probands had significantly higher Ireland, Scotland, and "British Isles" country of ancestry indices than Alabama control subjects. Our observations and those of others suggest that certain HLA genotypes may be markers for Celtic ancestry [ 8 , 14 ]. It has been reported that HLA phenotypes other than HLA-DRB1*04 occur in association with melanoma in various white national or ethnic groups [ 13 , 15 , 36 , 39 , 40 ]. This could be explained in part by the variation of HLA phenotype frequencies in these white national or ethnic groups or by other genes within the major histocompatibility complex in linkage disequilibrium that also play a major role in mediation of immune responses. Thus, various HLA phenotypes could be markers for white national or ethnic groups that also possess certain physical characteristics and immune response genes that increase their risk for developing cutaneous melanoma. Conclusions The present results support our hypothesis, and are also consistent with the association of cutaneous melanoma with Ireland, Scotland and "British Isles" ancestry, HLA-DR phenotypes, and estimations of a northern European somatic phenotype (eye color, hair color, amount of melanin in the skin) previously reported in white persons who reside in central Alabama [ 3 , 4 , 8 , 9 ]. The present observations also suggest that targeting white persons with relatively high "Celtic" or "British Isles" country of ancestry indices and HLA-DRB1*04 for cutaneous melanoma prevention and early diagnosis efforts would be an effective strategy to decrease the morbidity and mortality of this type of malignancy. Competing interests None declared. Authors' contributions RTA assessed the patients and participated in assessing the control population, conceived the study, participated in data collection, laboratory evaluation of the patients and controls, statistical evaluation, and wrote part of the manuscript. EHB participated in assessed the control population and in data collection. WWH participated in assessed the control population and in data collection. ALD participated in assessed the patient and control populations, in data collection and laboratory evaluation of the patients and controls. RCPG participated in conceiving the study. JCB participated in conceiving the study, assessing the control population, statistical evaluation and wrote part of the manuscript. All authors approved the final version of the manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC514609.xml |
506782 | "A calorie is a calorie" violates the second law of thermodynamics | The principle of "a calorie is a calorie," that weight change in hypocaloric diets is independent of macronutrient composition, is widely held in the popular and technical literature, and is frequently justified by appeal to the laws of thermodynamics. We review here some aspects of thermodynamics that bear on weight loss and the effect of macronutrient composition. The focus is the so-called metabolic advantage in low-carbohydrate diets – greater weight loss compared to isocaloric diets of different composition. Two laws of thermodynamics are relevant to the systems considered in nutrition and, whereas the first law is a conservation (of energy) law, the second is a dissipation law: something (negative entropy) is lost and therefore balance is not to be expected in diet interventions. Here, we propose that a misunderstanding of the second law accounts for the controversy about the role of macronutrient effect on weight loss and we review some aspects of elementary thermodynamics. We use data in the literature to show that thermogenesis is sufficient to predict metabolic advantage. Whereas homeostasis ensures balance under many conditions, as a general principle, "a calorie is a calorie" violates the second law of thermodynamics. | Review The recent awareness of an epidemic of obesity coincides with, and may have contributed to a dramatic increase in the popularity of a variety of low carbohydrate diets. This rapid switch in dietary habits of a significant part of the population, and the virtual revolution in the food industry, is unusual in that it stands in direct opposition to long-standing recommendations of the majority of the nutritional and medical establishment (e.g. [ 1 , 2 ]). Despite isolated examples, such as a recent editorial by Walter Willet pointing to the need to understand low carbohydrate diets [ 3 ], there is still little real acceptance by nutrition professionals or health organizations. One aspect of these diets that has been especially controversial is the so-called metabolic advantage – the idea that more weight may be lost calorie for calorie compared with diets of higher carbohydrate content. We recently reviewed the literature on metabolic advantage [ 4 ]. We showed that there is a sufficient number of reports in the literature to establish the existence of metabolic advantage and we tabulated results from ten or so studies demonstrating that low carbohydrate diets can lead to greater weight loss than isocaloric low fat diets. The reports we cited have frequently been met with the criticism that the data could not be right because they would violate the laws of thermodynamics ([ 5 , 6 ]). An example is the recent demonstration of metabolic advantage in a small, pilot study [ 7 ] which, despite its preliminary status, was extremely well controlled. Three groups were studied: A low carbohydrate group (LoCHO = 1800 kcal for men; 1500 kcal for women), a low fat group (LoFat, 1800 and 1500); a third group also consumed a low carbohydrate diet but an additional 300 kcalories were provided (LoCHO+300, 2100 and 1800). The order of average amount of weight lost was LoCHO = 23 lbs, LoCHO+300 = 20 lbs LoFat = 17 lbs. This work received a good deal of attention in the popular press. Media reports, however, included comments of experts that "It doesn't make sense, does it?" "It violates the laws of thermodynamics. No one has ever found any miraculous metabolic effects." ([ 5 ]). If this is an accurate quotation, it is odd indeed. Miraculous, or otherwise, a metabolic effect was found. In the absence of an identifiable methodological error, experimental data has to be accepted and numerous investigations, in fact, serve as precedents for Greene et al.'s findings (Reviews: [ 4 , 8 ]). In our previous review of metabolic advantage [ 4 ] we showed that there is, in fact, no theoretical violation of the laws of thermodynamics, and we provided a plausible mechanism. In general the pathways for gluconeogenesis that are required in order to supply obligate glucose (e.g. to brain and CNS), in combination with increased protein turnover, could account for the missing energy. Here, we simplify the thermodynamic argument and review some of the relevant principles. We show, moreover, that well-established data in the traditional nutritional literature predict metabolic advantage and no one should be surprised. The ironic conclusion is that the principle that weight gain on isocaloric diets must always be independent of macronutrient composition would violate the second law of thermodynamics. What do we mean by "a calorie is a calorie?" Because it is a colloquial phrase, it is important to understand exactly what it is meant by "a calorie is a calorie." The most common meaning is that is it impossible for two isocaloric diets to lead to different weight loss. Frequently, the concept is justified by reference to the "laws of thermodynamics", but an explicit connection has never been spelled out. More recently, Buchholz & Schoeller [ 10 ] appear to identify "a calorie is a calorie" with the first law of thermodynamics. They also admit that high protein /low carbohydrate diets can lead to greater weight loss than isocaloric low fat diets in agreement with our assessment [ 4 ]. Nonetheless they maintain that "a calorie is a calorie," now justifying it by their connection of the phrase to the concept of energy conservation. It is important to point out that no study of isocaloric diets has ever claimed that the first law of thermodynamics is not true. Buchholz & Schoeller [ 10 ] have limited themselves by only including the first law and, therefore, do not understand how the differential weight loss could occur and think it "deserves further study." Our major point here is that there is more than one law of thermodynamics and that a more accurate understanding of the role of the second law shows that differential weight loss is not inconsistent with any physical principle. Thermodynamics The idea that "a calorie is a calorie" comes from a misunderstanding of the laws of thermodynamics. There are two laws of thermodynamics. (The zeroth law that establishes the concept of temperature and the third law that describes absolute zero are not relevant here). When speaking of "the law s of thermodynamics" it is important to be sure that one is including the second law. The first law is very different in character from the second law [ 9 , 11 , 12 ]. The first law is a conservation law: it says that the form of energy may change, but the total is always conserved. The second law is a dissipation law: it defines a quantity, the entropy, S, which we traditionally identify with disorder or high probability. The second law says that in any (real) irreversible process, the entropy must increase (ΔS > 0); balance is not expected. Entropy is, in fact, identifiable with irreversibility. It is important to understand that it is the second law that drives chemical reactions. The first law is a bookkeeping law and tells us that the total energy attributed to work, heat and changes in chemical composition will be constant. It does not tell us whether such a reaction will occur, or if it does, what the relative distributions of the forms of energy will be. To predict the tendency of the reaction to occur, we must employ the second law that says the entropy must increase. In a chemical reaction, at constant temperature and pressure, the entropic and energetic effects are combined into the change in the Gibbs free energy, ΔG, whose sign predicts the direction of reaction, and whose magnitude indicates the maximum amount of work realizable from the reaction. Application of ΔG' To understand the implications of "a calorie is a calorie," that energy yield could be path-independent and the same for all diets consider that it implies that carbohydrate and protein are equivalent fuels as shown in Figure 1 . The diagram indicates that, because it is a state variable, the free energy (ΔG') for Path 1 must be equal to that for path 2 + 3. If the ΔG' values for path 1 and path 2 are taken to be their calorimeter values, they will be approximately equal (~4 kcal/g, path 1 corrected for ureagenesis). This means that ΔG' for path 3, the conversion of protein to carbohydrate (also corrected) must be about zero. There exists at least one condition where this is not true, the standard state; it is generally considered that gluconeogenesis from one mole of alanine requires about 6 ATP [ 13 , 14 ]. Of course free energies are concentration dependent, so in vivo values will differ from standard state values but they are continuous functions of the concentrations and there will be numerous conditions under which ΔG' is not zero. In other words, assuming that protein and carbohydrate are energetically equivalent leads to a contradiction. Figure 1 Pathways for oxidation of macronutrients. Inefficiency The second law was developed in the context of the industrial revolution and the attempt to understand the efficiency of machines. The law describes the theoretical limits on the efficiency of engines and applies as well to living (irreversible) systems. The second law says that no machine is completely efficient. Some of the available energy is lost as heat and in the internal rearrangement of chemical compounds and other changes in entropy. In other words, although the first law holds even in irreversible processes – energy is still conserved – the second law says that something is lost, something is unrecoverable. The efficiency of a machine is dependent on how the machine works and, for a biochemical machine, the nature of the fuel and the processes enlisted by the organism. A simple example is the inefficiency of low-test gasoline in high compression gasoline engines. If a "calorie is a calorie" were true, nobody would pay extra for high test gasoline. (The calorimeter values of a gasoline will be the same whether or not it contains an antiknock compound). In weight loss diets, of course, inefficiency is desirable and is tied to hormonal levels and enzyme activities Efficiency and thermogenesis In nutrition, one component of inefficiency is measured in thermogenesis (thermic effect of feeding), or the heat generated in processing food. There is a large literature on this subject and the general conclusion, as summarized in a recent review by Jéquier [ 15 ], is that thermic effects of nutrients is approximately 2–3 % for lipids, 6–8 % for carbohydrates, and 25–30% for proteins. It is interesting that this data itself might be enough to explain metabolic advantage. Here we took the average of Jéquier's values (2.5, 7 and 27.5 % for fat, CHO and protein) and calculated the effective energy yield for a 2000 kcal diet. If we assume a diet composition of CHO:fat: protein of 55:30:15, within the range of commonly recommended diets, the calculated effective yield is 1848 kcal. We now consider the effect of reducing carbohydrate progressively and substituting the calories removed equally between fat and protein. Figure 2 shows that the wasted calories due to thermogenesis increase as carbohydrate is reduced and reach 100 kcal at 21 % carbohydrate. This value of 100 kcal is recommended by several professionals as the goal for daily weight reduction (e.g. [ 16 ]). Notably, at 8 % CHO, the value for the early phase of the Atkins [ 17 ], South Beach [ 18 ] or Protein Power diets [ 19 ], 140 kcalories are lost as heat. Now, there will be metabolic accommodations and one can't predict that the ratios will stay the same over a long term diet, but the calculations show that the possibility of metabolic advantage should not come as a surprise. Figure 2 The dependence of effective calories on % carbohydrate in a 2000 kcal diet. Effective calories were determined by subtracting the losses due to thermogenesis as described in the text. Recommendations for fighting obesity frequently call for small reductions in calories [ 16 ]. In fact, given the resistance of steady state systems to small perturbations it is doubtful that this is a promising strategy. Nonetheless, taking the goal at face value, if it could be achieved by a simple change in macronutrient composition, such a method would seem worthy of serious consideration. The arguments above show that such a phenomenon is possible. There are plausible arguments for how it could take place and substantial experimental evidence for its occurrence [ 4 ]. Conclusions A review of simple thermodynamic principles shows that weight change on isocaloric diets is not expected to be independent of path (metabolism of macronutrients) and indeed such a general principle would be a violation of the second law. Homeostatic mechanisms are able to insure that, a good deal of the time, weight does not fluctuate much with changes in diet – this might be said to be the true "miraculous metabolic effect" – but it is subject to many exceptions. The idea that this is theoretically required in all cases is mistakenly based on equilibrium, reversible conditions that do not hold for living organisms and an insufficient appreciation of the second law. The second law of thermodynamics says that variation of efficiency for different metabolic pathways is to be expected . Thus, ironically the dictum that a "calorie is a calorie" violates the second law of thermodynamics , as a matter of principle. The analysis above might be said to be over-kill although it is important, intellectually, not to invoke the laws of thermodynamics inappropriately. There are also, however, practical consequences. The seriousness of the obesity epidemic suggests that we attack it with all the means at our disposal. Metabolic advantage with low carbohydrate diets is well established in the literature. It does not always occur but the important point is that it can occur. To ignore its possibilities and to not investigate the precise conditions under which it appears would be cutting ourselves off from potential benefit. The extent to which metabolic advantage will have significant impact in treating obesity is unknown and it is widely said in studies of low carbohydrate diets that "more work needs to be done." However, if the misconception is perpetuated that there is a violation of physical laws, that work will not be done, and if done, will go unpublished due to editorial resistance. Attacking the obesity epidemic will involve giving up many old ideas that have not been productive. "A calorie is a calorie" might be a good place to start. Competing interests None declared. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC506782.xml |
520751 | Scoliosis treatment using a combination of manipulative and rehabilitative therapy: a retrospective case series | Background The combination of spinal manipulation and various physiotherapeutic procedures used to correct the curvatures associated with scoliosis have been largely unsuccessful. Typically, the goals of these procedures are often to relax, strengthen, or stretch musculotendinous and/or ligamentous structures. In this study, we investigate the possible benefits of combining spinal manipulation, positional traction, and neuromuscular reeducation in the treatment of idiopathic scoliosis. Methods A total of 22 patient files were selected to participate in the protocol. Of these, 19 met the study criterion required for analysis of treatment benefits. Anteroposterior radiographs were taken of each subject prior to treatment intervention and 4–6 weeks following the intervention. A Cobb angle was drawn and analyzed on each radiograph, so pre and post comparisons could be made. Results After 4–6 weeks of treatment, the treatment group averaged a 17° reduction in their Cobb angle measurements. None of the patients' Cobb angles increased. A total of 3 subjects were dismissed from the study for noncompliance relating to home care instructions, leaving 19 subjects to be evaluated post-intervention. Conclusions The combined use of spinal manipulation and postural therapy appeared to significantly reduce the severity of the Cobb angle in all 19 subjects. These results warrant further testing of this protocol. | Background In the MEDLINE- indexed literature, chiropractic treatment has shown to be largely ineffective at significantly reducing scoliotic curvatures. Chiropractic treatment for scoliosis typically consists of spinal manipulation, electric stimulation, some form of isotonic, active exercises, and shoe lifts [ 1 ]. However, Lantz et al [ 2 ] has shown that these procedures, when applied over a one-year duration, were not sufficient to significantly reduce the Cobb angle of a scoliotic curvature. The treatment in this study focuses on the reduction of scoliosis by manipulative and rehabilitative methods not commonly used by most chiropractors. The major difference in this treatment compared to others is that stimulation of the involuntary postural reflexes is utilized in the clinic setting as well as in home care. Many of the proposed etiologies of idiopathic scoliosis are neurological in origin, including brain asymmetry [ 3 ], neural axis deformities [ 4 ], and central nervous system processing errors [ 5 ]. Additionally, many coexistent neurological alterations are present in scoliosis patients, such as visual deficiency [ 6 ] and decreased postural stability [ 7 , 8 ]. Therefore, the goals of the proposed treatment are not only to reduce the scoliotic curvatures, but also to rehabilitate any underlying postural and neurological weaknesses or imbalances. Previous chiropractic authors have investigated the effectiveness of various physiotherapeutic modalities in the treatment of scoliosis, such as Pilates [ 9 ], stretching and massage [ 10 ], therapeutic exercises [ 11 ], orthotics [ 2 ], and ultrasound or electric stimulation [ 1 ]. The purpose of the present study is to investigate any possible benefits from combining manipulative and rehabilitative techniques from a randomized sample collected from various chiropractic facilities. Preliminary evidence [ 12 ] suggests that these procedures may be beneficial for reducing the curvatures associated with scoliosis. Methods A nonrandomized set of 22 patients participated in the study. The age range of the subject group was 15–65 years of age. The patients were selected from 3 different chiropractic facilities in the United States. Patients were evaluated according to their chief complaint at initial presentation. Patients were excluded from the study if neoplasm, malignancy, fracture, scoliosis secondary to genetic disorders, or previous arthrodesis were identified. Each patient was examined radiographically for location and severity of scoliosis with standing anteroposterior full spine imaging. All patients removed their shoes for the imaging. Cobb angles were drawn on each radiograph to identify the degree of curvature present. A specific treatment plan was created based upon the results of each patient's radiographic measurements before and after a sample trial of the proposed clinical procedures. Initially, standing lateral cervical, nasium, lateral lumbar, and anteroposterior lumbopelvic views were taken. These views were taken to quantify forward head posture, cervical lordosis, lumbar lordosis, the sacral base angle, and the Cobb angle of the major lateral curvature. We decided to use the radiographic positioning and analysis outlined by Harrison et al [ 13 - 16 ], due to its previously published reliability. After these images were taken, each patient was fitted with a 4-lb anterior headweight. They were instructed to walk around with the headweight for 10 minutes. After 10 minutes, a follow-up lateral cervical radiograph was taken while wearing the anterior headweight. The purpose of this lateral stress view is to evaluate the potential improvement in cervical lordosis and reduction in forward head posture from using these procedures [ 17 , 18 ]. The basis for this aspect of the protocol is based upon the inherent properties of a curved column. In the spine, lateral spinal displacements may occur when the normal sagittal spinal curves [ 19 - 22 ] are flattened, reversed, or accentuated. These curves are necessary for the overall strength and flexibility of the curved spinal column, according to the Delmas Index [ 23 ]. Therefore, the proposed treatment is intended to restore a normal cervical and lumbar lordosis, and reduce forward head posture before the scoliotic curvatures are addressed. The specific manipulative and rehabilitative procedures used in this study are designed to both reduce the scoliotic curvature and theoretically retrain the involuntary neuromuscular, reflexive control of posture and balance. However, the specific neurological effects, if any, remain to be investigated. Some of the procedures have been separately introduced or tested [ 17 , 18 , 24 - 26 ]. The manipulative procedures included an upper cervical adjustment designed to mobilize the atlantal-occipital joint with the use of a percussive instrument. This technique is shown in Figure 1 . This technique is delivered to patients whose lateral cervical radiographs demonstrated atlanto-occipital flexion. If atlanto-occipital extension was present on the initial lateral cervical radiograph, a -Z drop piece was used to mobilize the occiput into flexion. This is also shown in Figure 1 . An anterior thoracic adjustment was administered with the patient's thoracic cage rotated opposite to the rotational displacement. A thoracic drop piece was also used to mobilize and correct the smaller upper thoracic curvature. Side posture lumbopelvic adjustments were delivered bilaterally to correct the rotational component of the pelvic misalignment. These side-posture manipulations were performed on a 30°-incline bench to help pre-stress the spine out of its existing scoliotic curvatures. Certain traction procedures are also employed. These procedures are delivered using high-density foam blocks to pre-stress the spine into specific positions so ligament deformation and stress-relaxation can take place. Supine pelvic blocking was performed on each patient for 15 minutes. The position of the blocks was determined by each patient's pelvic rotation on radiograph and posture analysis. One block is placed under the iliac crest of the posterior ilium, and the other block is placed under the femoral head of the opposite, anteriorly-rotated ilium. Figure 2 illustrates the position of the pelvic blocks. The rehabilitative procedures, demonstrated in Figure 3 , included the use of head, shoulder, and hip weighting devices. These devices may be used while simultaneously performing specific balancing exercises. These exercises include the use of a Pettibon Wobble Chair ® and a Posturomed ® [ 17 ]. Tjernstrom et al [ 27 ] showed that repeated performance of a postural alteration induces a long-term motor memory for achieving that novel postural position. The position of the body weighting was also determined radiographically for each patient. Initially, hipweights and shoulderweights were applied according to each patient's posture analysis. Anteroposterior cervicothoracic and lumbopelvic views were taken while wearing the head and body weighting. Since changes in spinal position are not reliably seen by visualization [ 28 , 29 ], these stress radiographs were taken to confirm their corrective effects. The attending physician treated each patient 3 times per week for the first 4–6 weeks. A total of 3 physicians performed the treatment intervention for all patients. However, each patient did not receive identical treatment at all visits. The physicians performed only those manipulative procedures that were deemed necessary based upon a visual posture analysis at the beginning of each treatment session. However, the rehabilitative procedures remained constant throughout the study for all patients. Specific home care exercise programs were taught to each patient. These exercises were performed on a daily basis. Each patient was instructed to wear the head and body weighting twice daily for 15 minutes each time. Secondly, each patient was given a set of triangular foam blocks to lie on once daily for 20 minutes, immediately prior to going to bed at night. The foam blocks were positioned under the cervicothoracic and thoracolumbar regions simultaneously. The position of these blocks is shown in Figure 4 . Patients participating in any weightlifting activities were required to cease those activities until further notice from the attending physician. Patients who failed to perform the home care more than 3 times were dismissed from the study. A total of 3 subjects were eventually dismissed, leaving 19 subjects to perform post-intervention evaluations. Results At the conclusion of the trial period, a post-intervention radiographic study was conducted. The same anteroposterior full spine view was taken, and Cobb angles were again measured at the same vertebral levels. The average starting Cobb angle was found to be 28°, while the post-intervention Cobb angles measured an average of 11°, for an overall average reduction of 17°. Every patient made at least a 25% improvement. The largest improvement measured 33°, and the smallest improvement measured 8°. Table 1 shows the results of all 19 patients that followed through with the entire treatment plan. Figure 5 is a sample of the improvements made by a few of the patients. It is important to mention that these patients were initially treated prior to this study. Because of this, the pre and post treatment radiographs had previously been analyzed for sagittal curve and Cobb angle measurements. For purposes of this study, however, all of the radiographs were sent to a single chiropractic physician to analyze each of the patient files. This physician did not participate in the treatment process, nor did this physician have contact with any of the patients. This was performed to separate examiner bias from the treatment results. While only radiographic procedures were reported for this study, other physiologic parameters were utilized to document patient progress. Unfortunately, since the patient files were extracted from 3 different spine clinics, a consistent functional or symptomatic measure was not used in all 22 cases. A functional rating index, a visual analog scale, and SF-36 were used on the patients here. As a result, these values are not reported to avoid variability in outcome interpretation. Discussion Scoliosis has recently been associated with a lower quality of life [ 30 - 32 ], lower scores on the SF-36 health questionnaire [ 33 ], and makes patients prone to developing chronic pain more often than the general population [ 34 ]. Therefore, reducing scoliotic curvatures, even in the absence of symptoms, seems to be a worthy outcome objective for clinical practice. This opinion is further supported by recent evidence of the deleterious effects of abnormal spinal loading [ 35 - 37 ]. Given that the average curvature progression in idiopathic scoliosis is 7.03° per year [ 38 ], the traditional method of regular observation without treatment seems to be reactionary rather than corrective or preventive. Spinal manipulation alone does not appear to significantly alter spinal structure when administered as a sole treatment modality [ 39 , 40 ]. Therefore, in the instance of scoliosis, treatment should include the use of both manipulative and rehabilitative procedures, so that structural changes can be attempted. It is important to stress that spinal manipulation was avoided, when possible, in the present study. Unpublished clinical observation by the authors has shown that over-manipulating or adjusting the spine seems to create a certain amount of instability, possibly leading to further buckling of the scoliotic curvature. The significance of home care to the results was not reported here. It is unknown how the omission of home care would have affected the outcome measurements, given that 3 subjects were dropped from the study for noncompliance in performing home care. Future research should account for this potential variable to determine its necessity and relevance. The outcome measures for this study are divided into a series of both short-term and long-term goals. The outcome of the initial stage of care is to reduce forward head posture and improve the sagittal cervical and lumbar curves. As the position of the head migrates forward, or away from the body's vertical axis, increased strain is placed upon the muscles of the head, neck and shoulders. Cailliet and Zohn indicated that an additional 10 inch/lbs of leverage is added to the spinal system in a forward head posture [ 41 , 42 ]. Additionally, this added leverage causes increased isometric contraction of various spinal muscles, such as the splenius capitis, trapezius, SCM, and levator scapula. Sjogaard et al [ 43 ] reported that blood flow through a given muscle is decreased as a muscles contraction increases, being virtually cut off at 50–60% contraction. The resultant lack of blood flow forces the muscle to rely on anaerobic metabolism. As anaerobic metabolism progresses, metabolites such as substance P, bradykinin, and histamine build up and excite chemosensitive pain receptors, causing a barrage of nociceptive afferent input [ 44 ], resulting in dysafferentation [ 45 ]. Being that postural control is largely dependant upon cervical joint mechanoreceptors and afferent input from ligament and musculotendinous sources [ 46 , 47 ], correcting the postural distortions responsible for this pathophysiologic process may be beneficial in patient populations, such as scoliosis, where postural control is significantly altered [ 48 ]. The effects of the loss of cervical and lumbar lordosis have been previously reported [ 19 , 35 - 37 ]. Rhee et al [ 49 ] noted that correction of the sagittal curves might be related to the long-term health of the spine in scoliosis management. Harrison et al [ 35 ] illustrated how a loss of the sagittal curve alters the mechanical properties of the spinal cord and nerve roots, which may change the firing patterns of involved neurons. Schafer illustrated how an increased demand is placed upon the cervical musculature when the cervical curve is straightened or reversed [ 50 ]. It is important that the cervical spine be in a normal structural alignment. A loss of the cervical lordosis and concomitant forward head posture may elicit the pelvo-ocular reflex, which causes an anterior pelvic translation to balance the head's center of gravity [ 51 ]. Wu et al [ 52 , 53 ] point out that in postural control, preference is given to the position of the head, neck, and trunk. Therefore, correction of the cervical spine becomes imperative so that the rest of the spine can be rehabilitated in relation to a normal reference point in space. Once the cervical and lumbar lordoses are corrected, coronal reduction of the scoliotic curvatures begins. Here this was accomplished by adding a shoulderweight to the right shoulder and a hipweight to the anterior right ilium and posterior left ilium. Wu and Essien [ 53 ] have previously reported the effects of adding external weight to the upper body via a shoulder weight. They identified predictable patterns in which the trunk would compensate for the amount and position of the weight. Wu and MacLeod [ 52 ] identified a shift in the center of mass toward the added weight when placed on the side of the pelvis. However, the trunk and head remained in the same position, while the pelvis and lower extremities shifted to counteract the weight while supporting the head and trunk [ 52 ]. In this protocol, we created an environment where external weight was added to the head, shoulder, and pelvic regions simultaneously. Knowing the predictable patterns of compensatory shifting to an altered center of gravity, we placed the headweight, shoulderweight, and hipweights in areas designed to reduce each patient's specific spinal distortion patterns. Learning a new motor coordination skill can be divided into 3 phases: cognitive, associative, and autonomous [ 54 ]. In the cognitive phase, the patient performs the motor task repetitively to learn until the task requirements are understood. As the patient progresses through the associative and autonomous phases, the task becomes easier to perform, and may ultimately be performed in a variety of practical contexts with decreased repetitions [ 54 ]. While Lantz et al [ 2 ] have shown that chiropractic management, consisting of a combination of manipulative procedures, electric stimulation, and orthotic inserts did not significantly reduce a scoliosis, this treatment does not incorporate these physiotherapeutic procedures. Instead, this treatment requires the use of specific rehabilitative equipment that theoretically recruits the use of head, neck, trunk, and extremity postural reflexes to create specific adaptation to an altered center of gravity and field of gaze. The study design used here does present specific limitations. Due to the lack of a control group, comparative data and conclusions cannot be made. Additionally, a retrospective design does not blind the practitioners to treatment. Although we attempted to select patient files at random from 3 separate spine clinics, nonrandomized sample populations such as ours do not necessarily reflect the potential outcomes in a general population. Therefore, future studies in this area should incorporate a control group and a randomized patient population. Follow-up studies should also focus on the potential long-term benefits of conservative scoliosis treatment, given the relative scarcity of biomedical literature available on long-term benefits from any scoliosis treatment. Conclusions Within the design limitations of the present study, the combined use of manipulative and neuromuscular rehabilitation seemed to reduce scoliotic curvatures in 19 subjects by an average of 17°. This reduction took place within a 4 to 6-week period. Although this treatment was not tested over the long term, the magnitude of the present results warrants further studies into its effectiveness. This treatment should also be tested on specific types of scoliosis in follow-up trials. A long-term investigation of this protocol is desirable. Competing interests This manuscript was submitted by Spinal Technologies, a BioMed Central institutional member. The rehabilitation equipment used in this study is patented by Burl R Pettibon, DC and Spinal Technologies. MWM is the Director of Research for the Pettibon Biomechanics Institute, and an active postgraduate instructor for Spinal Technologies. MWM does not receive monetary compensation for this position. Rather, he is granted funding from Spinal Technologies to obtain biomedical literature and statistician services. DW is a past postgraduate instructor for Spinal Technologies, founder and director of the CLEAR Institute, and CEO of the Flex Neck Company. GL is an active postgraduate instructor for Spinal Technologies. The authors receive lecture fees for each continuing education seminar conducted. All 3 authors maintain private chiropractic practices from where all of the patient files in this study were taken. None of the above companies donated, funded, or reimbursed any monies or equipment for this study. None of the authors have any ownership in Spinal Technologies or its subsidiary companies, and none will gain any financial interest as a result of this paper. Authors' contributions Each author worked on one-third of the patient population. The first author was responsible for collecting the data and putting our findings into written format. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC520751.xml |
553967 | Comparison of patterns of use, beliefs, and attitudes related to waterpipe between beginning and established smokers | Background To compare patterns of use, beliefs, and attitudes related to waterpipe smoking between university students (beginning smokers) and café customers (established smokers) in Aleppo Syria, in order to explore the evolution of this smoking method. Methods Two cross-sectional surveys were conducted among representative samples of university students (total 587, 48.4% men, mean age 22 years), and waterpipe users among cafe' customers (total 268, 60% men, mean age 30 years) in Aleppo, Syria. We used interviewer-administered questionnaire inquiring about pattern of waterpipe smoking (initiation, frequency), situational characteristics of use (partner, place, sharing), beliefs related to waterpipe smoking (harmful/addictive properties of waterpipe), attitudes related to waterpipe smoking (confidence in quitting, will to quit, motivation for quitting, past year quit attempt), and cigarette smoking. Results Daily and regular patterns of smoking become more prevalent with increased duration of smoking, but intermittent smoking remains the predominant pattern of waterpipe use. Women seem to be drawn later to the habit, which seem to escape the usual taboo against women's cigarette smoking. Patterns and context of waterpipe use tend to change with progress of the practice affecting frequency, setting, and sharing of waterpipe. Unlike beginners, established waterpipe smokers seem more smoking-method oriented, more hooked on the habit, less willing to quit, and less likely to foresee challenges to quitting. Conclusion Use patterns and attitudes related to waterpipe smoking evolve to accommodate the change in dependence and life circumstances of the smoker. Most of use features, beliefs, attitudes, as well as time-course seem unique to this smoking method requiring novel approach to intervention. | Background Waterpipe, is a generic name for tobacco use methods that share a common feature; passage of smoke through the water before inhalation. It is known under different names and shapes in different cultures and countries (e.g. shisha, narghile, hookah, hubble bubble) [ 1 ]. Although waterpipe smoking is centuries old, its use was declining until recent years when it witnessed a boom in popularity among Arab communities around the world [ 1 , 2 ]. Recent estimates show that this smoking habit has reached a quarter of some groups in the EMR, affecting youths and women among others [ 3 - 6 ]. Studies on the health effects of this smoking method are scanty, and often suffer from poor control of other confounding factors (e.g. cigarette smoking). We are just beginning to learn in a more methodological way the full spectrum of harmful effects of this smoking method and about shared and distinguishing features from cigarette smoking. For example, in regards to CO, one of the main cardiovascular risks of cigarette smoking, waterpipes are at least as efficient as cigarettes in delivering this toxic gas to smokers' lung [ 1 ]. On the other hand, tobacco smoke generated by this method seems to contain larger amounts of heavy metals such as arsenic, cobalt, chromium, and lead [ 7 ]. In addition, this smoking method is efficient in delivering nicotine to smokers, who tend to show signs of tobacco use dependence [ 8 , 9 ]. Because of the increasing trend of this smoking method, together with its health damaging and addictive potentials the need to develop ways to intervene with smokers and prevent initiation seems timely. Intervention efforts to deal with this emerging public health problem however, are hindered by the inadequacy of data characterizing various aspects related to patterns of waterpipe use and dependence [ 10 ]. We recently conducted a survey looking at prevalence, beliefs, and attitudes related to this smoking method among university students in Aleppo, Syria. Waterpipe was practiced by a quarter of male and 5% of female students and was characterized by intermittent and social use [ 3 ]. Also, smokers among university students seem to be at the initial stages of their waterpipe practice [ 3 ]. Informed by these results, a second survey was carried out among older, more established, smokers in cafés in Aleppo. In this report we compare the two groups in terms of pattern of use, beliefs, and attitudes related to waterpipe smoking in order to gain insights on the natural history of this smoking method and inform intervention development. Methods The methods of these surveys conducted in 2003 are described in details elsewhere [ 3 , 9 ]. Briefly, these surveys were conducted among representative samples of university students (total 587, 48.4% men, mean age 22 years, age range 18–30 years) of whom 86 (15%) were waterpipe smokers, and waterpipe users in coffee shops in Aleppo (total 268, 60% men, mean age 30 years, age range 18–68 years). The students survey was carried out at Aleppo University's dormitories (total 19), where four women's and four men's dormitories were selected randomly, and within each dormitory, a sampling frame was used to recruit about 75 participants. In the cafe' survey, 17 cafes were selected out of total 112 in Aleppo, and within each cafe' participants were recruited by random selection of pre-numbered waterpipes. Unlike the café survey, the university survey included both waterpipe smokers and non-smokers, and thus only waterpipe smokers (15%) among students were included in the current analysis (total 86, 82.5% men, mean age 22 years) (Table 1 ). Table 1 Basic indicators of the café survey population in comparison to University students University students (n= 86) n (%) Café customers (n = 268) n (%) Gender Male 71 (82.6) 161 (60,1) Female 15 (17.4) 107 (39.9) Residence City 43 (50.0) 264 (98,5) Country 43 (50.0) 4 (1.5) Religion Muslim 81 (94.2) 201 (75.3) Christian 5 (5.8) 66 (24.7) Marital status Married 1 (1.2) 128 (47.8) Single, divorced, or widowed 85 (98.8) 140 (52.2) Frequency of narghile use Daily 5 (5.8) 64 (24.0) Less than daily 81 (94.2) 203 (75.7) Mean ± SD Mean ± SD Age 22.3 ± 2.3 30.1 ± 10.2 Economic status (DI) 2.0 ± 0.7 1.2 ± 0.6 Number of years of education 14.7 ± 1.3 12.5 ± 3.6 The questionnaires used in these surveys were developed from standard instruments modified to suit waterpipe smoking [ 11 - 13 ]. The protocol and informed consent document were approved by the Institutional Review Boards at the Syrian Society Against Cancer and The University of Memphis. Generally, the two groups were compared on the following dimensions; 1- socio-demographics including age, gender, number of years of formal education, economic status assessed by the density index (DI = number of household members/number of rooms in the house), 2- smoking characteristics including frequency of use (daily, occasional), age of initiation of use, age of initiation of daily use, cigarette smoking status (past month cigarette smoking, ex-daily smoking for current non-smokers), 3- situational characteristics of waterpipe use including waterpipe smoking initiation (alone, with friends, with family), current waterpipe smoking (alone, with friends, with family), usual place of waterpipe smoking (open end question categorized later into home/dorms, cafes, other or no particular place), waterpipe sharing with others, usual sharing companion (as an open end question categorized later into friend, family), whether use is periodic/seasonal, the season of increased use (as an open end question categorized later into holidays, summer/spring, stress/exams, and other), 4- quit attitude (waterpipe, cigarette) including belief of own ability to quit, will to quit, past year quit attempt, quit motivation (open end question categorized later into health, cost, and other), perceived major challenge for quitting (open end question categorized into friends, addiction and habit, boredom and free time, no challenge, other or non specific answer), 5- perceived main health hazard of waterpipe smoking (open end question categorized into cardiovascular effects, respiratory effects, cancer, other bodily effects, none or don't know), harmful and addictive effects of cigarettes compared to waterpipe (categorized into cigarettes are more harmful, both methods are equally harmful, and waterpipe is more harmful, and the same categories were used for the comparison of addictiveness), and finally 6- family's attitude towards waterpipe and/or cigarette smoking categorized into (friendly, normal, not friendly), and smokers' perception of the extent they are hooked on waterpipe (categorized into not at all, somewhat, and very hooked) (tables 1 , 2 , 3 ). Participants were asked as well on the average time of each waterpipe smoking bout, and daily waterpipe smokers were asked about their average monthly expenditure on waterpipe smoking. Table 2 Waterpipe smoking patterns among café customers (established) compared to those of university students (beginners) Beginners (students, n = 86) n (%) Established (customers, n = 268) n (%) Frequency of waterpipe smoking Daily 5 (5.8) 64 (24.0%)* Occasionally 81 (94.2) 203 (76.0)* First experience with waterpipe smoking Alone 2 (2.3) 25 (9.3)* With friend 69 (80.2) 165 (61.6)* With family 15 (17.4) 75 (28.0) Current use of waterpipe Alone 1 (1.2) 31 (11.6)* With friends 69 (80.2) 175 (65.3)* With family 16 (18.6) 62 (23.2)* Place of usual waterpipe smoking Home 48 (55.8) 51 (19.2)* Café/restaurant 24 (27.9) 207 (77.8)* Other or no particular place 14 (16.3) 8 (3.0)* Share the same waterpipe 83 (96.5) 117 (43.8)* Individual most commonly waterpipe is shared with Friend 78 (94.0) 49 (41.9)* Family 5 (6.0) 66 (56.4)* Current cigarette smoking 41 (47.7) 71 (26.5)* Ex-cigarette smoking 8 (17.8) 24 (12.2) Seasonal increase in waterpipe smoking frequency 60 (69.8) 131 (48.9)* Period of increased waterpipe smoking frequency Holiday 15 (25.0) 25 (19.1) Summer 31 (51.7) 91 (69.5)* Stress/exams 12 (20.0) 5 (3.8)* Other or no specific answer 2 (3.3) 10 (7.7) Mean ± SD Mean ± SD Age of initiation of waterpipe smoking (years) 19.6 ± 2.6 24.3 ± 8.3* Age of initiation of daily waterpipe smoking (years) 21.8 ± 3.6 24.7 ± 8.8 Duration of waterpipe smoking (years) 2.7 ± 1.9 5.9 ± 6.2* Monthly cost of waterpipe smoking 680 ± 507 2366 ± 1918* Age of initiation of daily cigarette smoking (years) 18.8 ± 2.4 19.1 ± 4.1 * p < 0.05 using the Chi 2 test for dichotomous variables, and Student's t test or the Mann-Whitney test as appropriate for continuous variables (comparison is always between beginners and established smokers). Table 3 Attitudes and beliefs related to waterpipe quitting among café customers and university students Beginners (students, n = 86) n (%) Established (customers, n = 268) n (%) Belief can quit waterpipe any time 77 (89.5) 231 (86.5) Will to quit waterpipe 35 (40.7) 76 (28.4)* Last year quit attempt (waterpipe) 23 (65.7) 45 (59.2) Belief can quit cigarettes anytime 19 (46.3) 36 (50.7) Will to quit cigarettes 29 (70.7) 45 (63.4) Last year quit attempt (cigarettes) 26 (89.7) 32 (71.1) Main motivation for quitting waterpipe** Health 33 (91.6) 79 (103.8) Cost 3 (8.7) 3 (6.5) Other 3 (8.6) 4 (5.3) Main challenge for quitting waterpipe Friends 10 (28.6) 6 (7.9)* Addiction and habit 6 (17.1) 5 (6.6) Boredom and free time 3 (8.6) 7 (9.2) No challenge 13 (37.1) 47 (61.8)* Other or no specific 3 (8.6) 11 (14.5) Main health hazard of waterpipe Cardiovascular effects 5 (5.8) 25 (9.4) Respiratory effects 41 (47.7) 98 (36.7)* Cancer 25 (29.1) 96 (36.0) Other bodily effects 11 (12.7) 13 (4.8)* None, don't know, none specific 4 (4.7) 35 (13.0)* Belief about the addictive effects of waterpipe compared to cigarettes Cigarettes are more addictive 77 (89.5) 221 (82.8) Equally addictive 7 (8.1) 27 (10.1) Waterpipe is more addictive 2 (2.3) 19 (7.1) Belief about harmful effects of waterpipe compared to cigarettes Cigarettes are more harmful 32 (37.6) 124 (46.4) Equally harmful 11 (12.9) 49 (18.4) Waterpipe is more harmful 42 (49.4) 94 (35.2)* * p < 0.05 using the Chi 2 test ** Multiple responses were allowed (percentages add to more than 100) Statistical analysis Frequency analyses of main study indices were tabulated and differences between the two groups were assessed using the Chi 2 test for dichotomous variables. Measures of central tendency were calculated for continuous variables, expressed as mean ± standard deviation (SD), and were compared between the two groups using Student's t test or the Mann-Whitney test as appropriate. Spearman correlation coefficient was calculated for the relation between duration of waterpipe smoking and frequency of use (categorized as daily, weekly, monthly) in the café study. SPSS statistical software (release 11) was used for the analysis with p value <0.05 considered significant. Results Overall, the two samples combined show that daily waterpipe smoking was seen among 19.5% of participants (23.3% of smoking men and 12.3% of smoking women), while occasional smoking was seen among 80.2% of participants (76.3% of smoking men and 87.7% of smoking women). More than half of waterpipe users initiated use and currently use waterpipe with a friend (Table 2 ). About a quarter of waterpipe users (26.5%) smoke cigarettes as well. Duration of use and frequency of use (daily, weekly, monthly) were correlated (Spearman correlation coefficient 0.14, p = 0.02 ). Mean duration of waterpipe smoking session among café customers was 71.1 ± 35.8 minutes. Comparison of patterns of initiation and current use of waterpipe between café customers and university students is detailed in Table 2 . Mainly, café customers initiate waterpipe use and daily use later and have longer duration of waterpipe smoking compared to students. University students were more likely to start, currently smoke, and share the same waterpipe with their friends compared to their café counterparts (Table 2 ). Also, a marked difference between café customers and students is that cigarette smoking was more common among students waterpipe smokers than café ones ( p < 0.01 ), still ex-cigarette smoking did not differ between the two groups. Students as well were more likely to have a seasonal pattern of waterpipe use, which was more associated with exams and stress compared to café customers (Table 2 ). Quit attitude of café customers shows that the majority (86.5%) believe that they can quit waterpipe any time, but only a minority (28.4%) are interested in quitting, mainly out of health motivation (Table 3 ). In comparison, interest in quitting was higher among students and friends were more likely to be cited as the main challenge to quitting (Table 3 ). This was different from cigarette smokers, where only about a half of those (in both groups) believed they can quit anytime and the majority were willing to quit. Waterpipe users among café customers equally acknowledge respiratory disease and cancer as associated with waterpipe smoking. Attitude of families towards participants' waterpipe smoking for both café customers and students are depicted in figure 1 , showing more tolerance for women waterpipe smoking than for men's in general. Figure 2 illustrates participants (both groups) self perception of being hooked on waterpipe according to their frequency of smoking, were these dimensions were strongly associated among café customers but not students ( p < 0.001 , p = 0.1 , respectively). Figure 1 Family attitude towards participants (café customers, university students) waterpipe smoking showing more tolerant attitude for women's waterpipe smoking compared to men's. This unique observation marks the first incident in this region where a smoking method by women is more tolerated than by men. Figure 2 Self-perceived dependence among waterpipe smokers according to their frequency of waterpipe use (café customers compared to students). It shows that dependence is related to frequency of use, and that at comparable level of use, established smokers among café customers are more likely to perceive themselves as being hooked on the waterpipe compared to students. Finally, both mean education years and DI differed significantly between café customers (mean education years12.5 ± 3.6; mean DI 1.2 ± 0.6) and students (mean education years 14.7 ± 1.4, mean DI 2.0 ± 0.7) ( p < 0.01 for both). Discussion Obviously café customers are different from university students in many attributes that can reflect on waterpipe smoking and attitudes towards it. The longer duration of waterpipe smoking of café customers compared to students suggests that these two groups have been sampled at different stages of their waterpipe smoking practice. As such the two studied groups can provide an opportunity to look at features relevant to initiation and maintenance of waterpipe smoking. Such information can advance our understanding of this unique and increasingly popular smoking method and aid the design of intervention strategies. Generally, it looks that smoking patterns and quit attitude are mostly shaped by the stage of smoking, while beliefs and knowledge related to waterpipe tend to follow a more subtle roots likely reflecting cultural and socio-economic attributes. The first important characteristic that differentiates waterpipe use patterns in the two studied groups is that daily use was much higher among café customers practiced by about a quarter of waterpipe smokers. This is likely to reflect the time-course of waterpipe smoking, with regular use becomes more noticed as duration of smoking increases. In fact duration of waterpipe smoking was correlated to frequency of smoking (daily, weekly, and monthly) among café customers. Still, unlike cigarette smoking, which is practiced mainly on a daily basis in the two studied populations and in the Syrian society in general [ 13 ], intermittent use seems to be the predominant pattern of waterpipe smoking. This observation is supported by data from other countries in the EMR highlighting the unique social dimension of waterpipe use [ 4 , 14 ]. However, with average smoking bout of more than one hour and the potentially high levels of different toxicants in waterpipe smoke (arsenic, cobalt, chromium, lead) [ 7 ], even intermittent use can be associated with significant health hazards to smokers. Age of initiation of waterpipe differed significantly between men and women for both studied groups (analysis not shown). This may explain why the proportion of women among café customers is double that of students, although one should bear in mind the limitation of the samples for such a comparison. We have to be mindful as well that unlike cigarette smoking we are still at the expansion stage of the waterpipe epidemic, where people of different age groups are joining the new hype at relatively the same time period. An indication of this fact is that while age of initiation of waterpipe differed between the two studied groups, age of initiation of daily cigarette smoking was somehow identical. The tolerant attitude towards waterpipe smoking by women can aid its spread among them (Figure 1 ) [ 3 , 15 ]. Being perceived as closer to the local traditions, waterpipe smoking may escape the societal taboos of cigarette smoking by women in the EMR [ 15 , 16 ]. The social dimension, which is a salient feature of waterpipe smoking, seems also to differ between the two studied groups. So while students mainly initiate and currently practice the habit with their friends, more smokers among café customers are initiating or currently practicing the habit with family members or alone, despite being sampled in a social setting. It appears that the social context of waterpipe smoking changes with the change of both personal and smoking related characteristics (marriage, increased dependence). Our previous analysis of factors related to frequency of waterpipe use, as a marker of dependence, has lead us to suggest that as waterpipe smokers become more dependant the social pattern of smoking is gradually replaced with more individual one [ 9 ]. Sharing the same waterpipe and seasonal increase of waterpipe smoking, another important features of waterpipe smoking among students, were less noticed among café customers. It is likely that as waterpipe smokers move from the phase of experimentation to regular use they become more dependant, as such the habit is increasingly practiced in a way to satisfy dependence rather than as a pure social utility. Noticeably, at comparable level of use (daily) café customers perceive themselves to be more hooked on waterpipe than students (Figure 2 ). Waterpipe quit attitude and perceived challenges to quitting differed between café customers and university students, with café customers demonstrating lower interest in quitting and lower perception of challenges to quitting (Table 3 ). This can be related to the fact that while students waterpipe smokers come mainly from cigarette smokers (with higher cigarette smoking levels compared to their peers, 3), waterpipe smokers among café customers seem to be more smoking method-oriented (with lower cigarette smoking levels than that of adults in Syria, 13). The fact that ex-cigarette smoking did not differ between the two groups supports this argument. In addition, café customers were in disagreement with students about which smoking method is more harmful, favoring the waterpipe (less harmful). Finally, waterpipe smokers in cafés seem to be more hooked on this smoking method compared to students, and dependence among café customers was inversely related to their willingness to quit [ 17 ]. Thus, the difference of quit attitude between the two groups is likely to different stage and orientation along waterpipe smoking practice. The predominance of intermittent use among both groups on the other hand, may have created false perception of easiness of quitting waterpipe. This is likely to be a false perception as about two thirds of those willing to quit waterpipe in both groups made an unsuccessful quit attempt last year. In fact unlike the long held belief linking dependence and difficulty in quitting to daily/frequent use, new studies among youth cigarette smoking suggests that dependence and difficulty in quitting can develop at low levels of consumption [ 18 , 19 ]. As mentioned above a limitation of this study is that the café sample cannot be considered representative of adult smokers in the community. Also, the study consists of two cross sectional surveys in two populations presumably at two different stages of their waterpipe practice. Still, the two groups are not totally distinct from each other as they overlap in terms of age and smoking characteristics. However, as noted from the discussion we are mindful of these limitations throughout analyses and interpretation of the study's results. This study shows the predominance of intermittent use of this smoking method as well as its increasing popularity among women who are drawn to the habit relatively later than men. The social context of waterpipe smoking, which is a defining feature of this smoking method, tends to change to accommodate smoker's life and dependence progress. Our data suggest that at early stages of waterpipe use, users tend to come from cigarette smokers or people with liberal attitude towards smoking in general, but at more advanced stages of use smokers tend to be more smoking method-oriented and less keen on quitting. There is a general belief of easiness to quit waterpipe compared to cigarette, but this is likely to be an under-estimation as it is not translated into high success rate of quit attempts. Intervention or prevention strategies to curb this emerging epidemic should take into consideration the unique use features of this smoking method as well as smokers' perceptions and attitudes towards it. Competing interests The author(s) declare that they have no competing interests. Authors' contributions TA designed the study and wrote the first draft. KDW and TE participated in the study design and co-authored the manuscript. WM participated in the study design and wrote the final draft. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC553967.xml |
518999 | Identification of Birds through DNA Barcodes | Short DNA sequences from a standardized region of the genome provide a DNA barcode for identifying species. Compiling a public library of DNA barcodes linked to named specimens could provide a new master key for identifying species, one whose power will rise with increased taxon coverage and with faster, cheaper sequencing. Recent work suggests that sequence diversity in a 648-bp region of the mitochondrial gene, cytochrome c oxidase I (COI), might serve as a DNA barcode for the identification of animal species. This study tested the effectiveness of a COI barcode in discriminating bird species, one of the largest and best-studied vertebrate groups. We determined COI barcodes for 260 species of North American birds and found that distinguishing species was generally straightforward. All species had a different COI barcode(s), and the differences between closely related species were, on average, 18 times higher than the differences within species. Our results identified four probable new species of North American birds, suggesting that a global survey will lead to the recognition of many additional bird species. The finding of large COI sequence differences between, as compared to small differences within, species confirms the effectiveness of COI barcodes for the identification of bird species. This result plus those from other groups of animals imply that a standard screening threshold of sequence difference (10× average intraspecific difference) could speed the discovery of new animal species. The growing evidence for the effectiveness of DNA barcodes as a basis for species identification supports an international exercise that has recently begun to assemble a comprehensive library of COI sequences linked to named specimens. | Introduction The use of nucleotide sequence differences in a single gene to investigate evolutionary relationships was first widely applied by Carl Woese ( Woese and Fox 1977 ). He recognized that sequence differences in a conserved gene, ribosomal RNA, could be used to infer phylogenetic relationships. Sequence comparisons of rRNA from many different organisms led initially to recognition of the Archaea, and subsequently to a redrawing of the tree of life. More recently, the polymerase chain reaction has allowed sequence diversity in any gene to be examined. Genes that evolve slowly, like rRNA, often do not differ among closely related organisms, but they are indispensable in recovering ancient relationships, providing insights as far back as the origin of cellular life ( Woese 2000 ). On the other hand, genes that evolve rapidly may overwrite the traces of ancient affinities, but regularly reveal divergences between closely related species. Mitochondrial DNA (mtDNA) has been widely employed in phylogenetic studies of animals because it evolves much more rapidly than nuclear DNA, resulting in the accumulation of differences between closely related species ( Brown et al. 1979 ; Moore 1995 ; Mindell et al. 1997 ). In fact, the rapid pace of sequence change in mtDNA results in differences between populations that have only been separated for brief periods of time. John Avise was the first to recognize that sequence divergences in mtDNA provide a record of evolutionary history within species, thereby linking population genetics and systematics and establishing the field of phylogeography ( Avise et al. 1987 ). Avise and others also found that sister species usually show pronounced mtDNA divergences, and more generally that “biotic entities registered in mtDNA genealogies…and traditional taxonomic assignments tend to converge” ( Avise and Walker 1999 ). Although many species show phylogeographic subdivisions, these usually coalesce into single lineages “at distances much shorter than the internodal branch lengths of the species tree” ( Moore 1995 ). In other words, sequence divergences are much larger among species than within species, and thus mtDNA genealogies generally capture the biological discontinuities recognized by taxonomists as species. Taking advantage of this fact, taxonomic revisions at the species level now regularly include analysis of mtDNA divergences. For example, many newly recognized species of birds have been defined, in part, on the basis of divergences in their mtDNA (e.g., Avise and Zink 1988 ; Gill and Slikas 1992 ; Murray et al. 1994 ; AOU 1998 ; Banks et al. 2000 , 2002 , 2003 ). The general concordance of mtDNA trees with species trees implies that, rather than analyzing DNA from morphologically identified specimens, it could be used the other way around, namely to identify specimens by analyzing their DNA. Past applications of DNA-based species identification range from reconstructing food webs by identifying fragments in stomachs ( Symondson 2002 ) to recognizing products prepared from protected species ( Palumbi and Cipriano 1998 ) and resolving complexes of mosquitoes that transmit malaria and dengue fever ( Phuc et al. 2003 ). Despite such demonstrations, the lack of a lingua franca has limited the use of DNA as a general tool for species identifications. If a short region of mtDNA that consistently differentiated species could be found and accepted as a standard, a library of sequences linked to vouchered specimens would make this sequence an identifier for species, a “DNA barcode” ( Hebert et al. 2003a ). Recent work suggests that a 648-bp region of the mitochondrial gene, cytochrome c oxidase I (COI), might serve as a DNA barcode for the identification of animal species. This gene region is easily recovered and it provides good resolution, as evidenced by the fact that deep sequence divergences were the rule between 13,000 closely related pairs of animal species ( Hebert et al. 2003b ). The present study extends these earlier investigations by testing the correspondence between species boundaries signaled by COI barcodes and those established by prior taxonomic work. Such tests require the analysis of groups that have been studied intensively enough to create a firm system of binomials; birds satisfy this requirement. Although GenBank holds many bird sequences, these derive from varied gene regions while a test of species identification requires comparisons of sequences from a standard gene region across species. Accordingly, the barcode region of COI was sequenced in 260 of the 667 bird species that breed in North America ( AOU 1998 ). Results All 260 bird species had a different COI sequence(s); none was shared between species. COI sequences in the 130 species represented by two or more individuals were either identical or most similar to other sequences of the same species. Furthermore, with a few interesting exceptions discussed below, COI sequence differences between closely related species were far higher than differences within species (18-fold higher; average Kimura-2-parameter [K2P] differences between and within species, 7.93% and 0.43%, respectively) ( Figure 1 ). Figure 1 Comparison of Nucleotide Sequence Differences in COI among 260 Species of North American Birds Pairwise comparisons between 437 COI sequences are separated into three categories: differences between individuals in the same species, differences between individuals in the same genus (not including intraspecific differences), and differences between individuals in the same family (not including intraspecific or intrageneric differences). In most cases the neighbor-joining (NJ) tree showed shallow intraspecific and deep interspecific divergences ( Figure 2 ). However, in four exceptional cases, there were deep divergences within a species ( Tringa solitaria, Solitary Sandpiper; Sturnella magna, Eastern Meadowlark; Cisthorus palustris, Marsh Wren; and Vireo gilvus, Warbling Vireo). COI sequences in each of these polytypic species separated into pairs of divergent clusters in the NJ tree. The intraspecific K2P distances in these exceptional species were 3.7%–7.2%, 9- to 17-fold higher than the average distance ( Figures 2 , 3 , and S1 ). Figure 2 NJ Tree of COI Sequences from 30 Species in Family Scolapacidae (Sandpipers and Kin) The divergent pair of clustered sequences of Tringa solitaria is highlighted. An asterisk indicates a COI sequence from GenBank. Figure 3 NJ Tree of COI Sequences from 260 Species of North American Birds Intraspecific divergences were sampled in 130 species; these are marked in blue. Four species showed deep intraspecific divergence: (a) Sturnella magna, (b) Cistothorus palustris, (c) Vireo gilvus, and (d) Tringa solitaria. Higher-order classifications in families (gray) and orders (gold) are highlighted, and are labeled on the left and right of the figure, respectively. Gold numerals indicate the two species that appear as paraphyletic lineages at the family level: (1) Oenanthe oenanthe and (2) Hirundo rustica. Setting aside these polytypic species, the average intraspecific distance was very low, 0.27%, and the maximum average intraspecific difference was only 1.24%. Most congeneric species pairs showed divergences well above this value, but 13 species in four genera had interspecific distances that were below 1.25%. They included Larus argentatus, L. canus, L. delawarensis, L. glaucoides, L. hyperboreus, L. marinus, and L. thayeri (Herring Gull, Mew Gull, Ring-billed Gull, Iceland Gull, Glaucous Gull, Great Black-Backed Gull, and Thayer's Gull); Haematopus bachmani and H. palliatus (Black Oystercatcher and American Oystercatcher); Corvus brachyrhynchos and C. caurinus (American Crow and Northwestern Crow); and Anas platyrhynchos and A. rubripes (Mallard and American Black Duck) ( Figure S1 ). Although species were the focus of this study, we noted that the NJ tree of COI sequences generally matched avian classifications at higher levels, with most genera, families, and orders appearing as nested monophyletic lineages concordant with current taxonomy ( Figures 3 and S1 ). Discussion The simplest test of species identification by DNA barcode is whether any sequences are found in two species; none was in this study. Although sequences were not shared by species, sequence variation did occur in some species. Thus the second test is whether the differences within species are much less than those among species. In this study we found that COI differences among most of the 260 North American bird species far exceeded those within species. In order to conservatively test the effectiveness of COI barcodes as an identification tool, our sample must not have underestimated variability within species or have overestimated it among species. Our measures of intraspecific variation could be underestimates if members of a species show sequence divergence across their distribution that our study failed to adequately register. The two to three representatives of the 130 species used to examine this issue were collected from sites that were, on average, approximately 1,080 km apart, suggesting adequate representation of genetic diversity across their ranges. However, to further investigate this issue, we compared sequence differences within species to geographic distances between the collection points for their specimens and found these were unrelated ( Figure 4 ). Based on these results, high levels of intraspecific divergence in COI in North American birds appear uncommon, given that we analyzed 130 different species in a variety of orders. Our findings are supported by a review of 34 mostly North American birds which showed a similarly low average maximum intraspecific K2P divergence of mtDNA of 0.7% ( Moore 1995 ). Similarly, Weibel and Moore (2002) reported an average intraspecific divergence of 0.24% in their study of COI variation in woodpeckers. We conclude that our investigation has not underestimated intraspecific variation in any systematic fashion. Figure 4 Genetic Difference versus Geographic Distance For each same-species pair of specimens, the geographic distance between where specimens were collected is plotted against their COI divergence (K2P). On the other hand, our discovery of four polytypic species within a sample of 130 makes it likely there are other North American birds with divergent populations that may represent hidden species. Recent studies have identified marked mtDNA divergences within North American populations of Common Ravens ( Omland et al. 2000 ), Fox Sparrows ( Zink and Weckstein 2003 ), and Curve-billed Thrashers ( Zink and Blackwell-Rago 2000 ), leading to proposals to split each into two or more species. Species with Holarctic distributions are particularly good candidates for unrecognized species, and recent DNA and morphological investigations have led taxonomists to split several such species into two, including Wilson's and Common Snipes, American and Eurasian Three-toed Woodpeckers, and American and Water Pipits ( Zink et al. 1995 , 2002 ; Miller 1996 ; AOU 1998 ; Banks et al. 2000 , 2002 , 2003 ). Widespread application of COI barcodes across the global ranges of birds will undoubtedly lead to the recognition of further hidden species. Any critical test of the effectiveness of barcodes must also consider the possibility that our study has overestimated variability among species. We therefore looked at species individually, comparing their minimum distance to a congener with the maximum divergence within each species. This analysis included a number of well-recognized sibling species, including Calidris mauri and C. pusilla, Fraternicula arctica and F. corniculata, and Empidonax traillii and E. virescens. There were sufficient data to perform this analysis on three of the four polytypic species and on 70 of the 126 remaining species ( Figure 5 ). The average maximum K2P divergence within these 70 species was 0.29%, while the average minimum distance to a congener was 7.05% (24-fold higher), values comparable to those for the entire data set. Prior studies that looked exclusively at sister species of birds found an average K2P mtDNA distance of 5.1% in 35 pairs ( Klicka and Zink 1997 ) and 3.5% in 47 pairs ( Johns and Avise 1998 ). More generally, 98% of sister species pairs of vertebrates were observed to have K2P mtDNA divergences greater than 2% ( Johns and Avise 1998 ). Thus it appears that a COI barcode will enable the separation of most sister species of birds. Figure 5 Intraspecific Compared to Interspecific COI Distances (K2P) for Individual Species For each species in which this comparison was possible ( n = 73), maximum intraspecific variation is compared to minimum interspecific congeneric difference. For illustration purposes shown here, 2.0% is chosen as a cutoff between usual values for intra- and interspecific variation. This divides the graph into four quadrants that represent different categories of species: (I) Intraspecific distance, <2%; interspecific distance, >2%: concordant with current taxonomy; (II) Intraspecific distance, >2%; interspecific distance, >2%: probable composite species (i.e., candidate for taxonomic split); (III) Intraspecific distance, <2%; interspecific distance, <2%: recent divergence, hybridization, or synonymy; (IV) Intraspecific distance, >2%; interspecific distance, <2%: probable misidentification of specimen. There is a possibility that the North American bird fauna is not representative of the global situation. The recent and extensive glaciations in North America may have decreased within-species variability by inducing bottlenecks in population size or may have increased variation between species by pruning many sister taxa ( Avise and Walker 1998 ; Mila et al. 2000 ). This issue can only be resolved by evaluating the efficacy of barcodes in tropical and southern temperate faunas to ascertain if our results are general. We note that recent mtDNA studies in these settings have found both multiple sibling species in what were thought to be single species ( Ryan and Bloomer 1999 ) and geographically structured variation suggesting the presence of cryptic species ( Hackett and Rosenberg 1990 ; Bates et al. 1999 ). The diagnosis of species is particularly difficult when they are young. Moreover, hybridization is often common when the ranges of recently arisen species overlap, further complicating identifications. Such newly emerged species are sometimes referred to as superspecies ( Mayr and Short 1970 ), or species complexes, to indicate their close genetic similarity. For example, the white-headed gulls are thought to have diverged very recently, some less than 10,000 years ago ( Crochet et al. 2002 , 2003 ), and hybridization is common among many of them. It is thus not surprising that their COI barcodes and other gene loci are very similar. DNA barcodes can help to define the limits of such recently emerged species, but more gene loci need to be surveyed and more work is required to determine which analytical methods can best deduce species boundaries in such cases. The NJ method used here has the advantage of speed, and performs strongly when sequence divergences are low, so it is generally appropriate for recovering intra- and interspecies phylogeny. However, a library of COI barcodes linked to named specimens will provide the large data sets needed to test the efficacy of varied tree-building methods (for review, see Holder and Lewis 2003 ). Even between species that diverged long ago, hybridization will lead to shared or very similar sequences at COI and other gene loci. Because mitochondrial DNA is maternally inherited, a COI barcode will assign F 1 hybrids to the species of their female parent. Hybridization leading to the transfer of mtDNA from one species to another can result in a mtDNA tree that is incongruent with the species tree, but it will not necessarily prevent species from being distinguished, unless the mitochondrial transfer is so recent that their sequences have not diverged ( Moore 1995 ). However, recent hybridization will lead species to share COI barcodes, and we expect that more intensive study will reveal such shared sequences in species that are known to hybridize, such as the white-headed gulls ( Crochet et al. 2003 ) and Mallard/Black Ducks ( Ankney et al. 1986 ; Avise et al. 1990 ). In other cases, a lack of COI divergence may indicate that populations are part of a single species, helping to sort out misleading morphological classifications. For example, the blue and white morphs of Chen caerulescens, Snow Goose, were thought to be different species until recently ( Cooke et al. 1995 ). The close COI similarity of American and Black Oystercatchers revealed in this study is consistent with suggestions that these are allopatrically distributed color morphs of a single species ( Jehl 1985 ). Low COI divergences between American and Northwestern Crows similarly support earlier suggestions that these taxa are conspecific ( Sibley and Monroe 1990 ; Madge and Burn 1994 ). Just as COI similarities among species already questioned by taxonomists may reinforce these queries, deep COI divergences within species may reinforce suspicions of hidden diversity. For example, three of the four polytypic species in this study (Eastern Meadowlark, Marsh Wren, and Warbling Vireo) are split into two by some taxonomists ( Wells 1998 ), and the fourth, Solitary Sandpiper, contains two allopatric subspecies with morphological differences ( Godfrey 1976 ). In these cases, suspicions in the minds of taxonomists are reinforced by large COI divergences. If these species had not been the subject of prior scrutiny, COI barcoding would have flagged them as deserving of such attention. The importance of sampling multiple individuals within each species is highlighted by a recent review which found evidence of species-level paraphyly or polyphyly in 23% of 2,319 animal species, including 16.7% of 331 bird species ( Funk and Omland 2003 ). This review provides a clear discussion of possible causes (imperfect taxonomy, hybridization, incomplete lineage sorting) and indicates the need for the careful reexamination of current taxonomy and for the collection of genetic data across both geographic ranges and morphological variants. Barcoding, together with related developments in sequencing technology, is likely to provide an efficient approach to the assembly of such genetic data. We expect that the assembly of a comprehensive barcode library will help to initiate taxonomic investigations that will ultimately lead to the recognition of many new avian species. This process will begin with the discovery of novel COI barcodes. Some of these cases will simply represent the first barcode records for described but previously unanalyzed species, but taxonomic study will confirm that others derive from new species. We propose that specimens with barcodes diverging deeply from known taxa should be known by a “provisional species” designation that links them to the nearest established taxon. For example, the divergent clusters of Solitary Sandpiper specimens might be called T. solitaria PS-1 and T. solitaria PS-2, highlighting a need for further taxonomic study. What threshold might be appropriate for flagging genetically divergent specimens as provisional species? This threshold should certainly be high enough to separate only specimens that very likely belong to different species. Because patterns of intraspecific and interspecific variation in COI appear similar in various animal groups ( Grant and Bowen 1998 [sardines]; Hebert et al. 2003a [moths]; Hogg and Hebert 2004 [springtails]), we propose a standard sequence threshold: 10× the mean intraspecific variation for the group under study. If applied to the birds examined in this study (0.27% average intraspecific variation; 2.7% threshold), a 10× threshold would recognize over 90% of the 260 known species, as well as the four probable new species. As this result demonstrates, a threshold approach will overlook species with short evolutionary histories and those exposed to recent hybridization, but it will be a useful screening tool, especially for groups that have not received intensive taxonomic analysis. For 260 of the 667 bird species breeding in North America, our evidence shows that COI barcodes separate individuals into the categories that taxonomists call species. This adds to the evidence already in hand for insects and other arthropods that barcodes can be an efficient tool for species identification. Should future studies broaden this evidence, a comprehensive library of barcodes will make it easier to probe varied areas of avian biology. A DNA barcode will help, for example, when morphological diagnoses are difficult, as when identifying remnants (including eggs, nestlings, and adults) in the stomachs of predators. A DNA barcode could similarly identify fragments of birds that strike aircraft ( Dove 2000 ) and recognize carcasses of protected or regulated species ( Guglich et al. 1994 ). DNA barcodes could also reveal the species of avian blood in mosquitoes carrying West Nile virus ( Michael et al. 2001 ; Lee et al. 2002 ), help experts distinguish morphologically similar juveniles or nonbreeding adults in banding work, and allow expanded nonlethal study of endangered or threatened populations. The two essential components for an effective DNA barcode system (and thus a new master key to the encyclopedia of life [ Wilson 2003 ]) are standardization on a uniform barcode sequence, such as COI, and a library of sequences linked to named voucher specimens. The present study provides an initial set of COI barcodes for about 40% of North American birds. More detailed sampling of COI sequences is needed for these species, and barcodes need to be gathered for the remaining North American birds and for those in other geographic regions. This work could represent a first step toward a DNA barcode system for all animal and plant life, an initiative with potentially widespread scientific and practical benefits ( Stoeckle 2003 ; Wilson 2003 ; Blaxter 2004 ; Janzen 2004 ). Materials and Methods Existing data can only yield limited new insights into the effectiveness of a DNA-based identification system for birds. Two mitochondrial genes, cyt b and COI, are rivals for the largest number of animal sequence records greater than 600 bp in GenBank (4,791 and 3,009 species, respectively). However, COI coverage for birds is modest; 173 species share COI sequences with 600-bp overlap. As these records derive from a global avifauna of 10,000 species, they provide a limited basis to evaluate the utility of a COI-based identification system for any continental fauna, impelling us to gather new sequences. We employed a stratified sampling design to gain an overview of the patterns of COI sequence divergence among North American birds. The initial level of sampling examined a single individual from each of 260 species to ascertain COI divergences among species. These species were selected on the basis of accessibility without regard to known taxonomic issues. The second level of sampling examined one to three additional individuals from 130 of these species to provide a general sense of intraspecific sequence divergences, as well as a preliminary indication of variation in each species. When possible, these individuals were obtained from widely separated localities in North America. The third level of our analysis involved sequencing four to eight more individuals for the few species where the second level detected more than 2% sequence divergence among individuals. Our studies examined specimens collected over the last 20 years; 98% were obtained from the tissue bank at the Royal Ontario Museum, Toronto, Canada. Collection localities and other specimen information are available in the “Birds of North America” file in the Completed Projects section of the Barcode of Life website ( http://www.barcodinglife.com ). Taxonomic assignments follow the latest North American checklist ( AOU 1998 ) and its recent supplements ( Banks et al. 2000 , 2002 , 2003 ). Mitochondrial pseudogenes can complicate PCR-based studies of mitochondrial gene diversity ( Bensasson et al. 2001 ; Thalmann et al. 2004 ). We used protocols to reduce pseudogene impacts that included extracting DNA from tissues rich in mitochondria ( Sorenson and Quinn 1998 ), employing primers with high universality ( Sorenson and Quinn 1998 ), and amplifying a relatively long PCR product because most pseudogenes are short ( Pereira and Baker 2004 ). DNA extracts were prepared from small samples of muscle using the GeneElute DNA miniprep Kit (Sigma, St. Louis, Missouri, United States), following the manufacturer's protocols. DNA extracts were resuspended in 10 μl of H 2 O, and a 749-bp region near the 5′ terminus of the COI gene was amplified using primers (BirdF1-TTCTCCAACCACAAAGACATTGGCAC and BirdR1-ACGTGGGAGATAATTCCAAATCCTG). In cases where this primer pair failed, an alternate reverse primer (BirdR2-ACTACATGTGAGATGATTCCGAATCCAG) was generally combined with BirdF1 to generate a 751-bp product, but a third reverse primer (BirdR3-AGGAGTTTGCTAGTACGATGCC) was used for two species of Falco. The 50-μl PCR reaction mixes included 40 μl of ultrapure water, 1.0 U of Taq polymerase, 2.5 μl of MgCl 2 , 4.5 μl of 10× PCR buffer, 0.5 μl of each primer (0.1 mM), 0.25 μl of each dNTP (0.05 mM), and 0.5–3.0 μl of DNA. The amplification regime consisted of 1 min at 94 °C followed by 5 cycles of 1 min at 94 °C, 1.5 min at 45 °C, and 1.5 min at 72 °C, followed in turn by 30 cycles of 1 min at 4 °C, 1.5 min at 51 °C, and 1.5 min at 72 °C, and a final 5 min at 72 °C. PCR products were visualized in a 1.2% agarose gel. All PCR reactions that generated a single, circa 750-bp, product were then cycle sequenced, while gel purification was used to recover the target gene product in cases where more than one band was present. Sequencing reactions, carried out using Big Dye v3.1 and the BirdF1 primer, were analyzed on an ABI 377 sequencer. The electropherogram and sequence for each specimen are in the “Birds of North America” file, but all sequences have also been deposited in GenBank (see Supporting Information ). COI sequences were recovered from all 260 bird species and did not contain insertions, deletions, nonsense, or stop codons, supporting the absence of nuclear pseudogene amplification ( Pereira and Baker 2004 ). In addition to 429 newly collected sequences, nine GenBank sequences from five species were included (these were the only full-length COI sequences corresponding to species in this study). Sequence divergences were calculated using the K2P distance model ( Kimura 1980 ). A NJ tree of K2P distances was created to provide a graphic representation of the patterning of divergences among species ( Saitou and Nei 1987 ). Supporting Information Figure S1 Birds Appendix Complete NJ tree based on K2P distances at COI for 437 sequences from 260 species of North American birds. Entries marked with an asterisk represent COI sequences from GenBank. (100 KB PDF). Click here for additional data file. Accession Numbers Sequences described in Materials and Methods have been deposited in GenBank under accession numbers AY666171 to AY666596. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC518999.xml |
532396 | When Immune Defenses Turn Traitor | null | When pathogens enter your body, most wind up engulfed (by phagocytes), poisoned (by stomach acids), or flushed out of your system. These defenses kick in when a bacterium's toxic secretions bind to cellular receptors that trigger a chain of events ending in an innate response. A wide variety of cells and chemicals, including phagocytes and cytokines, initiate the inflammatory reaction that redirects innate defenses in the bloodstream to the infected site. If all goes well, the innate system successfully contains and eliminates a pathogen at the site of infection. But if infection persists, so will the inflammatory response, and these first responders turn traitor. It's thought that the immune response rather than the bacteria, for example, causes diarrhea in food poisoning. And when infection leads to death—as happens in septic shock—the immune response may be just as culpable as the infectious microbe. To better understand the interplay between pathogen and host in the onset of infection and disease, David Schneider and colleagues turned to the genetically compliant fruitfly Drosophila and the food-borne pathogen Salmonella . Working with mutant strains of both fly and bacteria, the authors identified genes important to the development of infection and disease and showed that the host's reaction can indeed be lethal. Pathogens possess various means to infect their host, unleashing toxins and secreting molecules that enhance virulence by breaching cell membranes and altering the intracellular environment. Fruitflies combat these intrusions with various innate responses, including phagocytosis. In this study, the authors investigated how Drosophila phagocytes find, engulf, and kill invading microbes and then alert the rest of the immune system—and how Salmonella circumvents these defenses to initiate disease. Adult fruit fly infected with flourescent Salmonella Schneider and colleagues used a Salmonella strain ( S. tyhphimurium ) that produces two pathogenicity complexes, called type III secretion apparatuses, which shuttle virulence molecules through the host's cell membrane and into its cytoplasm. One complex, SPI1, facilitates cell entry while the other, SPI2, retools the intracellular environment to suit bacterial growth. The authors created a series of less virulent Salmonella strains and examined their effects on wild-type (nonmutant) flies. In addition, they looked at infections of wild-type Salmonella in flies carrying mutations in two critical immune response pathways (called Toll and imd, for immune deficiency). Imd mutants infected with nonmutant Salmonella died much faster than the Toll mutants and had far more bacteria in their blood. When Schneider and colleagues correlated Salmonella population numbers with fly fate, they discovered something surprising. In other bacteria models, flies die when bacterial pathogen numbers reach a critical mass. Here, Salmonella populations hit a ceiling and the flies died with comparatively few bacteria. In flies infected with Salmonella strains lacking either one or the other virulence complex, the flies survived despite increased bacterial growth. The authors explain this surprising finding with a model in which the fly's immune response produces substances that ultimately engineer its own destruction. Their model is supported by experiments showing that flies carrying a mutation in the eiger gene—a homolog of the human TNF gene—live longer with Salmonella infections. This is the same phenomenon seen in TNF-induced septic shock, when patients die as much from their immune system's response to infection as from the bacteria itself. Since many of the proteins involved in the Drosophila imd pathway have counterparts in the mammalian antibacterial immune response, the model described here can help identify the genetic agents of metabolic collapse associated with bacterial infections in humans. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC532396.xml |
549081 | Associations between diet and disease activity in ulcerative colitis patients using a novel method of data analysis | Background The relapsing nature and varying geographical prevalence of ulcerative colitis (UC) implicates environmental factors such as diet in its aetiology. Methods In order to determine which foods might be related to disease activity in UC a new method of dietary analysis was developed and applied. Eighty-one UC patients were recruited at all stages of the disease process. Following completion of a 7 d diet diary, clinical assessment including a sigmoidoscopic examination (scale 0 (normal mucosa) to 6 (very active disease)) was conducted. Food weights for each person were adjusted (divided) by the person's calorific intake for the week. Each food consumed was given a food sigmoidoscopy score (FSS) calculated by summing the products of the (adjusted) weight of food consumed and sigmoidoscopy score for each patient and occurrence of food and dividing by the total (adjusted) weight of the food consumed by all 81 patients. Thus, foods eaten in large quantities by patients with very active disease have high FSSs and vice versa. Foods consumed by <10 people or weighing <1 kg for the whole group were excluded, leaving 75 foods. Results High FSS foods were characterized by high levels of the anti-thiamin additive sulfite (Mann-Whitney, p < 0.001), i.e. bitter, white wine, burgers, soft drinks from concentrates, sausages, lager and red wine. Caffeine also has anti-thiamin properties and decaffeinated coffee was associated with a better clinical state than the caffeine containing version. Beneficial foods (average intake per week) included pork (210 g), breakfast cereals (200 g), lettuce (110 g), apples and pears (390 g), milk (1250 ml), melon (350 g), bananas (350 g), bacon (120 g), beef and beef products (500 g), tomatoes (240 g), soup (700 g), citrus fruits (300 g), fish (290 g), yogurt (410 g), cheese (110 g), potatoes (710 g) and legumes (120 g). Conclusions The dietary analysis method described provides a new tool for establishing relationships between diet and disease and indicates a potentially therapeutic diet for UC. | Background Ulcerative colitis (UC) is a chronic, relapsing mucosal disorder that extends in continuous fashion proximally from the rectum and is limited to the colon. The aetiology of UC includes a genetic component possibly involving an abnormal cell-mediated immune response to commensal enteric bacteria in the large intestine. The relapse/remission pattern of the disorder and substrate driven nature of microbial metabolism in the large bowel implicate environmental factors such as diet. Apart from nutritional repletion, dietary measures do not play a role in the management of UC. Nonetheless, attempts to link the cause of UC with specific foods date back at least 50 years[ 1 ]. Many foods or food groups have been related to UC (table 1 – see additional file 1 ) [ 2 - 13 ] including sugar, eggs, soft drinks, fruit and vegetables, protein, carbohydrate and fat. However none have been proven to be of significant benefit or to contribute to the cause of UC. This may partly be because both the assessment of disease activity in UC and dietary intake are difficult to measure, or because the actual dietary component that is key to this relationship has not been measured. It has been proposed that sulfide, produced in the large bowel from either amino acid fermentation or sulfate reduction, may be a triggering factor in the inflammatory process of UC [ 14 - 16 ]. Recently, in a prospective dietary study where foods rich in sulfur compounds were quantitated, evidence that sulfur compounds may increase the likelihood of subsequent relapse in UC was found[ 17 ]. The main source of inorganic sulfur, predominantly sulfate, in the diet are the S (IV) family of additives; the sulfiting agents. Sulfites have been used as food preservatives since the 17th century and are amongst the most widely accepted and versatile of additives. Sulfiting agents, denoted by E220–228 in Europe and generally recognized as safe (GRAS) substances in the USA, include sulfur dioxide, hydrogen sulfites, sulfites and metabisulfites. Sulfiting agents are cheap, easy to use and extremely effective at preventing microbial growth and reducing spoilage[ 18 ]. They serve as antioxidants, inhibit enzymatic and non-enzymatic browning reactions and act as a texture modifier in biscuit dough. Sulfites improve color extraction from, and stabilization of grape must in winemaking and preserve lobsters and shrimps from discoloration during iced storage. However, there are some problems with sulfite use[ 19 , 20 ]. In the early 1980s ingestion or inhalation of sulfites was shown to cause bronchospasm in about 5 % of asthmatics. Sulfite sensitivity can pose a particular threat in the workplace where sulfiting agents are used, but may also occur with ingestion of sulfited foods such as potato products and wine. In addition, skin sensitivity has been reported and there are anti-nutritional effects particularly to thiamin which is readily cleaved by the sulfite ion[ 21 ]. The mechanism involves an initial nucleophilic attack to the methylene carbon activated by the positive charge on nitrogen, the reaction rate peaking between pH 5 and 6[ 18 ]. As a result of this anti-nutritional effect the GRAS status for sulfites was reviewed in the USA and in 1986 the use of sulfites in fresh and frozen fruit and vegetables revoked and a declaration on the label required[ 22 , 23 ]. Earlier (in the USA) their use in meat had been prohibited, because these foods are an important source of thiamin. A study of diet and disease activity in UC using a 7 d dietary diary, a full assessment of disease activity and a method of dietary data analysis that allows trends in food consumption not apparent using customary dietary software was therefore undertaken. Methods Subjects Eighty-one UC patients were recruited and informed consent obtained. Ethical permission was granted by Tayside Committee on Medical Ethics, Dundee, UK (ref 007/00). As it was important to have a range of disease activities present, recruitment included patients at all stages of the disease. Patients were excluded if clinical examination or histology indicated Crohn's disease or indeterminate colitis, if there was a positive stool culture for pathogens or if the patient had antibiotic treatment within 3 months preceding the start of the study. Dietary Assessment All the UC patients were asked to complete a 7 d diet diary[ 24 ]. The diet diary used has been validated for use in the European Prospective Investigation into Cancer study (EPIC). Following completion of the diet diary, subjects attended the research clinic and a full clinical assessment (see below) was carried out. The time interval between the first day of the diary and the clinical visit was on average 28 d. Thus the dietary data is prospective. 7d diet diaries were coded and analyzed using Tinuviel, WISP v3.0 nutritional analysis software (Warrington, UK). Due to the variation in the sulfiting protocols and widespread use of sulfiting agents, current tables of food composition do not contain inorganic sulfur values and cannot be used to quantify intake. Instead of quantitating the intake of particular dietary components, foods and food groups were assessed in their entirety using the method described in the dietary data analysis section (below). Clinical Assessment Clinical assessment included history, physical examination and global clinical grading, plus full blood count, liver function tests and inflammatory markers. Patients were examined by rigid sigmoidoscopy or flexisigmoidoscopy and graded on a scale 0–6 (integers and half integers used) according to the macroscopic appearances of the rectal mucosa at a distance 5–10 cm from the anal verge[ 25 ]. The clinical assessment of disease activity was confirmed in each case by histological examination, by a single histopathologist blinded to the clinical details, of a rectal biopsy taken from the posterior rectal wall 5–10 cm from the anal verge[ 26 ]. A simple clinical colitis score was assigned to patients on each visit following Walmsley's scoring system[ 27 ], together with blood parameters of disease severity (Hb, plasma viscosity, CRP, serum albumin). Dietary Data Analysis Patterns of dietary intake associated with disease activity became apparent through the study of the dietary diaries, e.g. high intakes of sulfite containing foods coupled with a modern processed, convenience diet was associated with a high sigmoidoscopy score. Traditional dietary coding (WISP) did not show any such clear associations between micro or macro nutrient intake and sigmoidoscopy score. Traditional dietary analysis was therefore thought to be missing important patterns in dietary data and a new method of dietary assessment was subsequently developed. This new method used the following procedure. To calculate the association of a particular food with clinical score, each food or food group consumed was given a food sigmoidoscopy score (FSS) calculated by summing the products of food weight and sigmoidoscopy score for each occurrence of the food or food group and dividing by the total weight of the food or food group contained in all diaries. In order for each diary to make equal contributions to the FSSs, the weight of each food was adjusted using the calorific intake for each person. This procedure was carried out separately for every food item recorded in the 7 d diet diaries but is explained below using the example of red wine. Red wine score = (Σv(i)s(i))/Σv(i) for i = 1 to 81 equation 1. Where: - i is the 7 d dietary diary number (n = 81). v(i) is the volume (divided by calorific intake for patient (i) of red wine recorded in 7 d dietary diary i. s(i) is the sigmoidoscopy score associated with 7 d dietary diary (i). Thus foods eaten in large quantities by patients with high levels of disease activity will have high scores and vice versa. The denominator in the above equation is the total volume of the food in question from all diaries (corrected for calorific intakes) so the food scores can be equated with the effect of a typical portion of the food in question on the sigmoidoscopy scores of the patients. This procedure is repeated for every food item. Foods or food groups were excluded from the analysis if 10 or fewer people consumed them or if they made up less than 1 kg of the total intake of the entire population. The decision as to where food group boundaries lay was made depending on the size of the group and whether the differences between the foods were considered important for this study. Statistics and Data Handling Dietary data was exported from WISP to Microsoft EXCEL 98 (Macintosh version, 1998). A worksheet containing the core headings; Patient ID, food description, weight and patient sigmoidoscopy score was completed. The data was then sorted by food description and each food copied to a separate EXCEL file. Equation 1 was then used to calculate food sigmoidoscopy scores for each food in a manner similar to the example in table 2 (see additional file 2 ). Correlation values for scatter plots were obtained using the linear regression function in EXCEL. The equation t = r √((n-2)/(1-r 2 )) combined with t tables provided corresponding significance levels. Results Of the 81 patients recruited 43 were male and 38 female. The average age (range) of the males and females were respectively 53 (26–78) y and 47 (19–74). The distribution of sigmoidoscopy scores is shown in fig 1 . One third of the patients had sigmoidoscopy scores of 0, 0.5 or 1. The mean sigmoidoscopy score for all 81 patients was 2.09. The correlation between the clinical activity indexes and sigmoidoscopy scores was r 2 = 0.25 (n = 81). Table 3 (see Additional file 3 ) shows the foods and food groups with associated sigmoidoscopy scores and average portion sizes. In total 75 foods (or food groups) were given FSS scores. The higher the FSS value the greater the association with disease activity and vice versa. The total weight of foods in all diaries was 1,681 kg. The average food sigmoidoscopy score (i.e. a food sigmoidoscopy score calculated for the entire dietary intake data set was 2.127). Foods excluded from the FSS table (Table 3), by virtue of contributing <1 kg or being consumed by <10 people, made up 8 % of the total weight of all foods and had a score slightly lower (2.001) than that of an average food (2.127). Standard errors are not quoted for the food scores as the data used to generate them (weight * sigmoidoscopy score) was not normally distributed due to the number of sigmoidoscopy scores of 0. The dietary diaries were assessed for completeness by comparing calorific intakes with expected values for the sexes. Expected (calculated from dietary reference tables using age and sex)[ 28 ] versus actual values for men and women were respectively 2481 kcal/d versus 2326 kcal/d and 1925 kcal/d versus 1887 kcal/d. Foods for which regulations exist in the EU permitting sulfite addition are shown in table 4 (see Additional file 4 ) [ 29 ]. Typically a manufacturer will add sulfite up to the maximum permitted level in order to achieve the longest shelf life for the product. A report on sulfite usage in the UK was produced in 2001[ 30 ]. Sweet wines, langoustines (prawns), dehydrated potatoes and dried fruit were not given FSS scores because their data quantity fell below the <10 people or <1 kg rule. Soft drinks were split into those known to contain sulfite (drinks made from fruit squash concentrates and lucozade) and the rest. In terms of intake (portion size*sulfite concentration), for this population, the major sources of sulfite (FSS, FSS table position) were bitter beer (3.91, 75), white wine (2.87, 73), burgers (2.84, 72), soft drink concentrates (2.79, 70), sausages (2.68, 68), lager (2.47, 64) and red wine (2.00, 29). A Mann-Whitney test on the FSS positions of these foods gave a significance of p < 0.001. The sulfite-containing, alcoholic beverages; wines and beers, were associated with increased UC disease activity, but spirits were not, which suggests a role for sulfite rather than alcohol in the disease process. A plot of alcohol consumption from wine and beer against sigmoidoscopy score revealed a significant positive correlation (n = 81, r 2 = 0.07, p < 0.02). Decaffeinated coffee appeared better for the UC patient than the caffeine-containing counterpart. Decaffeinated tea is not shown on table 3 because it was only drunk by 9 people but had a FSS of 1.71 versus 2.01 for the caffeine-containing product. Whole fruit consumption appeared better than the corresponding juice (e.g. fruit juice scored 2.43 compared to citrus fruits at 1.96 and apples at 1.67). An average thiamin concentration (mg / 100 g) for each food or food group is also shown in table 3. There is a significant correlation (p < 0.005) between this thiamin value and the food's sigmoidoscopy score. Discussion Ulcerative colitis is considered to have a genetic component. Twin studies[ 31 ] have shown a 10% concordance of UC in monozygotic and 3% in dizygotic twins suggesting about 90% environmental and 10% genetic contributions. The pool of genetically susceptible individuals is therefore at least 10 times greater than those diagnosed with the condition. A failure to date in identifying the gene(s) responsible points to a complicated genetic component featuring multiple polymorphisms. The first acute episode of UC must disrupt either, the ecology of, or the sensitivity and selectivity of the immune system to, the commensal enteric microflora sufficiently to cause the chronic condition. More extreme versions of the environmental conditions that lead to subsequent relapses could conceivably lead to the first acute episode. Of all the dietary components studied in relation to UC risk and disease severity, milk has probably received the most attention. Andreson[ 1 ] was the first to postulate that food allergy was the cause of UC in two-thirds of his patients, and by the use of elimination diets claimed to identify the offending food and remove it. In Andreson's experience, the most common provoking antigen was cow's milk. His views were confirmed by Rowe[ 32 ] and later by Truelove[ 33 ]. They all postulated that milk protein sensitivity was an aggravating cause of disease in up to 5% of colitic patients, who benefited from a milk-free diet. While able to demonstrate circulating antibodies to milk proteins more frequently and in higher titer than in matched controls, they were unable to correlate the occurrence and titer of these antibodies with the extent, severity, or duration of colitis, or with the response to a milk-free diet. Mishkin[ 34 ] concluded, in a review of the subject, that IBD patients avoid dairy products to a much greater extent than the prevalence of lactose malabsorption and/or milk intolerance in this population group would justify. This observation was probably due to the incorrect perceptions of patients and arbitrary advice of physicians and authors of popular diet books. In order to ascertain whether dietary antigens may sustain the mucosal inflammatory response, two prospective controlled trials have investigated the effectiveness of bowel rest and total parenteral nutrition as primary therapy in the management of acute UC[ 35 , 36 ]. Neither study found any benefit over conventional corticosteroid treatment alone and so the possibility of a dietary antigen driving the chronicity of the disease seems unlikely. These results are in agreement with work demonstrating[ 37 ] that a split ileostomy is of little benefit in the management of UC, but the latter observations may have been confounded by the development of diversion colitis[ 38 ]. The dietary analysis procedure proposed here has the potential to highlight trends in dietary data that would not be apparent using traditional dietary analysis software and could be useful in the study of other diseases with dietary associations. This system would highlight any possible dietary factors both positive and negative, not just sulfite. The proposed method is less reductionist than traditional coding as it assesses the risk of each food item or group rather than the risk from the foods' (quantitated) constituents. Part of the power of this study derives from the availability of a sigmoidoscopic grading (0–6) of the severity and extent of the disease. This grading provides the statistical variable that is normally obtained from a non-UC control group. Other alternative systems for analysis of disease risk for dietary components are; the use of disease occurrence odds ratios between the top and bottom quartiles of intakes, and assessing the correlation coefficients between disease activity and intakes. The odds ratio method loses data and data accuracy by characterizing intakes as high, high middle, low middle and low and then discarding the middle two quartiles. The correlation method is dependent on spread. The proposed system has neither of these disadvantages. The food sigmoidoscopy score calculation does rely on the assumption that the sigmoidoscopy score is an approximately linear scale, i.e. a sigmoidoscopy score of 6 is caused by the consumption of a double portion of a harmful food item of sigmoidoscopy score 3. This could be argued to be reasonable. Both the sigmoidoscopy grading and dietary analysis method are validated methodologies. The food sigmoidoscopy score is simply a mathematical function of these two variables. As all data is transformed according to the same simple rules any statistical treatment of the results is as valid as statistical treatment of the raw data. Whilst clinical activity indices were used to generate analogous scores to the food sigmoidoscopy scores, the results from these measurements are not included in this paper. Clinical activity index involves subjective measurements such as a feeling of well being. Thus, the food orders generated by these measurements were not thought to be as accurate as those generated by the sigmoidoscopy scores. The consensus of previous studies on diet and UC pointed to the modern, processed, highly refined, Western diets as being damaging. The results presented here linking diet with disease activity are broadly in agreement with this. Additionally they propose a new risk factor for UC, namely intake of sulfited foods. The involvement of diet in UC is controversial. Differences in dietary intake between patients and controls could be a result of changes in diet brought on by the symptoms of the disease process[ 4 ]. While this explanation is possible it does not seem likely that patients would increase their beer and wine intake as a consequence of feeling unwell. The relationship between sulfite intake and sigmoidoscopy score in this study was extremely strong and therefore an explanation for why sulfite should be a risk factor for UC is required. Sulfite has a number of effects that may be relevant to this discussion. Sulfite may be important because it is a precursor of sulfate. Sulfate can potentially be reduced to sulfide by sulfate reducing bacteria in the colon. Sulfide is a plausible metabolic toxin in UC. Supplementing patients with sulfate decreases the microbial incorporation of hydrogen into methane (as measured by breath methane) and increases the in vitro sulfide production rate of feces[ 39 ]. The end metabolic product of both sulfite and protein is sulfate. Sulfate from both sources can be reduced to sulfide in the gut. The absence of a significant relationship between protein intake and disease activity in this study does not support a mechanism for UC that involves a common pathway for sulfite and protein. Alternatively, the relevance of sulfite to UC may be because of its ability to degrade thiamin (particularly at colonic pH). Thiamin deficiency manifests itself in the nervous and cardiovascular systems. It is unlikely that it is the status of the patient that is important, but rather the amount of thiamin available to the gut microflora. An example of the importance of thiamin to the gut microflora is the requirement of the probiotic bacteria, lactobacilli, for thiamin. Thiamin status is influenced by a number of factors. Firstly, thiamin intake; in foods such as pork, fortified cereals and legumes which are good sources of thiamin, intakes were associated with improved clinical state. Traditional dietary analysis did not reveal a significant correlation between thiamin intake and sigmoidoscopy scores though no allowance is made in dietary coding software for the reduction in thiamin content caused by sulfite usage. Secondly, carbohydrate intake; Elmadfa et al . demonstrated that the thiamin status of adult humans depends on carbohydrate intake[ 40 ]. Carbohydrate (and sugar) intakes have previously been associated with UC relapse (table 1). Finally, thiamin status can be affected by caffeine's anti-thiaminergic properties. For both coffee and tea intake, the decaffeinated version was associated with better clinical state. However, there was a sub group (n = 8) of this population who recorded an intake of either vitamin B complex or multivitamins. This sub group did not have a mean sigmoidoscopy score significantly lower that the general UC population. It is likely that vitamin B1 is a factor in the disease process but not the only nutritional one. An additional possible interpretation for the experimentally determined food order is the carbohydrate nature and content of the foods. Carbohydrates, such as the α-amylase resistant starch (RS) and prebiotics, escape digestion in the small intestine and provide an energy substrate for the colonic microflora. Both prebiotics (found in chicory, legumes, artichokes alliums, and in small amounts in cereals) and resistant starch (potatoes, bananas, lentils and legumes) have been hypothesised to improve the colonic health of the host. For RS, resistance to digestion is a function of the morphology of the starch granules and their crystalline organisation, which is determined by the botanical source of the starch and the processing it has undergone before being eaten[ 41 ]. Prebiotics are non-digestible carbohydrates that selectively stimulate the growth of lactobacilli and bifidobacteria with benefit to health. Prebiotics are mainly fructose and galactose polymers with a degree polymerisation of between 2 and 60. Of the prebiotic sources; chicory and artichokes were not found in typical diets, legumes and cereals were seen to have probable benefits in this study and alliums were not. This study therefore provides only limited support for the use of prebiotics in UC. The foods containing RS were all found to be of benefit in this study and therefore the role of RS in UC is strongly supported. Any dietary advice provided to ulcerative colitis patients should be based on the FSS table. The table is of course imperfect because of experimental error, natural variation and the associations between foods. For example, milk and cereal are coded separately but are often consumed together. Thus the magnitude of the difference in the FSSs for these two foods is less than if they'd been independent variables. Suggestions have been made in this discussion as to the factors responsible for the FSS order and to distill these factors into the advice given in Table 5 (see Additional file 5 ). This table is speculation, as this diet has not been formally tested in the UC population. It does however represent the only comprehensive dietary advice available to ulcerative colitis patients at this time. The list of dietary risk factors for colon cancer[ 42 ] bears a similarity to the dietary risk factors presented here for UC. UC patients have an increased risk of colorectal cancer and it is probable that factors responsible for inflammation in UC patients are also responsible for neoplasia in the colon cancer population. Conclusion A dietary analysis method is described that provides a new tool for establishing relationships between diet and disease. This method has been applied to the study of ulcerative colitis and points to sulfite and caffeine as being harmful, with thiamin and resistant starch being potentially therapeutic. For the first time, dietary guidelines for ulcerative colitis patients, including food portion sizes have been developed. Abbreviations Ulcerative colitis (UC); Food sigmoidoscopy score (FSS); European Union (EU); Odds ratio (OD) and Confidence interval (CI). Competing interests The author(s) declare that they have no competing interests. Authors' contributions JHC, EAM and LME contributed to the study design and the writing of the manuscript. EAM was the co-ordinator of the study and along with ST had responsibility for managing the patients and coding the diaries. LME developed the dietary data analysis protocol assisted by RC. CK and JHC performed the sigmoidoscopic examinations of the patients. All authors read and approved the final manuscript. Supplementary Material Additional File 1 Review of studies of diet and ulcerative colitis (UC). Click here for file Additional File 2 Food sigmoidoscopy score (FSS) calculation example for red wine (NB incomplete data set used). Click here for file Additional File 3 Foods consumed in order of food sigmoidoscopy scores (FSS)[ 43 ]. Click here for file Additional File 4 Permitted levels of sulfite in the UK. Click here for file Additional File 5 Proposed dietary advice for ulcerative colitis patients[ 44 ]. Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549081.xml |
535348 | Effect of dietary palm olein oil on oxidative stress associated with ischemic-reperfusion injury in isolated rat heart | Background Palm olein oil (PO), obtained from refining of palm oil is rich in monounsaturated fatty acid and antioxidant vitamins and is widely used as oil in diet in many parts of the world including India. Palm oil has been reported to have beneficial effects in oxidative stress associated with hypertension and arterial thrombosis. Oxidative stress plays a major role in the etiopathology of myocardial ischemic-reperfusion injury (IRI) which is a common sequel of ischemic heart disease. Antioxidants have potent therapeutic effects on both ischemic heart disease and ischemic-reperfusion injury. Information on the effect of PO on ischemic-reperfusion injury is, however, lacking. In the present study, the effect of dietary palm olein oil on oxidative stress associated with IRI was investigated in an isolated rat heart model. Wistar rats (150–200 gm) of either sex were divided into three different groups (n = 16). Rats were fed with palm olein oil supplemented commercial rat diet, in two different doses [5% v / w (PO 5) and 10% v / w (PO 10) of diet] for 30 days. Control rats (C) were fed with normal diet. After 30 days, half the rats from each group were subjected to in vitro myocardial IRI (20 min of global ischemia, followed by 40 min of reperfusion). Hearts from all the groups were then processed for biochemical and histopathological studies. One way ANOVA followed by Bonferroni test was applied to test for significance and values are expressed as mean ± SE (p < 0.05). Results There was a significant increase in myocardial catalase (CAT), superoxide dismutase (SOD) and glutathione peroxidase (GPx) activities with no significant change in myocardial thiobarbituric acid reactive substances (TBARS) only in group PO 5 as compared to group C. There was no light microscopic evidence of tissue injury. A significant rise in myocardial TBARS and depletion of myocardial endogenous antioxidants (SOD, CAT and GPx) along with significant myocyte injury was observed in control rats subjected to ischemia-reperfusion (C IR). Hearts from palm olein oil fed rats subjected to ischemia-reperfusion (PO 5 IR and PO 10 IR) were protected from increase in TBARS and depletion of endogenous antioxidants as compared to C IR group. No significant myocyte injury was present in the treated groups. Conclusions The present study demonstrated for the first time that dietary palm olein oil protected rat heart from oxidative stress associated with ischemic-reperfusion injury. | Background Ischemic heart disease (IHD) is a major cause of death all over the world. Reduction in the blood flow to myocardium leads to IHD and its restitution (reperfusion), spontaneously or by drug / surgery, is essential for tissue/organ survival. However, reperfusion itself exacerbates myocardial injury, commonly known as myocardial ischemic-reperfusion injury (IRI) [ 1 ]. Therefore, IRI is considered as a common sequel of IHD. Oxidative stress has been largely implicated in the etiopathogenesis of IRI. Oxidative stress occurs due to increased production of reactive oxygen species (ROS) like, superoxide radical, hydrogen peroxide, hydroxyl radical at the time of reperfusion, which overwhelms the endogenous antioxidant defense [ 2 ]. Interaction of ROS with cell membrane and various other cellular components have deleterious effects on cellular functions and viability. Oxidative stress is evidenced by increased cellular accumulation of lipid peroxides and depletion of endogenous antioxidants [ 3 - 5 ]. Living organisms have developed antioxidant defense mechanisms against damage due to oxidative stress. These mechanisms in the heart have been extensively studied and the most active endogenous antioxidants involved in this process are superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPx) [ 2 , 6 ]. In addition to this, alpha-tocopherols or vit E, vitamin C and beta-carotene constitute important exogenous antioxidants present in diet [ 7 , 8 ]. The physiological actions of diet continue to be the focus of interest because of its major role in ischemic heart disease. Dietary antioxidants e.g., vitamin E, beta-carotene, vitamin C have beneficial effects in oxidative stress associated with various cardiovascular diseases, including ischemic heart disease [ 9 - 11 ]. Therefore, dietary antioxidants have potential therapeutic role in the prevention and treatment of ischemic heart disease. Palm oil, obtained from the fruit of the tropical plant Elaeis guineensis , is the second major edible oil used worldwide [ 12 ]. Palm olein oil (PO), a liquid fraction obtained from the refining of palm oil, is rich in oleic acid (42.7–43.9%), beta-carotene and vitamin E (tocopherols and tocotrienols). PO is used as dietary oil in many parts of the world including India. In some previous studies, palm oil has been reported to have antioxidant effects in hypertension [ 13 , 14 ] and arterial thrombosis [ 15 ] in rats. In addition to this, palm oil has been shown to increase prostacyclin (PGI 2 ) and reduce thromboxane A 2 (TXA 2 ) levels in tissues [ 16 ]. However scientific studies on antioxidant effects of palm olein oil on ischemic heart disease and ischemic-reperfusion injury are still lacking. Therefore, the present study was designed to evaluate the effects of dietary palm olein oil on myocardial endogenous antioxidants and on oxidative stress associated with ischemic-reperfusion injury in isolated rat heart model. Results There was no mortality, changes in body weight as well as food and water intake pattern of rats in any group. Biochemical parameters I. Changes in the basal level of myocardial lipid peroxidation and endogenous antioxidants (Table 1 ) Table 1 Effect of dietary palm olein oil on myocardial TBARS, catalase, SOD, and GPx levels in different groups Group TBARS (nmol/mg protein) CATALASE (U/mg protein SOD (U/mg protein GPx (U/mg protein) Control 7.8 ± 0.4 34.4 ± 2.1 3.5 ± 0.08 0.13 ± 0.01 Control IR 9.4 ± 0.3* 29.1 ± 0.8* 3.01 ± 0.15* 0.11 ± 0.002** PO 5 8.5 ± 0.8 50.2 ± 3.5** 5.6 ± 0.5** 0.18 ± 0.01** PO 10 8.2 ± 1.2 45.6 ± 4.0* 3.8 ± 0.3 0.16 ± 0.02 PO 5 IR 6.8 ± 0.4 +++ 56.9 ± 4.4 +++ 3.1 ± 0.2 0.12 ± 0.01 PO 10 IR 5.9 ± 0.6 +++ 36.3 ± 5.6 3.4 ± 0.3 0.15 ± 0.01 ++ All values are expressed as Mean ± SE (n = 6) p values: * < 0.05; ** < 0.01; vs Control; ++ < 0.01; +++ < 0.001 vs Control IR (one way ANOVA) I.a. Basal myocardial TBARS levels There were no significant changes in myocardial TBARS levels in both PO5 (8.5 ± 0.8 nmol / mg protein) and PO10 (8.2 ± 1.2 nmol / mg protein) groups when compared to that of control group (7.8 ± 0.4 nmol / mg protein). I.b. Basal myocardial catalase (CAT) activity There was a significant increase in myocardial CAT activity in both PO5 (50.2 ± 3.5 units / mg protein; p < 0.01) and PO10 (45.6 ± 4.0 units / mg protein; p < 0.05) groups as compared to that of control group (34.4 ± 2.1 units / mg protein). I.c. Basal myocardial superoxide dismutase (SOD) activity There were a significant increase in myocardial SOD activity in PO 5 (5.6 ± 0.5 units / mg protein; p < 0.01) group when compared to that of the control group (3.5 ± 0.08 units / mg protein). There was no significant change in myocardial SOD activity in PO10 (3.8 ± 0.3 units / mg protein) group. I.d. Basal myocardial glutathione peroxidase (GPx) activity There was a significant (p < 0.01) increase in myocardial GPx activity in PO5 (0.18 ± 0.01 units / mg protein) group when compared to that of the control group (0.13 ± 0.01 units / mg protein). There was no significant change in myocardial GPx activity in PO10 (0.16 ± 0.02 units / mg protein) group. II. Changes in myocardial lipid peroxidation and endogenous antioxidants following ischemic-reperfusion injury (Table 1) II.a. Myocardial TBARS levels after ischemic-reperfusion injury There was a significant (p < 0.05) increase in myocardial TBARS level in the C IR group (9.4 ± 0.3 nmol / mg protein) when compared to that of the control group (7.8 ± 0.4 nmol / mg protein). There was a significant (p < 0.01) decrease in myocardial TBARS levels in both PO5 IR (6.8 ± 0.4 nmol / mg protein) and PO10 IR (5.9 ± 0.6 nmol / mg protein) groups when compared to that of the C IR group. II.b. Myocardial CAT activity after ischemic-reperfusion injury There was a significant (p < 0.05) decrease in myocardial CAT activity (29.1 ± 0.8 units / mg protein) in C IR group when compared to that of the control group (34.4 ± 2.1 units / mg protein). There was a significant rise in CAT activity in PO5 IR (56.9 ± 4.4 units / mg protein; p < 0.001) group with no significant change in PO10 IR (36.3 ± 5.6 units / mg protein) group when compared to C IR group. II.c. Myocardial SOD activity after ischemic-reperfusion injury There was a significant decrease in myocardial SOD activity in C IR group (3.01 ± 0.15 units / mg protein; p < 0.05) as compared to control group (3.5 ± 0.08 units / mg protein). No significant changes in SOD activities were observed in both PO5 IR (3.1 ± 0.2 units / mg protein) and PO10 IR (3.4 ± 0.3 units / mg protein) groups as compared to C IR group. II.d. Myocardial GPx activity after ischemic-reperfusion injury There was a significant (p < 0.01) decrease in myocardial GPx activity in C IR group (0.11 ± 0.002 units / mg protein) as compared to control group (0.13 ± 0.004 units / mg protein). There was no significant change in myocardial GPx activity in PO5 IR group (0.12 ± 0.01 units / mg protein) with a significant (p < 0.01) increase in myocardial GPx activity in PO10 IR group (0.15 ± 0.01 units / mg protein) as compared to C IR group. Histopathological study Fig. 1 shows the H&E micrograph of control heart with normal architecture. In PO5 and PO10 groups there was no evidence of cellular injury (not shown). Focal loss of myocardial fibres and marked edema was observed in C IR group (Fig. 2 ). Mild to moderate edema was observed in PO10 IR group (Fig. 4 ). Degree of edema was reduced in PO5 IR with no evidence of focal necrosis (Fig. 3 ) Figure 1 Light micrograph of heart tissue. Control rat heart (C) showing normal architecture (H & E X 10). Figure 2 Light micrograph of heart tissue. Control rat heart subjected to 20 min ischemia and 40 min reperfusion (C IR) showing marked edema and focal destruction of myocardial fibres (H & E X 10). Figure 3 Light micrograph of heart tissue. Rat heart supplemented with 5% v/w of dietary palm olein oil subjected to 20 min ischemia and 40 min of reperfusion (PO5 IR) showing mild edema with occasional loss of myofibre (H & E X 10). Figure 4 Light micrograph of heart tissue. Rat heart supplemented with 10% v/w of dietary palm olein oil subjected to 20 min ischemia and 40 min of reperfusion (PO10 IR) with mild to moderate edema and occasional loss of myofibre (H & E X 10). Discussion In the present study, a significant increase in myocardial SOD, catalase and GPx activity was observed in the lower dose of palm olein oil fed rats. However, their further augmentation was not observed in the higher dose, i.e., a dose dependent effect was not observed. The finding correlates with the previous studies in which an increase in response was not observed with the increase in the dose of supplemented vitamin E [ 26 , 27 ]. The possible reasons behind the lack of dose response relationship may be a decrease in intestinal absorption as a result of increase in dose [ 28 ] and newly absorbed vitamin E in part replacing the alpha-tocopherol in circulating lipoproteins [ 29 ]. Augmentation of endogenous antioxidants (SOD, CAT, GPx) has been recognized as an important pharmacological property, present in natural as well as many synthetic compounds [ 30 - 33 ]. This constitutes a major mechanism of protection against oxidative stress, offered by them [ 30 , 35 , 37 ]. The most abundant reactive oxygen species generated in living system is superoxide radical which is acted upon by SOD to produce hydrogen peroxide which in turn is inactivated by catalase and / or GPx into water and oxygen. Thus an increase in both SOD and catalase along with GPx activity is considered to be more beneficial in the event of oxidative stress [ 34 ]. Increase in myocardial TBARS and depletion of myocardial endogenous antioxidants support the occurrence of oxidative stress in the control hearts following ischemia-reperfusion in the present study. It was also accompanied by tissue injury with marked edema and focal loss of myocardial fibres. Similar changes have been reported earlier to occur following brief period of ischemia followed by reperfusion in rat heart [ 35 - 37 ]. Hearts from palm olein oil fed rats in both doses were protected against oxidative stress, as evidenced by inhibition of increase in TBARS, depletion of catalase, GPx and tissue injury following ischemia-reperfusion. In a previous study, palm oil has been reported to prevent oxidative stress induced hypertension in rats [ 13 ]. The mechanism of such protection can be attributed to the augmented endogenous antioxidant reserve of heart in the lower dose. However, the higher dose, which did not cause any significant augmentation of endogenous antioxidants, also inhibited depletion of antioxidants, rise in TBARS and tissue injury. It is possible that direct antioxidant effects of palm olein oil may be attributable to the presence of alpha tocopherols and tocotrienols, which are known to protect against oxidative stress. Experimental as well as clinical studies with exogenous antioxidants supplementation have been shown to have protective effect in ischemic heart disease [ 38 , 39 ]. In this regard, the most commonly used exogenous antioxidants are vitamin E (tocopherols and tocotrienols), beta-carotene and vitamin C. Palm oil is also beneficial in conditions like hypertension [ 13 , 14 , 40 ], arterial thrombosis [ 15 , 41 ] and causes increase in PGI 2 /TXA 2 ratio [ 16 ]. Palm oil derived vitamin E rich in tocotrienols has shown beneficial effects against hypercholesterolemia [ 42 , 43 ] and is considered to be more potent than tocopherols [ 44 ]. The observations made in the present study have important nutritional significance for palm olein oil in relation to ischemic heart disease. However, further studies are required to establish the mechanism, underlying the augmentation of tissue antioxidants. Conclusions The present study, for the first time, demonstrated that long term oral supplementation of palm olein oil caused augmentation of endogenous antioxidants of heart, which were subsequently protected from developing oxidative stress following ischemia-reperfusion. Methods Preparation of diet Palm olein oil (Ruchi Gold, India) was obtained from the local market. Commercial rat diet (Ashirwad, India) containing protein: 24%, fat: 5%, fiber: 4%, carbohydrates: 55%, calcium: 0.6%, phosphorus: 0.3% w / w was supplemented with palm olein oil in two different doses [5 % v / w and 10 %v / w of diet]. The doses were selected from the previous studies [ 13 , 17 ]. Diet and water were provided ad libitum . Animals The study was approved by Institute Animal Ethics Committee (245 / IAEC / 04) and all animal care and experimental protocols were in compliance with the NIH guidelines for the care and use of the Laboratory Animals (NIH Publication #85–23, 1985). Laboratory bred Wistar rats (150–200 gm) of either sex were maintained under standard laboratory conditions at temperature 25 ± 2°C, relative humidity 50 ± 15% and normal photo period (12 h dark / 12 h light) was used for the study. Chemicals All chemicals were of analytical grade and chemicals required for sensitive biochemical assays were obtained from Sigma Chemicals (St. Louis, USA). Double distilled water (DDW) was used in all biochemical assays. Experimental protocol After one week of acclimatization, rats were randomly divided into three groups, each group containing 16 rats. In control group (C), rats were fed with normal diet for 30 days. In groups (PO5 and PO10), rats were fed with palm olein oil supplemented commercial rat diet for 30 days in two different doses; 5% v / w and 10% v / w of diet. Changes in body weight, food and water intake patterns of rats in all the groups were noted throughout the experimental period. At the end of the 30 days, rats were fasted overnight and half the rats from each group were subjected either to protocol I or to protocol II as described below. Rats were heparinised (375 IU / 200 gms, i.p), and 0.5 h later rats were anesthetized with sodium pentobarbitone (60 mg / Kg, i.p) and euthanised. Protocol I Basal level of biochemical and histopathological studies Immediately after euthanization, the hearts were rapidly harvested, washed in ice cold saline, frozen in liquid nitrogen and stored at -80°C until processed for estimations of biochemical parameters. For histopathological studies, heart was stored in 10% buffered formalin (pH 7.2). Group C: Normal diet fed rats (n = 8) Group PO 5: 5% v / w palm olein oil supplemented diet fed rats (n = 8) Group PO 10:10% v / w palm olein oil supplemented diet fed rats (n = 8) Protocol II Production of in vitro ischemic reperfusion injury Immediately after euthanization, hearts were rapidly harvested, washed in ice-cold saline, and perfused with the non-recirculating Langendorff's technique (Hufesco, Hungary), under constant pressure mode with modified Kreb Hensleit's buffer [ 18 ] containing [mM]: glucose 11.1; NaCl 118.5; NaHCO 3 25; KCl 2.8; KH 2 PO 4 1.2; CaCl 2 1.2; MgSO 4 0.6, with a pH of 7.4. The buffer solution equilibriated with 95% O2+ 5% CO2 was delivered to the aortic canula at 37°C and 65 mm Hg pressure. Following 10 min. of equilibration period, hearts were subjected to 20 min. of zero flow (global ischemia) and 40 min. of re-flow (reperfusion) [ 19 , 20 ]. Group C IR: Normal diet fed rats subjected to IR injury (n = 8) Group PO5 IR: 5% v/w palm olein oil supplemented diet fed rats subjected to IR injury (n = 8) Group PO10 IR: 10% v/w palm olein oil supplemented diet fed rats subjected to IR injury (n = 8) At the end of each experiment, heart was frozen in liquid nitrogen and stored at -80°C until processed for estimations of biochemical parameters. For histopathological studies, heart was stored in 10% buffered formalin (pH 7.2). Biochemical parameters Myocardial TBARS [ 21 ] Hearts were homogenized in 10% trichloroacetic acid (TCA) at 4°C. 0.2 ml homogenate was pipetted into a test tube, followed by addition of 0.2 ml of 8.1 % sodium dodecyl sulphate (SDS), 1.5 ml of 30% acetic acid (pH-3.5), 1.5 ml of 0.8% thiobarbituric acid (TBA) and volume was made upto 4.0 ml with DDW. Test tubes were boiled at 95°C for 60 min. and then cooled. 1.0 ml of DDW and 5.0 ml of n-butanol: pyridine (15:1 v / v) mixture was added to the test tubes and centrifuge at the 4,000 × g for 10 min. The absorbance of developed colour in organic layer was measured at 532 nm. Commercially available 1, 1, 3, 3-tetraethoxypropane (Sigma Chemicals) was used as a standard for MDA. Data is expressed as nmol / mg protein. Myocardial CAT [ 22 ] Hearts were homogenized at 4°C (1:10) in 50 mmol/l potassium phosphate buffer (pH- 7.4) and centrifuged at 3,000 × g for 10 min. Supernatant (50 μl) was added to a 3.0 ml cubette that contained 1.95 ml of 50 mM phosphate buffer (pH-7.0). Then1.0 ml of 30 mM hydrogen peroxide was added and changes in absorbance were measured for 30 sec. at 240 nm at an interval of 15 sec. Catalase activity is expressed as units / mg protein as compared to the standard. Myocardial SOD [ 23 ] Hearts were homogenized in 0.25 M tris sucrose buffer and centrifuged at 10,000 × g for 15 min at 4°C. The supernatant was fractionated by 50% ammonium sulphate and dialysed overnight. Aliquots of the supernatant (100 μl) was added to sodium pyrophosphate buffer (pH-8.3) followed by addition of 0.1 ml of 186 μM phenazine methosulphate, 0.3 ml of 300 mM nitroblue tetrazolium and 0.2 ml of 780 μM NADH. Reaction mixture was incubated for 90 sec. at 30°C and stopped the reaction by adding 1.0 ml of glacial acetic acid. 4.0 ml of n-butanol was then added and centrifuged at 3,000 × g for 10 min. The absorbance of organic layer was measured at 560 nm. SOD activity is expressed as units / mg protein as compared to the standard. Myocardial GPx [ 24 ] Hearts were homogenized at 4°C in 0.25 M phosphate buffer saline (pH-7.0). Homogenate was centrifuged at 15,000 × g for 60 min. at 4°C and supernatant were assayed for the GPx activity. GPx activity was in a 1.0 ml cubette containing 400 μl of 0.25 M potassium phosphate buffer (pH-7.0), 200 μl of sample, 100 μl of 10 mM GSH, 100 μl of 2.5 mM NADPH and 100 μl of glutathione reductase (6 U / ml). Hydrogen peroxide (100 μl of 12 mM) was then added and change in absorbance was measured at an interval of 1 min for 5 min at 366 nm. GPx activity is expressed as units / mg protein as compared to the standard. Protein concentration was measured by Bradford method [ 25 ]. Histopathological studies Heart tissue was fixed in 10% buffered formalin, routinely processed and embedded in paraffin. Paraffin sections (3 μm) were cut on glass slides and stained with hematoxylin and eosin (H&E), periodic acid Schiff (PAS) reagent and examined under a light microscope (Nikon, Japan). Histopathological study was carried out by one of the authors (AKD), blinded to the groups. Statistical analysis All values are expressed as mean ± SE. One way ANOVA followed by Bonferroni test was applied to test for significance of biochemical data of the different groups. Significance is set at p < 0.05. Abbreviations IRI: ischemic-reperfusion injury; PO: palm olein oil; PO5: rats fed with 5% v/w palm olein oil supplemented diet; PO10: rats fed with 10% v/w palm olein oil supplemented diet; C: rats fed with normal diet; C IR: rats fed with normal diet subjected to ischemic-reperfusion injury; PO5 IR: rats fed with 5% v/w palm olein oil supplemented diet subjected to ischemic-reperfusion injury; PO10 IR: rats fed with 10% v/w palm olein oil supplemented diet subjected to ischemic-reperfusion injury; TBARS: thiobarbituric acid reactive substances; SOD: superoxide dismutase; CAT: catalase; GPX: glutathione peroxidase; IHD: ischemic heart disease; ROS: reactive oxygen species; PGI 2 : prostacyclin; TXA 2 : thromboxane A 2 ; SE: standard error; ANOVA: analysis of variance; MDA: malondialdehyde; TCA: tricarboxylic acid; SDS: sodium dodecyl sulphate; DDW: double distilled water; NADH: nicotinamide adenine dinucleotide reduced; TBA: thiobarbituric acid; NADPH: nicotinamide adenine dinucleotide phosphate reduced; GSH: reduced glutathione; PAS; periodic acid Schiff reagent; H&E: hematoxylin and eosin. Authors' contributions DN carried out the animal experimentation, biochemical estimation and statistical analysis of results. SS and MT participated in the design of the study and statistical analysis. AKD carried out the light microscopic study. SKM conceived the study, participated in its design and coordination and drafted the manuscript. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC535348.xml |
449903 | Translating DNA into Synthetic Molecules | The unique ability of nucleic acids to replicate has recently been combined with the power of combinatorial chemistry, providing a new approach to the science of drug design | At some time almost 4 billion years ago, nature likely was faced with a chemical dilemma. Nucleic acids had emerged as replicable information carriers and primitive catalysts ( Joyce 2002 ), yet their functional potential was constrained by their structural homogeneity and lack of reactive groups. These properties rendered nucleic acids well suited for storing information, but flawed for mediating the diverse chemistries required to sustain and improve increasingly complex biological systems. It is tempting to speculate that translation emerged as the solution to this dilemma. Translation, defined here as the conversion of an informationcarrying molecule into a corresponding encoded structure, enabled the expanded functional potential of proteins to be explored using powerful evolutionary methods that depend on the unique ability of nucleic acids to replicate. A small but growing number of researchers have begun to tackle a modern version of this dilemma. While proteins and nucleic acids can be manipulated using powerful molecular biology techniques that enable their directed evolution, the size, fragility, and relatively limited functional group diversity of biological macromolecules make them poorly suited for solving many problems in the chemical sciences. Ideally, researchers would like to apply evolution-based approaches to the discovery of functional synthetic, rather than biological, molecules. A solution analogous to nature's translation of mRNA into protein could, in principle, address this contemporary problem ( Orgel 1995 ; Gartner and Liu 2001 ). If a laboratory system were developed that could translate amplifiable information carriers such as DNA into arbitrary synthetic molecules, the evolution of synthetic molecules using iterated cycles of translation, selection, amplification, and diversification would be possible. The translation of DNA into synthetic molecules is conceptually distinct from the use of DNA simply as a tag during the solidphase synthesis of a molecule that is part of a combinatorial library ( Brenner and Lerner 1992 ). The latter process uses DNA to record the history of a series of chemical reactions by cosynthesizing a portion of a DNA oligonucleotide during each step of a molecule's solidphase synthesis. As a result, the identity of compounds that pass screening can be inferred by PCR amplification and sequencing of the DNA associated with a given bead ( Needels et al. 1993 ). The resulting DNA, however, cannot redirect the synthesis of active compounds. In contrast, the translation of DNA into synthetic molecules uses the sequence of nucleotides in a strand of DNA to direct the synthesis of a nascent molecule. As a result, a complete cycle of translation, selection, and amplification can be applied to the discovery of synthetic molecules in a manner that is analogous to the processes that take place during biological evolution. DNA-templated organic synthesis (DTS) has emerged as one way to translate DNA sequences into a variety of complex synthetic small molecules ( Gartner and Liu 2001 ; Gartner et al. 2002 ; Li and Liu 2004 ). In this approach, starting materials covalently linked to DNA templates approximately 20–50 nucleotides in length are combined in very dilute solutions with reagents that are covalently linked to complementary DNA oligonucleotides. Upon Watson-Crick base pairing, the proximity of the synthetic reactive groups elevates their effective molarity by several orders of magnitude, inducing a chemical reaction. Because reactions do not take place between reactants linked to mismatched (noncomplementary) DNA, DTS generates synthetic products in a manner that is programmed by the sequence of bases in the template strand. In a series of three papers in this issue of PLoS Biology , Harbury and co-workers describe an elegant new approach to translating DNA into synthetic peptides called “DNA display.” Their approach uses DNA hybridization to separate mixtures of DNA sequences into spatially distinct locations. The first paper ( Halpin and Harbury 2004a ) reports the development of resin-linked oligonucleotides that efficiently and sequence-specifically capture DNA containing complementary subsequences. This immobilization process is efficient enough to be iterated, so that DNA sequences specifying multiple amino acids can be routed to the appropriate miniature resin-filled columns during each step. In the second paper ( Halpin and Harbury 2004b ), Harbury and coworkers detail solid-phase peptide synthesis performed on unprotected DNA 340mers bound to DEAE Sepharose. Optimization of amino acid side-chain-protecting groups and peptide coupling conditions enabled a variety of amino acids to undergo efficient peptide coupling to bound oligonucleotides containing an amine group. The third paper ( Halpin et al. 2004 ) integrates the routing and peptide synthesis described above into the translation of a library of 10 6 DNA 340mers into a corresponding library of up to 10 6 synthetic pentapeptides. To achieve chemical translation, the DNA library was subjected to iterated cycles of routing and solidphase peptide synthesis. After each routing step, the appropriate amino acid was coupled to each DNA-linked subpopulation. DNA routing was therefore used to achieve the splitting step of “split-and-pool” combinatorial peptide synthesis. The completed library of peptide–DNA conjugates was then subjected to in vitro selection based on the ability to bind an antibody with known affinity for the [Leu]enkephalin pentapeptide Tyr-Gly-Gly-Phe-Leu. After two rounds of routing, synthesis, and selection, followed by DNA sequencing, the remaining oligonucleotides predominantly encoded the Tyr-Gly-Gly-Phe-Leu sequence or close variants thereof. This result demonstrates that the DNA display method is capable of facilitating the discovery of functional molecules by enabling in vitro selection methods to be applied to molecules generated by split-and-pool combinatorial synthesis. The fundamental distinctions between DTS and DNA display approaches to chemical translation imply that these two strategies will be applicable to different types of synthetic structures. Because the DNA display approach separates the DNA hybridization step from the chemical synthesis step, it does not require the coupling of synthetic reagents to oligonucleotides (beyond the starting material), and can use reaction conditions such as high temperatures or high pH that may not be compatible with DNA hybridization. These features suggest that DNA display may be able to access structures that cannot be created by DTS. Likewise, because DTS approaches use effective molarity rather than intermolecular reactivity to direct organic synthesis, they enable modes of controlling reactivity (such as using otherwise incompatible reactions in a single solution [ Calderone et al. 2002 ]) and classes of chemical reactions (such as heterocoupling of substrates that preferentially homocouple) that cannot be accessed using split-and-pool synthesis. In principle, these two approaches are complementary, and it is tantalizing to envision the use of both DNA display and DTS to direct different steps during a single chemical translation. In order for either approach to fully realize its potential of truly evolving libraries of diverse synthetic molecules, rather than simply enriching libraries that already contain at the outset the “most fit” molecule, researchers must develop sophisticated library syntheses that generate remarkable complexity (vast numbers of different compounds) in a relatively modest number of DNA-compatible synthetic steps. True evolution takes place when the theoretical complexity of a population exceeds the number of different molecules that can be created in a single library translation step, and when diversification is required to access compounds in later generations that are more fit than any member of the starting pool. To my knowledge, no synthetic library to date contains this degree of complexity (indeed, the total size of the Chemical Abstracts Service database of known chemical substances is presently less than 10 8 compounds). However, because so few copies of a DNA-linked synthetic molecule are required for in vitro selection ( Doyon et al. 2003 )—compared with the relatively large quantities of material that are required for conventional screening approaches—these chemical translation methods offer the first hope of achieving such synthetic complexity without requiring an impractical amount of material or storage space. For comparison, a conventional-format synthetic library containing 100 µg of each of 10 8 different structures would represent 10 kg of material, not including the mass of beads or plates associated with the library, while a chemically translated library containing 10,000 copies of 10 8 different species represents less than 1 µg of total material. While significant remaining challenges face efforts to develop and apply chemical translation, the promise of marrying evolution and organic synthesis is an irresistible combination for some researchers. The work of Harbury and co-workers described in this issue represents the latest approach to the very ancient problem of translating replicable information into functional structures. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC449903.xml |
524523 | Activation of the hedgehog pathway in advanced prostate cancer | Background The hedgehog pathway plays a critical role in the development of prostate. However, the role of the hedgehog pathway in prostate cancer is not clear. Prostate cancer is the second most prevalent cause of cancer death in American men. Therefore, identification of novel therapeutic targets for prostate cancer has significant clinical implications. Results Here we report that activation of the hedgehog pathway occurs frequently in advanced human prostate cancer. We find that high levels of hedgehog target genes, PTCH1 and hedgehog-interacting protein (HIP), are detected in over 70% of prostate tumors with Gleason scores 8–10, but in only 22% of tumors with Gleason scores 3–6. Furthermore, four available metastatic tumors all have high expression of PTCH1 and HIP. To identify the mechanism of the hedgehog signaling activation, we examine expression of Su(Fu) protein, a negative regulator of the hedgehog pathway. We find that Su(Fu) protein is undetectable in 11 of 27 PTCH1 positive tumors, two of them contain somatic loss-of-function mutations of Su(Fu) . Furthermore, expression of sonic hedgehog protein is detected in majority of PTCH1 positive tumors (24 out of 27). High levels of hedgehog target genes are also detected in four prostate cancer cell lines (TSU, DU145, LN-Cap and PC3). We demonstrate that inhibition of hedgehog signaling by smoothened antagonist, cyclopamine, suppresses hedgehog signaling, down-regulates cell invasiveness and induces apoptosis. In addition, cancer cells expressing Gli1 under the CMV promoter are resistant to cyclopamine-mediated apoptosis. All these data suggest a significant role of the hedgehog pathway for cellular functions of prostate cancer cells. Conclusion Our data indicate that activation of the hedgehog pathway, through loss of Su(Fu) or overexpression of sonic hedgehog, may involve tumor progression and metastases of prostate cancer. Thus, targeted inhibition of hedgehog signaling may have significant implications of prostate cancer therapeutics. | Background The hedgehog (Hh) pathway plays a critical role in embryonic development and tissue polarity [ 1 ]. Secreted Hh molecules bind to the receptor patched (PTC-PTCH1, PTCH2), thereby alleviating PTC-mediated suppression of smoothened (SMO), a putative seven-transmembrane protein. SMO signaling triggers a cascade of intracellular events, leading to activation of the pathway through GLI-dependent transcription [ 2 ]. The hedgehog receptor PTCH1 is also a target gene of this pathway, which forms a negative feedback mechanism to maintain the pathway activity at an appropriate level in a given cell. Activation of Hh signaling through loss-of-function somatic mutations of PTCH1 in human basal cell carcinomas (BCCs) disrupts this feedback regulation, leading to uncontrolled SMO signaling. Activating mutations of SMO in BCCs, on the other hand, are resistant to PTCH1-mediated inhibition, leading to an outcome similar to PTCH1 inactivation [ 3 - 6 ]. More recently, abnormal activation of the sonic hedgehog pathway, through over-expression of sonic hedgehog, has been implicated in the development of subsets of medulloblastomas, small cell lung cancer and gastrointestinal tract (GI) cancers [ 7 - 10 ]. Development of prostate requires hedgehog signaling. Although the initial formation of prostate buds does not require sonic hedgehog signaling (shh), shh is critical for maintaining appropriate prostate growth, proliferation and tissue polarity [ 11 - 14 ]. In the adult prostate, however, the activity of the hedgehog pathway is quite low. It remains to be tested whether this hedgehog pathway is activated during development of prostate cancer, the second most prevalent cause of cancer death in American men. Activation of the hedgehog pathway is often indicated by elevated levels of PTCH1 and HIP. In addition to PTCH1 mutation, SMO activation and hedgehog over-expression, loss of Su(Fu) can result in activation of the hedgehog pathway. In the human, the Su(Fu) gene is localized at chromosome 10q24, a region with LOH in several types of cancer including prostate cancer, lung cancer, breast cancer and medulloblastomas [ 15 , 16 ]. As a negative regulator of the hedgehog pathway, Su(Fu) inhibits the function of Gli molecules, leading to inactivation of this pathway [ 17 - 19 ]. Su(Fu) is also reported to affect beta-catenin function [ 20 ]. In addition, over-expression of sonic hedgehog is shown to be involved in the development of GI cancers [ 9 , 10 ]. Here we report our findings that activation of the hedgehog pathway occurs frequently in advanced prostate cancers, possibly through loss of Su(Fu) protein or over-expression of sonic hedgehog. Results Elevated expression of hedgehog target genes in prostate cancer specimens As an important regulator of tissue polarity, active hedgehog signaling is required for ductual morphogenesis and proliferation during prostate development [ 11 - 14 ]. The adult prostate, on the other hand, does not contain active hedgehog signaling. Because hedgehog signaling is an important regulator for epithelial-mesenchymal interaction, an event critical during prostate cancer development, we examined whether the hedgehog-signaling pathway is activated in prostate cancer. Activation of hedgehog signaling causes elevated expression of target genes PTCH1 and HIP. Thus, increased protein expression of PTCH1 and HIP indicates activation of the hedgehog pathway. Using PTCH1 antibodies [ 10 ], we examined 59 prostate cancer samples for hedgehog signaling activation (see Table 1, Additional file 1 for details). We first tested the specificity of the PTCH1 antibodies in MEF cells. Ptch1 null MEF cells have no active Ptch1 gene, thus should not have positive staining with PTCH1 antibodies. Indeed, no staining was seen in Ptch1 null MEF cells (Fig. 1A ). After transfection of PTCH1 expressing plasmid, transfected cells showed positive staining (Fig. 1A ), indicating that the PTCH1 antibodies are specific to PTCH1. Furthermore, PTCH1 immunohistostining was abolished after addition of the specific peptide, from which the antibodies were raised (Fig. 1B,1c ). We found that percentage of PTCH1 positive staining tumors increased in high grade tumors (Table 1, Additional file 1 ). In prostate cancers with Gleason scores 3–6, 4 out of 18 specimens were positive for PTCH1 (22%), whereas 16 out of 22 undifferentiated carcinomas (Gleason Scores of 8–10) expressed PTCH1 (73%, see Table 1, Additional file 1 ), suggesting that the hedgehog pathway is frequently activated in advanced prostate cancer. To confirm this data, we found that all four available metastatic prostate cancer specimens were all positive for PTCH1 staining. Figure 1 Detection of PTCH1 expression in prostate cancers. Protein expression of PTCH1 was detected by immunostaining. PTCH1 antibodies (Santa Cruz Biotechnology Cat# 9149) were tested in Ptch1 -/- null MEF cells ( A ). While Ptch1 -/- null MEF cells had no positive fluorescent staining with PTCH1 antibodies, transfection of PTCH1 expressing plasmid lead to positive staining (green, indicated by an arrow, 400×). Immunohistochemistry of prostate cancer specimens with PTCH1 gave negative ( B-a, 200×) or positive (Red in B-b, 200×) signals. When PTCH1 antibodies were pre-incubated with the very peptide for raising the antibodies, no positive signals could be observed ( B-c ). To further confirm our data, we detected HIP protein expression, another marker of the hedgehog signaling activation. After transfection of HIP expressing plasmid into 293 cells, HIP antibodies recognize a single band around 75 KD (Fig. 3A ), and an endogenous HIP protein with a similar size was also detected in two cancer tissues, in which hedgehog signaling is known to be activated (Fig. 3B and data not shown here). In contrast, the matched normal tissue did not express detectable HIP. Thus, HIP expression appears to be a good marker for hedgehog signaling activation. Immunohistostaining with HIP antibodies in prostate cancer specimens revealed a similar pattern to prostate specific antigen (PSA) and PTCH1 (Fig. 3C and Table 1, Additional file 1 ), further confirming that hedgehog pathway is activated in advanced prostate cancers. Thus the hedgehog pathway appears to be frequently activated in advanced or metastatic prostate cancers. Figure 3 Detection of HIP in human cancer specimens. By Western blotting, HIP antibodies (R&D systems Cat# AF1568) recognized one band between 75 and100 KD ( A ). Expression of endogenous HIP was detected in two GI cancer tissues, which were known to contain activated hedgehog signaling (data not shown here), but not in the matched normal tissue ( B ). Immunohistostaining of HIP I prostate cancer showed a similar pattern to PSA ( C , 200×) Altered expression of Su(Fu) and Shh in prostate cancer specimens There are several mechanisms by which the hedgehog pathway in these prostate tumors can be activated, including loss of Su(Fu) or over-expression of hedgehog [ 6 - 10 ]. The Su(Fu) gene is localized at 10q24, a region with a frequent LOH in prostate cancer [ 15 , 16 , 18 ]. Mutations of Su(Fu) have been reported in other human cancers [ 6 ]. To test whether loss of Su(Fu) function is responsible for hedgehog signaling activation, we examined expression of Su(Fu) protein in these prostate cancer specimens. The antibodies of Su(Fu) recognize a single band at 52-kD in Western blotting analyses (Fig. 4A ), which was reduced following treatment with Su(Fu) SiRNA (Fig. 4B ), indicating the specificity of the antibodies. Furthermore, addition of the peptide, from which the antibodies were raised, prevented the antibody binding, further confirming the specificity of our Su(Fu) antibodies (data not shown). Of the 16 PTCH1 positive prostate cancer specimens with Gleason scores 8–10, 9 have no detectable Su(Fu) protein (Fig. 4C,4D,4E and Table 1, Additional file 1 ). In total, 11 of 27 PTCH1 positive prostate cancer specimens have no detectable Su(Fu) protein. Prostate cancers with low Gleason scores, however, frequently have detectable Su(Fu) protein (see Table 1, Additional file 1 ), suggesting that loss of Su(Fu) protein may be associated with prostate cancer progression. Figure 4 Detection of Su(Fu) in prostate cancer specimens. Su(Fu) antibodies (Santa Cruz Biotechnology Cat# 10933) recognized only one single band (54-Kd) in D283 cells ( A ). Following treatment of a specific SiRNA of Su(Fu), the endogenous Su(Fu) band was greatly reduced ( B ). Immunohistostaining with Su(Fu) antibodies in prostate cancer specimens revealed positive ( C , in red, 200×), negative ( D , 200×) or weak staining ( E , red, 200×). To confirm the immunohistochemistry data, we performed immunoblotting analyses using several dissected TURP (Transurethral resection of the prostate) specimens in which tumor portion can be as high as 70% of the tissue mass. Prostatectomy specimens (most of our tumors), however, often contain a small percentage (5–10%) of tumor tissue and are therefore not suitable for Western blotting or real-time PCR analyses. As shown in Fig. 5A , two tumors (PC48 and PC51) had no detectable Su(Fu) protein, which are consistent with our immunohistostaining, suggesting loss of Su(Fu) may be responsible for hedgehog pathway activation in these tumors. The matched normal tissues, however, retained expression of Su(Fu), indicating that alteration of Su(Fu) is a somatic event. Sequence analyses of these two tumors revealed genetic mutations in Su(Fu) , which are predicted to create STOP codons in the coding sequence (Fig. 5B and Table 1, Additional file 1 ). In PC48, a homozygous deletion of A1315 was detected, which results in a STOP codon at +1318 bp (Fig. 5B ). In PC51, we detected two types of mutations, one with a deletion of C255, which results in a STOP codon at +294 bp whereas another with a deletion of C198, create a STOP codon (Picture not shown here, see Table 1, Additional file 1 ). These mutations were confirmed with 6 independent clones from two separate experiments, which exclude the possibility of PCR errors. No mutations were detected from the matched benign tissues, indicating the somatic nature of the mutations. Real-time PCR analyses indicated that target genes of the hedgehog pathway, PTCH1 and Gli1, were all elevated in these tumors (Fig. 5C ), confirming activation of the hedgehog pathway in these tumors. Thus, Su(Fu) inactivation appears to contribute to activation of hedgehog signaling in these prostate tumors. Figure 5 Inactivation of Su(Fu) in prostate cancer. Two TURP (Transurethral resection of the prostate) tumors with loss of Su(Fu) expression were confirmed by Western blotting ( A ). One mutation of Su(Fu) found in prostate cancer PC48 is shown in B , which is predicted to create a STOP codon in the Su(Fu) coding sequence +1318. The levels of Gli1 and PTCH1 transcripts in prostate tissues were detected by real-time PCR (see methods for details) ( C ). Tumor tissues had higher levels of the target gene transcripts. For tumors with high level of PTCH1 expression, but no changes in Su(Fu) protein expression, we examined expression of sonic hedgehog. It is reported that expression of hedgehog may be responsible for hedgehog signaling activation in lung cancer and GI cancers. Immunohistostaining with sonic hedgehog antibodies indicate that sonic hedgehog is highly expressed in 24 of 27 advanced prostate tumors with elevated expression of PTCH1 and HIP (see Fig. 2 and Table 1, Additional file 1 ). Thus, activation of the hedgehog pathway, as indicated by elevated PTCH1 and HIP expression, is associated with loss of Su(Fu) expression or elevated hedgehog expression. Figure 2 Co-expression of PTCH1, PSA and Shh in prostate cancer specimens. Immunohistochemistry of prostate cancer specimens with PSA was used to confirm the cancer region. Positive staining was in red. Positive staining patterns of PTCH1 and Shh antibodies (Santa Cruz Biotechnology Cat# 9024) were similar to that of PSA staining. PC23 ( A-C ) was from tumors with Gleason score 7 (200×). PC38 ( D-F ) was a tumor from Gleason score 10 (400×) ( see Table 1, Additional file 1 for details). The role for activated hedgehog signaling for cellular functions of prostate cancer To demonstrate the role of hedgehog pathway in prostate cancer, we screen five available cell lines for the expression of Gli1, PTCH1 and HIP. TSU, LNCap, Du145 and PC3 are prostate cancer cell lines whereas RWPE-1 is a prostate epithelial cell line. We found that the hedgehog target genes were significantly elevated in all cancer cell lines (Fig. 6A ). Thus, we predicted that inhibition of the hedgehog pathway by smoothened antagonist, cyclopamine, would suppress cell proliferation and cell invasiveness. Figure 6 Cellular functions of the hedgehog pathway in prostate cancer cells . Expression of hedgehog target genes, PTCH1 and Gli1, were detected by real-time PCR ( A ). DNA synthesis was detected by BrdU labeling ( B ). Over 1000 cells were counted under fluorescent microscope for the percentage of BrdU positive cells, and the experiment was repeated twice ( C ). Following treatment with 5 μM cycloapmine in PC3 cells, expression of hedgehog target genes were dramatically inhibited (data not shown here), which was accompanied with a significant reduction of BrdU positive cells (see Fig. 6B for details). This effect is specific because addition of tomotidine, a non specific compound with a similar structure to cycloapmine, had no effects on either target gene expression or DNA synthesis (indicated by BrdU labeling in Fig. 6B and 6C ). The prostate epithelial RWPE-1 cells which have no activated hedgehog signaling, on the other hand, were not sensitive to cyclopamine (data not shown here), indicating that cyclopamine specifically affects cells with elevated hedgehog signaling. LN-CAP, Du145 and TSU cells, like PC3 cells were also sensitive to cyclopamine treatment (Fig. 6C ). Prostate cancer progression is accompanied by increased cell invasiveness. Because the hedgehog signaling activation occurs frequently in advanced prostate cancer, we examined if inhibition of the hedgehog signaling can reduce cell invasiveness. Using BD Bio-coat cell invasion chambers, we found that treatment of cyclopamine in PC3 cells reduced the percentage of invasive cells by 70% (Fig. 7A ). Similar data were also observed in Du145, LN-CAP and TSU cells (Fig. 7B ). Under the same condition, RWPE-1 cells were not very invasive. Thus, hedgehog signaling activation regulates both cell proliferation as well as cell invasiveness of prostate cancer cells. Figure 7 Effects of cyclopamine on cell invasiveness of prostate cancer cells. Cell invasion assay of prostate cancer cells was performed using BD Bio-coat cell invasion chambers ( A ). The rate of cell invasion was calculated by dividing cell numbers penetrated the matrigels by the number of cell in the control chambers (without matrigels) ( B ). It has been shown that cyclopamine induced apoptosis in cancer cells with activated hedgehog signaling [ 21 ]. We have shown that Gli1 down-regulation is necessary for cyclopamine-mediated apoptosis in basal cell carcinoma cells [ 21 ]. To test the significant role of Gli1, the down-stream effector and the target gene of the hedgehog pathway, in cyclopamine-mediated apoptosis, we first transfected Gli1 expressing plasmid in to PC3 cells, and then treated the cells with 5 μM cyclopamine for 36 h. Since Gli1 is expressed under the control of the CMV promoter, we predicted that ectopic Gli1-expressing cells should be resistant to apoptosis, which is detected by TUNEL staining. As shown in Fig. 8 , we found that all Gli1 positive cells (n = 500) were TUNEL negative, supporting our hypothesis that down-regulation of Gli1 may be an important mechanism by which cyclopamine mediates apoptosis in prostate cancer cells with activated hedgehog signaling. Figure 8 Cyclopamine induces apoptosis in prostate cancer cells. Cyclopamine-mediated apoptosis in prostate cancer cells was analyzed by TUNEL assay. TUNEL positive cells were indicated by arrowheads. Cells with expression of Gli1 under the CMV promoter (indicated by the arrows) did not undergo apoptosis (n = 500). All these data indicate that the hedgehog pathway is activated in advanced prostate cancers, as indicated by high expression of PTCH1 and HIP. Our results also indicate that hedgehog signaling is required for cell proliferation and cell invasion of prostate cancer cells. Thus, targeted inhibition of the hedgehog pathway may be effective in future prostate cancer therapeutics. Discussion Hedgehog signaling pathway regulates cell proliferation, tissue polarity and cell differentiation during normal development. Abnormal signaling of this pathway has been reported in a variety of human cancers, including basal cell carcinomas, medulloblastomas, small cell lung cancer and GI cancers [ 3 , 4 , 6 - 10 , 22 , 23 ]. Our findings in this report indicate a role of the sonic hedgehog pathway in prostate cancer. We detected a high expression of hedgehog target genes, PTCH1 and HIP, in advanced or metastatic prostate cancers. In contrast, only 22% of prostate tumors with Gleason scores 3–6 have elevated expression of PTCH1 and HIP. While our manuscript is being reviewed, three independent groups have recently reported similar results [ 24 - 26 ]. Thus, the hedgehog signaling pathway is frequently activated in advanced or metastatic prostate cancers. Alterations of genes in the hedgehog pathway in prostate cancer In our studies, we found that some prostate tumors had no detectable Su(Fu) protein expression while others contained high levels of Shh protein expression. We further identified inactivated mutations of Su(Fu) in two prostate cancers. In addition to inactivated mutations in the coding region, Su(Fu) may be inactivated through promoter methylation. The heterogeneous nature of prostate cancer makes it difficult to screen prostate cancer specimens for Su(Fu) mutations since the tumor content is often less than 5% of the specimens. Future improvement can be achieved using microdissection techniques for collecting pure population of tumor cells in gene mutation analysis. Since all available prostate cancer cell lines express Su(Fu) at a high level, the role of Su(Fu) on cellular functions of prostate cancer cannot be investigated in these cells. It appears that over-expression of sonic hedgehog may be responsible for hedgehog signaling activation in these cell lines [our unpublished data and [ 24 - 26 ]]. After screening over 30 human cancer cell lines, we identified non-prostate cancer cell line with elevated hedgehog target genes and no detectable Su(Fu) expression (data not shown here). The growth suppression effects of Su(Fu) was demonstrated in this cell line, in which Su(Fu) expression down-regulated hedgehog target genes, inhibited DNA synthesis and cell growth (data not shown here). Thus, inactivation of Su(Fu) can contribute to active hedgehog signaling in prostate cancer. Su(Fu) is reported to affect β-catenin signaling [ 27 , 28 ]. We analyzed expression of β-catenin and E-cadherin in our prostate cancer array and detected cytoplasmic distribution of E-cadherin and β-catenin only in PC51 (data not shown), indicating that Su(Fu) may be able to affect both the wnt pathway and the hedgehog pathway in prostate cancer. In addition to Su(Fu) inactivation, over-expression of Shh is another mechanism by which the hedgehog pathway is activated in cancer [ 7 - 10 ]. We noticed that sonic hedgehog expression varies from tumor to tumor, which may be resulted from the heterogeneity of prostate cancer. Our immunohistostaining also revealed that three tumors (PC14, PC20 and PC36) expressed PTCH1 and HIP at high levels, but had no alterations of Shh and Su(Fu). This could be due to elevated expression of indian hedgehog, or even alterations of other components of the pathway (such as Rab23 or Fused). Once hedgehog pathway is activated, the target gene expression will be up-regulated. Thus, analysis of target gene expression using immunohistochemistry will be an effective way to detect hedgehog pathway activation in prostate cancer. Currently, PTCH1, Gli1 and HIP are good markers for the hedgehog pathway. Perspectives on prostate cancer therapy Our findings not only provide novel basic understanding of prostate cancer, but also allow us to design new ways to treat prostate cancer. With a specific SMO antagonist, cyclopamine, it will be possible in the future to treat prostate cancers, which have over-expressed sonic hedgehog. However, as a downstream molecule, tumors with Su(Fu) inactivation may not respond to cyclopamine treatment. Therefore, additional small molecule inhibitors appear to be necessary to treat Su(Fu) inactivated prostate cancer. One possibility is to use Gli1 SiRNA since we have indicated that down-regulation of Gli1 may be an important mechanism by which inhibition of the hedgehog pathway by cyclopamine induces apoptosis (Fig. 8 ). Sanchez et al also indicated that Gli1 SiRNA down-regulated DNA synthesis in prostate cancer cells [ 24 ]. Conclusion Taken together, our findings suggest that activation of the hedgehog pathway involves prostate cancer progression. There might be several mechanisms by which the hedgehog pathway is activated in advanced prostate cancers, including loss of Su(Fu) protein expression, over-expression of sonic hedgehog or other alterations. We demonstrate that activation of the hedgehog pathway is associated with DNA synthesis and cell invasiveness in prostate cancer cells. Inhibition of the hedgehog pathway, on the other hand, causes apoptosis possibly through down-regulation of Gli1. Our studies predict that targeted inhibition of the hedgehog pathway may be an effective way to prevent prostate cancer progression. Materials and methods Tissue Microarray of Prostate Cancer A total of 55 paraffin-embedded tissue blocks from patients with prostate cancer were obtained from UTMB Surgical pathology with approval from the Institutional Review Board (IRB). Pathological reports, H#E staining of each specimen were reviewed to determine the nature of the disease and the Gleason scores. Of 55 specimens, 18 were from tumors with Gleason scores 3–6, 15 with Gleason score 7 and 22 with Gleason scores 8–10. The tumor area was first identified before tissue microarray (1.5 mm in diameter for specimens) was assembled with Beecher's Tissue arrayer-I ® according to manufacturer's instruction . Immunohistochemistry and Western blotting A standard avidin-biotin immunostaining technique was performed using a kit from Vector laboratories using specific antibodies to Su(Fu) (Santa Cruz Biotechnology Cat# 10933), PTCH1 (Santa Cruz Biotechnology Cat# 6149), HIP (R&D systems Cat# AF1568) and Shh (Santa Cruz Biotechnology Cat# 9024) and PSA (Vector laboratories). Positive staining was in red or brown. The specificity of antibodies was tested using the very peptide used for raising the antibodies, which abolished the specific staining. Hematoxylin was used for counterstaining (in blue). Protein was analyzed by Western analysis with appropriate antibodies [Su(Fu) antibodies were from Santa Cruz, beta-actin antibody was purchased from Sigma). The signals were visualized with the enhanced chemiluminescence detection system (Amersham). Cell lines and Cell invasion assay Cell lines (RWPE-1, Du145, PC3, LN-CAP were purchase from ATCC and cultured according to the suggested conditions. TSU was kindly provided by Dr. Allen Gao. Cell invasion assay was performed with BD Bio-coat cell invasion chambers according to manufacturer's instruction (BD Bioscience, Inc., Franklin Lakes, NJ), with triplicates for each sample and the experiment was repeated three times with the similar results. Cell were treated with 5 μM cyclopamine (or tomatidine) before (for 12 h) and during cell invasion assay (for 24 h). The rate of cell invasion was calculated by dividing cell numbers penetrated the matrigels by the number of cell in the control chambers (without matrigels). RT-PCR and sequencing analysis Total RNA was isolated using Trizol ® reagent (Invitrogen), and RT-PCR was performed using Promega's RT-PCR system according to the manufacturer's protocol. Two pairs of Su(Fu) primers were used (the first set with the forward primer 5'-cctacgcaccccgatggcg-3" and the reverse primer 5'-agccaaaaccactacctcca-3'; the second set with the forward primer 5'-tccaggttaccgctatcgtc-3' ad the reverse primer 5'-tagtgtagcggactgtcg-3'). PCR products were first purified using Qiagen's Gel Extraction Kit. Due to existence of possible Su(Fu) splicing isoforms in humans, Su(Fu) genetic mutations were screened after the PCR products were cloned into TOPO ® TA cloning vectors (Invitrogen). Several independent clones (from three experiments) of each PCR product were selected for sequencing analysis in UTMB sequencing facility. All mutations were confirmed by at least six independent clones. Real-time PCR We used Applied Biosystems' assays-by-demand 20× assay mix of primers and TaqMan probes (FAM™ dye-labeled) for the target genes (human Gli and PTCH1, the sequences have been patented by Applied Biosystems, Foster City, CA) and pre-developed 18S rRNA (VIC™-dye labled probe) TaqMan ® assay reagent (P/N 4319413E) for an internal control. The primers are designed to span exon-exon junctions so as not to detect genomic DNA and the primers and probe sequences were searched against the Celera database to confirm specificity. To obtain the relative quantitation of gene expression, a validation experiment was performed to test the efficiency of the target amplification and the efficiency of the reference amplification. All absolute values of the slope of log input amount vs. ΔC T were <0.1. Separate tubes (singleplex) one-step RT-PCR was performed with 20 ng RNA for both target genes and endogenous control. The reagent we used was TaqMan one-step RT-PCR master mix reagent kit (P/N 4309169). The cycling parameters for one-step RT-PCR was: reverse transcription 48°C for 30 min, AmpliTaq activation 95°C for 10 min, denaturation 95°C for 15 sec and annealing/extension 60°C for 1 min (repeat 40 times) on ABI7000. Triplicate C T values were analyzed in Microsoft Excel using the comparative C T (ΔΔC T ) method as described by the manufacturer(Applied Biosystems, Foster City, CA). The amount of target (2 -ΔΔCT ) was obtained by normalization to an endogenous reference (18sRNA) and relative to a calibrator. BrdU labeling and TUNEL assay BrdU labeling was performed using an in situ cell proliferation kit (Roche Molecular Biochemicals) [ 22 ]. Cells were treated with 5μM cyclopamine (or tomatidine) for 12 h before BrdU labeling (1 h at 37°C). The percentage of BrdU positive cells was obtained by counting over 1000 cells under microscope, and the experiment was repeated twice with similar results. TUNEL assay was performed using an in situ cell death kit (Roche Molecular Biochemicals) [ 21 , 29 ]. Cells were treated with 5 μM cyclopamine (or tomatidine) for 36 h before TUNEL assay). List of abbreviations PSA – prostate specific antigen; HIP – hedgehog-interacting protein; Su(Fu) – suppressor of fused; PTCH1 – human homologue of patched 1; Shh – sonic hedgehog; SMO – smoothened, BCC – basal cell carcinoma. Authors' contributions Tao Sheng contributed to Figures 6 , 7 , 8 , cellular functions of the hedgehog pathway in prostate cancer cells. Chegxin Li contributed to primary tumor protein expression, particularly on Su(Fu) expression. Xiaoli Zhang contributed to mutation analyses of Su(Fu) in prostate cancer and real-time PCR analyses. Sumin Chi contributed to HIP antibody test (Fig. 3A and 3B ). Nonggao He contributed to HIP antibody staining (Fig. 3C ). Kai Chen contributed to PTCH1 antibody test (Fig. 1A ). Frank McCormick involved in the initial project discussion. Zoran Gatalica contributed to prostate cancer histology and Gleason scores of the tumors. Supplementary Material Additional File 1 Table 1 Prostate cancer specimens and protein expression. Prostate cancer specimens and expression of several hedgehog signaling proteins are summarized in this table ( A ). A total of 55 specimens were used in this study. The Gleason scores and protein expression of Shh, PTCH1 and Su(Fu) are shown ( B ). Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC524523.xml |
524251 | Tuberculous Granuloma Formation Is Enhanced by a Mycobacterium Virulence Determinant | Granulomas are organized host immune structures composed of tightly interposed macrophages and other cells that form in response to a variety of persistent stimuli, both infectious and noninfectious. The tuberculous granuloma is essential for host containment of mycobacterial infection, although it does not always eradicate it. Therefore, it is considered a host-beneficial, if incompletely efficacious, immune response. The Mycobacterium RD1 locus encodes a specialized secretion system that promotes mycobacterial virulence by an unknown mechanism. Using transparent zebrafish embryos to monitor the infection process in real time, we found that RD1-deficient bacteria fail to elicit efficient granuloma formation despite their ability to grow inside of infected macrophages. We showed that macrophages infected with virulent mycobacteria produce an RD1-dependent signal that directs macrophages to aggregate into granulomas. This Mycobacterium -induced macrophage aggregation in turn is tightly linked to intercellular bacterial dissemination and increased bacterial numbers. Thus, mycobacteria co-opt host granulomas for their virulence. | Introduction Infection with pathogenic mycobacteria is thought to proceed through a series of defined steps. Mononuclear cells present at or recruited to sites of infection phagocytose bacteria and migrate deeper into tissues. Then additional macrophages and other immune cells are recruited to form complex, tightly aggregated structures called granulomas ( Adams 1976 ; Dannenberg 1993 ; Teitelbaum et al. 1999 ; Geijtenbeek et al. 2003 ; Peters and Ernst 2003 ; Tailleux et al. 2003 ; Cosma et al. 2004 ). Granuloma macrophages subsequently undergo differentiation into epithelioid cells, so called owing to their closely apposed cellular membranes ( Adams 1976 ; Dannenberg 1993 ; Cosma et al. 2004 ). Production and maintenance of granulomas is essential to the control of tuberculosis in murine models and humans ( Kaufmann 2000 ; Flynn and Chan 2001 ; Lawn et al. 2002 ). However, despite residence at the site of a robust focal immune response, the bacilli within granulomas are not always eradicated ( Cosma et al. 2003 ; Cosma et al. 2004 ). The relative contributions of host and pathogen determinants to the migration and aggregation of macrophages and the formation and maintenance of granulomas are not understood. While considerable progress has been made in identifying Mycobacterium virulence determinants ( Glickman and Jacobs 2001 ; Cosma et al. 2003 ; Smith 2003 ), our overall understanding of mycobacterial pathogenesis remains rudimentary. Virulence determinants are generally studied by examining mutant bacterial strains in cultured macrophage monolayers or by static assessment of bacterial numbers and tissue pathology at different time points during in vivo infection. During in vivo infection, spatially separated individual mononuclear cells are infected, migrate into tissues, and serve as a nidus for cellular aggregation ( Teitelbaum et al. 1999 ; Davis et al. 2002 ; Geijtenbeek et al. 2003 ; Tailleux et al. 2003 ). The steps and dynamics of bacterially mediated host-cell interactions that impact the outcome of infection through production of chemokines, cytokines, adhesion molecules, and their receptors cannot be elucidated by tissue culture studies and static assessments in vivo. To address these issues, we utilize a novel model of Mycobacterium infection in zebrafish. Zebrafish embryos and larvae (henceforth we will refer to both stages as embryos) are naturally susceptible to infection by Mycobacterium marinum, and we have previously shown that their optical transparency may be used to monitor the cellular dynamics of infection in real time ( Davis et al. 2002 ). Using differential interference contrast (DIC) video and fluorescence microscopy, we have monitored macrophage chemotaxis to M. marinum, its phagocytosis, transit of infected macrophages into tissues, and the recruitment of additional macrophages to initiate granulomas that have the pathological hallmarks and bacterial gene expression profile characteristic of tuberculous granulomas in adult animals. In this study, we use the zebrafish infection model to probe the cellular mechanisms of virulence of the Mycobacterium RD1 locus, an approximately 10-kb region missing from all attenuated bacille Calmette-Guérin (BCG) vaccine strains but present in virulent M. tuberculosis isolates ( Mahairas et al. 1996 ; Behr and Small 1999 ). The RD1 locus encodes a specialized secretion system for the putative virulence effector proteins ESAT-6 and CFP-10, also located within the locus ( Tekaia et al. 1999 ; Pallen 2002 ; Hsu et al. 2003 ; Pym et al. 2003 ; Stanley et al. 2003 ; Guinn et al. 2004 ). The M. tuberculosis RD1 deletion mutant has a growth defect in the mouse model of tuberculosis ( Lewis et al. 2003 ), and studies using cultured macrophages and other in vitro systems have identified complex phenotypes that may account for its in vivo attenuation ( Hsu et al. 2003 ; Stanley et al. 2003 ; Guinn et al. 2004 ). In vitro assays have suggested that RD1 may contribute to mycobacterial cytotoxicity to macrophages and epithelial cells, thus enabling bacterial spread between cells or transit across epithelial barriers ( Hsu et al. 2003 ; Guinn et al. 2004 ). Others have proposed that the RD1 region mediates dampening of host innate immune responses in macrophages ( Stanley et al. 2003 ). It is unclear how each of these individual in vitro phenotypes contributes to the complex sequence of events that ultimately lead to bacterial persistence in granulomas. We used the zebrafish- M. marinum infection model to elucidate the precise steps at which infection with wild-type (WT) and RD1 mutant bacteria differ. Our data suggest that the RD1 locus independently mediates macrophage aggregation and intercellular bacterial spread via host cell death within aggregates. These steps are associated with increased bacterial numbers and enhanced virulence, lending support to the idea that mycobacteria actually promote and exploit granuloma formation for the establishment of infection. Results The M. marinum RD1 Mutant Is Attenuated for Growth in Cultured Macrophages and Adult Frogs The genes in the M. marinum RD1 region are homologous to those in M. tuberculosis (for instance, their ESAT-6 and CFP-10 proteins are 97% and 91% identical, respectively) and the regions in the two organisms are syntenic ( http://www.sanger.ac.uk/Projects/M_marinum/ ; Figure 1 ). We derived an M. marinum RD1-deficient mutant with essentially the same deletion as the M. tuberculosis RD1 mutant described previously ( Figure 1 ) ( Lewis et al. 2003 ). Like the M. tuberculosis RD1 mutant ( Lewis et al. 2003 ), the M. marinum mutant (referred to as ΔRD1) was attenuated for growth in mouse and human monocyte/macrophage cell lines ( Figure 2 A and unpublished data). ΔRD1 was also attenuated for growth in an adult leopard frog infection model ( Figure 2 B) in which WT M. marinum causes chronic granulomatous infection ( Ramakrishnan et al. 1997 ). Specifically, significantly fewer ΔRD1 bacteria were recovered from spleens and livers of infected frogs at 2, 8, and 24 wk postinfection ( Figure 2 B and unpublished data). ΔRD1-infected frogs also had poorly formed macrophage aggregates at 8 wk postinfection, in contrast to the well-defined granulomas resulting from WT infection (unpublished data). Thus, by previously evaluated parameters, the RD1 region plays identical roles in the virulence of M. tuberculosis and M. marinum . Figure 1 The RD1 Regions in M. tuberculosis and M. marinum Are Homologous and Syntenic The white arrows represent the RD1 region deleted from M. tuberculosis . The black arrow represents a predicted open reading frame not present in M. tuberculosis. Rv3874 and Rv3875 are also known as cfp-10 and esat-6, respectively. Numbers represent the percent amino acid identities between the corresponding proteins of the two organisms. Figure 2 M. marinum ΔRD1 Is Attenuated In Vitro and In Vivo (A) Growth of M. marinum WT and ΔRD1 in J774 cells. Each time point represents the average of triplicate values. Error bars are ± standard error of the mean (SEM). (B) WT and ΔRD1 bacterial numbers in frog spleens. Each time point represents the average colony counts from 3–5 frogs. Error bars are ± SEM (* p ≤ 0.05, ** p = 0.016, unpaired Student's t -test). Infecting doses were 5.8 × 10 5 CFU for WT and 1.2 × 10 6 CFU for ΔRD1. M. marinum ΔRD1 Infection of Zebrafish Embryos Results in Reduced Macrophage Aggregation We injected fluorescent WT or ΔRD1 bacteria via the caudal vein directly into the bloodstream of 30 h postfertilization embryos ( Figure 3 A), which were then monitored for survival and bacterial load ( Figure 3 B and 3 C; Materials and Methods ) ( Davis et al. 2002 ). In contrast to WT bacteria, ΔRD1 failed to kill the embryos during the 12 d monitoring period ( Figure 3 B). Consistent with this difference in mortality, ΔRD1 bacterial growth was attenuated as compared to WT bacteria in the embryos ( Figure 3 C). Figure 3 ΔRD1 Is Attenuated in Zebrafish Larvae (A) Diagram of the zebrafish embryo/larva. Arrows indicate the two injection sites used in this study. (B) Survival of embryos infected with ΔRD1 (410 CFU) or WT bacteria (250 CFU) and null-injected embryos. (C) Whole embryo bacterial counts of WT- and ΔRD1-infected embryos. Infecting doses: 32 CFU for WT, 36 CFU for ΔRD1. Error bars are ± SEM (** p = 0.0075 comparing 7-d postinfection WT to 7-d postinfection ΔRD1; * p = 0.05 comparing 9-d postinfection WT to 9-d postinfection ΔRD1, unpaired Student's t -test). (D) Time of aggregate formation, showing delayed aggregation in the ΔRD1-infected embryos ( n = 13) as compared to WT-infected embryos ( n = 15). Infecting doses: 131 CFU for WT, 301 CFU for ΔRD1. (E) Whole embryo bacterial counts of WT- and ΔRD1-infected embryos on day of aggregate formation. Infecting doses: 36 CFU for WT, 78 CFU for ΔRD1. Error bars are ± SEM (*** p = 0.0008, unpaired Student's t -test; ΔRD1 n = 28, WT n = 29). (F) Fluorescent image of WT-infected embryo at 6 d postinfection with two aggregates (arrows). Scale bar, 200 μm. (G) WT-infected embryos with higher magnification overlay of fluorescent and DIC images showing an aggregate (arrow) with individual infected macrophages that are migrating toward aggregate (arrowheads). Scale bar, 50 μm. (H) Fluorescent image of ΔRD1-infected embryo at 6 d postinfection that has not formed any aggregates. Note the numerous infected macrophages throughout the head, body, and tail. Arrowhead and close-up insert (scale bar, 50 μm) show infected macrophages close to each other, but not aggregating. Scale bar, 200 μm. (I) ΔRD1-infected embryo under higher-magnification overlay of DIC and fluorescent images showing three individual infected macrophages (arrowheads). Scale bar, 50 μm. To understand the cellular basis of ΔRD1 attenuation, we undertook real-time microscopic monitoring of the infection process with WT and mutant bacteria. WT infection of the embryos is characterized by the transit of infected macrophages into tissues where macrophages are recruited to form granulomas within 3–5 d postinfection ( Figure 3 D, 3 F, and 3 G) ( Davis et al. 2002 ). In contrast, while ΔRD1 infected macrophages also migrated from the circulation to the tissues ( Figure 3 H), fewer, if any, aggregates formed, and the kinetics of their formation were delayed compared to the WT-infected cells ( Figure 3 D). Several highly infected individual macrophages were found scattered throughout the tissues, often close to each other ( Figure 3 H and 3 I). This is in sharp contrast to the case of WT infection, in which infected macrophages are nearly always found in aggregates ( Davis et al. 2002 ). Aggregates that formed in ΔRD1-infected embryos were more transient than those in WT-infected embryos, often dissociating into individual infected macrophages ( Figure 3 D and unpublished data). Also, the ΔRD1 aggregates remained small in contrast to WT aggregates, which often increased dramatically in size ( Figure 4 , compare images in [A] to those in [B]). This finding suggests that RD1 is required not only to initiate aggregate formation but for an ongoing recruitment of macrophages into the aggregate. Figure 4 Progression of Aggregates a WT Aggregate (A), and a ΔRD1 Aggregate (B) (A) WT aggregates shown on the first day of aggregate formation ( t = 0 h); 24 h after aggregate formation ( t = 24 h); and 48 h after aggregate formation ( t = 48 h). (B) ΔRD1 aggregates shown at the same time points as in (A). A 60× water lens was used for all photomicrographs except the image in (A) t = 48 h, which was taken with a 40× lens. Scale bar represents 50 μm. Since M. marinum ΔRD1 is attenuated for growth in the embryos (see Figure 3 C), we considered the possibility that this mutant strain did not replicate enough to reach the threshold bacterial numbers that might be required to stimulate host pathways for macrophage aggregation. In that case, the inability of ΔRD1 to induce macrophage aggregation would be a simple consequence of its primary replication defect in the embryos. This scenario would predict that the number of bacteria required for aggregate formation would be similar in WT and ΔRD1 infection. To investigate this possibility, we infected embryos with similar numbers of the two bacterial strains and examined them daily for aggregate formation. On the day each embryo developed an aggregate(s), it was lysed and bacterial colony-forming units (CFU) determined. The bacterial load at which macrophage aggregation first occurred was over 4-fold higher in ΔRD1- than WT-infected embryos (see Figure 3 E). These results show that ΔRD1 infection is associated with a primary aggregation defect that is not a consequence of its decreased replication in macrophages. The RD1 Locus Specifically Mediates Macrophage Aggregation Macrophages are rapidly recruited to the site of WT M. marinum infection, where they phagocytose the bacteria, migrate to the tissues, and form aggregates ( Davis et al. 2002 ). Since macrophage migration and aggregation are likely mediated by as yet ill-defined chemotactic networks, we asked if the RD1 locus also affected other chemotactic macrophage functions. Macrophage recruitment is most stringently assessed by injecting bacteria into the hindbrain ventricle, an isolated cavity devoid of macrophages in the absence of bacteria (see Figure 3 A) ( Herbomel et al. 1999 ; Davis et al. 2002 ). Similar numbers of macrophages migrated to the hindbrain ventricle in response to the injection of WT and ΔRD1 bacteria at 4 h postinfection, and most of the bacteria had been phagocytosed in both cases ( Figure 5 A; unpublished data). Therefore, the RD1 locus does not affect macrophage chemotaxis to the bacteria or phagocytic capabilities. Figure 5 Normal Macrophage Chemotaxis to Initial Sites of ΔRD1 Infection Overlay of DIC and fluorescent images showing the hindbrain ventricle (hv) of infected embryos. The hindbrain ventricle/brain (hv/b) boundary indicated by a white dashed line. (A) ΔRD1-infected embryo 4 h postinfection with individual infected macrophages marked by arrowheads. (B) ΔRD1-infected embryo 5 h postinfection in which individual infected macrophages (arrowheads) have migrated from the hindbrain ventricle and into the brain. (C) WT-infected embryo 24 h postinfection with macrophages beginning to aggregate (white arrow) in the hindbrain ventricle. A second out-of-focus aggregate is to the left (yellow arrow). Scale bar, 100 μm. The abundance of ΔRD1-infected macrophages in tissues following bloodstream infection (see Figure 3 H and 3 I) suggested that the RD1 locus is not required for tissue migration following infection. However, tissue migration can also be examined more stringently following the ventricle injection assay ( Davis et al. 2002 ). 5 h after infection of the ventricle, many of the ΔRD1-infected macrophages had entered the brain tissue ( Figure 5 B). By 24 h, most of the macrophages were widely disseminated throughout the tissues (unpublished data). Indeed, the lack of aggregation by RD1-infected macrophages led to their enhanced tissue dissemination compared to WT-infected macrophages, which had often formed aggregates within the ventricle itself by 24 h postinfection ( Figure 5 C). Likely as a result, the WT-infected macrophages were slower to migrate out of the ventricle than ΔRD1-infected macrophages. Even when they did migrate out of the ventricle, they often formed aggregates in the adjacent brain tissue and did not disseminate into the trunk and tail as rapidly as did the ΔRD1-infected macrophages. In summary, the ventricle infections showed that the RD1 locus is not required for macrophage chemotaxis to the site of infection, bacterial phagocytosis, or tissue migration of infected macrophages. Furthermore, this assay highlighted the difference in the aggregation of WT- and ΔRD1-infected macrophages from very early in infection. ΔRD1-Infected Macrophages Can Receive, but Not Send, Signals That Promote Aggregation The aggregation defect of ΔRD1-infected macrophages suggests that they lack the capacity to either produce or receive signals that mediate aggregation of macrophages during WT infection. To begin to dissect the nature of the missing signal(s), we infected embryos with red-fluorescent WT bacteria and allowed aggregates to form. These embryos were then superinfected with green-fluorescent ΔRD1 or WT bacteria ( Figure 6 ). Both superinfecting strains were phagocytosed by individual macrophages, which migrated in similar numbers to preexisting aggregates within 4 h ( Figure 6 A and 6 B). These data indicate that ΔRD1-infected macrophages can receive signals produced by WT-infected macrophages and migrate rapidly toward aggregates. Figure 6 Superinfection with WT Bacteria Rescues ΔRD1 Aggregation Defect (A and B) Embryos with aggregates at 3d postinfection with 85 CFU red-fluorescent WT bacteria are shown 4 h after superinfection with green-fluorescent strains of either ΔRD1 (134 CFU) (A) or WT (169 CFU) (B) bacteria. Superinfecting strains were injected at sites distant from the aggregates, and pictures were taken outside of injection regions. Arrowheads indicate macrophages infected with superinfecting strain. Scale bar, 100 μm. (C) Embryo infected with 171 CFU green-fluorescent ΔRD1 for 4 d shown 4 h post-superinfection with 364 CFU of red-fluorescent ΔRD1. Arrowheads point to macrophages infected with each of the bacterial strains. Scale bar, 200 μm. (D) Embryo infected with 171 CFU green-fluorescent ΔRD1 for 4 d shown 4 h after superinfection with 363 CFU of red-fluorescent WT bacteria. Arrow points to macrophage aggregate. Scale bar, 200 μm. (E) Higher magnification image of aggregate (arrow) in (D) showing green fluorescent ΔRD1 and red fluorescent WT bacteria. Arrowhead points to WT-infected macrophage outside the aggregate. Scale bar, 50 μm. (F and G) Embryo infected with green fluorescent ΔRD1, superinfected with red fluorescent WT (as in D and E) shown at 24 h post-superinfection (F), and the same aggregate at 48 h post-secondary infection (G). Scale bars, 50 μm. All panels are fluorescent images. In a reciprocal experiment, embryos were first infected with green-fluorescent ΔRD1 bacteria, and after 4 d, when there were abundant individual infected macrophages (but no aggregates) in the tissues, the embryos were superinfected with either WT or ΔRD1 red-fluorescent bacteria. As expected, both superinfecting strains were rapidly phagocytosed by uninfected macrophages. ΔRD1 superinfection did not cause aggregate formation, and individual macrophages carrying both the original or superinfecting ΔRD1 were scattered throughout the tissues ( Figure 6 C). In contrast, WT-infected macrophages induced the aggregation of preexisting ΔRD1-infected macrophages as early as 4 h after superinfection ( Figure 6 D, 6 E, and 7 ). These newly formed aggregates were often composed mostly of ΔRD1-infected macrophages with only a few WT-infected macrophages in them ( Figure 7 ). All aggregates had at least some WT-infected macrophages. Furthermore, ΔRD1/WT aggregates that formed developed normally, increasing in size and recruiting both ΔRD1- and WT-infected macrophages ( Figure 6 F and G ). Thus, this experiment confirmed that ΔRD1-infected macrophages receive but do not send aggregation signals and have no intrinsic chemotactic defects. Furthermore, it appears that WT-infected macrophages are required to serve as a nidus for each aggregate, suggesting that they create a chemotactic gradient that recruits macrophages. Figure 7 Superinfection with WT Bacteria Rescues ΔRD1 Aggregation Defect over Time Embryos were injected with fluorescent ΔRD1 (green) at 1 d postfertilization. 3 d post-primary infection, embryos were injected with fluorescent WT (red) and followed for 24 h post-secondary infection. Approximate injection sites are shown with green and red arrows for ΔRD1 and WT bacteria, respectively. Box in top panel indicates the magnified field in fluorescent images. Inset panel at 24-h time point is a magnified image of the starred aggregate. Scale bar, 125 μm. Macrophage Aggregation Is Tightly Linked to Intercellular Bacterial Spread Having demonstrated that ΔRD1 infection results in both reduced aggregation and lower bacterial numbers, we next pursued experiments to determine the relationship between these two phenotypes. In contrast to the notion that a primary reduction in bacterial numbers obviated the need for aggregation (see Figure 3 E), we found that more ΔRD1 than WT bacteria were required for aggregates to form. Therefore, we sought to determine if, conversely, the aggregation defect resulted in reduced bacterial numbers. One way that aggregation could impact bacterial numbers is by facilitating the spread of bacteria to uninfected macrophages that are recruited to the aggregates. If so, then aggregate formation should correlate with a dramatic increase in the number of infected macrophages and bacterial burdens. To test this hypothesis, we assessed the number of infected macrophages and bacterial numbers in relation to the time of aggregate formation ( Figure 8 ). We enumerated daily by microscopy the number of infected macrophages during the course of infection starting at 1 d postinjection of bacteria and continuing up to 2 d after aggregates formed ( Figure 8 A). We counted as day 0 the first day of aggregation. During WT infection, the number of infected macrophages did not change significantly until aggregates formed ( Figure 8 A). However, upon aggregation, the number of infected macrophages increased dramatically ( Figure 8 A). Similarly, the number of viable bacteria also did not increase until after aggregation occurred 3–5 d postinfection ( Figure 8 B). Taken together, these data suggest that during WT infection, macrophage aggregation promotes intercellular bacterial spread and an increased bacterial burden. Figure 8 Macrophage Aggregation Correlates with Bacterial Dissemination during WT Infection (A) Enumeration of infected macrophages in embryos by fluorescent and DIC microscopy after infection with green-fluorescent bacteria. Infecting doses: 151 CFU for WT, 301 CFU for ΔRD1. Time points are in reference to day of aggregate formation, which is set at 0. 15 WT infected embryos and 13 ΔRD1 embryos were monitored. The graph represents all 15 WT embryos, but only the 7/13 ΔRD1 infected embryos that formed aggregates over the course of the experiment. Error bars are ± SEM. (* p = 0.0136 comparing WT day 0 and WT day –2; ** p = 0.0053 comparing WT day 1 and WT day –2, unpaired Student's t -test). (B) Whole embryo bacterial counts following WT infection (*** p ≤ 0.0003, 5 d postinfection and 8 d postinfection, respectively, compared to 3 d postinfection, unpaired Student's t -test). In the case of ΔRD1 infection, macrophage aggregation did not result in an increase in the number of infected macrophages ( Figure 8 A). This difference could be solely due to the ongoing defect in macrophage recruitment by the ΔRD1-containing aggregates. It could also involve additional pathways that result in decreased bacterial spread to uninfected macrophages in the aggregates. In either case, this result suggests that while aggregation is required for intercellular bacterial spread, it is not sufficient. Additional RD1-mediated events must occur to facilitate spread after aggregation. WT Aggregates Have More Cell Death Than ΔRD1 Aggregates Having established that RD1 is involved in macrophage recruitment to aggregates, we sought to determine if it also affects intercellular bacterial spread by additional means. The M. tuberculosis RD1 locus is thought to promote intercellular bacterial spread in confluent cultured macrophage monolayers by promoting death of infected cells and subsequent phagocytosis of the released bacteria by surrounding cells ( Guinn et al. 2004 ). Therefore, we hypothesized that RD1 might operate similarly in the embryo aggregates to mediate bacterial spread by facilitating cell death. For this assessment, we achieved comparably sized aggregates with the two strains by infecting embryos with 6.6-fold more ΔRD1 than WT bacteria, and performed terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick end labeling (TUNEL) staining on whole infected embryos to visualize dead and dying cells. The TUNEL reaction labels double-stranded DNA breaks that can occur during apoptosis and certain forms of necrosis ( Gavrieli et al. 1992 ). WT infected embryos had more aggregates with TUNEL-positive cells (15 of 23) than did ΔRD1-infected embryos (6 of 22) ( Figure 9 ). Contingency table analysis revealed that WT aggregates were 2.3 times more likely to contain TUNEL-positive cells than ΔRD1 aggregates ( p = 0.017, Fisher's exact t -test). The TUNEL-positive cells were often M. marinum -infected ( Figure 9 A). Furthermore, TUNEL-positive infected cells were found almost exclusively within aggregates. Taken together, our data suggest that the RD1 locus first mediates macrophage aggregation and subsequently promotes cell death within the aggregates. Figure 9 WT Aggregates Are More Likely to Have TUNEL-Positive Cells Than ΔRD1 Aggregates Representative fluorescent images of aggregates following TUNEL staining of 6-d postfertilization embryos infected with 71 green-fluorescent WT (A), or 474 green-fluorescent ΔRD1 (B) bacteria. TUNEL staining is imaged with red fluorescence, and colocalization with green-fluorescent bacteria appears yellow. Scale bar, 100 μm. The RD1 Locus Continues to Mediate Granuloma Formation During Long-Term Infection We have previously used the zebrafish embryo model to demonstrate that granulomas can be initiated solely by interactions of mycobacteria with innate immunity ( Davis et al. 2002 ). However, their maturation and an enhancement in their mycobacteriocidal potential likely requires the participation of adaptive immunity ( Flynn and Chan 2001 ; Davis et al. 2002 ). Thus, the zebrafish embryo model has been useful to separate the effects of innate and adaptive immunity on granuloma formation ( Davis et al. 2002 ). Since both the M. tuberculosis and M. marinum RD1 mutants exhibit a sustained attenuation during chronic infection of adult animals (see Figure 2 B) ( Lewis et al. 2003 ), we wished to determine if the early aggregation defect we had discovered using the zebrafish embryo model impacts granuloma formation and maturation later in infection. We infected embryos with low doses of either WT or ΔRD1 bacteria and confirmed infection microscopically at 6 d postfertilization. Using these low infection doses, we could raise a very few WT-infected embryos to adulthood. In contrast, the mortality of the ΔRD1-infected embryos was no different from that of uninfected embryos during a 32 d observation period (see Figure 3 B) (unpublished data). At 32 d, the fish were assessed by tissue histopathology. Mycobacteria were identified within the granulomas in all of the surviving fish, showing that they were chronically infected ( Figure 10 ). WT-infected fish had highly organized granulomas, both caseating and noncaseating in roughly equal proportions ( Figure 10 A and 10 C), with bacteria located predominantly in the caseum ( Figure 10 B and 10 D). These granulomas appeared identical to granulomas resulting from infection of adult zebrafish ( Cosma et al. 2004 ). In contrast, ΔRD1-infected fish had only a few granulomas ( Figure 10 E and 10 F), all of which were noncaseating and markedly different even from WT noncaseating granulomas ( Figure 10 C and 10 E). The WT-induced granulomas were compact and were composed of tightly packed cells that displayed the indistinct cytoplasmic borders and abundant eosinophilic cytoplasm characteristic of epithelioid cells ( Figure 10 C) ( Adams 1976 ; Bouley et al. 2001 ). In contrast, the ΔRD1-induced aggregates had more loosely aggregated cells, with evidence of epithelioid transformation only in the centers of some ( Figure 10 E). Figure 10 ΔRD1 Infection Is Associated with Persistent Defects in Granuloma Organization Tissue histology of 32-d postfertilization fish infected with either 21 WT (A–D) or 9 ΔRD1 (E–F) bacteria (doses were not significantly different p = 0.15) at 1 d postfertilization. Arrows indicate granulomas and loose aggregates, arrowheads indicate caseum. Hematoxylin and eosin staining are shown in (A), (C), and (E), and modified acid-fast staining is shown in (B), (D), and (F). (A) Organized caseating WT granulomas (arrow) with central caseum (arrowhead). (B) WT granuloma showing mycobacteria predominantly in caseum with a few within epithelioid cells. (C) Noncaseating but highly organized WT M. marinum -induced granulomas showing the expected few bacteria within cells in (D). (E) Large, loose, and poorly organized macrophage aggregate of ΔRD1-infected fish with evidence of epithelioid transformation only in the center (denoted by *). (F) A few mycobacteria in the ΔRD1 aggregates. Scale bar, 100 μm. Images in (A–D) were taken with a 40× lens, whereas those in (E) and (F) were taken with a 20× lens. In summary, the ΔRD1-induced macrophage aggregates in infected embryos raised to adulthood had the same lack of organization as the lesions resulting from infection of adult animals (unpublished data) ( Sherman et al. 2004 ). These studies link our early real-time observations of phenotypes in the context of innate immunity alone to these seen later in infection. It appears that while some macrophage aggregation does occur in the absence of RD1, this locus continues to mediate aspects of macrophage chemotaxis and/or differentiation that contribute to granuloma architecture even as the infection becomes chronic. Discussion We used the zebrafish- M. marinum infection model to identify the steps at which the Mycobacterium RD1 virulence locus impacts the infection process. We found that two steps are independently affected: macrophage aggregation into granulomas and intercellular spread therein. Promotion of cell death within these aggregates appears to be at least one of the means by which RD1 affects intercellular spread. By comparing WT and ΔRD1 infection in real time, we have uncovered the promotion of granuloma formation as a mechanism of Mycobacterium virulence. The RD1 locus has been the recent subject of attention following its discovery as a major factor in the attenuation of BCG ( Pym et al. 2002 ; Lewis et al. 2003 ) and the potential of an M. tuberculosis RD1-defective strain as a candidate vaccine for tuberculosis ( Hsu et al. 2003 ; Pym et al. 2003 ). Subsequent studies suggested that it encodes a novel specialized secretion system for specific virulence effectors ( Hsu et al. 2003 ; Pym et al. 2003 ; Stanley et al. 2003 ; Guinn et al. 2004 ). Deletion of the RD1 locus in M. tuberculosis and M. marinum results in reduced bacterial numbers during infection of cultured macrophages and adult animals ( Hsu et al. 2003 ; Lewis et al. 2003 ; Stanley et al. 2003 ; Guinn et al. 2004 ; this study). Different in vitro studies have implicated the RD1 locus and its effectors ESAT-6 and CFP-10 in a variety of functions including lysis of cultured macrophage in confluent monolayers to promote intercellular spread ( Guinn et al. 2004 ), disruption of artificial membranes (taken as a surrogate for lysis of host epithelial cell layers) ( Hsu et al. 2003 ), and dampening of macrophage proinflammatory responses ( Stanley et al. 2003 ). Whether these in vitro activities operate in vivo and how they impact virulence are not known. As is the case with virtually all Mycobacterium virulence determinants, the precise steps at which RD1 impacts virulence have not been elucidated. The Mycobacterium infection model used here allows monitoring of the earliest individual stages of infection in real-time ( Davis et al. 2002 ). Using this model, we were able to confirm that RD1 promotes macrophage death in vivo. Additionally, we have shown that the RD1 locus affects an unanticipated earlier step in pathogenesis, macrophage aggregation. Aggregation of infected cells is independent of bacterial replication within individual macrophages and distinct from other macrophage functions such as their chemotaxis to the bacteria and migration back to deeper tissues. We speculate that RD1 mediates aggregation via its secreted effectors, which presumably interact with components of host macrophage signaling pathways to modulate macrophage aggregation. Some combination of chemokines, cytokines, and adhesion molecules is likely to be affected. Our superinfection experiments suggest a model by which RD1 impacts cellular signaling and aggregation. Because the ΔRD1-induced aggregates that form upon superinfection with WT bacteria always have at least one WT-infected macrophage, the RD1-induced signal likely diffuses from the infected macrophage to attract other macrophages to it to form aggregates. As ΔRD1-infected macrophages can receive but not send signals for aggregation, RD1 is likely required to induce expression of a chemotactic molecule but not its receptor. The formation of the tuberculous granuloma requires a complex cascade of interrelated signals that mediate cell recruitment, adhesion, and differentiation. In our model, ΔRD1 infection results in alterations of all three processes. The initial lack of aggregation suggests a specific defect in the ability of ΔRD1-infected macrophages to recruit additional macrophages to form aggregates. Our real-time monitoring showed that ΔRD1-infected macrophages fail to aggregate even when they are in close proximity, suggesting that the primary defect in ΔRD1-infected macrophages is in the ability to send chemotactic signals for aggregation. On the other hand, our finding that ΔRD1-infected macrophage aggregates are more transient than WT ones may implicate both chemotactic and adhesion defects. This idea is further supported by the finding that the ΔRD1-induced lesions in the adult fish are composed of loosely aggregated macrophages with little epithelioid differentiation. However, a primary defect in the ability of ΔRD1-infected macrophages to send chemotactic signals could affect subsequent expression of adhesion molecules ( Peters and Ernst 2003 ). Therefore we propose that the Mycobacterium RD1 locus induces infected macrophages to send chemotactic signals for aggregation of macrophages, which in turn affect adhesion and other downstream events that result in granuloma formation. Granuloma formation is not completely blocked upon infection with a ΔRD1 strain, as has been shown previously in the murine model of M. tuberculosis infection and in patient studies of disseminated BCG infection ( Emile et al. 1997 ; Sherman et al. 2004 ). However, our data indicate that RD1 influences early aggregation events that seem to extend into later stages of infection. Our data further suggest a model in which bacterial dissemination is facilitated by recruitment into the aggregates of uninfected macrophages that provide new habitats for further bacterial growth. Some ways in which incoming macrophages become infected could include transfer of bacteria between macrophages along membranous tethers ( Davis et al. 2002 ), actin-based motility of extravacuolar bacteria leading to intercellular transfer ( Stamm et al. 2003 ), and release of bacteria from dying infected cells. These dead cells could either release bacteria for phagocytosis by neighboring cells or be engulfed in their entirety ( Ramakrishnan and Falkow 1994 ; Davis et al. 2002 ). All of these modes of bacterial transfer are likely to be enhanced by the close juxtaposition of macrophages within aggregates. Our data are consistent with earlier reports suggesting that RD1 mediates bacterially induced toxicity to host cells ( Hsu et al. 2003 ; Guinn et al. 2004 ). While TUNEL staining is not a conclusive indication of apoptosis, it is most often associated with programmed cell death. In vitro studies indicate that apoptosis leads to bacterial cell death; however, our experiments indicate a correlation between host cell death and bacterial dissemination ( Fratazzi et al. 1999 ). Since we did not observe TUNEL-positive infected cells prior to aggregation or outside of the aggregates, we believe that aggregation precedes cell death. Mechanistically, it is possible that RD1 effectors modulate impinge upon distinct signaling pathways for cell aggregation and death. Alternatively, RD1 may impact a common molecule, such as tumor necrosis factor, that affects both processes ( Locksley et al. 2001 ). Other modes of bacterial transfer may also contribute to bacterial dissemination, and these may or may not be mediated by RD1. Our examination of early infection events in vivo may serve to identify relevant findings from the in vitro studies. For instance, our data do not support the model that RD1 contains a cytolysin for epithelial cell barriers that allows mycobacteria to penetrate directly into deeper tissues ( Hsu et al. 2003 ). Rather, these findings corroborate previous work from our laboratory and others showing that systemic dissemination of mycobacteria is effected mainly by trafficking of infected host mononuclear cells ( Teitelbaum et al. 1999 ; Davis et al. 2002 ; Geijtenbeek et al. 2003 ; Tailleux et al. 2003 ; Cosma et al. 2004 ). We show, furthermore, that RD1 is not required for this early event. Another in vitro study describes the dampening of several macrophage innate immune responses, including the cytokine tumor necrosis factor, by WT but not RD1-mutant M. tuberculosis ( Stanley et al. 2003 ). While this may be true in vivo as well, our functional approach shows that there is not a global dampening of chemotactic and innate immune responses by WT mycobacteria. Rather there is an RD1-mediated enhancement of macrophage aggregation and death. Ultimately, our studies reveal that Mycobacterium expresses specific virulence factors that enhance macrophage aggregation into granulomas, starting very early after infection. This effect correlates with bacterial dissemination and an increase in the bacterial burden. Granulomas are thought to be primarily protective host immune structures that provide a focused immune response to restrict mycobacteria. According to prevailing models, recruitment and activation of additional macrophages provide a concentrated source of immune effectors that thwart the bacteria. The specific differentiation of macrophages into epithelioid cells with tightly interdigitated intercellular membranes helps sequester the infection. While there is clear evidence that granulomas are necessary for protection, there is increasing evidence that they are incompletely effective ( Flynn and Chan 2001 ; Cosma et al. 2003 ; Cosma et al. 2004 ). We have recently shown that superinfecting mycobacteria traffic rapidly into preestablished granulomas, yet can survive therein ( Cosma et al. 2004 ). The present study showing that mycobacteria promote the formation of these structures to enhance their dissemination reveals an even greater complexity in the granuloma's role in the pathogenesis of tuberculosis. Materials and Methods Construction of M. marinum strains To generate a M. marinum RD1 mutant, PCR fragments immediately upstream (1,004 bp) and downstream (1,296 bp) of the region to be deleted (see Figure 1 ) were amplified from genomic DNA and cloned into the plasmid pKO, to flank a kanamycin resistance determinant ( Sherman et al. 2001 ). The resulting plasmid, pJC2, was used to generate a RD1 deletion mutation in M. marinum as described ( Ramakrishnan et al. 2000 ; Lewis et al. 2003 ). Both WT and ΔRD1 strains were transformed with plasmids containing transcriptional fusions of genes encoding either red-fluorescent protein (dsRed2) or green-fluorescent protein (gfp) to a constitutive M. marinum promoter as described ( Chan et al. 2002 ; Cosma et al. 2004 ). Macrophage infection assays J774 mouse macrophage-like cells and THP1 human macrophage-like cells were grown and prepared for infection as described ( Chan et al. 2002 ; Guinn et al. 2004 ). Infection with M. marinum and determination of intracellular bacterial counts was done as described ( Chan et al. 2002 ). Frog infections Frogs were injected intraperitoneally with M. marinum, and tissue bacterial counts were obtained as described ( Ramakrishnan et al. 1997 ). Zebrafish embryo infections Zebrafish embryos were maintained and injected with M. marinum strains as described ( Davis et al. 2002 ). Microscopy of embryos DIC and video microscopy were performed using a Nikon E600 (Nikon, Tokyo, Japan) equipped with 10×, 20×, and 40× magnifications, or a Nikon DIC 60× water “fluor” objective. Fluorescent as well as black and white images were collected with a Photometrics CoolSnap “cf” camera (Roper Scientific, Trenton, New Jersey, United States). Overlays of DIC and fluorescent images and video compilations were produced by using Metamorph software as described ( Davis et al. 2002 ). Determination of whole embryo bacterial counts Individual embryos were placed in microcentrifuge tubes containing 100 μl of embryo medium containing 20 μg/ml kanamycin for 1 h at room temperature. This medium was removed by aspiration and replaced with 150 μl of 0.25% Trypsin-EDTA. After incubation for 6–8 h at room temperature, Triton X-100 was added to a 0.1% final concentration, and the tubes were sonicated for 10 min in an ultrasonicator (Bransonic Ultrasonic Cleaner 1510R-NT; Branson Ultrasonics, Danbury, Connecticut, United States). The entire sample from each tube was plated onto individual 7H11 solid media plates containing 20 μg/ml kanamycin. TUNEL assay 5 d following infection, embryos were fixed in 4% paraformaldehyde in PBS overnight, dehydrated in methanol for a minimum of 24 h at 4 °C, rehydrated in PBS in a graded series of 5-min washes (in 75% methanol in PBS, 50% methanol in PBS, and 25% methanol in PBS), and washed four or five times in PBST (0.5% Tween 20 in PBS). Embryos were permeabilized using 10 μg/ml proteinase K in PBST for 30 min at 37 °C, postfixed in 4% paraformaldehyde in PBS for 20 min, washed five times for 5 min each in PBST, and twice for 5 min each in TTase Buffer (25 mM Tris-HCl [pH 6.6], 0.2M sodium cacodylate, 0.25 mg/ml BSA, and 0.2% Tween 20) plus 1 mM CoCl. Embryos were then incubated with TUNEL enzyme (#1767305; Roche, Basel, Switzerland) and TUNEL label mix (#1767291, Roche) according to the manufacturer's specifications. Primary antibody staining with sheep anti-fluorescein (#1426338 at 1/10,000; Roche) was done in Western blocking solution (#1921673, Roche) overnight at 4 °C. Secondary antibody staining with horseradish peroxidase-conjugated rabbit anti-sheep (#313035047 at 1/500; Jackson Immunoresearch, Bar Harbor, Maine, United States) was done for 2 h at RT. Detection was done with Tyramide Amplification Signal kit with AlexaFluor 555 (Molecular Probes #T30953; Molecular Probes, Eugene, Oregon, United States) according to the manufacturer's specifications. Tissue histology of adult fish Fish were fixed in Dietrich's fixative (30% ethanol, 10% formalin, and 2% glacial acetic acid in deionized water) and sectioning and staining were performed by Histo-Tec (Hayward, California, United States). Statistics. Statistics were calculated using GraphPad InStat version 3.05. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC524251.xml |
544599 | The association of Alu repeats with the generation of potential AU-rich elements (ARE) at 3' untranslated regions. | Background A significant portion (about 8% in the human genome) of mammalian mRNA sequences contains AU (Adenine and Uracil) rich elements or AREs at their 3' untranslated regions (UTR). These mRNA sequences are usually stable. However, an increasing number of observations have been made of unstable species, possibly depending on certain elements such as Alu repeats. ARE motifs are repeats of the tetramer AUUU and a monomer A at the end of the repeats ((AUUU) n A). The importance of AREs in biology is that they make certain mRNA unstable. Proto-oncogene, such as c-fos, c-myc, and c-jun in humans, are associated with AREs. Although it has been known that the increased number of ARE motifs caused the decrease of the half-life of mRNA containing ARE repeats, the exact mechanism is as of yet unknown. We analyzed the occurrences of AREs and Alu and propose a possible mechanism for how human mRNA could acquire and keep AREs at its 3' UTR originating from Alu repeats. Results Interspersed in the human genome, Alu repeats occupy 5% of the 3' UTR of mRNA sequences. Alu has poly-adenine (poly-A) regions at its end, which lead to poly-thymine (poly-T) regions at the end of its complementary Alu. It has been found that AREs are present at the poly-T regions. From the 3' UTR of the NCBI's reference mRNA sequence database, we found nearly 40% (38.5%) of ARE (Class I) were associated with Alu sequences (Table 1 ) within one mismatch allowance in ARE sequences. Other ARE classes had statistically significant associations as well. This is far from a random occurrence given their limited quantity. At each ARE class, random distribution was simulated 1,000 times, and it was shown that there is a special relationship between ARE patterns and the Alu repeats. Table 1 Defined ARE classes. (Symbol marks are used in this study instead of full sequences.) Symbol ARE sequence Class I (AUUU)5A AUUUAUUUAUUUAUUUAUUUA Class II (AUUU)4A AUUUAUUUAUUUAUUUA Class III U(AUUU)3AU UAUUUAUUUAUUUAU Class IV UU(AUUU)2AUU UUAUUUAUUUAUU Class V U4AUUUAU4 UUUUAUUUAUUUU Class VI W3UAUUUAUW3 WWWUAUUUAWWW Conclusion AREs are mediating sequence elements affecting the stabilization or degradation of mRNA at the 3' untranslated regions. However, AREs' mechanism and origins are unknown. We report that Alu is a source of ARE. We found that half of the longest AREs were derived from the poly-T regions of the complementary Alu. | Background Varying more than ten-fold, messenger RNA degradation is essential for the regulation of gene expression [ 1 , 2 ]. Differential mRNA decay rates were determined by specific cis -acting sequences within mRNA. For example, the mRNA sequences of yeast, many mammalians, and other eukaryotes contain AU-rich elements or AREs at their 3' untranslated regions (UTR) [ 3 , 4 ]. For example, in yeast, AREs stimulated the shortening of poly adenine (poly A), and two kinds of degradation pathways followed. One is 5'-to-3' exonuclease access by removal of the 5' cap structure. The other is 3'-to-5' digestion by a complex of exonucleases called exosome [ 5 , 6 ]. Genes required for these steps have been identified in yeast and were found to be conserved among eukaryotes. Although the mechanisms of AREs enhanced mRNA degradation are unknown, several groups provided evidence that 3'-to-5' degradation by the exosome may be the major pathway of decay for at least some mammalian mRNAs, including ARE-containing mRNA sequences [ 7 - 9 ]. The length of AREs also affected the half-life of mRNA. The nonamer UUAUUUAUU is a typical ARE, and the simple repeats, (AUUU) n A motif, is the well-known pattern of AREs. It has been shown that the number of ARE motifs correlated with the turnover of ARE-mRNAs such as GM-CSF [ 10 , 11 ]. Because of this, AREs are usually classified according to the number of the repeats [ 12 ]. It is known that the stabilization factor, such as HuD, is able to bind to AREs [ 13 ] and most AREs seem to function as destablizing factors. The overall importance of AREs in biology is that they can make certain critical gene products unstable. They include proto-oncogenes such as c-fos [ 14 ], c-myb [ 15 ], c-myc [ 16 ], and Pim-1 [ 17 ]. Another class of ARE-associated genes are immune response genes such as interferon [ 15 , 18 ] and interleukin [ 15 , 19 - 21 ]. Growth factors, such as Gro-α [ 22 ] and the vascular endothelial factor [ 23 ] in humans, are also known to be associated with AREs. AREs consist of a great number of thymine (or uracil) and a few adenines. Alu repeats can be a source of poly-T regions in mRNA. Therefore, there is a possible link between ARE and Alu repeats. Alu repeats are sequences of approximately 300 nucleotides (nt) transcribed by RNA polymerase III. The Alu region is then reverse-transcribed and inserted into a new location in the genome [ 24 ]. It can reach a copy number in excess of 500,000 in the human genome [ 25 ]. Alu repeats were thought to be inserted very early in primate evolution, approximately 65 million years ago (mya). Alu amplification appears to have reached a maximum rate between 35 and 60 mya, and is currently amplifying at only 1% of the maximum rate [ 26 ]. Statistical analyses have identified key diagnostic nucleotide positions in Alu sequences that define 12 subfamilies. J class is the oldest one, S class is intermediate, and Y class is the newest. The majority of Alu retrotranspositions were completed at least 30 mya when the Alu-Sx subfamily, which accounts for half of all human Alu sequences, and the Alu-Sp and Alu-Sq subfamilies became unable to replicate [ 27 - 30 ]. Alu repeats account for 6–13% of the human genome [ 31 ] and were identified in 5% of 1,616 human full-length cDNA. Of the 5%, 82% were found in the 3' UTR, while 14% were located in the 5' UTR, and very rarely in the coding region [ 32 ]. The common role of Alu at 3' UTR has not been reported, although there is one specific case that the chemical, PMA, can bind to Alu at 3' UTR and increased mRNA half-life [ 43 ]. We investigated the link between Alu sequence and the potential AREs (that have not been experimentally verified but contain ARE sequence patterns), and suggest that the complementary poly-adenine regions of Alu is one of the sources of AREs at the 3' UTR of mRNA. Figure 1A shows that the poly-adenine regions of Alu contained in the anti-sense strand on DNA complemented the poly-thymine regions in the sense strand; therefore, the poly-thymine regions on DNA transcribed the poly-uracil regions on mRNA (Figure 1B ). We propose a mechanism on how Alu has been converted to AREs gradually. When adenine was inserted at a regular interval in the poly-T(U) regions, it eventually led to the generation of potential AREs. It is not clear why such a regular insertion occurs, but the phenomenon has also been found in other ARE-like sequences. Figure 1C shows transcribed ARE on mRNA [ 33 , 34 ]. Figure 1 The schematic diagram of poly-thymine (poly-T) generation by Alu. (A) Alu contains poly-adenine (poly-A) region at the end. It is shown as 'aaaaaaaa'. The poly-A of Alu at anti-sense becomes poly-T (complement of poly-A) at sense strand on DNA. It is shown as 'tttttttt'. (B) mRNA now contains a poly-uracile (poly-U) region after the transcription of poly-T region. (C) AU-rich elements are found in this poly-U region in (B). Results The results from the method are shown in Figure 2 . In the ARE class I, marked as (AUUU)5A pattern in Table 1 , 26 AREs were found in all 21,121 mRNA 3' UTR. 38.5% of 26 AREs included in the class I, were detected in Alu sequences at 3' UTR. When we did a simulation test for the 26 AREs and 1,504 Alu sequences by 1,000 times, with a 95% confidence interval (C.I.) threshold, it was statistically significant (see the statistical analysis of the search results in the Methods section). In other words, 38.5% occurrences were out of the likelihood for random overlaps of Alu and ARE patterns in the human genome. In the ARE class II (Table 1 , (AUUU)4A pattern), 41 were found in all 3' UTR, and 7 were detected in Alu sequences among them (17.1%). The simulation results showed the 17.1% was less than the maximum random range of 7.3%. Therefore, class II data also showed a significance between ARE patterns and Alu. In class III (Table 1 ), 94 AREs were discovered from all 3' UTR. 15 out of 94 AREs were located in Alu sequences (16.0%). 16% was also statistically significant with the given sample size. In classes IV and V, 5% and 6.1% of ARE were found in Alu, respectively. These results were still out of the random chance distribution, although they were relatively less significant than the previous classes. In class VI, only 85 out of 8,649 AREs were detected in Alu (1%), and it is an insignificant hypothesis that the class VI pattern is associated with Alu sequences. Figure 2 ARE found in Alu at each class (Table 1). The numbers of ARE found in all 3' UTR, the number of ARE found in the Alu sequence, the ratio between them, and the randomly simulated results among 1,000 times at each ARE class (Table 1). Only the maximum possible ratios of the randomly simulated range at 95% confidence interval (C.I.) were shown. X-axis is for ARE patterns in all the classes. The left Y-axis is for the number of AREs, and the right Y-axis is for the overlap ratios. Discussion The possible mechanism of how AREs originated from Alu is as follows: Alu is a special sequence that contains a poly-adenine (poly-A) region at its end. The poly-A region plays an important role in the retroposition mechanism of Alu [ 35 ]. It is known that the products of LINE (L1) transposon bind the poly-A of Alu. This enables Alu to retroposition [ 36 , 37 ]. When Alu with poly-A are inserted as above, it is in the double helix form with the complementary poly-T. Therefore, the poly-T regions produce poly-uracil (poly-U) regions in mRNA when transcribed (Figure 1 ). We hypothesized that the poly-U regions generated from the Alu are the source of AREs after either random or directed mutation. With this hypothesis, we suggest a new role for Alu was involved in the 3' UTR. It is well known that Alu affected gene expression at the 5' of genes and alternative splicing at the intron region [ 38 , 39 ]. However, no Alu role at the 3' UTR has been suggested yet. We could have applied the same test to Alu at 5' UTR region, but there were too few data sources [ 32 ]. Conclusion AREs are mediating sequences that affect the stabilization or degradation of biologically important genes' mRNA. However, their origin in evolution has not been clear. This report presents a hypothesis and statistical evidence that Alu was one of the sources of ARE generation or origin. A possible mechanism of ARE generation from Alu via retroposition and regular pattern mutation is suggested. Methods Human 3' UTR sequences We used the RefSeq database from the National Center for Biotechnology Information (NCBI) for human 3' UTR sequences [ 41 ]. We extracted 3' UTR of CDS (coding sequence) from all the annotated mRNA sequences (mRNA_Prot, 2004.9.13). The number of 3' UTR was 21,121 and the average length was 996 bp. We used the Biojava package [ 42 ] to extract only 3' UTR with Genbank's feature information. The number of 3' UTR was 21,121 and the average length was 996 bp. Alu sequence and AU-rich element (ARE) pattern detection AREs were searched for in the all 3' UTR (Table 1 ). An in-house java program was used to search for these AREs. While the number of AUUUA repeats decreased, the T flank region increased to 21 bp. Each ARE was allowed within one base mismatch. This is a stricter mismatch criterion than the one of AU-rich elements database (ARED) (the ARED trained experimental ARE data allow 10% of ARE length mismatch [ 24 ]). The RepeatMasker program was used for finding Alu. It is a program for finding repeat sequences [ 25 ]. After finding Alu sequences using RepeatMasker at 3'UTR, for each Alu, we recorded the position information (RefSeq ID, start and end position) for the next step analysis. Comparison between two search results We compared the positions of 3' UTR Alu and ARE sequences. If an ARE was discovered within an Alu sequence, this ARE was regarded found in 3' UTR Alu. For example, if an Alu was found between 100–400 bp and an ARE was found between 99–129 bp, this ARE was in 3' UTR Alu in the same 3' UTR. If less than 50% of an ARE length was discovered in an Alu, we further check if there is 7 bp TSD (Target Site Duplication) between the Alu's end and the ARE's end [ 4 ]. For example, if an Alu is between 100–400 bp and an ARE between 80–110 bp, about 10 bp (33%) of the ARE belongs to the Alu. In this case, we check if there is 7 bp TSD between upstream region from 80 bp and downstream from 400 bp. Statistical analysis of the search results To validate the significance of the searches, we calculated the random chance of the ARE and Alu sequence overlap at each class (Table 1 ). Hypothesis H0: ARE occurs in human 3'UTR independently from Alu. Random sequence generation for statistical validation The average length of 3' UTR of 21,121 human sequences was 996 bp. Within the long theoretical sequence of 21,121 × 996 bp, we generated 1,504 Alu (300 bp) and ARE sequences (21–13 bp). For example, 1,504 Alu and 26 (21 bp) AREs in ARE class I (Table 1 ) were generated following a uniform distribution as a control set. 1,504 and the number of AREs for ARE classes were the actual numbers of Alu and AREs found by our method. This random sequence generation was done 1,000 times with a 95% significance threshold. Test results In the ARE class I (Table 1 ), the significance range at a 5% error range was 0.0–11.5% (Figure 2 ) for the random chance of association between ARE patterns and Alu sequences. The results in other ARE classes are also shown in Figure 2 . Our result of a 38.5% – 6.1% overlap between AREs and Alu, depending on ARE classes, was statistically significant. Therefore, hypothesis H0 was rejected. Authors' contributions HJA conceived of this study, carried out the tests, and drafted the manuscript. JB participated in the design of the study and drafted the manuscript. KWL and DL amended and improved the design of the study. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544599.xml |
549042 | HIV-1 Tat, apoptosis and the mitochondria: a tubulin link? | The Tat protein of HIV-1 is a powerful activator of viral gene expression. Besides this essential function at the HIV-1 promoter, the protein also exerts a remarkable number of other biological activities, among which the induction of cellular apoptosis. Two papers now published in Retrovirology provide possible molecular mechanisms for the pro-apoptotic effect of Tat, which involve the cell's microtubular network and the mitochondrial pathway of apoptosis. | Although more than 20 years have passed since the identification of HIV as the cause of AIDS, several essential questions about its pathogenicity remain as yet unanswered. In particular, a central, still unresolved issue is the mechanism underlying the progressive development of immunodeficiency. It is now well established that HIV infection determines a rapid turnover of infected CD4 cells [ 1 , 2 ]; consistent with this finding, multiple molecular pathways triggered by different HIV proteins are known to lead to cell apoptosis [ 3 , 4 ]. However, the capacity of the immune system to regenerate its cells by far exceeds the number of dying HIV infected cells. Thus, the extension of the apoptotic message to neighboring, bystander cells has long been recognized as a potential mechanism sustaining the immunodeficiency that accompanies HIV disease progression [ 5 ]. In this context, the finding that the virus-encoded Tat protein is released by the infected cells and can be taken up by neighboring, uninfected cells via an endocytic mechanism [ 6 , 7 ] has long suggested the possibility that some of the bystander apoptotic effects exerted by HIV might be mediated by this protein. Over ten years ago different investigators did indeed show that extracellular Tat can trigger apoptosis in T-cell lines and primary T-cells [ 8 , 9 ]. The classical apoptotic pathway, involving the cell's mitochondria, is regulated by the Bcl-2 family of proteins. This family contains both anti-apoptotic (Bcl-2, Bcl-XL) and pro-apotpotic (Bax, Bid, Bim) members that exert their function primarily at the mitochondrion by either preventing or inducing mitochondrial dysfunction. Upon receiving a death signal, the pro-apoptotic proteins translocate from the cytoplasm to the outer mitochondrial membrane, where they interact with their pro-apoptotic partners. This occurrence is followed by mitochondrial dysfunction, release of pro-apoptotic proteins out of the mitochondrion (among which, a prominent role can be ascribed to cytochrome c), and subsequent caspase activation [ 10 ]. One of the cellular events that trigger the mitochondrial pathway of apoptosis is the disturbance of the dynamic formation of microtubules in the cell. This event can be triggered by a variety of microtubule-targeted, tubulin-polymerizing agents (MTPAs), which include paclitaxel (Taxol) and several other anticancer drugs [ 11 ]. Following intracellular uptake, MPTAs bind β-tubulin and promote tubulin polymerization, which interferes with the function of the mitotic spindle resulting in mitotic arrest at the metaphase-anaphase transition and subsequent induction of the mitochondrial pathway of apoptosis. A link between microtubule polymerization and the pro-apoptotic effect of Tat has first been suggested a few years ago in the observation that Tat directly interacts with the αβ-tubulin dimers and polymerized microtubules in the cytoplasm of the cell [ 12 ]. The functional consequence of this interaction, which requires the integrity of four amino acids in the conserved Tat core domain, is the stabilization of microtubules and the consequent prevention of microtubule depolymerization. This disturbance in the microtubular network is a powerful inducer of the mitochondrial pathway of cellular apoptosis, an event that is transduced by the pro-apoptotic Bcl-2 relative Bim. These findings supported previous observations that had already shown that Tat causes changes in mitochondrial membrane permeability [ 13 , 14 ] and that it interferes with the polymerization of microtubules [ 15 ]. Two papers now published in Retrovirology extend the link between the microtubule network, the mitochondrial pathway of apoptosis, and Tat. De Mareuil and coworkers show that Tat enhances tubulin polymerization into microtubules, an effect similar to that exerted by the MTPAs, and physically associates with the polymerized microtubuli [ 16 ]. As opposed to paclitaxel, however, Tat only increases the rate of tubulin polymerization while it does not permanently affect the organization of the microtubule network, nor does it blocks cell cycle progression. Most notably, the ability of different Tat variants to induce tubulin polymerization correlates with their capacity to induce apoptosis. Similar to paclitaxel and other microtubuli damaging agents, the pro-apoptotic effect of Tat parallels the induction of cyctochrome c release from the mitochondria, a critical event triggering apoptosis. The accompanying manuscript by Epie and coworkers describes the identification of a microtubule-associated protein, LIS1, which specifically binds Tat [ 17 ]. In the course of a biochemical project entailing the fractionation of T-cell extracts searching for Tat-associated kinases that phosphorylate the C-terminal domain of RNA polymerase II – a known biochemical activity associated to Tat -, these authors found that LIS1 co-purifies with a complex of proteins including one of the CTD kinases, CDK7, its cyclin partner, cyclin H and the MAT1 co-factor. Of note, out of the four purified proteins, only LIS1 directly bound Tat, as shown by GST-pulldown and co-immunoprecipitation experiments, and by the yeast two hybrid assay. LIS1 is known to regulate microtubule dynamics by interacting with dynein and additional components of the dynein motor [ 18 ]. What might be the relevance of these findings in the context of HIV-1 infection? They clearly provide a mechanism for CD4 T-cell apoptosis and for the extension of the apoptotic effect to bystander, uninfected cells in the lymph node. Moreover, the interaction of Tat with the microtubular network might explain the occurrence of neuropathogenesis accompanying the progression of HIV disease, since many human neurodegenerative conditions are elicited by a reorganization of the neuronal cytoskeleton [ 19 ]. Thus, the disturbance of the microtubular network induced by Tat adds to other potentially pro-apoptotic mechanisms induced by the protein, such as the upregulation of FasL [ 9 ], TRAIL [ 20 ], Bax [ 21 ] and caspase 8 [ 22 ] and the downregulation of Bcl2 [ 21 ]. As commonly happens in biology, the findings reported in these manuscripts raise more questions than answers. First, the Tat domains involved in the described interactions are different, a surprising finding given the very small size of Tat. This observation might possibly suggest that Tat is part of a large multi-molecular complex associated with the tubular network, making multiple contacts with different proteins. This issue can be experimentally addressed biochemically, or even within the cell, by taking advantage of the biophysical techniques available to investigate protein-protein interactions in vivo [ 23 ]. Secondly, the role of LIS1, if any, in the Tat-triggered mitochondrial pathway of apoptosis or in the functions of CDK7 and its partners, with which it unexpectedly co-purifies is unclear. Third, and most importantly, it remains to be seen whether the concentration at which Tat binds tubulin and exerts its pro-apoptotic effects is compatible with the concentration at which the protein is expressed in the infected cells and diffuses to neighboring cells. As a matter of fact, the measurement of the extracellular concentration of Tat still remains a holy grail in the HIV research field [ 24 ], partly due to the weak avidity of the currently available anti-Tat antibodies, partly because of the biological property of extracellular Tat that is sequestered by extracellular matrix proteoglycans [ 25 ]. Until more reliable methods are developed to determine the levels of extracellular Tat in vivo, the full biological implications of Tat-induced apoptosis cannot be entirely appreciated. Abbreviations MTPAs: microtubule-targeted, tubulin-polymerizing agents CTD: carboxy-terminal domain Competing interests The author(s) declare that they have no competing interests. Figure 1 To form microtubules, α- and β-tubulin molecules join to form a heterodimer. These dimers then attach to other dimers forming oligomers that elongate into protofilaments; eventually, the oligomers will join to give rise to a ringed microtubule. Microtubules or unpolymerized tubulin bind microtubule-associated proteins (MAPs), which regulate polymerization, facilitate assembly, stabilize the microtubules and regulate microtubular transport of macromolecules and vesicles. The HIV-1 Tat protein binds to αβ-tubulin dimers and microtubules thus enhancing microtubule polymerization, and to the microtubule-associated protein LIS1, which is also known to facilitate assembly of microtubules. Disturbance of the dynamics of microtubular network formation activates the intrinsic mitochondrial apoptotic pathway. Pro-apoptotic Bcl2 family members – in particular, Bim – are recruited to the mitochondrion; as a consequence, the mitochondrial membrane potential collapses, and pro-apoptotic factors are released into the cytoplasm. These include reactive oxygen intermediates (ROIs), apoptosis-inducing factor (AIF), and cytochrome c, among others. Release of cytochrome c is a point of no return as it leads to autoactivation of caspase 9, which in turn proceeds to cleave the downstream effector caspases (caspase 3, 6, etc.). | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549042.xml |
553973 | Novel hepatocyte growth factor (HGF) binding domains on fibronectin and vitronectin coordinate a distinct and amplified Met-integrin induced signalling pathway in endothelial cells | Background The growth of new blood vessels in adult life requires the initiation of endothelial cell migration and proliferation from pre-existing vessels in addition to the recruitment and differentiation of circulating endothelial progenitor cells. Signals emanating from growth factors and the extracellular matrix are important in regulating these processes. Results Here we report that fibronectin (FN) and vitronectin (VN) modulate the responses of endothelial cells to HGF (Scatter Factor), an important pro-angiogenic mediator. Novel binding sites for HGF were identified on both FN and VN that generate molecular complexes with enhanced biological activity and these were identified in the supernatants of degranulated platelet suspensions implicating their release and formation in vivo . In the absence of co-stimulation with an ECM glycoprotein, HGF could not promote endothelial cell migration but retained the capacity to induce a proliferative response utilising the Map kinase pathway. Through promoting Met-Integrin association, HGF-FN and HGF-VN complexes coordinated and enhanced endothelial cell migration through activation of the PI-3 kinase pathway involving a Ras-dependent mechanism whereas a Ras-independent and attenuated migratory response was promoted by co-stimulation of cells with HGF and a non-binding partner ECM glycoprotein such as collagen-1. Conclusions These studies identify a novel mechanism and pathway of HGF signalling in endothelial cells involving cooperation between Met and integrins in a Ras dependent manner. These findings have implications for the regulation of neovascularization in both health and disease. | Background The generation and repair of blood vessels in adult life requires the regulation of endothelial cell survival, migration, proliferation and their differentiation from lineage-committed progenitors by the coordinated action of several classes of vaso-active agents including growth factors, cytokines, and the extracellular matrix (ECM) [ 1 - 4 ]. Elucidating the molecular mediators of these signals and their mechanism of action is vital to understanding the fine regulation of neo-vessel development and maintenance. There is growing evidence pointing to a close collaboration between growth factors and the ECM in several biological processes including vasculogenesis and post-natal revascularization. Studies have shown that the response of cells to growth factors such as vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF) and epidermal growth factor (EGF) are potentiated by integrin ligation to specific ECM glycoproteins [ 5 - 8 ]. In a previous report, we showed that VEGF-induced endothelial cell migration was augmented by fibronectin (FN) [ 9 ]. We also presented evidence that the VEGF/VEGFR-2 pathway is coupled to the integrin α 5 β 1 through a mechanism involving the promotion of an integrin α 5 β 1 -VEGFR-2 signalling moiety generated as a consequence of receptor ligation by a VEGF-FN complex. These events promoted the sustained activity of Erk kinase, which was coupled to the migratory response. More recently, we presented data demonstrating that FN significantly enhanced VEGF-mediated migration of CD34 + cells and their differentiation into endothelial cells [ 10 ]. In addition to the VEGF pathway, in vitro studies have highlighted the importance of hepatocyte growth factor (HGF) as a pro-angiogenic mediator. HGF, also termed scatter factor, has a well-established role in tumourogenesis but may be an important mediator of neovascularization since studies show that HGF induces the expression of VEGF in endothelial cells in vitro and that HGF synergises with VEGF to promote capillary-tube assembly in collagen matrices [ 11 , 12 ]. In addition, neovascularization in the rat cornea was also elevated by co-administration of HGF and VEGF compared to either growth factor in isolation [ 11 ]. The emerging significance of HGF as a pro-angiogenic mediator was further highlighted by a recent study of a large cohort of patients (1090 patients, CAPTURE trial) with acute coronary syndromes and identified serum levels of HGF as a positive indicator of patients' prognosis associated with a significantly lower event rate and increased collateralization of the target vessel [ 13 ]. Although the pro-angiogenic effects of HGF are known, the detailed mechanism of HGF action on the vascular cells, including the identity of intracellular mediators remains poorly understood. In the present work we show that HGF forms a specific physical complex with FN and VN and that these complexes are present in degranulated platelet suspensions implicating a putative role in vivo . Significantly, we show that HGF-FN and HGF-VN molecular complexes induce a unique and enhanced intracellular signal employing Ras, thereby highlighting an important mechanism of growth factor receptor tyrosine kinase and integrin cooperation in promoting pro-angiogenic responses. Results Identification of novel HGF binding domains on FN and VN We recently identified binding domains on FN for VEGF, which played an important role in promoting the activity of VEGF [ 9 ]. Since HGF is also an important angiogenic factor, experiments were designed to establish whether HGF had specific ECM binding partners. Using a solid phase assay, we measured the binding of 125 I-labelled HGF to a variety of ECM molecules immobilized on plastic wells. As shown in Fig. 1A , 125 I-labelled HGF bound to both FN and VN specifically with residual binding observed to either collagen-1 or laminin. Further experiments were performed to locate the HGF binding site on the FN molecule using purified FN proteolytic fragments immobilised onto the polystyrene microtiter wells. In these experiments 125 I-labelled HGF bound to the 70 kDa N-terminal fragment and the 40 kDa C-terminal fragment. No significant binding was observed to the 120 kDa fragment that harbours the internal cell binding domain (Fig. 1B ). To further analyse the association between HGF and FN, the interaction of HGF with the FN fragments was measured in real time by surface plasmon resonance analysis (SPR). As shown in Fig. 1C & 1D , HGF bound to the 70 kDa N-terminal FN fragment immobilized on the sensor chip in a specific and saturable manner with a K d of approximately 300 ± 93 nM for a one-site model. The data shown in Fig. 1D could be applied to a two-site model with equal probability showing K d values for the high and low affinity sites of 15 nM ± 2 nM and 4 μM respectively. HGF binding to the 40 kDa fragment could not be measured directly by SPR, as immobilization of the 40 kDa fragment on the sensor chip appeared to mask the HGF binding site (data not shown). Figure 1 Identification of HGF Binding Domains on FN and VN and the presence of HGF-FN and HGF-VN complexes in platelet supernatants. Panel A -Plastic wells coated with various matrix proteins or BSA (10 μg/ml) were incubated with 125 I-labelled HGF to equilibrium and then washed and counted in a γ-counter. Panel B -Plastic wells were coated with FN and FN derived proteolytic fragments or BSA. 125 I-labelled HGF was incubated in these wells to equilibrium binding, washed and the bound HGF levels determined using γ-counter. Panels C & D -Binding of HGF to FN 70 kDa fragment in real time by SPR. Panel C -Real time binding isotherms of increasing concentrations of HGF (0, 16.5, 31.25, 62.5, 125, 188, 250, 315, 375, 500 nM)with association and dissociation phases for a single representative experiment with the arrows indicating injection start and finish. Panel D -The data from panel C is plotted as a function of HGF concentration (30–500 nM). Inset shows the data analysed by the method of Scatchard showing a single binding site with a K dapp = 300 nM. The data could also fit equally well to a two-site model (see text for details). n = 3. Panel E -Platelet suspensions (20.0 × 10 8 /ml) were stimulated with either saline (-) or 1 U/ml thrombin (+) for 5 min at room temperature to allow degranulation. Cellular and membranous material was cleared by centrifugation (100,000 × g) and the supernatants were immunoprecipitated with monoclonal antibodies to FN or an isotype matched control reagent (control IgG). Immune complexes were analysed by SDS-PAGE and Western blotting probing with an antibody to HGF (Santa Cruz). The blot was stripped and re-probed with antibodies to FN to confirm the specificity of the primary immunoprecipitation step. Panel F -Supernatants derived from thrombin-stimulated platelets were immunoprecipitated with antibodies with specificity for either FN, VN or an isotype matched control reagent. The immune complexes were analysed as in panel E. Lower panel shows the same blot stripped and re-probed with a monoclonal antibody to VN to confirm the primary precipitation. Blots were developed by chemiluminescence. The data shown are representative blots of experiments repeated three times. Platelets release HGF complexed to FN and VN To establish whether HGF-FN and HGF-VN molecular complexes occur in vivo we examined platelets, a rich source of growth factors, for the presence of these complexes. Washed human platelet suspensions were stimulated with thrombin (1 U/ml) to promote degranulation and the derived supernatants were immunoprecipitated with antibodies directed to FN or VN. The resulting immune complexes were analysed for co-precipitation of HGF (Fig. 1E & 1F ). Immunoprecipitation of FN from thrombin-stimulated platelet supernatants resulted in significant co-precipitation of HGF (Fig. 1E ). In contrast, minimal levels of HGF was observed in samples derived from unstimulated platelet supernatants or from samples derived from thrombin-stimulated platelet supernatants when an isotype-matched control antibody was employed in the experiment. Probing of the same blot with antibodies to FN confirmed that the primary precipitation of FN was responsible for the co-precipitation of HGF (Fig. 1E , lower panel). In a parallel experiment, immunoprecipitation of VN also co-precipitated HGF to a similar if not greater extent than FN (Fig. 1F ). These experiments demonstrate that HGF is released from platelets and is found in the form of soluble molecular complexes with both FN and VN, confirming the results of the ligand binding studies in vitro . HGF-Induced endothelial cell migration is dependent upon co-stimulation with ECM We next sought to determine whether the responses of endothelial cells to HGF could be modulated by its ECM binding partners. In cell migration assays, human microvessel endothelial cells (HMVEC) were incubated with an optimal concentration of HGF alone or in combination with fixed concentrations of FN, VN or collagen-1 (Fig. 2A ). Significantly, little or no endothelial cell migration above basal levels (control) was observed when cells were stimulated with HGF (10 ng/ml) in the absence of ECM. A moderate migratory response of endothelial cells to HGF was observed in the presence of collagen-1 (non-HGF binding ECM), which was less than 2-fold above basal levels. When HGF was co-administered with either FN or VN, endothelial cell migration was significantly enhanced by 4–5 fold. The differences in magnitude of the migration in the presence of these ECM glycoproteins was not related to variable degrees of cell adhesion upon the transwell filters as HGF-stimulated endothelial cells adhered equally well to ECM glycoprotein-coated transwells (Fig. 2B ). The migratory response to HGF was dose responsive with a maximal response observed at a concentration of 10–20 ng/ml (data not shown). In addition, a negligible migratory response was observed when HMVEC were stimulated with these ECM molecules in the absence of HGF consistent with our previous report (data not shown, ref .9). Figure 2 Effect of ECM molecules on HGF induced Endothelial cell migration. Panel A -Dye-loaded HMVEC suspensions (9 × 10 4 /ml) were treated with HGF (10 ng/ml) in the presence or absence of ECM molecules, FN or VN or collagen-1 in a modified Boyden chamber assay. Migration of cells was measured at 3 hours. The data is presented as relative migration with the migratory response presented as ratio to the basal migration in the absence of stimulus. Panel B -The levels of cell adhesion to Fluorblok transwell filters coated with either poly-L-lysine (PLL) or the various ECM glycoproteins was determined. Dye loaded HMVEC (10 5 cells) were treated with HGF (10 ng/ml) for 30 min prior to application to upper chamber of the transwell. Following extensive washes the number of adherent cells was determined using a fluorescence plate reader. The role and identity of integrins mediating the migration response in cells stimulated with HGF and FN (panel C) and HGF and VN (panel D) was demonstrated by the pre-treatment of HMVEC suspensions with anti-integrin monoclonal reagents (10 μg/ml) with specificity for integrins α 5 β 1 (JBS5), α v β 3 (LM609) and α v β 5 (LM142) for 30 min at room temperature prior to application into the upper transwell chamber. The data is presented as specific migratory response in fluorescence units with the basal migration subtracted from the total migratory response. (n = 2). To further characterize the degree and identity of integrin involvement in the observed migratory response, we investigated the consequences of blocking integrin receptors on HMVEC with specific integrin antibodies prior to HGF-ECM stimulation. Antibodies directed to the integrin α 5 β 1 completely inhibited HGF-FN-induced endothelial migration (Fig. 2C ). In contrast, an antibody with specificity for the αv-subunit (LM142) had no inhibitory effect on endothelial cell migration. However, antibodies to the α v β 3 integrin (LM609) did inhibit endothelial cell migration to HGF-FN by 20% suggesting an ancillary role for this integrin in mediating HGF-FN responses. When endothelial cell migration was induced by HGF-VN complexes, the integrin dependence shifted as expected (Fig. 2D ). Under these conditions endothelial cell migration was predominantly dependent on αv-integrins for mediating the migratory signal with some apparent involvement of the integrin α 5 β 1 (approximately 30%). This latter effect may be a consequence of integrin signal cross-talk (transdominant integrin regulation), as reported previously [ 14 , 15 ]. These experiments demonstrate that for HMVEC, HGF induced cell migration is dependent upon the ligation of integrins by ECM molecules. Met associates with α v β 3 and α 5 β 1 integrins Previous work [ 5 - 7 , 9 ] has demonstrated that the physical association of growth factor receptor tyrosine kinases and integrins promote enhanced cellular responses. We, therefore, postulated that the elevated cell migration induced by HGF-FN and HGF-VN in the present study could be due to a signalling mechanism involving the physical association between Met and integrins on endothelial cells. As shown in Fig. 3A , endothelial cell lysates derived from samples exposed to collagen-1, FN or VN, in the presence of HGF, when immunoprecipitated with antibodies to integrins α 2 β 1 , α 5 β 1 and α v β 3 respectively, predominantly co-precipitated Met with the integrins α 5 β 1 and α v β 3 . In contrast, Met co-precipitation with the integrin α 2 β 1 was minimal for lysates derived from cells stimulated with HGF and collagen-1. The level of Met expression in these samples was not altered by treatment of the cells with various combinations of HGF and ECM molecules (Fig. 3A lower panel) discounting the possibility that the differences in the level of Met co-precipitation was due to differences in the expression levels of its antigen. In the absence of HGF, co-precipitation of Met with the integrins α 5 β 1 and α v β 3 was minimal despite the presence of the ECM glycoprotein, indicating that ligation of the integrin with its cognate ligand was not sufficient to induce an association with Met. Figure 3 Association of Met with integrins is driven by HGF-FN and HGF-VN complexes. Panel A -HMVEC (1 × 10 6 /ml) were seeded onto collagen-1, FN or VN coated plastic wells and were stimulated with either saline or HGF (10 ng/ml) for 60 min at room termperature. Cells were lysed and the corresponding lysates (500 μg protein) were immunoprecipitated with antibodies to integrins α 2 β 1 (lanes 1 & 4), α v β 3 (lanes 2 & 5) and α 5 β 1 (lanes 3 & 6). Immunocomplexes were analysed by SDS-PAGE and Western blotting probing with an anti-Met antibody using chemiluminescence detection (top panel). The bottom panel shows the antigen levels of Met in the corresponding cell lysates after analysis by SDS-PAGE (40 μg/lane) and Western blotting probing with an antibody to Met were essentially unaltered during the duration of the experiment. Panel B -HMVEC (5 × 10 6 ) were stimulated with HGF or HGF-FN or HGF-VN complexes at the indicated time periods at room temperature. Cells were lysed and immunoprecipitated with a polyclonal antibody to phosphotyrosine Met and the immune complexes analysed by SDS-PAGE and Western blotting using a monoclonal antibody to Met. The Met antigen was detected using chemiluminescence. Panel C -HMVEC (5 × 10 6 ) were stimulated with HGF or HGF-FN or HGF-VN complexes for 15 minutes at room temperature. Cells were lysed and the lysates immunoprecipitated with monoclonal antibodies to the integrins α 5 β 1 and α v β 3 and the immune complexes analysed by SDS-PAGE and Western blotting probing with a polyclonal antibody to Met. To elucidate the role of Met activation in the formation of the Met-integrin signalling complex, endothelial cells were treated with HGF in the absence of ECM glycoprotein and with HGF-FN and HGF-VN complexes and the kinetics of Met tyrosine phosphorylation investigated (Fig. 3B ). These experiments demonstrated that HGF in the absence of ECM glycoprotein could activate Met transiently with a strong signal present at 15 min but absent at 1 hour. In contrast, cells stimulated with HGF-FN and HGF-VN showed strong activation of Met at 15 min, which was sustained at 1 hour and was evident, although reduced, at 2 hours post-stimulation. Cell lysates derived from samples stimulated for 15 mins were also assessed for the presence of a Met-integrin complexes. As shown in Fig. 3C , HGF in the absence of FN or VN did not promote a significant association of Met with the integrins α5β1 or αvβ3. However, cells treated with HGF-VN and HGF-FN for 15 min contained significant levels of Met in a physical association with these integrins. These studies show that Met activation by HGF is insufficient to promote a physical association with integrins. HGF binding domains on FN and VN promote enhanced intracellular signals We next investigated whether the association of Met with integrins modulated HGF/ECM-induced intracellular signalling, focussing on the ERK and the PI-3 kinase pathways. Analysis of the phosphorylation kinetics of Erk-1/2 in response to HGF alone or HGF/ECM combinations showed distinct patterns of activation (Fig. 4A ). With HGF alone, Erk 1/2 phosphorylation showed kinetics with a peak signal at 60 min post-stimulation and significant reduction by 120 min although phosphorylation was still apparent. A distinct activation profile was observed when cells were stimulated with HGF and collagen-1 (non-HGF binding ECM), with Erk 1/2 levels peaking at 30 min and returning to near basal phosphorylation levels by 120 min. However, stimulation of endothelial cells with either HGF-FN or HGF-VN complexes promoted a rapid but sustained phosphorylation of Erk 1/2 with levels near maximal at 120 min post-stimulation. Analysis of the activation of the PI-3 kinase pathway was assessed by measurement of the phosphorylation status of Akt/PKB on Ser 473 (Fig. 4B ). Interestingly, both distinct levels and kinetics of Akt phosphorylation were observed in these samples. When endothelial cells were stimulated with HGF in the absence of ECM co-stimulation, little phosphorylation of Akt above basal levels was observed. However, when cells were treated with HGF plus collagen-1, Akt phosphorylation was rapidly detected at 5 min and peaked at 30 min with significant reduction by 120 min. In contrast, with cells treated with either HGF-FN or HGF-VN complexes, Akt phosphorylation kinetics appeared to mirror Erk 1/2 phosphorylation kinetics implying a common regulatory mechanism for both pathways. As with Erk 1/2 phosphorylation, Akt phosphorylation peaked by 30 min post stimulation and this level of activation was sustained even at 120 min. Significantly, Akt phosphorylation levels in these samples were elevated approximately 3-fold, (assessed by densitometry) compared with the levels in observed in cells stimulated with HGF plus collagen-1 (data not shown). Figure 4 HGF-FN and HGF-VN complexes augment HMVEC migration via the PI3 kinase pathway. Panel A -HMVEC suspensions (1 × 10 6 /ml) were stimulated with HGF (10 ng/ml) in the absence of ECM molecules or in the presence of collagen-1, FN or VN (2 μg/ml) for varying time intervals (2–120 min) at room temperature. The reaction was stopped by rapid centrifugation and lysis of the cell pellet. Samples of cell lysate were analysed by SDS-PAGE and Western blotting probing with an antibody with specificity for phosphorylated Erk 1/2 (top panel). The blots were reprobed with an antibody to Erk 2 (bottom panel). Panel B -The samples from panel A were also probed simultaneously with an antibody with specificity for phosphorylated Akt (top panel) and these blots were stripped and re-probed with an antibody to Akt. Visualization was by chemiluminescence. Panel C & D -Effect of inhibitors on HMVEC migration in response to HGF-FN and HGF-VN respectively. The data is a representative experiment using triplicate samples which was performed three times giving essentially similar results. HGF/ECM-Induced endothelial migration is coupled to the PI-3 kinase pathway To determine the intracellular pathway(s) that were coupled to the migratory response, HMVECs were treated with specific inhibitors of MEK and PI-kinase. In cell migration assays, LY294002 but not U1026 inhibited endothelial cell migration induced by HGF-FN (Fig. 4 C ) and HGF-VN (Fig. 4 D ), clearly demonstrating that the PI-3 kinase pathway was predominantly coupled to the migratory response and not the Map kinase pathway. Other inhibitors of potential down stream effectors were also tested. HGF-FN stimulated cells pre-treated with PP1, U73122, and piceatannol showed maximal migratory responses indicating that Src, PLCβ and Syk were not components of the migratory signal (data not shown). HGF-Induced endothelial proliferation is coupled to the Erk-pathway The effect of the co-administration of ECM molecules with HGF on endothelial cell proliferation was also investigated. In contrast to cell migration, HGF, in the absence of ECM molecules, induced a significant proliferative response (Fig. 5A ). However, in the presence of FN or VN, HGF-induced endothelial proliferation was enhanced compared to HGF alone or in combination with collagen-1. As with the migratory response, endothelial cell proliferation was dose responsive to HGF with an observed maximal response at a concentration of 10–20 ng/ml (data not shown). Chemical inhibitors were then used to determine the signalling pathways involved in HGF-induced endothelial cell proliferation. In these studies, the MEK inhibitor, U1026 significantly impaired HGF-induced endothelial proliferation (50–80% inhibition) irrespective of co-stimulation with or without ECM molecules. This suggests that unlike migration, which was shown above, to be PI-3 kinase dependent, the Erk-pathway plays an important role in mediating HGF-induced endothelial cell proliferation. While both LY294002 and FPT-III blocked HMVEC proliferation, this appeared to be due to apoptosis (data not shown). This observation is consistent with the role of PI-3 kinase in promoting cell survival and a role for Ras in regulating PI-3 kinase in these cells [ 16 , 17 ]. Figure 5 HGF induced HMVEC proliferation requires the Map kinase pathway . Panel A -HMVEC in MCDB-131 medium was plated on poly-D-lysine coated 48-well plates at a density of 2.5 × 10 3 cells/well and were stimulated with HGF (10 ng/ml) in the presence and absence of ECM molecules FN, VN or collagen-1. Basal proliferation was measured with cells treated with non-supplemented medium. Cell numbers were quantified 48 hours post-stimulation using CyQuant reagent. Data is presented as cell numbers with the increase of proliferation of HGF-FN and HGF-VN treated samples significant compared with cells treated with HGF alone as determined by a one-way ANOVA (p < 0.05) n = 3. Panel B -HMVEC were plated onto poly-D-lysine coated wells (1 × 10 4 /well) overnight in supplemented medium. Cells were then incubated with medium comprising 0.1% FBS plus HGF (20 ng/ml) in the presence or absence of ECM proteins (10 μg/ml) as shown containing no inhibitor (white bars) or the MEK inhibitor U1026 (10 μM, black bars). Cells numbers were quantified after a further 48 h using CyQuant reagent. Ras is a specific, upstream regulator of Erk and PI-3 kinase pathways in cells stimulated with HGF-FN and HGF-VN complexes The data shown above indicate that HGF-induced endothelial cell migration and proliferation were mediated by PI-3 kinase and Erk pathway respectively. We next investigated the role of Ras in regulating these two pathways induced by HGF-FN and HGF-VN complexes. Since Ras is a well-documented regulator of p85 PI-3 kinase and Erk and as well as a down stream effector of both the Met and integrin receptors, we assessed the activation status of Ras by measuring the comparative levels of GTP-loaded Ras after endothelial cells were stimulated with HGF in the presence and absence of ECM molecules (Fig. 6A & 6B ). Endothelial cells stimulated with HGF alone showed high levels of GTP-Ras at 60 min post-stimulation and this was sustained even at 120 min (Fig. 6A ). In contrast, cells co-stimulated with HGF and collagen-1 showed activation of Ras at 60 min post-stimulation but to a significantly lower degree (approx 50% compared to HGF alone Fig. 6B ), with the signal diminished by 120 min. With HGF-FN and HGF-VN co-stimulation, GTP-Ras levels were more than two-fold higher than observed when cells were co-stimulated with HGF-collagen-1 (Fig. 6B ). Significantly, GTP-Ras levels were sustained at 120 min consistent with the observations of the activation profiles for the MAP kinase and PI-3 kinase pathways. These studies suggested that inhibiting Ras in cells stimulated with HGF-FN and HGF-VN complexes would exhibit reduced migration responses. To test this hypothesis, cells were treated with the membrane permeable farnesyltransferase inhibitor FPT-III, which inhibits Ras function as a consequence of the loss of membrane localization in the absence of farnesylation. Upon stimulation with HGF-FN, endothelial cells treated with FPT-III (100 μM) showed little activation of Ras following HGF-FN stimulation compared to basal levels in unstimulated cells (Fig. 7A ). In comparison, pre-treatment of cells with the geranylgeranyl transferase inhibitor GGTI (2 μM) had little inhibitory effect on HGF-FN induced Ras activation. The effect of these inhibitors was tested in endothelial migration assays. Endothelial cells pre-treated with FPT-III displayed a profound reduction in cell migration of 50% and 73% when stimulated with HGF-FN and HGF-VN complexes respectively compared to cells pre-treated with GGTI (Fig. 7B ). In contrast, FPT-III had little inhibitory effect on migration induced by HGF plus collagen-1 indicating that Ras has no significant role in the regulation of the migratory signal with this stimulus. Figure 6 Enhanced and sustained activation of Ras by HGF-FN and HGF-VN Complexes Panel A -Kinetic GTP-Ras pull down analysis. HMVEC suspensions (3 × 10 6 /ml) were stimulated with HGF (10 ng/ml) in the absence and presence of ECM molecules (2 μg/ml) as shown for 5, 60 and 120 min at room temperature. Lane Rest. represents resting (unstimulated) levels of GTP-Ras. Cells were pelleted and lysed in cold MLB buffer (see methods). Samples of cell free lysates were incubated with of RBD-Sepharose and then analysed by SDS-PAGE and Western blotting probing for Ras. Visualization was by chemilumininescence using a Kodak imaging station. Panel B -The levels of GTP-Ras at the 60 min time point from the gel shown in panel A were quantified by densitometric analysis using ImageQuant software (Kodak). Figure 7 HGF-FN and HGF-VN induced HMVEC migration is Ras dependent. Panel A -GTP-Ras pull down assay showing the affects of FPT III (100 μM) and GGTI (2 μM) inhibitors upon GTP-Ras levels. HMVEC suspensions (3 × 10 6 /ml) were pre-treated with FPT III and GGTI inhibitors for 45 min at room temperature prior to stimulation with HGF-FN complexes for 60 min at room temperature. Samples were analysed as mention in the legend to Fig 2. Panel B -Effect of Ras inhibition upon cell migration. Calcein M loaded HMVEC suspensions (1 × 10 5 /ml) were pre-treated with the inhibitors as mentioned above prior to application to the top chamber of the transwell filter. Cells were stimulated with HGF plus ECM proteins (placed in the bottom chamber) as shown. Cell migration was measured at 3 hours post stimulation using a fluorescence plate reader. The data is shown as relative migration and is the combined data from two independent experiments with sample wells in triplicate. To further characterise the role of Ras in regulating endothelial cell responses to HGF/ECM, the effect of the FPT-III inhibitor on the phosphorylation levels of Erk 1/2 and Akt was investigated. In cells stimulated with HGF alone, Erk 1/2 was activated and significantly inhibited by the FPT-III inhibitor and to a lesser extent by GGTI (Fig. 8A ). A similar inhibitory profile was observed for cells stimulated with HGF-FN (Fig. 8A ) and HGF-VN (not shown). In contrast, cells stimulated with HGF and collagen-1 showed no apparent reduction in Erk 1/2 phosphorylation levels when pre-treated with the FPT-III inhibitor suggesting little involvement of Ras in the activation of Erk 1/2. These samples were also assessed for Akt phosphorylation as an indication of PI-3 kinase activity. Consistent with our observations, stimulation of HMVEC with HGF in the absence of ECM did not lead to a significant activation of Akt (Fig. 8B ). However, in the presence of collagen, Akt activation was observed but this was not affected by pre-treatment of the cells with FPT-III implying that Ras was not an upstream regulator of the activation of PI-3 kinase. In contrast, HGF-FN complexes promoted a 3-fold enhancement of Akt phosphorylation and this was inhibited by approximately 50 % by treating the cells with FPT-III (Fig. 8B ). These observations suggest that the inhibition of Ras activation reduces the activation of PI-3 kinase for cells stimulated with HGF-FN complexes and not cells stimulated with HGF and collagen-1. The data would therefore predict that when HMVEC are stimulated with HGF-FN and HGF-VN complexes, specific integrins are utilized to recruit Ras, which in turn would regulate the activation of PI-3 kinase. To test this hypothesis, we immunoprecipitated integrins α5β1 and α2β1 from cells stimulated with HGF-FN and HGF with collagen-1 respectively and analysed these integrin immune complexes for the co-precipitation of Ras. High levels of Ras was specifically associated with α5β1 immune complexes and this appeared be independent of HGF stimulation. Little or no Ras was co-precipitated with the integrin α2β1 (Fig. 8C ). Figure 8 HGF-FN Induced Activation of Map Kinase and PI-3 Kinase Pathways is Ras dependent Panel A & B -HMVEC suspensions were pre-treated with buffer (no inhib.) or inhibitors FPT III (100 μM) and control GGTI (2 μM) for 45 min at room temperature prior to stimulation with HGF with no ECM proteins, or HGF plus collagen-1 or HGF-FN complexes for 60 min at room temperature. Cells were pelleted and lysed in an ice-cold lysis buffer. Samples were analysed by SDS-PAGE and Western blotting probing for phospho Erk 1/2 (panel A) and phospho AKT (panel B). The relative band intensities measured by image analysis software have been placed above each band as a ratio of the signal obtained in the sample where no inhibitor was present. Blots were stripped and re-probed with antibodies to Erk 2 and Akt to confirm equal loadings. Panel C-HMVEC suspensions (5 × 10 6 /ml) were incubated with either FN or collagen in the presence or absence of HGF as shown for 60 min at room temperature. Cells were pelleted and lysed in lysis buffer. Immunoprecipitation analyses of the lysates were performed with antibodies to integrins α5β1 and α2β1 and the immune complexes analysed by SDS-PAGE and Western blotting probing with antibodies to Ras. Visualization was by chemiluminescence. Discussion The major finding of the present report is that HGF-induced endothelial cell responses are significantly augmented through the formation of molecular complexes between this growth factor and the ECM glycoproteins FN and VN. The significance of this finding is highlighted by the observation that HGF-induced endothelial cell migration, a PI-3 kinase coupled response, did not occur in the absence of additional signals originating from the ECM. However, HGF-induced endothelial cell proliferation was evident in the absence of signals emanating from the ECM. These observations have led us to propose a model for the mechanisms of HGF-induced responses in endothelial cells (Fig. 9 ). This model shows that HGF alone can induce endothelial cell proliferation through its receptor Met via activation of the Ras-Erk kinase pathway (Fig. 9A ). However, this signal is insufficient to promote significant cell migration for which an additional signal(s) from the ECM via specific integrin ligation appears necessary for activation of the PI-3 kinase pathway (Fig. 9B & 9C ). Uniquely, in cells stimulated with HGF-FN or HGF-VN complexes, which promotes the association of Met with integrins, an enhanced and unique intracellular signal is generated by the recruitment and sustained activation of Ras, which presumably, concomitantly activates both p85 PI-3 kinase and Raf (Fig. 9C ). This is in contrast to the mechanism of activation of the PI-3 kinase pathway induced by HGF in the presence of collagen-1, which is Ras independent (Fig. 9B ). Figure 9 Mechanisms of HGF induced cellular responses in endothelial cells. Panel A -HGF in the absence of ECM molecules can activate the Map kinase pathway through the Met receptor tyrosine kinase leading to a proliferative response. Panel B -HGF and co-stimulation with collagen induces activation of both the Map kinase and PI-3 kinase pathways through an unknown mechanism presumably through integrin ligation, which is Ras independent. Panel C -HGF-FN (and by analogy HGF-VN) complexes promote enhanced cellular responses by promoting the association of integrins with the Met receptor leading to the recruitment and enhanced activation of Ras, Erk 1/2 kinases and PI-3 kinase promoting both elevated proliferative and migratory responses. The model in Fig. 9C is supported by the following observations. The enhanced responses of HMVEC to HGF-FN and HGF-VN complexes is consistent with the observation that in these cells the activity of Ras, PI-3 kinase (AKT phosphorylation) and Erk 1/2 phosphorylation were sustained, and in the case of Ras and Akt, were 2–3 fold higher than observed in cells stimulated with HGF and collagen-1 (Fig. 6 ). The distinct signalling mechanisms induced by the co-activation of endothelial cells with HGF in the presence of a binding and non-binding ECM glycoprotein partner was also supported by the observation that treatment of cells with the inhibitor of Ras farnesylation, FPT III, reduced the phosphorylation of both Erk 1/2 and Akt kinases in cells stimulated with HGF-FN but not HGF plus collagen-1 (Fig. 8A & 8B ). Furthermore, Ras co-precipitated with the integrin α 5 β 1 derived from endothelial cell lysates stimulated with HGF-FN complexes but not with the integrin α 2 β 1 derived from cells stimulated with HGF and collagen-1 (Fig. 8C ). These results are consistent with the pioneering work by Rodriquez-Viciana and colleagues who demonstrated the regulation of p85 PI-3 kinase by Ras via direct molecular interaction. It is now known that the regulatory subunit of all type 1, PI-3 kinases contain a Ras binding domain that associates with activated Ras (GTP-Ras) [ 17 - 20 ]. Therefore, our data and model demonstrating the sustained activation of Ras and PI-3 kinase by stimulation of endothelial cells with HGF-FN and HGF-VN complexes is consistent with previous work showing Ras to be a key regulator of PI-3 kinase. The identity of the Ras subtypes mediating the regulation of PI-3 kinase in our cell system is currently under investigation. The results of the present study both support and extend our previous observations of the enhanced endothelial cell migration induced by VEGF-FN molecular complexes [ 9 ]. In that study, VEGF binding domains identified on FN drove the formation of VEGF-FN complexes that upon receptor ligation promoted the association of the integrin α 5 β 1 with VEGFR-2. This co-receptor activation gave rise to a sustained activation of the Erk kinase activity, which promoted an enhanced migratory response. Similarly, the present work has shown that HGF-FN and HGF-VN molecular complexes induce the formation of Met-integrin signalling complexes promoting the transduction of a unique Ras-dependent signal. Several studies have illustrated the significance of the cooperation between integrins and growth factor receptor tyrosine kinases in mediating cellular responses. For example, the proliferation and migration of fibroblasts in response to PDGF-BB was enhanced in the presence of VN and was accompanied by the physical association of the α v β 3 integrin with the PDGF-β receptor [ 5 , 7 ]. Furthermore, it was recently demonstrated that HGF in combination with FN prolongs the survival of GM-colony-forming cells [ 21 ] and enhanced the adhesion and motility of MTLn3 breast carcinoma cells [ 22 ]. In addition, integrins α v β 3 and α v β 5 were shown to be necessary for mediating FGF-2 and VEGF mediated angiogenesis respectively by the differential regulation of components of the Erk kinase pathway [ 23 ]. However, the present study extends these observations and is, to our knowledge, the first description of a distinct signalling pathway employed by the activity of growth factor-ECM molecular complexes as opposed to growth factors and ECM proteins functioning independently through ligation of their respective receptors. The identification of a Ras-dependent pathway in endothelial cells specifically activated with HGF-FN and HGF-VN complexes as opposed to HGF in the presence of collagen-1 is significant and correlates with Met-integrin association. Although the precise nature of the interaction between the Met tyrosine kinase and integrins was not elucidated, the role of Ras in this system appears important for the sustained and enhanced activation of the PI-3 kinase and Erk kinase pathways. In contrast to the migratory signals promoted by VEGF-FN molecular complexes [ 9 ], HGF-FN and HGF-VN complexes induce a response in endothelial cells characterized by a tight coupling of the PI-3 kinase pathway to cell migration. Several additional pro-angiogenic mediators such as sphingosine 1-phosphate and NO, or the activation of CD40 and Eph B4 receptors by their counter ligands, promote endothelial cell migration through activation of the PI-3 kinase pathway [ 24 - 28 ]. In addition, HGF on its own was shown to stimulate smooth muscle cell migration in a PI-3 kinase dependent manner [ 29 ]. However, the lack of a significant migratory response, coupled with the absence of Akt phosphorylation observed in the present study, suggests that in primary endothelial cells the Met receptor is unable to activate PI-3 kinase without cooperative signals from the ECM/integrins. This observation is intriguing bearing in mind that Met has been shown to activate PI-3 kinase in epithelial cells via recruitment and activation of Gab-1, which directly interacts with the p85 subunit [ 30 ]. Consistent with our observation of an integrin dependency for signal transduction, Trusolino et al showed that in carcinoma cell lines Met induced signals were considerably amplified as a consequence of its constitutive association with the integrin α6β4. Intriguingly, in this system the authors showed that the role of the integrin α4 subunit was independent of extracellular integrin ligation since a truncated α4 construct lacking its extracellular portion could mediate HGF/Met responses and signals to downstream effectors provided that its ability to recruit the adaptor Shc was not affected [ 31 ]. In contrast, our studies using primary endothelial cells showed that integrin ligation was essential for generating a significant migratory signal via PI-3 kinase and in the case of HGF-FN and HGF-VN complexes, for promoting the association of Met with the integrins α 5 β 1 and α v β 3 respectively. Indeed, Met association with the integrins α5β1 and αvβ3 was dependent upon the activation of both Met and integrins through ligation of their cognate ligands since tyrosine phosphorylation of Met by HGF alone could not induce integrin association (Fig. 3C ). These observations support the contention of a signalling mechanism requiring the formation HGF-ECM molecular complexes as a prerequisite for Met-integrin association and consequent signal amplifiation as proposed in Fig. 9C . However, the importance of integrin cytoplasmic domains in recruiting Ras and Ras-binding partners appears to reflect a common mechanism of HGF signal transduction between these cellular systems. Conclusions The results of the present work demonstrate an important mechanism by which integrins collaborate with growth factor receptor tyrosine kinases on endothelial cells and predict that HGF binding domains on both FN and VN may play a significant role in promoting wound healing and post-natal neovascularization. In support of this contention, HGF-FN and HGF-VN complexes were identified in the supernatants derived from degranulated platelet suspensions indicating that these complexes do exist in vivo and may be deposited at sites of vessel perturbation or injury. This observation is similar to the identification of VEGF-FN molecular complexes in platelet supernatants in our previous report [ 9 ] and suggests that HGF and VEGF may act synergistically in vivo . Indeed, recent studies have shown that HGF synergises with VEGF to promote capillary-tube assembly in collagen matrices and neovascularization in the rat cornea [ 11 ]. Furthermore, HGF positively regulates VEGF expression and down regulates TSP-1, an inhibitor of angiogenesis, thereby promoting angiogenesis [ 32 ]. It is noteworthy that the HGF binding domains for FN were located in the same proteolytic fragments as those of VEGF, namely the N-terminal 70 kDa and C-terminal 40 kDa fragments. Further studies involving the fine mapping and characterization of the binding domains for VEGF and HGF on FN and VN should help decipher the mechanism of interplay between these important pro-angiogenic mediators. Methods Solid phase assay and Surface Plasmon Resonance Analysis (SPR) ECM proteins and FN peptides were purchased from Sigma and Gibco and were further purified by gel filtration and ion exchange chromatography. The assay was performed as described previously [ 9 ]. 125 I-HGF (NEN) in binding buffer (PBS containing 2% BSA) were added to the microtitre plates and incubated for 30 min at room temperature (RT) before washing and counting to determine bound radioactivity. SPR analysis was performed on the BIAcore X (Biacore Herts UK) as described previously [ 9 ]. HGF (30–500 nM) was injected across the FN 70 kDa fragment immobilised on a CM5 chip in HEPES saline (pH 7.4) supplemented with 1 mM MgCl 2 , 2 mM CaCl 2 and the sensograms recorded. The data was analysed by the ASSAY programme (Biosoft, UK) in order to determine the EC 50 value and K d . Migration and proliferation assays Human dermal microvessel endothelial cells (HMVEC) were maintained in EBM-2 growth medium (Clonetics Corp). Migration studies were carried out essentially as described previously [ 9 ] using serum starved Calcein AM-loaded HMVEC in a modified Boyden chamber assay using Fluorblok transwell chambers (BD Bioscience) as described by the manufacturer. Cell migration was detected by fluorescence measurement (within the lower chamber compartment). Membranes of transwell chambers were coated with either FN or VN or collagen-1 (10 μg/ml) overnight at 4°C and preliminary experiments were performed to assess the optimal dosage of both HGF and ECM protein. With antibody inhibition studies, the transwell chamber was coated with poly-L-lysine (Sigma) to facilitate cell attachment to the filters as opposed to adhesion using ECM glycoproteins. HMVEC were pre-treated with α v β 3 and α 5 β 1 integrin blocking antibodies for 30 min at room temperature prior to application to the upper transwell chamber. The levels of cell adhesion to ECM-coated transwell filters were determined by allowing HGF-stimulated HMVEC to adhere to transwells (coated overnight with ECM glycoprotein (10 μg/ml) and then blocked by incubation in 3.5 mg/ml BSA in basal culture medium) for 1 hour at room temperature followed by extensive washes with basal culture medium. The remaining cells were measured using a fluorescence plate reader (measuring fluorescence in the upper transwell compartment). For proliferation experiments, cell division was measure by fluorescence labeling of DNA (CyQuant, Molecular Probes). HMVEC was plated on poly-D-lysine coated 48-well plates and cultured overnight in MCDB-131 medium containing 5% FBS. After washing plates with PBS, endothelial cells were then cultured in MCDB-131 medium + 0.1% FBS containing HGF (10 ng/ml) in the presence or absence of VN, FN or Collagen-1 (10 μg/ml). Cells incubated for 48 h and HGF/ECM was added every 24 hours. Cell proliferation was quantified using a fluorescence plate reader. Phosphorylation analysis and ras activation HMVEC were assessed for the activation profiles of Erk1/2 and Akt using phosphospecific antibodies (Cell Signalling Technology) to Erk (Thr 202 /Tyr 204 ) and Akt (Ser 473 ) respectively by Western blotting. These studies were performed with both cells in suspension and with adherent populations. Cells were grown to 80% confluence and serum starved for 2 hours prior to harvesting. Cells were resuspended in serum-free MCDB-131 medium (BioWhittaker) supplemented with 0.1% BSA (resuspension buffer) at a concentration of 1–5 × 10 6 cells /ml. The cell suspensions were challenged 10 ng/ml HGF supplemented with 2 μg/ml collagen-1, or FN or VN for various durations ranging from 2 to 120 min at room temperature. Cells were harvested by centrifugation at 4°C and lysed in 10 mM Tris pH 7.4, 145 mM NaCl supplemented with 0.1% Triton X-100 and protease inhibitors. For inhibitor studies, serum-starved HMVEC suspensions were pre-treated with the inhibitor for 45 min prior to stimulation with HGF and ECM molecules for a further 60 min at room temperature. The cells were pelleted, washed in ice-cold resuspension buffer without BSA and lysed in a lysis buffer containing 1% (v/v) Triton X-100. Cell lysates were analysed by Western blotting using protocols specific for the phosphospecific antibodies according to the manufacturer's recommendations. Blots were cut along appropriate marker divides and probed with antibodies to phopho Erk 1/2 and Akt (Ser 473 ) (Cell Signaling Technologies) simultaneously. For GTP-Ras pull down assays, serum-starved HMVEC were stimulated with HGF and ECM molecules for a desired time point and the cells were spun down and washed in ice-cold resuspension buffer without BSA. Cell pellets were lysed in MLB buffer (25 mM HEPES pH 7.5, 150 mM NaCl, 1 % Igepal CA-630, 10% Glycerol, 25 mM NaF, 10 mM MgCl 2 , 1 mM EDTA, 1 mM sodium orthovanadate and protease inhibitor cocktail) and 500 μg of cell lysate was mixed with a 10 μl suspension of RBD-Sepharose (Upstate Biotechnology) for each reaction at 4°C for 60 min. Sepharose beads were spun down and washed in MLB prior to solubilization and analysis by Western blotting probing for Ras using a monoclonal antibody (Upstate Biotechnology). For Ras inhibition studies cells were pre-incubated with FPT-III (100 μM) and GGTI (2 μM) (approx 40 × IC 50 values) for 45 min at room temperature prior to cell stimulation for 60 min with HGF and ECM. Met-Integrin immunoprecipitation Human microvessel endothelial cells (HMVEC) in serum-free MCDB-131 medium (BioWhittaker) supplemented with 0.1% BSA were plated on collagen, FN and VN coated petri dishes in the absence or presence of HGF (50 ng/ml) for 15 min to 1 hour at room temperature. Cells were then harvested as described previously [ 9 ] and integrin immunoprecipitation was performed with monoclonal antibodies (Chemicon) to α 2 β 1 (clone JBS2), α 5 β 1 (clone JBS5) and α v β 3 clone (LM609). After analysis by SDS-PAGE and protein transfer, the blot was then probed with a monoclonal to Met (clone DL-21, Upstate Biotechnology) and developed by chemiluminescence. For Met tyrosine phosphorylation analysis, cells were stimulated with HGF alone or HGF-FN and HGF-VN complexes for various time points ranging from 15 mins to 2 hours at room temperature. Lysed samples were immunoprecipitated with a polyclonal anti-phosphoMet antibody (Cell Signalling) and the immune complexes analysed by SDS-PAGE and Western blotting using a monoclonal ant-Met antibody (Upstate). Met was visualised using chemiluminescence technology (Pierce). Immunoprecipitation of FN-HGF and VN-HGF complex from platelet supernatants Supernatants from non-stimulated and thrombin-stimulated washed platelet suspensions were prepared as previously described [ 9 ]. Supernatants were immunoprecipated with an antibody to FN or VN (Chemicon) or an isotype matched control reagent (IgG). Following SDS-PAGE and immunoblotting, HGF was detected with a polyclonal antibody (Santa Cruz) by chemiluminescent development. Abbreviations HGF, hepatocyte growth factor, FN fibronectin, VN vitronectin, Col-1, collagen-1, HMVEC, human microvessel endothelial cells. Authors' contributions SR, study design, HMVEC migration studies, SPR ligand binding analysis, HMVEC signalling studies, co-immunoprecipitation studies (integrin-Ras), manuscript preparation. ESW, study design, HMVEC proliferation studies, solid-phase ligand binding studies, co-imunoprecipitation (integrin-Met), manuscript preparation. YMP, study design, co-imunoprecipitation studies (HGF-FN/VN). JM, protein purification and preparation, KVP, GTP-Ras pull-down assays, RS, technical support. MS, study design. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC553973.xml |
552304 | Response of SII cortex to ipsilateral, contralateral and bilateral flutter stimulation in the cat | Background A distinctive property of SII is that it is the first cortical stage of the somatosensory projection pathway that integrates information arising from both sides of the body. However, there is very little known about how inputs across the body mid-line are processed within SII. Results Optical intrinsic signal imaging was used to evaluate the response of primary somatosensory cortex (SI and SII in the same hemisphere) to 25 Hz sinusoidal vertical skin displacement stimulation ("skin flutter") applied contralaterally, ipsilaterally, and bilaterally to the central pads of the forepaws. A localized increase in absorbance in both SI and SII was evoked by both contralateral and bilateral flutter stimulation. Ipsilateral flutter stimulation evoked a localized increase in absorbance in SII, but not in SI. The SII region that responded with an increase in absorbance to ipsilateral stimulation was posterior to the region in which absorbance increased maximally in response to stimulation of the contralateral central pad. Additionally, in the posterior SII region that responded maximally to ipsilateral stimulation of the central pad, bilateral central pad stimulation approximated a linear summation of the SII responses to independent stimulation of the contralateral and ipsilateral central pads. Conversely, in anterior SII (the region that responded maximally to contralateral stimulation), bilateral stimulation was consistently less than the response evoked from the contralateral central pad. Conclusions The results indicate that two regions located at neighboring, but distinctly different A-P levels of the anterior ectosylvian gyrus process input from opposite sides of the body midline in very different ways. The results suggest that the SII cortex, in the cat, can be subdivided into at least two functionally distinct regions and that these functionally distinct regions demonstrate a laterality preference within SII. | Background There is general agreement that in cats and monkeys (and presumably in humans) the spike discharge activity a mechanical stimulus sets up in rapidly adapting (RA), slowly adapting (SA), and Pacinian (PC) skin mechanoreceptors is projected centrally, at short latency and with relatively minor transformation, to primary somatosensory cortex (both SI and SII) in the contralateral hemisphere. There is no consensus, however, about the way in which the stimulus-evoked response in the ipsilateral hemisphere contributes to cerebral cortical somatosensory information processing and somatosensation. A distinctive property of SII is that it is the first cortical stage of the somatosensory projection pathway that integrates information arising from both sides of the body. Similar to SI, SII possesses a clear topographic organization [ 1 ], but unlike SI, a significant fraction of SII neurons possess bilateral receptive fields (RFs). The fraction of SII neurons with bilateral RFs varies from one topographic region of SII to the next. While ipsilateral input to SII is widely accepted, the role(s) of this input in somatosensory information processing remains uncertain. To investigate the effects on SII input deriving from mechanoreceptors in ipsilateral skin regions, the technique of optical intrinsic signal (OIS) imaging was used to assess the impact on SII of ipsilateral input on the response of SII to a contralateral input. The responses to contralateral, ipsilateral and bilateral vibrotactile stimulation (25 Hz – "flutter") of the forepaw of the cat were quantified and compared to make this assessment. Although evoked responses in both SI and SII were imaged in the studies, the primary focus of this report is the response of SII to the aforementioned stimuli. Results Figure 1 shows the OIS responses evoked in SII of two exemplary subjects by contralateral, ipsilateral, and bilateral central pad stimulation. Visual inspection of the images for the three stimulus conditions in each subject shows that: (1) the optical response to contralateral stimulation occurs in a region more anterior in SII than does the response to ipsilateral stimulation; (2) the SII optical response to bilateral stimulation occupies both the anterior and posterior regions that responded to independent stimulation of the contralateral and ipsilateral central pads; and (3) the optical response to ipsilateral stimulation does not evoke a large absorbance change in SI. To more accurately characterize the spatial disparity between SII loci activated by contralateral vs. ipsilateral stimulation, the absorbance values were obtained and plotted along the posterior-anterior axis of SII. Figure 2 shows an absorbance vs. distance plot for each of the 2 subjects whose images are shown in Figure 1 . Note that the distance between the peak absorbance values obtained under the ipsilateral and contralateral conditions is approximately 2 mm for Subject 1 (plots on left) and approximately 3 mm for Subject 2 (plots on right). The average across-subject (n = 6) distance between the peaks of SII activation evoked by contralateral and ipsilateral stimulation is 2.4 +/- 0.46 mm [ 2 ]. Interestingly, for both subjects, in the posterior region of SII, the magnitude of the response to bilateral stimulation exceeds that of the response to either contralateral or ipsilateral stimulation, but, in the anterior region of SII, the magnitude of response to contralateral stimulation is greater than that of the response to bilateral stimulation. The center of the distribution of the peaks of absorbance evoked by contralateral and ipsilateral stimulation, such as those shown in Figure 2 , were used to define the anterior and posterior regions of SII in subsequent analyses (e.g., the peaks were used as the center point of the sampled regions of interest). A more comprehensive view of the SII response to the contralateral and bilateral stimulus conditions can be better appreciated with a multi-dimensional surface plot of the data. Figure 3 compares the stimulus evoked response of SII in the two subjects to contralateral and bilateral stimulation, and from these plots, it is quite apparent that a large area (in the medial-lateral dimension) of the anterior region of SII is suppressed in the bilateral stimulus condition, relative to the response evoked by the contralateral stimulus. The surface plots of Figure 3 enable direct comparison of specific modules that exhibit a particular profile in one condition (e.g. the contralateral response of Subject 2 – see locations marked 1, 2 and 3) with that of another condition (compare with modules marked 1', 2' and 3' for the bilateral response). In this comparison, the absorbance values at loci 1 and 2 (in anterior SII) are clearly larger than those at locus 3 (in posterior SII) in the contralateral response, and the absorbance values at locations 1' and 2' (from the bilateral response) are much smaller than those at loci 1 and 2 of the contralateral response. Additionally, the response at locus 3 (posterior SII) in the contralateral response is very weak, but is larger in the response to the bilateral stimulus. Thus, the increase in activity observed at locus 3' (i.e., 3'>>3), in the posterior region of SII, evoked by bilateral stimulation parallels a decrease in activity at loci 1' and 2', in the anterior region of SII. In the case of Subject 1, the decrease in activity in anterior SII is not as pronounced as that seen in Subject 2 (also note difference in Figure 2 ), although the concurrent increase in posterior SII activity (compare module P with P') with decreased anterior SII activity (compare module A with A') is consistent with the shift in activation along the posterior-anterior axis observed in Subject 2. Responses evoked by ipsilateral stimulation are also displayed in Figure 3 . Note that in both Subjects 1 and 2, absorbance values in the posterior region are much greater than those in the anterior region of SII. To directly compare the time course of the response of the anterior and posterior regions of SII to the three stimulus conditions, we determined the time course of the absorbance changes in each region under each stimulus condition (Figure 4 ). The plots in Figure 4 show that in each of these 2 subjects and under each stimulus condition, the magnitude of the absorbance change evoked in either the posterior or anterior region of SII by ipsilateral stimulation was less than that evoked in the same region by contralateral or bilateral stimulation. Moreover, in both subjects, the magnitude of the response of the posterior region to bilateral stimulation was greater than that evoked by contralateral stimulation, whereas in the anterior region of SII, the magnitude of the response evoked by bilateral stimulation is either less than or approximately equal to the response evoked by contralateral stimulation. Cluster plots were used to directly compare the response of SII to the different conditions of ipsilateral and contralateral stimulation. In each plot in Figure 5 , the absorbance value obtained at each pixel to the 2 different stimulus conditions is plotted against each other – i.e., the x-axis is the absorbance value evoked by the contralateral stimulus and the y-axis is the absorbance value evoked by the ipsilateral stimulus. The clusters reveal a distinct differentiation in the population of SII neurons to the different stimulus conditions. Additionally, it appears that this could be a time dependent process, as there is little difference in the behavior of the pixels localized to the SII region in the early stages of the response (t = 1 sec), there is some grouping after 2 seconds, and there are two distinct clusters formed after several seconds (t = 5 sec). It should be emphasized that this type of graphic does not necessarily reflect spatial differences in the responses of two different stimulus conditions, but rather, it emphasizes whether or not different members of a set respond differently to different stimulus conditions. Thus, the information demonstrated by the cluster plots in Figure 5 can be summarized by stating that with an increase in stimulus duration (from 1 to 5 seconds), there is an increase in the segregation of the behavior of the different groups of pixels whose values are more predominantly affected by a contralateral vs. an ipsilateral stimulus. The contralateral and bilateral stimulus conditions can also be compared using cluster analysis. However, because the difference between the contralateral and ipsilateral responses is more robust than the difference between the contralateral and bilateral responses, Figure 6 displays the results of cluster analysis of the anterior and posterior regions of SII independently. Independent analysis of the two regions allows for better resolution of shifts in the behavior of activity within particular regions. For each subject, the peak response was identified in both the anterior and posterior regions. Cluster plots were obtained by plotting the response (contralateral along the x-axis, bilateral along the y-axis) at each location within a 1 × 1 mm 2 boxel surrounding the peak. In anterior SII, the majority of the responses to contralateral stimulation are stronger than the responses to bilateral stimulation – hence, the majority of the points plotted fall below the reference line (which has a slope of 1). In the posterior region, the majority of the points plotted are above the reference line – indicating that the response to the bilateral stimulus was greater than the response to the contralateral stimulus. Thus far, the results suggest that the anterior and posterior regions of SII are differentially activated by contralateral, ipsilateral and bilateral stimulation. To determine the across-subject consistency of these findings, the average absorbance values evoked by the 3 different stimulus conditions were determined for all 6 of the subjects (Figure 7 ). Clearly, in the anterior region of SII, the contralateral stimulus condition evoked the largest magnitude of response, and therefore, the values of the absorbance increase obtained in this and the posterior region under each stimulus condition were normalized to this absorbance value (thus, standard error for the contralateral/anterior region condition = 0). Whereas in the anterior region of SII, the response evoked by bilateral stimulation is approximately 35% less than that evoked by contralateral stimulation, in the posterior region of SII, the bilateral stimulus evoked a response approximately 25% larger than that evoked by the contralateral stimulus. Analysis of variance showed that, at a 95% confidence interval, the average bilateral:contralateral response ratio was between 0.42 and 0.83 in the anterior region of SII. In the posterior region of SII, the same analysis showed the bilateral:contralateral response ratio to be between 0.97 and 1.43 at a 95% confidence interval. Although the response of the posterior region to the bilateral stimulus is larger than that evoked by the contralateral stimulus, it is less than that predicted by summation of the responses to the ipsilateral and contralateral stimuli (computed values, Figure 7 ). Furthermore, while the response of neither the anterior or posterior regions of SII to bilateral stimulation approximate a linear summation of the responses to independent stimulation of each central pad, the approximation of the bilateral response by summation of responses evoked by independent stimuli is much closer to being accurate in the posterior region of SII. Discussion The findings of this study demonstrated clearly that the anterior and posterior regions of SII process bilateral inputs very differently. At the locus of the maximal OIS response evoked in the posterior region by an ipsilateral stimulus, bilateral stimulation evoked a response that was, on average, about 25% larger than that evoked from the contralateral stimulus site. Conversely, at the locus of the maximal OIS response evoked by contralateral stimulation in the anterior region, bilateral stimulation evoked a response that was, on average, 35% lower than the activity evoked by a contralateral stimulus. This discrepancy between the optical responses of the anterior and posterior regions could be related to neurophysiological observations reported in earlier studies. For example, Carreras and Andersson [ 3 ] found that for a sizable fraction of the cat SII neurons in their study, ipsilateral mechanical skin stimulation inhibited the response to contralateral stimulation, whereas in contrast, Picard et al. [ 4 ] found in their study of neurons in the distal forelimb regions of cat SII that simultaneous delivery of contralateral and ipsilateral mechanical skin stimuli led to strong facilitation of SII neuron response. In the study of Picard et al. [ 4 ], the responses of cells to bilateral stimulation were found to exceed the stronger of the responses to unilateral stimulation by, on average, 230%. Their study was limited, however, to the very low numbers of SII neurons that had bilateral RFs on the distal limbs. Burton, et al. [ 5 ], similar to Carreras and Andersson [ 3 ], reported that SII cells with bilateral receptive fields (monkey) exhibited a reduction in mean firing rate of 30% when the contralateral stimulus was preceded by an ipsilateral stimulus. Finally, other workers have found that callosally-transmitted inputs tend to have excitatory effects on SII neurons that have bilateral RFs, and exert inhibitory effects on SII neurons that have exclusively contralateral RFs [ 6 - 8 ]. Simoes et al. [ 9 ] showed significant suppression of the MEG SII response in humans, with simultaneous inputs delivered to the same skin sites, and Hoechstetter et al. [ 10 ] described "interactions" in SII cortex (a response that was not the summation of the ipsilateral and contralateral response) to simultaneous bilateral stimuli. Definitive establishment of the relationship between stimulus-evoked SII neuroelectrical and OIS activation, however, must await the performance of combined imaging and neurophysiological investigations which utilize both methodologies in the same subjects and under the same stimulus conditions. The main, although not the only, route for ipsilateral input to SII is through the corpus callosum, from cells located in SI and SII of the opposite cerebral hemisphere [ 4 , 7 , 8 , 11 ]. Even those regions in SII that represent most distal parts of the limbs receive significant numbers of connections from the homologous zones of the contralateral SI and SII [ 12 - 15 ]. Graziosi [ 16 ] showed that separate populations of cells in SI provide callosal projections to SI and SII in the opposite hemisphere and ipsilateral projections to SII. Some separation within SII of the responses to ipsilateral and contralateral stimulation was also shown by Friedman et al. [ 17 ] and Juliano et al. [ 18 ]. The neurons in the distal limb regions of SII do receive substantial callosal connections, but these neurons have been reported to lack ipsilateral RFs [ 1 ], indicating that callosal inputs are not strong enough to generate action potentials (at least under the conditions used in RF mapping studies). This suggests that SII neurons do not use their sensory inputs from the ipsilateral side of the body to construct functional properties dependent on bilateral inputs; in other words, to extract information about higher-order properties of bi-manually contacted objects from coordinated patterns of sensory stimulation of the two hands. Instead, it could be postulated that neurons in the distal limb regions of SII use their ipsilateral peripheral inputs to modulate the responses to contralateral peripheral stimulation. On the other hand, Bennett et al. [ 19 ] found that bilateral convergence on SII neurons varies markedly with the different classes of tactile neurons, and modulation of the SII response by ipsilateral inputs may vary from one cortical area to another with different stimulus modalities. A number of interactions between stimuli applied to both hands have been demonstrated in human psychophysical studies. Gilson [ 20 ] found that the threshold for detection of vibrotactile stimuli applied to a fingertip is elevated by parallel stimulation of the other hand's fingers. In addition, Gescheider and Verrillo [ 21 ] reported that the magnitude of vibrotactile sensation, elicited by brief 25 or 300 Hz stimuli applied to thenar eminence, was decreased by stimuli applied simultaneously to the opposite hand, but was enhanced when the contralateral stimulus was applied 150 msec prior to the test stimuli. Essick and Whitsel [ 22 ] reported that the perception of the direction of motion of brushing stimuli on the skin is enhanced by the presence of a simultaneous contralateral brushing stimulus when the two stimuli move in the same direction, but is weakened when the contralateral stimulus moves in a direction opposite to that on the other arm. While the above described reports provide possible perceptual correlates for bilateral interactions that might occur in SII, such as those identified in the present study, it will remain uncertain until anterior or posterior SII cortical activity is studied under conditions that permit direct correlations of perceptual performance and cortical activity under precisely controlled conditions of contralateral vs. bilateral skin stimulation. A recent report [ 23 ] demonstrated 3 separate functional cortical fields along the anterior-posterior axis in the macaque. These functional fields were defined based on differential neural responses from three distinct cortical fields, and their report was unique in that it described cortical areas within SII based on functional properties of cortical areas. In this report, we demonstrate at least two functional subdivisions within SII in the cat based on functional properties as well. However, the modes of stimulation used to distinguish the functional differences along the anterior-posterior axis of SII were very different in this study (contralateral/ipsilateral/bilateral vs. proprioceptive/cutaneous inputs in the Fitzgerald study), and subsequent investigations using other stimulus modalities could reveal that SII of the cat is organized in a very similar fashion to SII of primates. The multiple fields found in SII, based on functional differences, could be, as suggested by Fitzgerald, et al. [ 23 ], indicative of the existence of a number of distributed processing streams. The significance of the presented work is that the response of these different cortical areas, which could represent information from so-called separate information streams, changes in a manner dependent upon the activity of neighboring cortical areas. Distinction of cortical areas within SII, identified by functional characteristics, demonstrates the nonlinearity of the integration of information from different sources (or information streams). One question that the results suggest is whether or not SII can be segregated by laterality preference, in a manner similar to that observed in other sensory systems. Laterality has been demonstrated in the primary sensory cortex of both the visual system and the auditory system of both primates and cats, and the data in this report strongly suggest that there are cortical areas within SII that exhibit preference to ipsilateral or contralateral inputs. In terms of processing information from simultaneous contralateral and ipsilateral stimuli, there could be further similarities between the somatosensory, auditory and visual systems that have yet to be described. Future investigations will aim to further clarify the role of SII in integration of information from inputs across the body midline. Conclusions The responses evoked by contralateral and ipsilateral flutter stimulation of the central pad of the cat forepaw define functional subdivisions in SII: the two modes of stimulation maximally activate cortical regions that are anterior and posterior to one another, respectively. Bilateral stimulation, or providing simultaneous contralateral and ipsilateral stimulation, reveals, additionally, that the two adjacent cortical areas process bilateral inputs differently. In the posterior region, where ipsilateral stimulation evokes a maximal response, bilateral stimuli evoke a response that is greater than the response evoked by either the individual ipsilateral or contralateral response. In the anterior region of SII, where the contralateral stimulus evokes a maximal response, bilateral stimuli evoke responses that are smaller in magnitude than the responses evoked by the contralateral stimulus. Methods Subjects & preparation Adult cats (males and females; n = 6) were subjects. All surgical procedures were carried out under deep general anesthesia (1 – 4% halothane in a 50/50 mixture of oxygen and nitrous oxide). After induction of general anesthesia the trachea was intubated with a soft tube and a polyethylene cannula was inserted in the femoral vein to allow administration of drugs and fluids (5% dextrose and 0.9% NaCl). For each subject, a 1.5 cm diameter opening was made in the skull overlying somatosensory cortex, a chamber was mounted to the skull over the opening with dental acrylic, and the dura overlying anterior parietal cortex was incised and removed. Following the completion of the surgical procedures all wound margins were infiltrated with long-lasting local anesthetic, the skin and muscle incisions were closed with sutures, and each surgical site outside the recording chamber was covered with a bandage held in place by adhesive tape. Subjects were immobilized with Norcuron and ventilated with a gas mixture (a 50/50 mix of oxygen and nitrous oxide; supplemented with 0.1 – 1.0% halothane when necessary) delivered via a positive pressure respirator 1–3 hours prior to the data acquisition phase of the OIS imaging experiments. Respirator rate and volume were adjusted to maintain end-tidal CO 2 between 3.0 – 4.0%; EEG and autonomic signs (slow wave content; heart rate, etc.) were monitored and titrated (by adjustments in the anesthetic gas mixture) to maintain levels consistent with light general anesthesia. Rectal temperature was maintained (using a heating pad) at 37.5°C. Euthanasia was achieved by intravenous injection of pentobarbital (45 mg/kg) and by intracardial perfusion with saline followed by fixative (10% formalin). Following perfusion fiducial marks were placed to guide removal, blocking, and subsequent histological sectioning of the cortical region studied. All procedures were reviewed and approved in advance by an institutional committee and are in full compliance with current NIH policy on animal welfare. Stimuli and stimulus protocols Results were obtained during stimulation of the contralateral central pad of the forepaw and/or the ipsilateral central pad of the forepaw. The stimuli always consisted of sinusoidal vertical skin displacements (25 Hz, 400 microns, stimulus duration 5 – 20 sec, inter-stimulus interval 60 sec) and were applied using a servocontrolled transducer (Cantek Enterprises, Canonsburg, PA) that is capable of delivering sinusoidal stimuli in the range of 1–250 Hz at amplitudes in the range of 0–1000 microns. The stimuli were delivered independently to the ipsilateral and contralateral skin sites, and also were applied simultaneously to both sites (bilateral stimulation). The stimulus probes were positioned 500 microns beyond the point at which skin contact was detected (via force transducer on the Cantek). The bilateral stimulus protocols reported in this paper were synchronized to start and stop at the same time. The contralateral, ipsilateral and bilateral stimuli were interleaved on a trial-by-trial basis. This approach was used to control for temporal changes in cortical "state" unrelated to stimulus conditions which, if unrecognized, might obscure or modify any differences between the optical responses evoked by the contralateral, ipsilateral and bilateral stimulus conditions. OIS imaging Near-infrared (IR; 833 nm) OIS imaging was carried out using an oil-filled chamber capped with an optical window [ 24 ]. Images of the exposed cortical surface were acquired 200 msec before stimulus onset ("reference" or "prestimulus" images) and continuously thereafter ("poststimulus" images; at a resolution of one image every 0.5 to 1.5 sec) for 15–20 sec following stimulus onset. Exposure time was 200 msec. Absorbance images were generated by subtracting each prestimulus (reference) image from its corresponding poststimulus image and subsequently dividing by the reference image. Averaged absorbance images typically show regions of both increased absorption of IR light and decreased absorption of light (to a depth of approximately 1400 microns) which have been shown to be accompanied by increases and decreases in neuronal activation, respectively [ 24 - 29 ]. Histological procedures/identification of cytoarchitectural boundaries At the conclusion of the experiment, the imaged cortical region was removed immediately following intracardial perfusion with saline and fixative. The region then was blocked, postfixed, cryoprotected, frozen, sectioned serially at 30 μm, and the sections stained with cresyl fast violet. The boundaries between adjacent cytoarchitectonic areas were identified by scanning individual sagittal sections separated by no more than 300 μm and were plotted at high resolution using a microscope with a drawing tube attachment. The resulting plots then were used to reconstruct a two-dimensional surface map of the cytoarchitectonic boundaries within the region studied with optical and neurophysiological recording methods. The locations of microelectrode tracks and electrolytic lesions evident in the histological sections were projected radially to the pial surface and transferred to the map of cytoarchitectonic boundaries reconstructed from the same sections. As the final step, the cytoarchitectonic boundaries (along with the locations of microelectrode tracks and lesions whenever present) identified in each brain were mapped onto the images of the stimulus-evoked intrinsic signal obtained from the same subject, using fiducial points (made by postmortem applications of india ink or needle stabs) as well as morphological landmarks (e.g., blood vessels and sulci evident both in the optical images and in histological sections). Locations of cytoarchitectonic boundaries were identified using established criteria [ 30 - 32 ]. Abbreviations A-P = anterior-posterior RA = rapidly adapting SA = slowly adapting PC = Pacinian RF = receptive field OIS = optical intrinsic signal IR = near infrared EEG = electro-encephalogram MEG = magneto-encephalogram Authors' contributions BW and OF participated in the design of the experiments, the data collection, and drafting of the manuscript. SS, JC, and VT made significant contributions to the data collection and the analysis of the data. MT played a major role in all aspects of the development of the manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC552304.xml |
546417 | The burden of non communicable diseases in developing countries | Background By the dawn of the third millennium, non communicable diseases are sweeping the entire globe, with an increasing trend in developing countries where, the transition imposes more constraints to deal with the double burden of infective and non-infective diseases in a poor environment characterised by ill-health systems. By 2020, it is predicted that these diseases will be causing seven out of every 10 deaths in developing countries. Many of the non communicable diseases can be prevented by tackling associated risk factors. Methods Data from national registries and international organisms are collected, compared and analyzed. The focus is made on the growing burden of non communicable diseases in developing countries. Results Among non communicable diseases, special attention is devoted to cardiovascular diseases, diabetes, cancer and chronic pulmonary diseases. Their burden is affecting countries worldwide but with a growing trend in developing countries. Preventive strategies must take into account the growing trend of risk factors correlated to these diseases. Conclusion Non communicable diseases are more and more prevalent in developing countries where they double the burden of infective diseases. If the present trend is maintained, the health systems in low-and middle-income countries will be unable to support the burden of disease. Prominent causes for heart disease, diabetes, cancer and pulmonary diseases can be prevented but urgent (preventive) actions are needed and efficient strategies should deal seriously with risk factors like smoking, alcohol, physical inactivity and western diet. | Background For centuries, communicable diseases were the main causes of death around the world. Life expectancy was often limited by uncontrolled epidemics. After the second World War, with medical research achievements in terms of vaccination, antibiotics and improvement of life conditions, non communicable diseases(NCDs) started causing major problems in industrialized countries. Heart diseases, cancer, diabetes, chronic pulmonary and mental diseases became a real burden for health systems in developed countries. For a while, these diseases were associated with economic development and so called diseases of the rich. Then, by the dawn of the third millennium, NCDs appeared sweeping the entire globe, with an increasing trend in developing countries (Table 1 ) where, the transition imposes more constraints to deal with the double burden of infective and non-infective diseases in a poor environment characterized by ill-health systems. In 1990 the leading causes of disease burden were pneumonia, diarrhoeal diseases and perinatal conditions. By 2020, it is predicted that NCDs will account for 80 percent of the global burden of disease, causing seven out of every 10 deaths in developing countries, compared with less than half today[ 1 , 2 ]. Table 1 Evolution of NCDs in developing countries (in million) [1,8,9] Non-Communicable Diseases Communicable Diseases + Maternal + Perinatal + Nutritional Injuries total 1990 18.7 (47%) 16.6 (42%) 4.2 (11%) 39.5 (100%) 2000 25.0 (56%) 14.6 (33%) 5.0 (11%) 45.0 (100%) 2020 36.6 (69%) 09.0 (17%) 7.4 (14%) 53.0 (100%) Efficient (preventive)strategies are needed and urgent measures should be taken to control risk factors like tobacco, alcohol, obesity, blood pressure diet and inactivity. Otherwise, developing countries will be unable to provide their people with standard health care. The costly and prolonged treatment of NCDs raises the equity problem between and within countries. As expressed by the WHO Director-General in his overview to the annual report[ 1 ], If a Japanese woman develop chronic diseases, excellent treatment and rehabilitation services will be available and she can expect to receive, on average, medications worth about US$ 550 per year and much more if needed. Meanwhile, a woman in Sierra Leone can expect, on average, medicines worth about US$ 3 per year and, if she survives middle age and develop chronic diseases then she will die prematurely as a consequence of inadequate treatment. The contrasts in opportunities of treatment exist also within developing countries; between poor and rich, cities and rural areas and also between men and women. In previous papers, the authors proposed mathematical models dealing with the burden of diabetes and its complications[ 3 ], Dynamics of a disabled population[ 4 ], the effect of physical exercise[ 5 ] and a model of dengue fever[ 6 ]. The present paper is devoted to the burden caused by NCDs in developing countries. In order to reverse the increasing trend of this burden (or at least to control it), the focus is made on the risk factors associated with these diseases. Different methods can be considered to quantify the burden of NCDs. In order to overcome the specific problems of each country, the most used method is the approach that measures the global burden of NCDs in terms of Disability Adjusted Life Years (DALYs) which is a combination of Years of Life Lost(YLL) through premature death, and Years Lived with Disability(YLD). Thus, DALY is thought of as one lost year of healthy life [ 7 - 9 ]]. For example, deaths from underweight every year rob the world's poorest children of an estimated total of 130 million years of healthy life[ 10 ]. According to this approach, the burden of adult NCDs account for 80% in developed countries and for 70% in middle-income countries. Even in the high-mortality regions of the world, almost 50% of the adult disease burden is attributable to NCDs. Methods Data from national registries and international sources are collected, compared and analyzed in order to show the trend of NCDs. Four diseases or cluster of diseases(Cardio-Vascular Diseases(CVDs), diabetes, cancers and chronic respiratory diseases) are considered to illustrate the growing burden of NCDs in developing countries. The main sources of data are the annual reports and regular publications released by the World Health Organization(WHO), World Heart Federation(WHF), Pan American Health Organization(PAHO), International Diabetes Federation (IDF), International Agency for Research on Cancer(IARC), Centre for Chronic Disease Prevention and Control(CCDPC), International Task Force for Prevention of Coronary Heart Disease and a multitude of websites and papers dealing with NCDs. The literature associated with these diseases in developed countries is abundant. However, despite the encouraging programmes and joint projects proposed by WHO and other organisms in the form of collaborative research agreements to developing countries, in order to support national registries, unreliable and insufficient data are still prevailing in most of these countries. Moreover, the release of health data is shadowed by the security vision in some countries. Extrapolations are needed in the case of missing or incomplete data. Consequently, more efforts are needed to convince health decision makers in low- and middle-income countries of the necessity to develop epidemiological studies that allow for preventive strategies making health policy at the centre of sustainable development. Results According the World Health Organization's statistics, chronic NCDs such CVDs, diabetes, cancers, obesity and respiratory diseases, account for about 60% of the 56.5 million deaths each year and almost half of the global burden of disease. In 1990, 47% of all mortality related to NCDs was in developing countries, as was 85% of the global burden of disease and 86% of the DALYs attributable to CVDs. An increasing burden will be born mostly by these countries in the next two decades. The socio-economic transition and the ageing trend of population in developing countries will induce further demands and exacerbate the burden of NCDs in these countries. If the present trend is maintained, it is predicted that, by 2020, NCDs will account for about 70 percent of the global burden of disease, causing seven out of every 10 deaths in developing countries, compared with less than half today. In 1990, approximately 1.3 billion DALYs were lost as a result of new cases of disease and injury, with the major part in developing countries. In 2002, these countries supported 80% of the global YLDs due to the double burden of communicable and non communicable diseases. Consequently, their people are not only facing higher risk of premature life(lower life expectancy) but also living a higher part of their life in poor health[ 1 ]. These remarks indicate that NCDs are exacerbating health inequities existing between developed and developing countries and also making the gap more profound between rich and poor within low and middle-income countries. CVDs in developing countries CVD is the name for the group of disorders of the heart and blood vessels and include hypertension (high blood pressure), coronary heart disease (heart attack), cerebrovascular disease (stroke), peripheral vascular disease, heart failure, rheumatic heart disease, congenital heart disease and cardiomyopathies. These diseases constitute the major contributor among NCDs (Table 2 ). Table 2 Deaths caused wordwide by specific diseases (× 10 3 ) Deaths & % Disease 2002 [1] 1990 [8] Ischaemic heart disease 7000 (12.6%) 6260 (12.4%) Cerebrovascular disease 5400 (09.6%) 4380 (08.7%) Lower Respiratory Diseases 3700 (06.6%) 4300 (08.5%) COPD 2700 (04.8%) 2211 (04.4%) Cancer(all types) 7100 (12.6%) 6200 (11.2%) Diabetes 3200 (05.6%) 2400 (05.0%) Worldwide, an estimated 17 million people die of these diseases, particularly heart attacks and strokes, every year. Once associated with industrialized countries, CVDs are now emerging or rapidely increasing in developing countries. Indeed, in 1998, 86% of the DALYs caused by CVDs were attributed to developing countries and in 1999 CVDs contributed to a third of global deaths with 78% in low- and middle-income countries. The trend is increasing, indicating that by the year 2010 CVDs will be the leading cause of death in developing countries as a consequence of lifestyle changes brought about by industrialization and urbanization in developing countries engaged in the socio-economic transition. CVDs are promoted by risk factors like tobacco use, alcohol, physical inactivity and unhealthy diet. Unfortunately, the harm caused by these risk factors affects the rise of life expectancy in developing countries[ 1 , 11 , 12 ]. The costly and prolonged care of CVDs in low-and middle-income countries often divert the scarce family and societal resources to medical care. Consequently, the lower socio-economic groups have greater prevalence of risk factors, higher incidence of disease and higher mortality. Diabetes The recent statistics released by the World Health Organization and the International Diabetes Federation are alarming[ 1 , 12 ](Table 3 ). The number of diabetes in the world is expected to increase from 194 Million in 2003 to 330 in 2030 with three in four living in developing countries. Moreover, in developed countries most people with diabetes are above the age of retirement, whereas in developing countries those most frequently affected are aged between 35 and 64 which makes the burden in terms of DALYs and YLDs heavier in poorer countries. Indeed, in some countries of the Middle East, one in four deaths in adults aged between 35 and 64 years is attributable to diabetes. The burden is exacerbated by the complications such as blindness, amputations and kidney failure for which diabetes is the leading cause, and the interfering action of CVDs which are responsible for between 50 and 80% of deaths in people with diabetes. The burden of premature death from diabetes is similar to that of HIV/AIDS, yet the problem is largely unrecognised [ 13 ]. Table 3 Diabetes prevalence (× 10 6 ) [13] Country 2000 2030 India 31.7 India 79.4 China 20.8 China 42.3 United States 17.7 United States 30.3 Indonesia 8.4 Indonesia 21.3 Japan 6.7 Pakistan 14.9 Pakistan 5.2 Bangladesh 11.8 Russia 4.6 Brazil 11.3 Brazil 4.5 Japan 8.9 Italy 4.2 Italy 5.4 Bengladesh 3.2 Russia 5.3 Studies in different countries have shown that diabetes is a costly disease accounting for between 2.5 and 15% of the total healthcare expenditure[ 3 ]. For the age category 20–79, the world annual direct cost is estimated to be over $153 billion and expected to double in 2025. According to the National Institute of Diabetes and Digestive Kidney Disease(NIDDK) and the American Diabetes Association, diabetes was the sixth leading cause of death in 1999 with a direct cost of $44 billion and an indirect cost of $54 billion annually. In 2002, the direct and indirect cost totalled $132 billion. In France, an estimation of $5.7 billion was given for the direct cost of diabetes, whereas, an equivalent cost of 5.2 billion, representing approximately 9% of the annual NHS budget, was given for UK in 2000. The burden affects more and more developing countries as stressed by the different authors who attended the seventh congress of the Pan-African diabetes study group in 2001[ 14 ] and the Metabolic syndrome, type II diabetes, and atherosclerosis congress in 2004[ 15 ]. Cancer Cancer is now a major cause of mortality throughout the world (Table 4 ). In the developed world, it is generally exceeded only by CVDs but developing countries are responsible for the globally increasing trend. Over 10 million new cases and over 7 million deaths from cancer occurred worldwide in 2000[ 1 , 2 , 16 - 19 ]]. The contribution of developing countries was 53% for incidence and 56% for deaths (Table 5 ). From 1990 to 2000, the incidence and deaths increased by 2.4% per annum. Table 4 Cancer by types and numbers worldwide (× 10 3 ): Cancer 2000 [18] Incidence % 2000 deaths % 1990 Incidence % 1990 [19] deaths % Lung 1239 12.3 1103 17.8 1037 12.8 921 17.8 Breast 1050 10.4 373 6.0 796 9.8 314 6.1 Colorectal 945 9.4 492 8.0 783 9.7 437 8.4 Stomach 876 8.7 646 10.0 798 9.9 628 12.1 Liver 564 5.6 546 8.8 437 5.4 427 8.2 Prostate 543 5.4 204 3.3 396 4.9 165 3.2 Cervical 471 4.7 233 3.7 371 4.6 190 3.7 Oesophagus 413 4.1 337 5.4 316 3.9 286 5.5 Head&neck 390 3.9 207 3.3 306 3.8 162 3.1 Bladder 336 3.3 132 2.1 261 3.2 115 2.2 Other 3228 32.2 1934 31.0 2582 32.0 1537 30.0 Total 10055 100% 6209 53% 8083 100.0 5182 100.0 Table 5 Cancer in developing countries :incidence & deaths (× 10 3 ) in 2000 [18] Cancer Developing Incidence % Developing deaths % Lung 792 14.7 522 14.6 Breast 471 8.8 184 5.6 Colorectal 334 6.2 252 7.0 Stomach 543 10.1 417 11.7 Liver 457 8.5 443 12.4 Prostate 127 2.4 76 2.1 Cervical 379 7.0 194 5.4 Oesophagus 341 6.3 274 7.7 Head&neck 262 4.9 154 4.3 Bladder 124 2.3 65 1.8 Other 1546 71.2 992 27.8 Total 5376 100% 3563 57.4 Between 2000 and 2020, the total number of cases of cancer in the developed world is predicted to increase by 29% whereas, in developing countries an increase by 73% is expected (largely as a result of an increase in the number of old people and as a result of urbanization and change in dietary habits). The incidence of cancers of the lung, colon and rectum, breast and prostate generally increases in parallel with economic development, while the incidence of stomach cancer usually declines with development[ 2 ]. Lung cancer This is currently the most common cancer in the world. In developed countries, smoking causes over 80% of such cancers and generally, heavy smoking increases the risk by around 30-fold making lung cancer a major problem in developing countries where the consumption of tobacco is flourishing. Breast cancer According to the International Agency for Research on Cancer (IARC), there were over a million new cases in the world in the year 2000, making it the second most common in the world and the most common among women with 47% in developing countries. Although rates are five times higher in industrialized countries, the burden of disease is heavier in poorer countries because breast cancer is highly curable if detected early and, unfortunately, about 80% of the cases are detected at advanced stages in developing countries. Colorectal cancer Ranking at the third place, with incidence rates tenfold higher in developed than in developing countries, this type of cancer is assumed to be mainly related to dietary factors which account to up to 80% of the between-country differences in rates. Stomach cancer 20 years ago, this cancer used to be the most common in the world. At the moment, it is the fourth most common in the world but the second most common in developing countries. Substantial evidence suggests that risk is increased by high intakes of some traditionally preserved salted food and that risk is decreased by high intakes of fruit and vegetables. Liver cancer Approximately 75% of cases occur in developed countries, the rate vary over 20fold between countries. In developing countries, ingestion of contaminated food is an important risk factor together with active hepatitis virus infection whereas, alcohol consumption is the main diet-related risk factor in the world. Cervical cancer 80% of the new cases and deaths are occurring in developing countries where it constitutes a major health problem. In developed countries, screening programmes and early detection have led to a noticeable decline in cervical cancer incidence and mortality, whereas, the trend is stable or increasing in low- and middle-income countries owing to their limited health care resources but also to their ill-health systems generating inefficient (or no)strategies[ 20 ]. Oral cavity, pharynx and oesophagus In developed countries these types of cancer are mainly correlated to alcohol and tobacco(up to 75% of such cancers are attributable to these two lifestyle factors). In developing countries, around 60% of such cancers are thought to be a result of micronutrient deficiencies related to a restricted diet that is low in fruit and vegetables and animal products. There is also consistent evidence that consuming drinks and foods at a very high temperature increases the risk for these cancers[ 2 ]. Pancreatic, endometrial, prostate and kidney cancers These types of cancer are more common in industrialized countries. However, the fact that overweight/obesity is an established risk factor, their incidence is expected to increase in developing countries engaged in the socio-economic transition[ 2 ]. Chronic respiratory diseases Chronic respiratory diseases represent a major burden for the health systems worldwide. Most developing countries have no standard protocols for assessing and managing chronic non communicable respiratory diseases such as Chronic Obstructive Pulmonary Disease (COPD) and Asthma. In these countries, the population afflicted by poverty and illiteracy, having very little (or no) access to health services, will die before the age of 40 years. They comprise 15% of the population in Latin America, 34% in Arab world, 45% in Sub-Saharan Africa and south-east Asia[ 21 , 22 ]. Respiratory diseases cause 15% of the global burden of disease. Worldwide, it is estimated that 600 million people suffer from COPD and 2.5 million deaths were attributed to these diseases in 2000. By 2020, COPD is expected to become the third most common cause of mortality in the world. Discussion Risk factors: the enemies In the previous sections, we considered four classes of non communicable diseases, namely, CVDs, diabetes, cancer and chronic respiratory diseases. Despite some differences between these classes and into each class, they do have a common denominator which is the risk factors. Indeed, Tobacco, alcohol, high blood pressure, diet and physical inactivity were indicated, at different levels, as risk factors in the four classes of NCDs. Moreover, these risk factors are seen to affect people worldwide with an increasing tendency. (Table 6 ) Table 6 Burden of disease and risk factors worldwide:year 2002 [1] Risk factor Deaths (× 10 3 ) % of total death DALYs (× 10 3 ) % of total DALY Hypertension 7141 12.8 64270 04.5 Tobacco 4907 08.8 59081 04.1 High cholesterol 4415 07.9 40437 02.8 Low fruit & veg 2726 04.9 26662 01.9 Overweight 2591 04.6 33415 02.3 Alcohol 1804 03.2 58323 04.0 Phys. inactivity 1922 03.4 19092 01.3 Globally, many of the risk factors for heart disease, diabetes, cancer and pulmonary diseases are due to lifestyle and can be prevented. Physical inactivity, western diet and smoking are prominent causes[ 23 ]. Tobacco is the enemy number one. It is the most important established cause of cancer but also responsible in CVDs and chronic respiratory diseases. Tobacco and diet are the principal risk factors, responsible for more than 40% of cancer deaths and incidence. Obesity and dietary habits are the principal risk factors for diabetes of type 2. Tobacco[ 1 , 2 , 24 ] In the 20 the century, approximately 100 million people died worldwide from tobacco-associated diseases such as cancers, chronic lung disease, diabetes and CVDs. While tobacco consumption is falling in most developed countries, it is increasing in developing countries by about 3.4% per annum. Today, 80% of the 1.2 billion smokers in the world live in poorer countries where smoking prevalence among men is nearly 50% (48%) and 50% of the 5 million deaths attributed to smoking in 2000 occurred in developing countries, also responsible for the increase in deaths by more than one million during the last decade. Tobacco remains the most important avoidable risk for the four classes of NCDs. It increases the risk of dying from coronary heart disease and cerebrovascular disease 2–3 fold. It increases the risk of many types of cancer, for lung cancer the risk is increased by 20–30fold. According to studies conducted in Europe, Japan and North America, 83–90% of lung cancers in men and 57–80 in women, are imputable to tobacco. Between 80 and 90 % of oesophagus, larynx and oral cavity are caused by tobacco and alcohol [ 17 ]. In developing countries, an estimated one-third of all cancer deaths was attributable to smoking in 1995. Finally, tobacco exacerbates the conditions of people living with COPD and asthma. Lifestyle[ 2 , 25 - 27 ] Up to 80% of cases of coronary heart disease, and up to 90% of cases of types 2 diabetes, could potentially be avoided through changing lifestyle factors. One-third of cancers could be avoided by eating healthily, maintaining normal weight, and exercising throughout life. It was estimated that in high-risk populations, an optimum fish consumption of 40–60 grams per day would lead to approximately a 50% reduction in death from coronary heart disease. A recent study based on data from 36 countries, reported that fish consumption is associated with a reduced risk of death from all causes as well as CVD mortality. Unfortunately, the fish consumption is very low even in some countries known for their large fish stock like the north African region. Daily intake of fresh fruit and vegetables in adequate quantity (400–500 grams per day), is recommended to reduce the risk of coronary heart disease, stroke and high blood pressure. But, once more, this is thwarted by the western lifestyle invading developing countries. Overweight/Obesity[ 2 , 28 ] Overweight and Obesity lead to adverse metabolic changes such as insulin resistance, increasing blood pressure and cholesterol. Consequently, they promote CVDs, diabetes and many types of cancer. Worldwide, overweight affects 1.2 billion of which 300 million are clinically obese. In some developed countries like USA, the prevalence reaches 60% but developing countries like Kuwait have also a very high prevalence. More and more children are suffering from overweight and obesity. However, the most contrasting phenomenon is to find Overweight/Obesity and malnutrition side by side in low- and middle-income countries and hence contributing to the growing burden afflicting these countries. According to the International Obesity Task Force (IOTF) and the WHO World Health report 2002, about 60% of diabetes globally can be attributable to overweight and obesity. In other respects, it is estimated that 60% of world's population do not do enough physical activity. Alcohol[ 2 ] Alcohol consumption has also increased in the last decades, with the major part of this increase imputable to developing countries. In 2000, Alcohol was responsible for nearly 2 million deaths in the world, representing 4% of the global disease burden. Moreover, alcohol was estimated to cause 20 to 30 % of oesophagus cancer, liver disease, epilepsy, motor vehicle accidents and other hazards. Conclusions Non communicable diseases are more and more prevalent in developing countries. These diseases are highly correlated to risk factors like smoking, alcohol, obesity, diet and inactivity. The World Health Organization and many other organisms and associations are urging health decision makers to develop efficient preventive strategies to halt the growing trend of NCDs through the control of risk factors. However, although most of developed countries have reacted by pragmatic measures, the trend remain globally passive mainly because developing countries have been, so far, satisfied with adopting national conventions and adhering to international recommendations instead of pragmatic decisions such as prohibiting smoking in public areas, controlling alcohol abusers, encouraging physical activity, promoting healthy diet and improving primary health care for screening and early detection of chronic diseases. In these countries, 2.8 billion people live with less than 2 dollars, 1.2 billion live with less than one dollar and 1.3 billion live on fragile and often remote rural ecosystems[ 29 ]. So, the behaviour can be partly explained by lack of means and poor budget affected to health care but, in general, bad management and absence of goodwill assume a large part of responsibility. For instance, many developing countries have signed the Framework Convention on Tobacco Control(FCTC) and voted laws that prohibits smoking in public areas but the laws are not executed. Also, in the absence of early detection, many people are diagnosed at advanced stages of cancer, cardiovascular diseases and diabetes complications. Also, in these countries, until recently, it was widely believed that economic development was a necessary prerequisite for improving a population health status and the health was often classified as a non productive sector. Now, politicians and health policy makers are timidly recognizing that investing in people's health is a necessary condition for economic development but energetic decisions are needed for the adoption of urgent and consequent strategies. The need for such strategies is enhanced by the fact that risk factors like cholesterol, tobacco, blood pressure, and obesity are no more a specificity of industrialized countries, they are becoming more prevalent in developing nations, where they double the burden of infectious diseases that have always afflicted poorer countries. Moreover, multinational companies have been competing fiercely to expand their sales in developing countries and western lifestyle is invading middle-income countries [ 2 , 30 ]. Adhesion to the Framework Convention on Tobacco Control(FCTC) and other international strategies must be taken seriously by developing countries facing the pandemics of NCDS. Competing interests The author(s) declare that they have no competing interests. Authors' contributions AB contributed by the collection of data concerning CVDs and diabetes and to English writing SB contributed by the collection of data concerning cancer and chronic pulmonary diseases. The two authors contributed equally to the final version of the paper. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC546417.xml |
554775 | p27Kip1 is expressed in proliferating cells in its form phosphorylated on threonine 187 | Background G1/S cell cycle progression requires p27 Kip1 (p27) proteolysis, which is triggered by its phosphorylation on threonine (Thr) 187. Since its levels are abundant in quiescent and scarce in cycling cells, p27 is an approved marker for quiescent cells, extensively used in histopathology and cancer research. Methods However here we showed that by using a specific phosphorylation site (pThr187) antibody, p27 is detectable also in proliferative compartments of normal, dysplastic and neoplastic tissues. Results In fact, whereas un-phosphorylated p27 and MIB-1 showed a significant inverse correlation (Spearman R = -0.55; p < 0,001), pThr187-p27 was positively and significantly correlated with MIB-1 expression (Spearman R = 0.88; p < 0,001). Thus proliferating cells only stain for pThr187-p27, whereas they are un-reactive with the regular p27 antibodies. However increasing the sensitivity of the immunocytochemistry (ICH) by the use of an ultra sensitive detection system based on tiramide signal amplification, simultaneous expression and colocalisation of both forms of p27 was shown in proliferating compartments nuclei by double immunofluorescence and laser scanning confocal microscopy studies. Conclusion Overall, our data suggest that p27 expression also occurs in proliferating cells compartments and the combined use of both regular and phospho- p27 antibodies is suggested. | Background Immunocytochemistry (ICH) is an important method for identification of proteins in cells and in tissues. Since the biological activity of many proteins is dependent on their phosphorylation status, a challenge for immunocytochemistry is to characterize the protein form and not just the total amount [ 1 ]. p27 Kip1 (p27) is a key inhibitor of cell division that protects tissues from excessive cell proliferation [ 2 ]. As a consequence of an altered balance between synthesis and degradation, the amounts of this protein are abnormally low in advanced and poorly differentiated neoplasms [ 3 ]. Since p27 expression is readily assessed by ICH, this protein is a prognostic marker quite popular in histopathology [ 4 ]. However, little is known on its in vivo regulation. p27 cellular levels, copious in quiescent cells undergoing terminal differentiation, are scanty in cycling cells [ 2 ]; in these cells p27 is phosphorylated on Thr 187 by cyclin-dependent kinase (cdk) 2 in late G1 [ 5 ]. This event leads to enhanced ubiquitination and p27 proteolysis by the proteasome, which marks the restriction point and promotes cell proliferation [ 5 ]. Therefore un-phosphorylated ("plain") p27 is representative of the total protein amounts only in quiescent cells, whereas in cycling cells a fraction of p27 is transiently present in the pThr187 form before degradation [ 5 ]. Recently, Montagnoli et al raised an antibody (Ab), specific for pThr187-p27 that was reactive in immunoprecipitates from proliferating cells and negative in quiescent cells [ 6 ]. Even more recently this Ab was shown to be reactive also on paraffin sections [ 7 , 8 ]. The present study was undertaken to assess pThr187-p27 Ab staining pattern in a wide range of normal, dysplastic and neoplastic tissues; its expression was correlated to those of MIB-1, a standard marker of proliferation, and of "plain" p27. The relationship between the two forms of p27 was also studied at a sub cellular level by double immunofluorescence (IF) and laser scanning confocal microscopy (LSCM). Here we show that p27 expression is not restricted to quiescent cells but that it also occurs in proliferating cellular compartments, where it is detectable by regular ICH only in its pThr187 form. Therefore, to fully assess p27 tissue expression both antibodies should be used. Methods Antibodies pThr187-p27 was detected by the 71–7100 polyclonal antibody (PcAb) (Zymed Laboratories, San Francisco, CA, USA) and by the sc-16324 PcAb (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA). These Ab's were raised against a short peptide corresponding to the portion of human p27 containing phosphorylated Thr-187, in order to detect only phospho-p27 and to be unreactive with "plain"-p27. The 71–7100 Ab was previously employed in immuno-precipitation experiments by Western blot, [ 6 , 9 , 10 ] and to immunostain neoplastic and degenerative human tissue [ 7 , 8 ]. "Plain" p27 protein levels were detected with the K10125 monoclonal Ab (McAb) from (Transduction Laboratories, Lexington, Ky, USA), and with the rabbit PcAb (C-19) (Santa Cruz Biotechnology). These antibodies were previously shown to share the same staining pattern [ 11 ]. MIB-1 McAb from Novocastra (Newcastle upon Tyne, UK) was used to stain proliferating cells and as control of antigenic preservation and of successful antigenic retrieval [ 12 ]. Tissues A wide range of normal, dysplastic and neoplastic tissues was obtained from surgical specimens. At least five samples from each type of normal tissue were processed (Table 1 ). Dysplastic diseases included five cases of colonic adenoma, five cases of low- and five cases of high grade- squamous intraepithelial lesions (SIL) of the uterine cervix. Several carcinomas were also evaluated, including different tumour types in which p27 down regulation had previously been described, such as invasive squamous cell carcinoma (ISCC) of the oral cavity (n = 7) [ 13 ], of the lung (n = 10) [ 14 ] and of the uterine cervix (n = 9) [ 15 ]; ductal cell carcinoma of the breast (n = 12) [ 16 ]; invasive adenocarcinoma of the colon (n = 6) [ 17 ] and of the prostate (n = 5) [ 18 ]; papillary (n = 6) and anaplastic (n = 5) thyroid carcinoma, [ 19 ], glioblastomas (n = 5) [ 20 ], and choriocarcinoma (n = 2) [ 21 ]. Table 1 pThr187-p27, "plain"-p27 and MIB-1 expression in normal tissues. Tissue type Phospho-p27 "plain"-p27 MIB-1 Normal squamous epithelium skin, tonsil, larynx, cervix. Parabasal layer +++ - ++++ intermediate - ++++ + Granular layer - +++ - Germinal centers Tonsil, lymphnode, apendix Mantle cells - ++++ - Centroblasts +++ - ++++ Centrocytes - +++ - Bowel Crypt cells Deep +++ - ++++ superficial - ++++ - Placenta trophoblast +++ - ++++ syncytiotrofoblast - ++++ - Kidney +/- +++ +/- Lung +/- +++ +/- Thyroid +/- +++ +/- Prostate +/- +++ +/- Key: ++++ >80% of positive cells. +++ 50–80% of positive cells ++ 10–50% of positive cells + <10% of positive cells +/- Only occasional staining - Negative staining Immunostaining techniques Xylene dewaxed and alcohol rehydrated paraffin sections were placed in Coplin jars filled with a 0.01 M tri-sodium citrate solution, and heated for 3 minutes in a conventional pressure cooker [ 15 ]. After heating, slides were thoroughly rinsed in cool running water for 5 minutes. They were then washed in Tris-Buffered Saline (TBS) ph 7.4 before incubating overnight with the specific Ab, diluted as follows: anti-pThr187-p27 (Zymeed) 1:1000; anti-pThr187-p27 (Santa Cruz) 1:200; anti-p27 (Transduction) 1:4000; anti-p27 (Santa Cruz) 1:50; anti-Ki-67 1:50. After incubation with the primary Ab, tissue sections were covered with biotinylated anti-mouse or anti-rabbit immunoglobulins, followed by peroxidase labelled streptavidine (LSAB-DAKO); the signal was developed by using diaminobencidine (DAB) chromogen as substrate. Immunostaining controls pThr187-p27 antibody specificity was controlled by Western blot analysis of MDA MB 468 (breast cancer) and NPA (thyroid papillary cancer) cell lines. These were lysed (50 mmol/L Tris-HCl, pH 7.4, 150 mmol/L NaCl, 0.1% Triton X-100, 5 mmol/L ethylenediaminetetraacetic acid, 1 mmol/L Na 3 VO 4 , and 1 mmol/L phenyl methyl sulfonyl fluoride and protease inhibitors) and 20 to 30 μg of proteins were electrophoresed in sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel and transferred onto nitrocellulose membranes. The membranes were first blocked and then incubated with the primary antibody as it follows: anti-pThr187-p27 (Zymeed), 1:2000 and anti "plain" p27 (Transduction), 1:3000 for 1 hour at room temperature. To confirm equal loading, membranes were immunoblotted with monoclonal anti b-tubulin antibody (1:1000, Santa Cruz). After three washes, filters were incubated with horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit antibodies (1:2000; Amersham, Arlington Heights, IL, USA) for 1 hour at room temperature. Detection of immunocomplexes was performed with an enhanced chemiluminescence system (ECL, Amersham). pThr187-p27 immunostaining specificity was assessed by several control experiments performed in parallel, in which the primary Ab was either replaced by a similarly diluted normal rabbit serum, or adsorbed with increasing concentrations of its phospho and dephosphopeptides (up to 0.14 mg mL -1 ). Quantitative study and statistical analysis In the neoplastic cases examined, labelling indices for pThr187-p27, "plain" p27 and Ki67/MIB-1 were determined. Adjacent sections were used and counting was performed in similar areas; quantitative analysis performed with a computerised analyser system (Ibas 2000, Kontron, Zeiss) was used to score the nuclei of individual cells for expression of these proteins. As already described, nuclear boundary optical density and Ab threshold were adjusted for each case examined [ 15 ]. A minimal threshold was established by counting at least 1000 cells per sample and the results were expressed as a percentage of the total cell population. Statistical analysis was performed by means of SPSS Inc. package. The range of expression of pThr187-p27, "plain" p27 and Ki67/MIB-1 for each neoplastic type is reported in table 2 . The relationships among these were analyzed by calculating the nonparametric Spearman R coefficient. Table 2 Range of positive cells for pThr187-p27, "plain"-p27 and MIB-1 expression in different type of carcinomas Histotype pThr187-p27 "plain" p27 MIB-1 ISCC cervix n = 9 45–60 5–30 50–78 ISCC oral n = 7 40–70 5–15 35–80 ISCC lung n = 10 15–60 5–70 30–70 Thyroid, papillary carcinoma n = 6 1–3 5–60 2–4 Thyroid, anaplastic carcinoma n = 5 23–45 5–11 41–55 Breast carcinoma n = 12 3–55 5–70 5–80 Colonic carcinoma n = 6 30–45 5–80 45–80 Prostate carcinoma n = 5 2.5–10 30–75 4–15 Glioblastoma n = 5 30–50 10–27 50–63 Chorioncarcinoma n = 2 45–55 3–7 60–80 Double Immunofluorscence (IF) staining and Laser Scanning Confocal Microscopy (LSCM) Tissue sections from both normal and neoplastic specimens were also stained using double IF labelling for pThr187-p27 and "plain" p27, according to previous studies with minor modifications [ 22 ]. The primary anti pThr187-p27 (Zymeed) was incubated (1/1000) for 1 hour, followed by incubation with swine anti rabbit HRP (1/200) (Dako) and Cy3 tyramide amplification (Perkin Elmer Life Sciences). To avoid cross-reactivity due to residual HRP, sections were incubated with 0.3% H 2 O 2 for 1 hour. The primary mouse McAb anti p27 (1/4000) was detected by overnight incubation at 4°C, followed by goat anti mouse HRP (1/200, DAKO). After washing in TBS, sections were incubated with fluorescein isotiocyanate (FITC) tyramide amplification (Perkin Elmer Life Sciences) and mounted with Vectashield-DAPI mounting medium (Vector). Control sections in which the second primary Ab was omitted were included to ascertain destruction of peroxidase activity. Single IF stained sections as well as colorimetric immunostaining were used as controls of signal specificity. Slides were examined with a Leica TCS SP2 UltraSpectral LaserScan Confocal microscope. FITC was excited at 488 and detected with a bandpass 500 to 550 nm. Cy3 was excited at 514 nm and detected with a bandpass 580 to 655 nm. Series of images were processed with the Leica confocal soft package. Confocal images were captured and imported into Adobe Photoshop 7 (Adobe Systems, Mountain View, CA) and processed with an eMac personal computer. Results pThr187-p27 expression in normal tissues Similar results were obtained with both anti pThr187-p27 Ab's, with only proliferating tissue compartments being stained. This pattern overlapped to that of Ki67/MIB-1, whereas it was different to that shown by both Mc and Pc Ab's directed to "plain" p27 (Table 1 ). Stratified squamous epithelium showed the same staining pattern in skin, tonsil, larynx and uterine cervix. In analogy to the Ki67/MIB-1 staining, pThr187-p27 was expressed by parabasal cells (Figure 1A ); on the contrary, nuclei of the more superficial layers showed intense "plain" p27 nuclear labelling, as already described [ 13 , 15 ]. In small intestine and colon (Figure 1B ), in analogy to MIB-1, pThr187-p27 showed positivity in nuclei of deep crypt cells. On the contrary, the upper half of the crypts was stained by "plain" p27 [ 17 ]. In germinal centres of normal lymph nodes, tonsil and appendix the expression of the phosphorylated form of p27 was clearly detectable (Figures 1A and 1C ). Within the germinal centres, pThr187-p27 showed a signal distribution similar to that of the Ki-67 protein; the outer rim of centroblast and mitotic cells were strongly positive, whereas centrocytes were less frequently labelled. This pattern of expression was opposite to the one identified for "plain" p27: negative staining for centroblasts and strong staining in most centrocytes, mantle cells and interfollicular small lymphocytes [ 11 ]. Trophoblastic villi of placenta showed pThr187-p27 intense staining only in the cytotrophoblastic layer and not in the syncytiotrophoblastic layer (Figure 1D ). On the contrary "plain" p27 showed intense staining only in the syncytiotrophoblast overlying the villus, as already described [ 21 ]. Figure 1 p27 expression was detected by the pThr187 Ab p27 only in the proliferative compartments of normal tissues. In tonsil (A) both parabasal squamous cells and lymphoid germinal cells were stained. In intestinal epitehlium only the nuclei of deep crypt cells were stained (B). Within the germinal centres, the outer rim of centroblasts and mitotic cells were strongly positive for pThr187-p27, whereas centrocytes were less labelled (C). Trophoblastic villi of placenta show pThr187-p27 intense staining only in the cytotrophoblastic layer and not in the syncytiotrophoblastic layer (D). In other normal tissues, comprising breast, lung, kidney, pancreas, prostate, thyroid and parathyroid, "plain" p27 was expressed by the vast majority of cells, whereas pospho-p27 single labelled cells could be made out by meticulous scrutiny; in these tissues also MIB-1 staining was sporadic. pThr187-p27 expression in dysplastic and neoplastic tissues As examples of dysplastic lesions, the expression of the pThr187-p27 was assessed in low-and in high grade squamous intraepithelial lesions (SIL) of the uterine cervix and in colonic adenoma. According to the relationship between phospho-p27 expression and proliferation, in low grade SIL the dysplastic basal and parabasal cells were in most instances positive for the expression of both pThr187-p27 (Figure 2A ), and Ki-67; these were instead negative for "plain" p27 whose staining was confined to the intermediate and superficial squamous cells, as described [ 15 ]. In high grade SIL, the expression of phospho-p27 correlated well with the extent of the dysplastic cell population (Figure 2B ), whereas "plain" p27 reactivity was restricted to very superficial layers displaying squamous differentiation. In colonic adenoma pThr187-p27 positive cells were randomly located throughout the crypts and in contrast to that seen in normal mucosa, also superficial cells were stained. Figure 2 The relationship between phospho-p27 expression and proliferation was evident both in low- (A) and high- (B) grade SIL, with its expression correlating well with the extent of the dysplastic cell population. Intense staining for phospho-p27 was also observed in the neoplastic cells of colonic (C) and lung (D) adenocarcinoma, in cervical squamous invasive carcinoma (E) and in choriocarcinoma (F) A wide range of different tumours was examined, in order to assess whether staining for phospho-p27 may yield diagnostic information in addition to those provided by "plain p27" (Table 2 ). As a general rule, the expression of the two forms of p27 was alternative and dependent on the degree of tumour differentiation, recapitulating the pattern featured by normal and dysplastic tissues; phospho-p27 strongly labelled aggressive tumours, whereas "plain p27" staining was only retained by well differentiated tumours. Thus, poorly differentiated and highly proliferating adenocarcinoma of the breast, colon and prostate were strongly labelled by the pThr187-p27 Ab, (Figure 2C–D ) whereas "plain p27" staining was only retained by well differentiated neoplasms. The above staining differences between the two forms of p27 were also evident in squamous cell carcinomas. Poorly differentiated neoplasms, composed of nests of small undifferentiated cells with high MIB-1 index and minimal keratinisation, showed intense staining for pThr187-p27 (Figure 2E ) whereas "plain p27" labelled those neoplasms showing abundant keratin, squamous pearl formation and low mitotic activity. Similarly, anaplastic thyroid carcinoma showed pThr187-p27 staining stronger than papillary carcinomas, while in this latter "plain p27" was prevalent (Table 2 ). An intense pThr187-p27 expression was also found in choriocarcinoma (Figure 2F ), in which the level of "plain-p27" staining was instead low. The Spearman's correlation coefficient for continuous variables revealed a positive and a significant correlation between pThr187-p27 staining and MIB-1 expression (Spearman R = 0.88; p < 0,001). On the contrary both pThr187-p27 and "plain" p27 (Spearman R = -0.61; p < 0,001) and MIB1 and "plain" p27 (Spearman R = -0.55; p < 0,001) showed significant inverse correlations. Immunostaining controls Western blot analysis of MDA MB 468 (breast cancer) and NPA (thyroid papillary cancer) cell lines revealed a single anti-pThr187-p27 band, whose molecular weight (27 kDa) corresponded to that showed by "plain"-p27 (Fig. 3A ). pThr187-p27 immunostaining specificity was confirmed by: (i) the disappearance of the signal when the primary Ab was replaced by a similarly diluted normal rabbit serum; (ii) the progressive signal quenching due to the competitive inhibition between the increasing concentrations of the phosphopeptide and the phosho-p27 Ab. (Fig. 3 B–C ). Figure 3 Western blot analysis of MDA MB 468 (breast cancer) and NPA (thyroid papillary cancer) cell lines revealed a single anti-pThr187-p27 band, corresponding to that shown by the regular p27 antibodies. (A). pThr187 staining of parabasal squamous cells (B) abolished by absorption of the antibody with the immunizing peptide. Double IF & LSCM studies Tissue sections from both normal and neoplastic specimens were also stained using double IF labelling for pThr187-p27 and "plain" p27 and analyzed by LSCM. The use of an ultra sensitive detection system based on tiramide signal amplification for IF staining, revealed a more precise pattern of protein expression and distribution, slightly different to results above shown of regular colorimetric IHC. In proliferative compartments of normal tissues the cells frequently showed simultaneous expression of the two forms of p27; in these cells, pThr187-p27 was more localized as an inner rim along the nuclei membrane, whereas "plain" p27 was more centrally located in the same nuclei with an intermediate area of colocalization (Fig. 4A ). Similarly neoplastic nuclei also showed simultaneous expression of the two p27 forms, frequently displaying high amounts of pThr187-p27 throughout the whole nucleoplasm, colocalizing with "plain" p27. (figure 4B ). Figure 4 Coexpression of the forms of p27 in lymphoid germinal center (A) and in glioblastomas (B). Simulataneous detection of p27 by both regular (green) and pThr187 (red) antibodies. Signal generated by the two antibodies Yellow areas (arrows) showed the colocalitation of the signal generated by both antibodies. Discussion p27 is an approved marker for quiescent cells, extensively used in histopathology and cancer research. However here we showed that an unusual p27 expression pattern may be obtained when only its portion phosphorylated on threonine 187 is stained. By using a specific phosphorylation site antibody, p27 expression may be detected in the proliferative compartments of squamous and intestinal epithelia, of lymph node germinal centres, of trophoblastic villi, of cervical and colonic pre-invasive lesions and of several carcinomas types. These results are similar to those reported in vitro by Montagnoli et al , who also detected p27 expression in proliferating cells by the pThr187-p27 Ab [ 6 ]. The same antibody, here, employed to stain a wide range of normal, dysplastic and neoplastic tissues yielded a pattern similar to that of Ki-67 (Spearman R = 0.88; p < 0,001), underlining the close association between cell proliferation and the expression of that portion of p27 targeted for degradation. Since pThr187-p27 Ab should react with cells from late G1 through G2-M, whereas MIB-1 stains throughout the entire cell cycle, it is also conceivable that the pThr187-p27 Ab stained a lower percentage of cells than MIB-1/Ki-67 in both normal (Table 1 ) and neoplastic tissue. (Table 2 ) In this study we showed that p27 expression does not feature only one pattern, but that two different staining may be obtained by the use of either regular or pThr187 antibodies; in fact, when the overall data relative to the expression of p27 by both antibodies in tumours were analysed, a significant inverse correlation between the two forms was found (Spearman R = -0.61; p < 0,001). This data reflects the different relationship between each form of p27 and tissue differentiation:aggressive tumours were strongly labelled by phospho-p27 Ab, whereas plain-p27 was prevalent in well differentiated tumours. Thus the ratio between the two forms of p27, more than the total protein amounts, could yield information on tumour differentiation and behaviour. Since cdk2 activity is low in differentiating cells and p27 is only present in its un-phosphorylated form, it is reasonable that squamous superficial layers, upper half of the intestinal crypts, lymphoid mantle cells and syncytiotrophoblastic cells only react with the regular antibodies. More intriguing is the search for the reasons explaining why in proliferating cells p27 is detected by the pThr187-Ab and missed by the regular antibodies. The possibility of an aberrant cross-reactivity by the pThr187-Ab was excluded both by Western blot analysis and by phosphopeptide adsorbtion. Montagnoli et al , who also confirmed antibody specificity by biochemical assays, showed that p27 expression may be better detected by using both antibodies than either alone and that there is a portion of p27 that is only recognizable by the anti-phospho Ab [ 6 ]. Thus, in proliferating cells anti-pThr187 Ab may be more efficient than the regular p27 antibodies, being protein levels low and mainly present in their phosphorylated form. To test this hypothesis, the sensitivity of the p27 detection by the regular antibodies was increased by the use of an ultra sensitive detection system based on tiramide signal amplification and observed at high resolution by laser scanning confocal microscopy. The results obtained by this approach were slightly different to those obtained by colorimetric IHC, as p27 was simultaneously detected by both antibodies in the same nuclei of the proliferating compartments; in particular the signal generated by both antibodies showed specific colocalitation by double immunofluorescence. Conclusion In this study we have reported the following observations: (1) pThr187-p27 Ab selectively stains proliferating cells; (2) this staining does not identify the "plain" p27 protein, expressed by quiescent cells. Therefore, the pThr187-p27 antibody is a useful tool to study p27 in vivo regulation. Indeed, the pattern of expression observed in this study, strongly suggest that the anti-pThr187-p27 Ab identifies proliferating cells. Considering the importance of the different biological functions of p27: regulator of cell growth, cell differentiation, contact inhibition, apoptosis, protection against immunological aggression, and protection against environmental stress, among others, the combined use of both "plain" and pThr187 p27 antibodies will be a useful tool to better characterize the biology of these conditions. Competing interests The author(s) declare that they have no competing interests. Authors' contributions GT conceived the study and wrote the manuscript; JCM carried out the LSCM study; AC scored the immunostaining; AI coordinated the study; MR performed the peptide experiments; PZ evaluated the tumour pathological features; IM participated in its design and coordination and helped to draft the manuscript; MLM carried out the double immunostainings; DC carried out the Western blot analysis; EAP performed the statistical analysis. LP participated in the design of the study and evaluated the results. All authors read and approved the final manuscript Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC554775.xml |
509314 | Inhibition of the DNA Damage Pathway by a Telomere-Binding Protein | null | To maintain the integrity of their genetic content, cells closely monitor the state of their chromosomal DNA. Any break or pairing anomaly in the double helix is perceived as damage that must be repaired. Mechanisms that sense DNA damage enlist the help of cell cycle checkpoint proteins, which stall cell division until the damage has been repaired. The exposed ends of linear chromosomes of eukaryotic cells (cells with nuclei) resemble the double-strand breaks of damaged DNA, which should make them vulnerable to unnecessary manipulations by repair enzymes. But chromosomes have special DNA sequences at their tips, called telomeres, that are coated with telomere-binding proteins that appear to protect ends from the unwanted attentions of repair enzymes. For instance, removing a telomere-specific DNA-binding protein, called TRF2, from telomeres leads to a rapid attack on chromosome tips, associated with the activation of the DNA damage pathway. The biochemical basis for TRF2's shielding effects remains obscure. In this issue of PLoS Biology , Jan Karlseder et al. propose that TRF2 is a direct inhibitor of an early mediator of the DNA damage signal. Their observations offer new insights into how telomeres resist the inappropriate interventions of the DNA repair machinery. Ataxia telangiectasia mutated (ATM) kinase is an enzyme that induces the activation of DNA repair enzymes as well as regulators of the cell cycle and apoptosis. Its enzymatic function is triggered by DNA damage sensors. Several observations suggest an antagonism between TRF2 and ATM: ATM is activated in aging cells with shortened telomeres and participates in ushering the cell into replicative senescence. (Cells divide only so many times during their lives before forgoing the process altogether.) In contrast, TRF2 overexpression protects shortened telomeres from decay and delays cell entry into senescence. Here the authors examine the effect of TRF2 overexpression in cells subjected to radiation-induced DNA damage. This treatment is expected to lead to cell cycle arrest, an outcome mediated for the most part by ATM activation. Irradiated cells that overexpress TRF2, however, continue to enter cell division. Known targets of ATM's enzymatic activity are activated to a lesser degree in these cells, which harbor only about half the amount of active ATM detected in controls. This suggests that TRF2 may interfere with the DNA damage pathway early on, when ATM is activated. Do these findings reflect a direct interaction between ATM and TRF2? Karlseder et al. suggest they might. Though only a small fraction of TRF2 associates with ATM in normal cells, ATM and TRF2 proteins can form a complex in a test tube. The region of the ATM protein that interacts with TRF2 is the very region required for ATM's activation by DNA damage sensors. The authors propose that TRF2 binds to inactive ATM, and in so doing, prevents ATM's transition to an active state. The authors integrate their results in an elegant proposal that relates a cell's health to the length of its telomeres. They have previously shown that TRF2 binds another protein called Mre11, a DNA damage sensor known to activate ATM. Thus, by inhibiting ATM, TRF2 may nip in the bud any misguided attempts by Mre11 to “repair” DNA breaks in telomeres. But in aging cells, whose shortened telomeres no longer retain large amounts of TRF2, ATM activation, no longer muffled, eventually allows the onset of senescence. While TRF2 is abundant in normal cells, the authors note that its strict association with telomeres should locally limit its effect on ATM, leaving the DNA damage pathway intact at other chromosomal locations. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC509314.xml |
544566 | A functional hierarchical organization of the protein sequence space | Background It is a major challenge of computational biology to provide a comprehensive functional classification of all known proteins. Most existing methods seek recurrent patterns in known proteins based on manually-validated alignments of known protein families. Such methods can achieve high sensitivity, but are limited by the necessary manual labor. This makes our current view of the protein world incomplete and biased. This paper concerns ProtoNet, a automatic unsupervised global clustering system that generates a hierarchical tree of over 1,000,000 proteins, based solely on sequence similarity. Results In this paper we show that ProtoNet correctly captures functional and structural aspects of the protein world. Furthermore, a novel feature is an automatic procedure that reduces the tree to 12% its original size. This procedure utilizes only parameters intrinsic to the clustering process. Despite the substantial reduction in size, the system's predictive power concerning biological functions is hardly affected. We then carry out an automatic comparison with existing functional protein annotations. Consequently, 78% of the clusters in the compressed tree (5,300 clusters) get assigned a biological function with a high confidence. The clustering and compression processes are unsupervised, and robust. Conclusions We present an automatically generated unbiased method that provides a hierarchical classification of all currently known proteins. | Background The explosive growth in the number of sequenced proteins has created a glut of proteins that are sequenced but whose structure and function are as yet unknown. A common way to tackle this problem is to use database searches to find proteins similar to a newly discovered protein, thus inferring protein function. This method is generalized by protein clustering or classification where databases of proteins are organized into groups or families in a manner that attempts to capture protein similarity. Such classification into families is a critical component in structural and functional genomics [ 1 - 4 ]. The number of protein families comprising the entire protein-space has been conjectured to range between 6,000-30,000, excluding rare and peculiar single proteins [ 5 - 8 ]. Various expert-based databases provide a good description of certain selected families but are limited in scope to thoroughly studied proteins (i.e. [ 9 , 10 ]). Other methods for classification strongly rely on 3D-structural information as in the case of SCOP [ 11 ], CATH [ 12 ], FSSP [ 13 ] and others. Classifying the entire protein space into families serves not only as a method for large-scale protein annotations but also to support functional and structural genomic initiatives [ 14 ]. Some prominent examples for protein classification efforts are ProtoMap [ 15 ], Picasso [ 16 ], SYSTERS [ 17 ], iProClass [ 18 ] and ProtoNet [ 19 ]. These systems are based on a variety of algorithmic paradigms, each yielding a distinct hierarchical classification of proteins into families. Amongst the clustering methods listed above only ProtoNet attempts to generate a global hierarchical arrangement of the entire protein space via agglomerative hierarchical clustering. The sequence similarity between every pair of protein sequences is taken as the BLAST [ 20 ] E-value between a given pair of proteins. Next, the proteins are clustered using a given merging strategy. The strategy used is Unweighted Pair Group with Arithmetic Mean (UPGMA), whereby in each iteration, the two most similar clusters (in terms of their average pairwise distance for every protein pair spanning the two clusters) are merged. ProtoNet (version 4.0) [ 21 ] provides a classification hierarchy of over 1,000,000 proteins including the SwissProt and TrEMBL protein databases [ 22 ]. Most proteins included in the SwissProt database are manually validated and furthermore, the degree of biological knowledge associated with them is much higher in comparison to the proteins archived in TrEMBL. Thus, this work concerns only the 114,033 proteins in the SwissProt database (version 40.28). An extended version that includes over one million protein sequences is available in the form of an interactive website at . For the SwissProt-based tree, there are 227,436 clusters (including the proteins as singletons). The classification provided by ProtoNet provides the full range of cluster granularity: from single proteins to huge protein clusters that carry no biological relevance (the root clusters). We test the biological validity of ProtoNet, by its examination from different perspectives, using external-defined protein keyword annotations. Four different annotation sources are used (InterPro [ 23 ], GO [ 24 ], SCOP [ 11 ] and ENZYME [ 25 ]) in order to be able to validate different biological aspects. First, we demonstrate that it is possible to match the majority of such external-defined protein families to specific clusters within the ProtoNet clustering. Second, we show that the hierarchy of the ProtoNet tree represents a valid functional hierarchy and correlates well with the GO hierarchical structure. As mentioned, ProtoNet contains 227,436 clusters, which is obviously much more than the upper estimate of 30,000 protein families [ 8 , 26 ]. Therefore, we seek to cleverly discard those clusters that have less biological relevance. Compression of the protein space offers many advantages. It can yield a smaller set of biologically meaningful clusters, which will allow for a more manageable handling of the entire protein space. Furthermore, if this compression's correspondence to external, independent annotation sources can be validated, then this compression can be used to replace the original hierarchical structure, without sacrificing much information originally present in the whole system. In this paper we describe methods for the unsupervised compression of the ProtoNet tree, by using intrinsic tree-based parameters of the clusters that correlate well with biological validity. By preserving the unsupervised nature of the ProtoNet data, we prevent biasing towards previously discovered findings and better allow for future generalizations, in addition to maintaining the automation of the process. Finally, automatic functional annotation to proteins is of great importance. In ProtoNet, an automatic method for assignment of biological annotation to the protein clusters is used, yielding high-confidence functional assignments for a large majority of the proteins' clusters. Results and discussion Correspondence of clusters to external biological sources In order to measure the correspondence between a given cluster and a specific annotation, and allow for supervised validation of the ProtoNet clusters, we define the notion of a correspondence score (CS). The CS for a certain cluster and a given keyword measures the correlation between the cluster and the keyword, using the well-known intersect-union ratio. Let C be a cluster in the ProtoNet tree, and K be a keyword (from a specific source) that annotates (some of) the proteins in the system; Let c be the set of annotated proteins in cluster C; Let k be the set of proteins in the system annotated by keyword K; We define: The cluster receiving the maximal score for keyword K is considered the cluster that best represents K within the ProtoNet tree (K's best cluster ). The score for a given cluster on keyword K ranges from 0 (no correspondence) to 1 (the cluster contains exactly all of the proteins with keyword K, i.e. maximally corresponds to the keyword). In order to assess the clustering's biological validity, the mean best CS on all annotations was examined for each of the following sources: InterPro, SCOP (Family, SuperFamily, and Fold levels), GO (Molecular Function) and ENZYME (subclass, sub-subclass and entry). The results (Table 1 ) show a high level of correspondence between the ProtoNet clusters to the various keyword sets of each of the external sources. It can be argued that a good fit between a set of keywords and the ProtoNet set of protein clusters could happen by chance. In order to assess the statistical significance of these results, the mappings of the keywords to the proteins were randomized, creating a new group of random keyword sets that have the same size distribution but do not represent any biological features. For each random keyword set, the mean best CS was calculated. This randomized test showed a normal distribution, allowing the calculation of an approximate p-value for the mean best CS obtained by ProtoNet for the external sources. The results showed an extremely high level of statistical significance for all sources (all had P-values smaller than 10 -100 ). Note that even for the SCOP fold level, which is associated with proteins that may be extremely remote in sequence, ProtoNet's relative success is extremely high (for details on ProtoNet's performance vis-à-vis structural entities, see [ 27 ]). To avoid trivial correspondences between a keyword and a cluster, such as the assignment of a keyword that annotates only one protein to its singleton cluster, we tested our success only with keywords that annotated at least two proteins (for SCOP and ENZYME keywords). For InterPro and GO, we selected a threshold of 20 proteins per keyword, as the majority (85% in InterPro; 98% in GO) of the annotations is included above this threshold, thus allowing the test to focus on the more significant keywords. Correspondence of ProtoNet hierarchy to external biological sources In order to validate the hierarchical structure of ProtoNet, we compare it with the hierarchical structure of GO as described in Figure 1 . To do this, we select, for each GO term, the best matching cluster in ProtoNet according to the CS. The subset of all terms that have highly matching clusters (best CS>0.5) was selected. In graph-theoretic terminology, this set of terms can be represented as vertices in a graph. We consider two possible sets of directed edges between the vertices: those defined by GO as the parent-child relationship of the clusters' respective terms, and those of the ProtoNet hierarchy. Thus we wish to compare these two sets of graph edges. We use a very conservative test, counting the number of edges that are common to both graphs. A total of 1577 GO terms were selected as described, with 1798 edges between them according to the GO hierarchy. 771 out of 1291 (60%) edges that were produced by the ProtoNet hierarchy appear in the GO hierarchy. This number is highly significant considering the fact that there exist over 1,200,000 possible edges between the 1577 vertices in the graph (considering it as a DAG). It should be noted that there are some terms in GO that are connected to many other vertices. These vertices may bias the results of this test. In order to confirm that ProtoNet performs well without these vertices as well, the vertices were removed manually and the test was repeated, with similarly significant results (33% of the edges were correct). Compression by using an intrinsic parameter In order to allow unsupervised automatic compression of the ProtoNet tree, we searched for an intrinsic parameter of the clustering process that would specify clusters of biological validity. By applying such a parameter one could dispose of clusters that do not pass a certain threshold value, remaining with clusters of high biological validity. Once we remain with a subset of biologically valid clusters, the new hierarchy amongst them can be reconstructed according to the original tree hierarchy. The agglomerative hierarchical clustering scheme defines a set of terms that are intrinsically associated with the process. In such a scheme, each cluster is created from smaller clusters, which are captured as its descendants in the clustering tree. The ProtoLevel (PL) ranges from 0-100 and is used as a standard quantitative measure of the relative height of a cluster in the merging tree. The PL of a cluster is defined as the arithmetic average of the BLAST E-score of the pairs of its proteins. The PL of the leaves of the tree is defined as 0, whereas the PL of a root equals 100. The larger the PL, the later the merging that created the cluster took place. Therefore, the PL scale is considered as an internal monotonic timer of merging, during the clustering process. As mentioned above, a cluster is said to be created when the merging of its two children clusters forms it. The PL at this point is said to be the birthtime of this cluster. The deathtime of a cluster is defined as the PL at its termination, i.e. the point at which it merges into its parent cluster (or 100 if it has no parent). The lifetime (LT) of a cluster is defined as: LT = deathtime - birthtime Therefore, the LT of a cluster reflects its remoteness from the clusters in its "vicinity" in protein sequence space. We examined the LT distribution of the set of InterPro best clusters in comparison with the LT distribution of all clusters in ProtoNet (Figure 2 ). The results suggest that the best clusters have a substantially higher LT than other ProtoNet clusters. This poses the LT as a possible candidate that could allow a biologically-valid tree compression by disposing of all clusters with LT below a certain threshold value. In order to search for a reasonable LT threshold value (that would eliminate a large number of clusters while maintaining biological validity), several threshold values were examined (Figure 3 ). The results show that by using a LT threshold for cluster elimination, in addition to removing the singleton clusters, 87.8% of the clusters may be eliminated with only a minimal reduction in performance (i.e., a reduction of 2.7% in mean best CS), leaving only 27,823 clusters. Furthermore, we compare the LT threshold scheme with a random elimination of similar amounts of clusters. The LT threshold convincingly outperforms the random elimination. The mean best CS was examined for all four external sources (Table 2 ). The results show that the mean CS of ProtoNet were only slightly reduced, while the random mean CS are significantly reduced due to the much smaller amount of clusters. Automatic functional annotation of clusters The following scheme was used to annotate the protein clusters: For each cluster C and keyword K we define the following score: Where TP is the amount of true positives (proteins in C that have the keyword K), FN is the amount of false negatives (proteins not in C that have the keyword K) and FP is the amount of false positives (proteins in C that do not have the keyword K). For each cluster, we search against all keywords of GO and InterPro for the keyword with the highest AS. If the AS of the cluster is greater than 0.25, the cluster is assigned that keyword as its annotation. The logic behind the score and the threshold is as follows: the score is determined by two parameters, the specificity and the sensitivity; let us consider the two worst-case limit cases. In the first case, specificity>0.5 and sensitivity = 1: a majority of the proteins of the cluster share the keyword, and there exist no other known proteins that have the keyword. In the second case, specificity = 1 and sensitivity>0.25: all proteins of the cluster share the keyword and they constitute more than 1/4 of the total proteins known to have this keyword. In both cases, the keyword can be assigned to the protein cluster with a high degree of confidence. All other clusters fall in between these cases. Using this method, all 6,879 clusters that contain 20 or more proteins and that remain after the compression were tested. 5,355 (77.8%) clusters passed the high confidence threshold and were therefore given an annotation. Figure 4 shows the plot of the highest AS score for each of the clusters and the threshold function. Naturally, by relaxing the threshold it would be possible to obtain a higher level of annotation. Cation channels: a biological example Figure 5 shows one of the trees that appear in ProtoNet after compression. The root cluster contains 249 proteins and is annotated as "Cation Channel". There appears to be a correct division between potassium channels to non-potassium channels. Furthermore there is an apparent inner division of the potassium channels into two-pore channels and voltage dependent channels, and of the non-potassium cation channels into sodium channels and TRP channels. Notably, an unannotated cluster of 2 proteins is categorized as potassium channel, but does not appear to be voltage-dependent or two-pore. Closer inspection shows that this cluster contains the 2 orthologs of the LctB bacterial protein. Experimental results suggest that LctB is a new type of non-voltage-mediated potassium channel [ 28 ]. This corresponds well to the fact that ProtoNet did not assign an annotation to this cluster and separated it from the other potassium channels. Conclusions The challenge of protein classification by using sequence similarity has been addressed extensively by several different methods. In order to assign function to proteins, advanced methods (such as Hidden Markov Models implemented in Pfam) have been used to learn sequence-based patterns on "seeds", manually validated alignments of known protein families. The widely-used BLAST algorithm is considered to be a reliable tool for sequence alignment, but has been shown to lack sensitivity for weak similarities that may be detected by signature detection methods. We show here that by using an unsupervised bottom-up clustering method based on BLAST E-values, we have been able to construct a global hierarchy of the SwissProt proteins that can be validated by external biological sources, merely by undertaking a global view of the protein space. The four different external sources that were used for validation reflect different aspects of the protein space: InterPro annotation is predictive and is based on various signature detection methods; GO annotation assignments are both based on prediction and from known research, while the GO hierarchy was constructed completely manually; SCOP is a semi-manual classification of structures that is not necessarily reflected in sequence; the ENZYME database supplies Enzyme Commissions, which constitute a hierarchy that is based on the enzymes' chemical reactions. ProtoNet successfully constructs clusters that correspond highly to all four of the sources. Even high levels of SCOP (such as the Fold classification), which are considered to have no detectable sequence similarity, are partially matched (also see discussion in [ 27 ]). Notably, the correspondence of ProtoNet to InterPro is generally higher than the correspondence to the other sources. This is not surprising, considering the fact that InterPro is based on prediction from sequence. However, it is worthwhile to note that the InterPro families may be reconstructed almost perfectly without using the various sensitive detection methods that InterPro uses, and more importantly without using the manually constructed seeds. After validating the biological relevance of the ProtoNet clusters by using external sources, we examined the hierarchy of ProtoNet. The test showed that the hierarchy presented by ProtoNet significantly corresponds to the manually-constructed biological hierarchy of GO. It is important to note that the method used by ProtoNet is not expected to fully recapture the GO hierarchy due to the fact that ProtoNet is structured as a collection of trees while GO is structured as a DAG. In this sense, the approach of ProtoNet may be limited in the portrayal of evolutionary complexity (as in cases of multiple domains). However by using a domain-based clustering approach, allowing multiple entities of each protein in the hierarchy, a substantial improvement in the CS quality measure may be achieved (unpublished results). An intrinsic parameter that reflects the stability of clusters during the clustering process was used in order to compress the cluster sets, leaving 16.4% of the clusters; removing the singletons clusters as well leaves 12.2% of the clusters. As mentioned above (see Methods), the entire ProtoNet scaffold contains 227,436 clusters that are represented by 630 roots; following this condensation, there are only 27,823 clusters that are represented by 2,236 roots. We show that although a massive portion of the clusters is discarded, very little performance is lost by this condensation process. It is obvious that prior to the condensation process, ProtoNet holds within it both clusters that correctly represent biological groups and clusters that are irrelevant artifacts of the clustering process (e.g. the large root clusters that are constituted of tens of thousands of proteins). Therefore, by allowing a major reduction without significant loss of biological coherence ProtoNet seems to present a more correct view of the protein world. An automatic unsupervised method for the classification of proteins has some important advantages over supervised methods (such as signatures based on manually validated seeds): First, an unsupervised method is unbiased in automatic assignment of function to proteins, a major goal in bioinformatics. Also, it allows high-throughput analysis of whole genomes and enhances understanding of global biological systems without the need for the manual annotation of every protein. Using an automatic annotation method, we are able to successfully annotate 77.8% of the major protein clusters (of size 20 or more) that remain after the compression of the ProtoNet tree. The annotation uses a relatively conservative threshold and therefore yields high-confidence annotations. This further suggests that the clusters remaining after the condensation process are relevant biological clusters and not mere artifacts. One aspect that we have rigorously examined is the robustness of the ProtoNet tree: given a different set of proteins to cluster or a different clustering method, would the resulting tree be significantly different, or are its properties maintained? Interestingly, changing the underlying protein databases (ranging in size from 30,000 to over 1,000,000 proteins), the substitution matrices used for the preliminary BLAST, or the merging strategy [ 19 ] produced very similar trees (unpublished results), suggesting that the performance of ProtoNet is not due to a specific computational method but perhaps to the robust properties of the protein sequence space. Methods ProtoNet version 2.4 which was used for the analyses described in this paper is based on classification of the SwissProt database (version 40.28) that contains 114,033 proteins. The entire ProtoNet scaffold contains 227,436 clusters that are contained in 630 trees. Most trees (611) are singletons and only one contains most (>99%) of the proteins. For more details on the construction of the ProtoNet hierarchy see [ 19 ]. ProtoNet version 4.0 [ 21 ] which is available online contains a wider classification of over 1,000,000 proteins (a union of the SwissProt and TrEMBL databases). Several external sources were used as a biological reference for validation of the ProtoNet tree: InterPro [ 23 ] is an extensive family and signature archive that integrates several different databases: PRINTS, Pfam, PROSITE, ProDom, Smart, TIGRFAMs, and recently also PIR SuperFamily and SUPERFAMILY. Each of these databases relies on a different detection method. Many of these signatures and family keywords are considered to be undetectable by a routine BLAST search. InterPro (version 5.2) contains 5,551 signatures. Gene Ontology (GO) [ 24 ] is a collaborative project of creating a hierarchy of biological terms. GO is represented as a directed acyclic graph (DAG), which is divided into three parts: Molecular Function, Cellular Localization and Cellular Process. In this study only the Molecular Function aspects of GO were used. GO's Molecular Function subdivision (July 2002) contains 5,947 biological terms. SCOP [ 11 ] is a hierarchical representation of protein structures. SCOP uses a tree-like hierarchy of 4 levels: Class, Fold, SuperFamily and Family. SCOP (version 1.57) contains 2,927 structures terms. The ENZYME database (as part of Swissprot data) indicates the EC number of a protein [ 25 ]. EC (Enzyme Commission) numbers are a classification scheme for enzymes, based on the chemical reactions they catalyze. The EC number includes 4 fields (for example, 1.2.3.4 represents the enzyme class, subclass, sub-subclass and entry number, respectively). ENZYME (updated July 2002) contains 3,958 enzyme classifications. We have used EBI mappings of InterPro and GO to SwissProt proteins. List of abbreviations Annotation Score (AS), Correspondence Score (CS), Directed Acyclic Graph (DAG), Enzyme Commission (EC), Gene Ontology (GO), Lifetime (LT). Authors' contributions ML conceived of the compression of the ProtoNet tree. NK, MF and MF participated in implementation of the condensation and in the design and implementation of the various tests. All authors participated in the analysis of the results. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544566.xml |
547906 | Irradiation specifically sensitises solid tumour cell lines to TRAIL mediated apoptosis | Background TRAIL ( t umor necrosis factor r elated a poptosis i nducing l igand) is an apoptosis inducing ligand with high specificity for malignant cell systems. Combined treatment modalities using TRAIL and cytotoxic drugs revealed highly additive effects in different tumour cell lines. Little is known about the efficacy and underlying mechanistic effects of a combined therapy using TRAIL and ionising radiation in solid tumour cell systems. Additionally, little is known about the effect of TRAIL combined with radiation on normal tissues. Methods Tumour cell systems derived from breast- (MDA MB231), lung- (NCI H460) colorectal- (Colo 205, HCT-15) and head and neck cancer (FaDu, SCC-4) were treated with a combination of TRAIL and irradiation using two different time schedules. Normal tissue cultures from breast, prostate, renal and bronchial epithelia, small muscle cells, endothelial cells, hepatocytes and fibroblasts were tested accordingly. Apoptosis was determined by fluorescence microscopy and western blot determination of PARP processing. Upregulation of death receptors was quantified by flow cytometry. Results The combined treatment of TRAIL with irradiation strongly increased apoptosis induction in all treated tumour cell lines compared to treatment with TRAIL or irradiation alone. The synergistic effect was most prominent after sequential application of TRAIL after irradiation. Upregulation of TRAIL receptor DR5 after irradiation was observed in four of six tumour cell lines but did not correlate to tumour cell sensitisation to TRAIL. TRAIL did not show toxicity in normal tissue cell systems. In addition, pre-irradiation did not sensitise all nine tested human normal tissue cell cultures to TRAIL. Conclusions Based on the in vitro data, TRAIL represents a very promising candidate for combination with radiotherapy. Sequential application of ionising radiation followed by TRAIL is associated with an synergistic induction of cell death in a large panel of solid tumour cell lines. However, TRAIL receptor upregulation may not be the sole mechanism by which sensitation to TRAIL after irradiation is induced. | Background TRAIL ( T umour necrosis factor r elated a poptosis i nducing ligand) is one of the most promising anti-cancer agent being currently under investigation (for review see [ 1 - 5 ]). Initially it was shown that TRAIL specifically induces tumour cell apoptosis and tumour regression in nude mice, even when applied as single agent [ 6 , 7 ]. In the meantime rare reports questioned the tumour cell specificity of TRAIL since it was shown that TRAIL induced apoptosis also occurred in normal human liver cells [ 8 , 9 ]. However, subsequentially it was shown that the biochemical preparation of TRAIL rather than TRAIL itself was responsible for the observed toxic effect on hepatocytes [ 10 ]. This is strongly supported by the lack of toxicity of TRAIL receptor agonists in humans in first phase I studies [ 11 ]. One possible explanation for the tumour specifity of TRAIL could lie in its potential role as a mediator of tumour immune surveillance in vivo . In this regard it has been shown that mice lacking TRAIL display a significantly reduced capacity to eliminate syngenic tumour cells in the liver [ 12 , 13 ]. Although TRAIL is a member of the death receptor ligand family certain differences exist to the well characterised ligands TNF ( T umour necrosis factor) or CD95. Most important in this regard is the fact that at least five, instead of only two resp. one, different TRAIL receptors have been identified: The proapoptotic DR4/Trail-R1 [ 14 ] and DR5/TRAIL-R2/TRICK2 [ 15 - 18 ] as well as TRAIL-R3/DcR1/TRID [ 16 , 33 , 20 ] and TRAIL-R4/DcR2/TRUNDD [ 21 , 22 ], which lack any pro-apoptotic function. The latter were shown to protect cells from TRAIL induced apoptosis by competing with the agonistic receptors for TRAIL binding. In addition, TRAIL-R4/DcR2 is able to induce NF-κB activation, which might upregulate a wide array of anti-apoptotic proteins [ 19 - 21 , 23 , 24 ]. The role of the fifth TRAIL binding protein, the soluble osteoprotegerin (OPG) is still unclear, since it displays only low binding affinity to TRAIL at physiological temperatures [ 25 , 26 ]. In analogy with the signalling events triggered by CD95, multimerisation of the agonistic TRAIL receptors induces the recruitment of the FADD adapter molecule to the receptor, leading to a subsequent autoproteolytic activation of initiator caspase-8 [ 27 , 28 ]. Active caspase-8 in turn triggers the proteolytic activation of downstream caspases including caspase-3. Downstream caspases ultimately degrade a broad range of cellular proteins and apoptosis is finalized (for review see [ 29 ]). Up to now, caspase-8 was shown to be the most crucial mediator of TRAIL induced apoptosis [ 30 , 31 ]. However, it has been shown that caspase-10 may act as a surrogate for caspase-8 in some cell systems [ 32 , 33 ]. In contrast to receptor mediated apoptosis, DNA damage triggers apoptosis mainly via mitochondrial death pathways (for review see [ 34 , 35 ]). Key step of mitochondrial apoptosis pathways is the mitochondrial release of pro-apoptotic mediators including cytochrome c. This release is generally controlled by a complex interplay of pro-apoptotic members of the Bcl-2 family namely Bax, Bak, Noxa and Puma. Activation of either of those molecules may occur directly via conformational changes [ 36 ] or transcriptional upregulation [ 37 - 39 ]. Cytochrome c released from the mitochondria triggering the activation of caspase-9 by association with APAF-1 in an ATP dependent manner [ 39 , 40 ]. Caspase-9 subsequently activates the downstream effector caspase cascade including caspase-3 and, in analogy to receptor mediated apoptosis, cell death is finalised. With only very few exceptions [ 41 ], apoptosis induction via mitochondrial death pathways is abrogated by anti-apoptotic members of the Bcl-2 family. Anti-apoptotic proteins of the Bcl-2 family interfere with the cytochrome c release from mitochondria on multiple stages [ 42 - 44 ]. Interestingly, both (death receptor mediated and mitochondrial) pathways are interconnected on several levels. In case of death receptor activation, the propagation of the apoptotic signal is enhanced by caspase-8 mediated activation of Bid [ 46 ]. Bid like other BH3-only molecules triggers the release of cytochrome c from mitochondria ultimately resulting in activation of caspase-9 [ 45 , 40 ]. Thus, receptor mediated death pathways are directly connected to mitochondrial death pathways. Vice versa, activation of caspase-9 via the mitochondrial pathway results in secondary activation of caspase-8 and Bid also leading to an amplification of the intracellular death signal [ 47 - 49 ]. Although TRAIL induces apoptosis when given alone, it has been shown that combination of TRAIL with cytotoxic drugs as 5-fluorouracil, etoposide, paclitaxel, actinomycin C and cisplatin has an even higher apoptotic efficacy [ 50 - 59 , 6 ]. Whereas abundant data therefore support the combination of TRAIL with cytotoxic drugs, only limited studies of TRAIL combined with ionising radiation have been performed [ 31 , 33 , 60 - 62 ]. Except for breast and renal cancer, no data supporting the application of TRAIL in the field of radiation oncology are available. Up to now, statistical analysis of a synergistic efficacy have been presented rarely. In addition, potentially harmful effects of a combination on normal cells have not sufficiently ruled out. Methods Chemicals All biochemicals were obtained from Sigma-Aldrich chemicals (Deisenhofen, Germany) unless otherwise specified. Hoechst 33342 was purchased from Calbiochem and dissolved in distilled water as 1.5 mM stock solution. Cell culture The tumour cell lines MDA-MB 231, HCT-15, Colo 205, NCI H460, FaDu DD and SCC-4 cells were purchased from ATCC (Bethesda, MD, USA). HSF6/ HSF7 fibroblasts, HUVEC and SMC were kindly provided from H-P. Rodemann and R. Kehlbach (Tübingen, Germany). Respectively. All cells were grown in RPMI 1640 medium (Gibco Life Technologies, Eggenstein, Germany) and maintained in a humidified incubator at 37°C and 5% CO 2 . Normal human epithelial cells (HMEC, PrEC, RPTEC, SAEC) and hepatocytes were obtained from Clonetics/Cambrex (Taufkirchen, Germany). Cell culture was performed according to the manufacturer's protocols. TRAIL stimulation TRAIL induced apoptosis was induced with recombinant human TRAIL/TNFSF10 (R&D Systems, Wiesbaden-Nordenstadt, Germany) in concomitant or sequential application with irradiation. Irradiation Cells were irradiated with 6 MV Photons using a Siemens Mevatron linear accelerator with a dose rate of 4 Gy per min at room temperature. Quantification of apoptosis induction Apoptosis induction was quantified by counting of cells with a characteristic apoptotic morphology after DNA staining with Hoechst 33342. Cells were stained by incubation with Hoechst 33342 at a final concentration of 1.5 μM for 15 min. Microscopy was performed using a Zeiss Axiovert 200 microscope (Carl Zeiss, Jena, Germany) using an excitation wavelength filter of 380 nm. All apoptotic rates were means of at least three independent experiments. The given error bars represent the standard error of the mean from independent measurements of the same cell batch. Westernblotting Cells (1 × 10 6 ) were lysed for 30 min in a lysis buffer containing 25 mM HEPES, 0.1% SDS, 0.5% deoxycholate, 1% Triton X-100, 10 mM EDTA, 10 mM NaF and 125 mM NaCl on ice. After removing insoluble material by centrifugation for 10 min at 12.000 g, 20 μg lysate was separated by SDS-PAGE. Blotting was performed employing a tank blotting apparatus (Biorad, Munich, Germany) onto Hybond C membranes (Amersham, Braunschweig, Germany). Equal protein loading was confirmed by Ponceau S staining (Sigma). Blots were blocked in PBS buffer containing 0.05 % Tween 20 and 5% bovine serum albumin at 4°C over night. Primary antibodies were detected after repeated washings with PBS/Tween 20 (0.05%) of the membrane, using a secondary antibody (anti IgG-AP 1:10.000, Santa-Cruz-Biotech, Heidelberg, Germany) diluted in PBS/Tween and incubated for 3 hours at room temperature and washed three times with PBS/Tween. Detection of antibody binding was performed employing enhanced chemoluminescence (CSPD ® -Solution Tropix, Applied Biosystems, MA, USA). PARP cleavage was tested using a polyclonal antibodies for cleaved and uncleaved PARP from Boehringer (Mannheim, Germany) in a 1: 1000 dilution. Monoclonal antibodies for caspase 8 were a gently gift from Prof. K. Schultze-Osthoff and used in a 1:45 dilution. β-Actin (Santa Cruz, Heidelberg, Germany) antibody was used in a 1: 5000 dilution. Receptor expression Cells (0,2 × 10 6 ) were washed twice with PBS and incubated for 30 min with PE-labeled anti-R1/DR4 or -R2/DR5-antibody (R&D Systems, Wiesbaden-Nordenstadt, Germany) at a dilution of 1: 400 in 0,5% FCS/PBS. FACS analyses of superficial receptor expression was performed according to manufacturer's protocol with the Quantibrite™ kit from BD (Heidelberg, Germany). Statistical analysis Efficacy of the combined modalities were evaluated by the isobolic method [ 63 ]. Results Ionising radiation sensitises solid tumour cells to TRAIL induced apoptosis Rates of apoptosis induction in response to ionising radiation or TRAIL alone and after combination were determined. Since previous studies on Jurkat T cells or breast cancer cells demonstrated an upregulation of TRAIL receptor R2/DR5 after combined treatment with TRAIL and irradiation [ 54 , 31 ] two different application schedules were tested. TRAIL was either applied directly after cell irradiation or 12 hours later, to allow for receptor upregulation. As shown in figure 1 , ionising radiation alone (10 Gy) at 48 h induced apoptosis in all cell systems from e.g. 16,0 % in FaDu cells and 34,1% in NCI H460 up to 58,0 % in Colo 205 cells, whereas 0.1 ng/ml TRAIL had a very limited activity in FaDu (3,0%) and Colo 205 cells (14,5%) and up to 30,7 % in NCI H460 tumour cells. In contrast, combination of irradiation (10 Gy) with immediate TRAIL application (0,1 ng/ml) was associated with a much higher apoptotic response (e.g. FaDu 23,3 %, NCI H460 51,9% and Colo 205 80,3%). This effect was even more pronounced when TRAIL was applied 12 hours after irradiation (e.g. FaDu 30,0 %, NCI H460 65,2% and Colo 205 88,0%). In order to test whether the effect of TRAIL and radiation was additive or synergistic an isobologram analysis was performed. Even when applied concomitantly the interaction was synergistic in some cell systems (figure 2a , Colo205, HCT 15 and FaDu) but additive in others (MDA MB 231 and SCC-4). When TRAIL was applied 12 hours after irradiation the interaction was synergistic in all but one cell system (figure 2b ). Processing of caspase 8 and PARP In order to substantiate the findings on apoptosis induction, caspase activation was verified by analysis of caspase-8 and the processing of the caspase-3 substrate PARP 24 hours after TRAIL application. In keeping with the above results, the most prominent effects were found for Colo 205 and NCI H460 cells with strongly increased caspase-8 and PARP processing after combined treatment. Colo 205 cells were particularly sensitive to sequential application of irradiation and TRAIL, whereas less intensive caspase-8 and PARP processing was found for MDA MB231, SCC4 and FaDu (fig. 3 ). These data correlate well with the kinetics of apoptosis induction as determined by fluorescence microscopy of Hoechst stained cells (fig. 1 ). Irradiation induced regulation of TRAIL receptors In order to analyse the role of TRAIL receptor regulation in combined therapy with irradiation and TRAIL receptor expression was quantified by flow cytometry using the Quantibrite™ kit system from Becton Dickinson (Heidelberg, Germany). No upregulation of DR4/R1 was found in all tested tumour cell lines. Instead, a subtle downregulation of DR4/R1 after irradiation was observed (fig. 4A ). Fig. 4B demonstrates upregulation of TRAIL receptor DR5/R2 12–18 h after irradiation with 10 Gy in four of six tested cell lines. Colo 205 cells showed the most pronounced receptor upregulation of 196,8%. In HCT-15 and NCI H460 cells an upregulation of R2/DR5 of 118,0% resp. 96,4% could be measured. In MDA MB231 cells and SCC-4 cells no significant upregulation of receptors was found. FaDu cells do not express TRAIL-receptor R2/DR5 and therefore no upregulation of R2/DR5 could be observed after treatment. Combined treatment of TRAIL and ionising radiation do not damage normal tissue As stated above, hardly any data on normal tissue toxicity after combined treatment are available. We therefore analysed the effect of irradiation plus TRAIL on human hepatocytes, fibroblasts (HSF6 and 7), epithelial cells from prostata (PrEC), kidney (RPTEC), breast (HMEC) and small airways (SAEC), endothelial cells from umbilical cord (HUVEC) and small muscle cells (SMC). Normal tissue cells were treated with 10 Gy and a tenfold higher dosage of TRAIL (1 ng/ml) as used for tumour cells. For evaluation of apoptosis strict morphologic criteria as condensation of chromatin and nuclear fragmentation were used. 48 h after combined treatment no relevant sensitisation of normal tissue cells to TRAIL induced apoptosis could be detected in all cultures. Moreover, even preirradiation did not sensitise normal cells to TRAIL induced apoptosis as observed in tumour cells (table 1 and fig. 5 ). Discussion Based on the rationale that radiation and TRAIL induce cell death via distinct but overlapping cell death pathways, tumour cell lines and normal tissue cultures were subjected to either radiation or TRAIL alone or combined with varying application schedules. Our data show that combining radiation with TRAIL induces apoptosis in a significantly higher percentage than either treatment alone. It is important to note that for TRAIL stimulation in our experiments on tumour cells very low concentrations of TRAIL were used (0.1 ng/ml). The pharmacodynamic properties of TRAIL, especially the peak plasma levels in humans, are not known. Since it is likely that toxicity rises with higher doses of TRAIL, the observation of pronounced effects on tumour cell kill at such low doses is particularly intriguing. Any combination of radiation with TRAIL proved to be more effective than either one alone; however depending on dose level and schedule of stimulation less than additive, additive and synergistic effects were detectable. When TRAIL was applied simultaneously with irradiation three of the six cell systems reacted synergistically. In contrast, five of the six cell systems reacted with synergistic effects when TRAIL was given sequentially 12 hours after irradiation. One of the cell lines displayed only less than additive effects. Statistical analysis confirmed, that synergistic effects are more pronounced and occur with greater likelihood after sequential application of TRAIL after irradiation when compared to concomitant treatment schedules. Possible explanations for the positive interaction of TRAIL and DNA damaging agents including ionising radiation are being disputed. Since the sensitisation was associated with upregulation of the DR5 receptor in some experimental settings [ 31 , 54 , 64 , 68 - 70 ], it is thought that the synergy is based on the increased surface density of the death receptors. Our data support the notion that DNA damage leads to an increased surface expression of the DR5 receptor. However, no tight correlation between receptor upregulation and magnitude of cell kill was observed. It has been suggested, that intact p53 is essential for upregulation of R2/DR5 death receptor expression by ionising radiation [ 64 , 68 ]. However, at least in one of our cell lines (HCT-15) known to harbour a non-function p53 DR5 upregulation was clearly upregulated. Thus, alternative p53 independent pathways for the upregulation of DR5 may exist. This finding is in accordance with data on p53 -/- HCT-116 cells and mouse embryonic fibroblasts showing that NF-κB may also be important for a irradiation induced upregulation of death receptors [ 65 ]. Recently, it was shown in overexpression experiments of NF-κB that its subunits c-Rel and RelA regulate expression of cell death molecules in a differential manner. This suggests that RelA, in contrast to c-Rel, acts as a survival factor by inhibiting expression of DR4/DR5 and caspase-8 and up-regulating cIAP1 and cIAP2 [ 71 ]. This process depends also on cell type and microenvironment [ 54 , 65 ]. Therefore NF-κB subunits seem to play an ambiguous role in regulation of apoptotic pathways. To know its exact role in radiation induced cell death further research is necessary. Additionally, conflicting data regarding the role of Bcl-2 in death receptor-mediated apoptosis have been provided in the past few years. Interestingly, new data point to a complex relationship between Bcl-2-mediated inhibition of apoptosis and the Bcl-2 protein expression level, the strength and the duration of the death receptor stimulus [ 66 ]. Therefore, Bcl-2 expression levels might play a critical role in the modulation of TRAIL sensitivity of tumour cells. Recently, it has been shown that the interaction of 5-FU with TRAIL is strictly Bax but not Bak dependent in HCT 116 cells [ 67 ]. Thus, these experiments suggest also a critical role for the proapoptotic Bcl-2 homolog Bax in linking the TRAIL death receptor pathway to the mitochondrial apoptosis signalling cascade. However, the general mechanism of the positive interaction of TRAIL with irradiation remains unclear. It may ultimately turn out, that manifold mechanisms exist and only some mechanisms will be operative in a single cell system [ 68 ]. The second part of our experiments the toxicity of TRAIL combined with radiation on normal tissues. In order to be able to draw relevant conclusions for normal tissue cells [ 8 ], 10 fold higher doses of TRAIL were used in these experiments. The first important finding is that TRAIL did not induce apoptosis in any of our cell systems including human liver cells. Thus, our experiments confirm the high tumour cell specificity of TRAIL [ 6 , 7 ]. Prior to embarking upon a phase I trial combining TRAIL with irradiation we wished to show a lack of sensitisation to TRAIL by pre-irradiation in normal tissues, compared to tumour cells. Using a wide array of normal cell systems we could not detect any effects of a pre-irradiation on TRAIL sensitivity. Combination of TRAIL with irradiation showed no increase in toxicity over radiation alone. These results are in good accordance with data from Shankar and coworkers showing that TRAIL induced apoptosis rates are only mildly enhanced by a preirradiation of non-malignant human prostate epithelial cells [ 68 ]. Conclusions Our data suggest that TRAIL has great potential in cancer treatment, especially in sequential combination with radiotherapy. We did not observe any sensitising effect of the sequential treatment on TRAIL sensitivity in normal tissue cells. Xenograft experiments designed to answer the questions regarding the short and long term efficacy of a combination of radiation with either TRAIL or TRAIL specific antibodies are underway in our laboratory. Abbreviations APAF, apoptosis protease activating factor; FADD, Fas-associated death domain protein; NFκB, nuclear factor κB; PARP, poly-ADP-ribosyl polymerase; TNF, tumor necrosis factor; TRAIL, TNF-related apoptosis inducing ligand. HMEC: human mammary epithelial cells, PrEC: human prostate epithelial cells, RPTEC: epithelial cells of renal proximal tubule, SAEC: small airways epithelial cells, HUVEC: human epithelial cells of umbilical vein, SMC: smooth muscle cells, HSF 6,7: human fibroblasts Competing interests The author(s) declare that they have no competing interests. Authors' contributions PM performed FACS-analysis and microscopic evaluation of Hoechst stained cells, analyzed the data, participated in the conception of the trial and participated in the preparation of the manuscript. AS participated in microscopic evaluation of tumour cell apoptosis and accomplished analysis of the resulting data. VJ participated in receptor analysis and microscopic evaluation of Hoechst stained cells. HF carried out Western blotting. PTD participated on the preparation of the manuscript. WB performed isobologramm analysis. CB participated in the conception, design of the study, coordination of the study as well as preparation of the manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC547906.xml |
552310 | Optimal sampling of MRI slices for the assessment of knee cartilage volume for cross-sectional and longitudinal studies | Background MRI slices of 1.5 mm thickness have been used in both cross sectional and longitudinal studies of osteoarthritis, but is difficult to apply to large studies as most techniques used in measuring knee cartilage volumes require substantial post-image processing. The aim of this study was to determine the optimal sampling of 1.5 mm thick slices of MRI scans to estimate knee cartilage volume in males and females for cross-sectional and longitudinal studies. Methods A total of 150 subjects had a sagittal T1-weighted fat-suppressed MRI scan of the right knee at a partition thickness of 1.5 mm to determine their cartilage volume. Fifty subjects had both baseline and 2-year follow up MRI scans. Lateral, medial tibial and patellar cartilage volumes were calculated with different samples from 1.5 mm thick slices by extracting one in two, one in three, and one in four to compare to cartilage volume and its rate of change. Agreement was assessed by means of intraclass correlation coefficient (ICC) and Bland & Altman plots. Results Compared to the whole sample of 1.5 mm thick slices, measuring every second to fourth slice led to very little under or over estimation in cartilage volume and its annual change. At all sites and subgroups, measuring every second slice had less than 1% mean difference in cartilage volume and its annual rate of change with all ICCs ≥ 0.98. Conclusion Sampling alternate 1.5 mm thick MRI slices is sufficient for knee cartilage volume measurement in cross-sectional and longitudinal epidemiological studies with little increase in measurement error. This approach will lead to a substantial decrease in post-scan processing time. | Background Osteoarthritis (OA) is the most common form of arthritis and a leading cause of musculoskeletal disability in most developed countries [ 1 ]. The knee is one of the most frequently affected joints with a prevalence of 30% in people older than 65 years [ 2 ] and high resultant disability [ 3 ]. Defects in cartilage are widely considered to be the initial problem in OA [ 4 , 5 ], although this viewpoint is not shared by all investigators [ 6 ]. Detection of cartilage morphological change is critical in the evaluation, diagnosis, and monitoring of OA. Conventional radiography is used in evaluating the progression of OA but is limited by its inability to directly visualise cartilage. Magnetic resonance imaging (MRI) offers the distinct advantage of detecting morphologic changes in articular cartilage and is a sensitive and accurate test for evaluating articular cartilage non-invasively [ 7 - 11 ]. The correlation coefficient is 0.99 between knee cartilage volumes measured by MRI and the true volumes by means of water displacement [ 9 ]. This method uses 1.5 mm thick MRI slices and has high reproducibility with coefficients of variation of 2–3% [ 12 ] and has been used in both cross sectional and longitudinal studies of OA [ 12 - 15 ]. However, the method is difficult to apply to large studies as most techniques used in measuring knee cartilage volumes require substantial post-image processing [ 12 ] and the process has not yet been automated. One possible solution is to select a sample from within the 1.5 mm thick slices to reduce the post-image processing time, as has been reported for the estimation of brain compartment volume [ 16 ] and fetal volume[ 17 ]. The aim of the study, therefore, was to determine the optimal sampling of 1.5 mm thick MRI slices required to estimate the volumes of and rate of change in lateral, medial tibial and patellar cartilage with minimal increase in measurement error. Methods Subjects The present study consisted of two datasets; one was part of the Tasmanian Older Adult Cohort Study (TASOAC), an ongoing prospective population-based study aimed at identifying the environmental, genetic, and biochemical factors associated with the development and progression of OA at multiple sites (hand, knee, hip, and spine), which commenced in 2002. Subjects aged between 50 and 79 years were selected randomly from the electoral roll of Southern Tasmania, with an equal number of males and females. Another dataset was a younger adult sample from the Knee Cartilage Volume study (KCV) as previously reported [ 15 ]. Both studies were approved by the Southern Tasmanian Health and Medical Human Research Ethics Committee and all subjects provided informed written consent. MRI An MRI scan of the right knee was performed on all subjects. Knee cartilage volume was determined by means of image processing on an independent work station using the software program Osiris as previously described [ 12 , 15 ]. Two observers were utilised. Knees were imaged in the sagittal plane on a 1.5-T whole body magnetic resonance unit (Picker) with use of a commercial transmit-receive extremity coil. The same machine scanned all knees and Philips Quality Procedure (Philips ACR Support Program, XJR153-2922.03) was utilised for MRI slice thickness quality assurance. The following image sequence was used: a T1-weighted fat saturation 3D gradient recall acquisition in the steady state; flip angle 55 degrees; repetition time 58 msecs; echo time 12 msec; field of view 16 cm; 60 partitions; 512 × 512 matrix; acquisition time 11 min 56 sec; one acquisition. Sagittal images were obtained at a partition thickness of 1.5 mm and an in-plane resolution of 0.31 × 0.31 (512 × 512 pixels). The image data were transferred to the workstation. The volumes of individual cartilage plates (medial tibial, lateral tibial and patella) were isolated from the total volume by manually drawing disarticulation contours around the cartilage boundaries on a slice-by-slice basis. All individual slice areas for each cartilage site and each subject were subsequently transferred to and recorded on a spreadsheet. The total area of each individual cartilage was then multiplied by the slice thickness to produce a volume estimate. This "all slice" estimate of cartilage volume (based on slice thickness of 1.5 mm) was used as the gold standard for other comparisons. Then, the volumes of all individual cartilage plates were recalculated based on different sampling intervals from 1.5 mm thick slices by extracting one in two, one in three, and one in four slice areas from the individual data file. These were then summed and the total was multiplied by the corresponding slice distance. Femoral cartilage volume was not assessed in this study as it is strongly correlated with tibial cartilage volume and thus adds little extra information [ 18 ], tibial cartilage volume is the parameter that is most frequently examined in the literature [ 12 , 19 - 23 ], and femoral cartilage volume has worse reproducibility than tibial cartilage volume [ 11 ]. Other measurements Weight was measured to the nearest 0.1 kg (with shoes, socks and bulky clothing removed) using a single pair of electronic scales (Seca Delta Model 707) which were calibrated using a known weight at the beginning of each clinic. Height was measured to the nearest 0.1 cm (with shoes and socks removed) using a stadiometer. Body Mass Index (BMI) was calculated as weight (kg) / height (m 2 ). A standing AP semi-flexed view of the right knee was performed in all subjects. Radiographs were then assessed utilising the Altman atlas[ 24 ]. Each of the following was assessed: medial joint space narrowing (0–3), lateral joint space narrowing (0–3), medial osteophytes (femoral and tibial combined) (0–3) and lateral osteophytes (femoral and tibial combined) (0–3). Each score was arrived at by consensus with two readers simultaneously assessing the radiograph with immediate reference to the atlas. Any knee ROA was defined as total score ≥ 1. The total score could vary from 0–12. This method had high reproducibility in our hands with ICCs >0.98 [ 25 ]. Statistics Descriptive statistics of the characteristics of the study subjects were tabulated. The annual change in knee cartilage volume was calculated as percent change by means of dividing absolute volume change by baseline cartilage volume. Intraclass correlation coefficient was utilized to assess the measurement agreement. The difference in cartilage volume measured with different samples extracting one in two, one in three, and one in four 1.5 mm thick slices of MR image compared to that measured using 1.5 mm thickness was calculated and expressed as percent absolute difference. Desirable agreement was defined as an ICC ≥ 0.98 with ≤ 1% difference between two measurements. In addition, Bland & Altman plots [ 26 ] were also utilized. Desirable agreement was defined as the mean difference between two measurements close to zero with 95% of individual differences being within 2 SD. All analyses were performed using the SPSS statistical package (version 12.1, SPSS, Chicago, IL). Results A total of 150 subjects took part in this study: 100 subjects with cross-sectional data (female: 48, male: 52) were from the TASOAC study and 50 subjects with longitudinal data (female: 31, male: 19) were from the KCV study. Characteristics of the study sample are presented in Table 1 . Subjects from TASOAC were older, heavier and had a higher prevalence of ROA than those from KCV. Most of participants with ROA were mild with a total ROA score ≤ 3 out of 12. Lateral and medial tibial cartilage volumes were lower in subjects from KCV than those from TASOAC. Table 1 Characteristics of the study population* TASOAC dataset N = 100 KCV dataset N = 50 Age (year) 62.3(7.6) 42.8(6.1) Sex (female %)† 48 62 Height (cm) 167.4(8.7) 168.6(7.9) Weight (kg) 76.0(15.0) 73.9(13.7) BMI (kg/m 2 ) 27.1(4.3) 25.9(4.1) Any knee ROA (%)† 62 18 Knee ROA total score (0- 12 ) 1.3 (1.7) 0.2(0.7) Lateral tibial cartilage volume (ml)‡ 3.0(0.7) 2.6(0.5) Medial tibial cartilage volume (ml)‡ 2.7(0.5) 2.2(0.5) Patellar tibial cartilage volume (ml)‡ 3.5(1.0) 3.5(0.9) Lateral tibial cartilage volume change (%) per year‡ - -1.2(3.4) Medial tibial cartilage volume change (%) per year‡ - -2.9(3.9) Patellar cartilage volume change (%) per year‡ - -3.8(3.4) *Values are mean (SD) except for indicated. BMI: body mass index. ROA: radiographic osteoarthritis. † Percentage. ‡ Measured with the whole sample of 1.5 mm thick MRI slices. In cross-sectional analysis, compared to the cartilage volume measured using 1.5 mm thickness, decreasing the number of the slices by extracting one in two to one in four led to a very little underestimation in the magnitude of the average cartilage volume at lateral, medial tibial and patellar sites with ICCs of 0.98–1.00 (Table 2 ). The maximum underestimation was 3.3% at the medial tibial site with one in four slices (Table 2 ). Similar results were obtained when the analysis was done separately for people with and without ROA (Table 3 ) although the differences tended to be larger in the ROA group. The difference also tended to be larger for medial tibial cartilage in the TASOAC sample and lateral tibial cartilage for the KCV sample (Table 2 ). At all sites and subgroups, cartilage volume measured with one in two slices had less than 1% difference compared to that measured with all 1.5 mm slices with an ICC of 1.0 (Table 2 & 3 ). Bland & Altman plots showed that the mean difference was zero for lateral tibial cartilage and -0.01 ml for medial tibial and patellar cartilage with 95% of individual differences within ± 2SD. The variability was random and uniform throughout the range of cartilage volume (Figure 1 ). Table 2 Agreement analysis of knee cartilage volume measured with different samples of 1.5 mm thick MRI slices* Whole sample (n = 150) TASOAC sample (n = 100) KCV sample (n = 50) %Difference (SD) ICC† %Difference (SD) ICC† %Difference (SD) ICC† Lateral tibial cartilage The whole sample Reference Reference Reference Reference Reference Reference 1/2 whole sample‡ -0.04(1.5) 1.00 0.35(1.4) 1.00 -0.84(1.4) 1.00 1/3 whole sample‡ -0.61(2.3) 1.00 0.11(2.1) 1.00 -2.09(1.8) 1.00 1/4 whole sample‡ -1.12(3.4) 1.00 -0.11(3.0) 1.00 -3.18(3.3) 0.99 Medial tibial cartilage The whole sample Reference Reference Reference Reference Reference Reference 1/2 whole sample‡ -0.50(1.7) 1.00 -0.98(1.4) 1.00 0.46(1.7) 1.00 1/3 whole sample‡ -1.70(3.3) 0.99 -2.97(2.8) 0.99 0.83(2.9) 1.00 1/4 whole sample‡ -3.27(5.0) 0.98 -5.09(3.9) 0.97 0.38(4.9) 0.99 Patellar cartilage The whole sample Reference Reference Reference Reference Reference Reference 1/2 whole sample‡ -0.36(1.2) 1.00 -0.40(1.2) 1.00 -0.29(1.3) 1.00 1/3 whole sample‡ -0.91(2.0) 1.00 -0.93(2.0) 1.00 -0.86(1.9) 1.00 1/4 whole sample‡ -2.24(3.0) 1.00 -2.12(2.9) 1.00 -2.50(3.3) 1.00 * SD: standard deviation. ICC: intraclass correlation coefficient. † All P < 0.001. ‡ Derived by extracting one in two, one in three, or one in four of the 1.5 mm thick MRI slices. Table 3 Agreement analysis of cartilage volume measured with different samples of 1.5 mm thick MRI slices in people with and without ROA* ROA absent (n = 76) ROA present (n = 68) Difference (SD) ICC† Difference (SD) ICC† Lateral tibial cartilage The whole sample Reference Reference Reference Reference 1/2 whole sample‡ -0.30(1.4) 1.00 0.24(1.6) 1.00 1/3 whole sample‡ -1.14(2.3) 1.00 -0.01(2.1) 1.00 1/4 whole sample‡ -1.85(3.4) 0.99 -0.29(3.4) 1.00 Medial tibial cartilage The whole sample Reference Reference Reference Reference 1/2 whole sample‡ -0.39(1.7) 1.00 -0.77(2.2) 1.00 1/3 whole sample‡ -1.20(3.3) 0.99 -2.13(3.4) 0.99 1/4 whole sample‡ -2.56(5.3) 0.98 -3.77(4.5) 0.98 Patellar cartilage The whole sample Reference Reference Reference Reference 1/2 whole sample‡ -0.38(1.2) 1.00 -0.40(1.2) 1.00 1/3 whole sample‡ -0.87(1.9) 1.00 -1.10(2.0) 1.00 1/4 whole sample‡ -2.02(2.9) 1.00 -2.50(3.2) 1.00 *Six subjects had missing values for ROA. Difference in cartilage volume measured with different thick slices of MR images is expressed as percentage. ICC: intraclass correlation coefficient. ROA: radiographic osteoarthritis. SD: standard deviation. † All P < 0.001. ‡ Derived by extracting one in two, one in three, or one in four 1.5 mm thick MRI slices. Figure 1 Bland & Altman plots of cartilage volume measured by every second 1.5 mm thick MRI slice compared to that measured by the total sample at lateral (a), medial tibial (b), and patellar (c) sites. The x-axis represents average values of two measurements while the y-axis represents the individual difference between two measurements, and the three horizontal lines stand for mean individual difference ± 2 SD. Similarly, in longitudinal analysis, compared to the cartilage volume change using 1.5 mm thick slices, decreasing the number of the slices by extracting one in two to one in four slices led to very little over or under estimation of the mean changes in cartilage volume at lateral, medial tibial and patellar sites (Table 4 ). The mean difference ranged from -0.05% to 0.14% with the maximum difference at the patellar site. ICCs ranged from 0.85 to 0.99 (Table 4 ). The difference became larger but all were ≤ 1% in subjects with and without ROA (Table 4 ). At all sites, the annual change in cartilage volume measured with one in two slices had an ICC ≥ 0.98 with less than 0.3% difference compared to that measured using all the slices. Bland & Altman plots showed that 95% of the individual differences were within ± 2 SD and the variability was random and uniform throughout the range of cartilage volume (Figure 2 ). Table 4 Agreement analysis of the annual change in knee cartilage volume measured with different samples of 1.5 mm thick MRI slices* Whole sample (n = 50) ROA present (n = 9) ROA absent (n = 41) Difference (SD) ICC† Difference (SD) ICC† Difference (SD) ICC† Lateral tibial cartilage The whole sample Reference Reference Reference Reference Reference Reference 1/2 whole sample‡ 0.06(0.9) 0.99 0.23(1.1) 0.99 0.02(0.9) 0.98 1/3 whole sample‡ 0.05(1.5) 0.96 -0.65(1.4) 0.98 0.20(1.5) 0.95 1/4 whole sample‡ -0.03(2.2) 0.92 -0.04(2.4) 0.95 -0.02(2.2) 0.91 Medial tibial cartilage The whole sample Reference Reference Reference Reference Reference Reference 1/2 whole sample‡ -0.05(1.1) 0.98 -0.29(1.0) 0.99 0.00(1.1) 0.98 1/3 whole sample‡ -0.03(1.8) 0.95 0.24(1.8) 0.97 -0.10(1.8) 0.95 1/4 whole sample‡ 0.02(3.0) 0.85 -1.04(2.7) 0.92 0.25(3.1) 0.83 Patellar cartilage The whole sample Reference Reference Reference Reference Reference Reference 1/2 whole sample‡ 0.10(0.8) 0.99 -0.07(0.7) 1.00 0.13(0.8) 0.99 1/3 whole sample‡ 0.10(1.5) 0.96 -0.18(1.4) 0.98 0.16(1.5) 0.95 1/4 whole sample‡ 0.14(1.8) 0.93 0.61(1.5) 0.97 0.03(1.9) 0.92 * Difference in the annual change in cartilage volume was expressed in percentage. SD: standard deviation. ROA: radiographic osteoarthritis. ICC: intraclass correlation coefficient. † All P < 0.001. ‡ Derived by extracting one in two, one in three, or one in four 1.5 mm thick MRI slices. Figure 2 Bland & Altman plots of the annual change in cartilage volume measured by every second 1.5 mm thick MRI slice compared to that measured by the total sample at lateral (a), medial tibial (b), and patellar (c) sites. The annual change in cartilage volume was expressed as a percentage. The x-axis represents average values of two measurements while the y-axis represents the individual difference between two measurements, and the three horizontal lines stand for mean individual difference ± 2 SD. Discussion This study suggests that lateral, medial tibial and patellar cartilage volumes measured from up to one in four 1.5 mm thick slices are quite comparable to those obtained from 1.5 mm thick slices. If the agreement is defined at high levels expected to lead to minimal measurement error, then knee cartilage volume can be measured sufficiently and accurately with one in two slices both cross-sectionally and longitudinally regardless of ROA status and./or reader. This approach will lead to a substantial decrease in post-scan processing time and make large-scale studies of knee cartilage volume more feasible. Currently, there is no reported information on the number of the slices of MRI scans to measure cartilage volume apart from a recent paper from our own group which had similar findings to this study with different readers and geographic location [ 27 ]. In a study estimating fetal volume by MRI, Roberts et al reported that using the same thickness of MRI slices (10 mm), volume measured from the low sampling intensity (the distance between scan section midplanes T = 4.5 cm) was virtually identical to those obtained with the high sampling intensity (T = 1.5 cm) with a coefficient of error (CE) < 5% [ 17 ]. In the study estimating brain compartment volume from MR Cavalieri slices [ 16 ], irrespective of slice thickness, a minimum of 3, 5, and 10 slices provided estimates of the true total volume of grey matter and white matter in the cerebrum with coefficients of error (CEs) of 10, 5, and 3%. For a given number of slices CE decreases rapidly when the slices are thicker than the gaps between them; when the slices are thinner than the gaps, then CE is similar to that in the situation when the slice thickness is zero. The current study demonstrates similar results for knee cartilage. Decreasing the number of slices by extracting up to one in four 1.5 mm slices resulted in a very little underestimation in average volume of lateral, medial tibial and patellar cartilage. The maximum mean difference in cartilage volume obtained from one in four slices to that obtained from all slices was 3.3%, which is substantially smaller than the difference of 9% between cartilage volume obtained from 1.5 mm thick slices of MR image and that measured by means of water displacement [ 9 , 19 , 28 , 29 ]. The difference increased slightly when we analysed the data separately for people with and without ROA, but the results were similar for both groups, suggesting ROA within the range we report has very limited effect on the cartilage volume measured with subsamples of MRI slices. If we arbitrarily define an ICC ≥ 0.98 with ≤ 1% difference as optimal as it is expected to minimise the measurement error and only slightly increase the variance, then cartilage volume and its rate of change can be measured accurately with one in two 1.5 mm thick slices for lateral, medial tibial and patellar cartilage. Bland & Altman plots confirmed this with a random scatter about zero as would be expected if there is no difference between two measurements and uniform variability throughout the range of measurements. Of note, for longitudinal data even decreasing the number of slices by extracting up to one in four resulted in a maximum difference of 0.14% in mean annual change in cartilage volume which is very small when compared to the 5% cartilage loss annually we have reported in patients with OA [ 30 ]. Thus, a subsample of MRI slices could also be utilised with marked decreases in processing time allowing greater numbers of subjects to be studied offsetting the accompanying increase in measurement error. Ideally, the more slices used, the more accurate the estimation of the object's volume, as they may contain more information. However, for a completely regular structure, such as a cylinder, the area of a single slice with length gives an exact volume. It is therefore reassuring but not surprising that the current study demonstrates a minimum reduction in the knee cartilage volume and volume change over time as tibial and patellar cartilages have a relatively regular structure. A different interpretation may apply to femoral cartilage and we do not have data on this imaging site. The current study simply examined the effect of decreasing the number of slices on the estimation of knee cartilage volume and volume change while all other variables were kept constant. We did not re-scan the study subjects but simply estimated the cartilage volume by using one in two, one in three, or one in four slices. This has an advantage of allowing us to examine the single effect of sampling intensity in the situation where all other variables such as re-positioning the subject and measurement were kept constant. The effect of these errors on measurement have been well-documented [ 9 , 31 ]. For longitudinal analysis, all the MR images were processed by a single observer. For cross sectional analysis, two observers processed the MR images, one for TASOAC data, and another for the KCV study. However, the difference was even smaller in the whole sample than in the two separate samples providing reassurance that our results may be generalisable to different observers as documented with different readers and machines in Melbourne [ 27 ]. The current study has a number of potential limitations. Firstly, which sampling intensity should be used in the MRI scan of knee cartilage depends on the purpose of the measurement. Our results cannot be applied to individual cartilage volume, but only for mean cartilage volume in groups as the individual difference in cartilage volume increases with decreasing sampling intensity. Secondly, decreasing sampling intensity will increase measurement error as the remaining slices focus on different portions of the irregularly shaped cartilage. Depending on what particular surfaces remain, however, the overall volume may be increased or decreased. If this is random, then the mean will remain the same as demonstrated in the current study. Thirdly, the ICC can be influenced by traits in the sample in which it is assessed. Age, sex and BMI have been reported to be associated with knee cartilage volume [ 32 ]. These may result in a higher ICC in the current study, as between-subject variance would become larger. However, subgroup analyses by sex, BMI (< 25, >= 25), and age (<50, >= 50 yr) did not change the results (data not shown). Further analysis using the Bland & Altman method confirmed the good agreement and interchangability between thick and thin slices, indicating that the result of the current study should be applicable to other populations regardless of the demographic factors related to cartilage volume. Fourthly, the participants in the study had only mild ROA, and these conclusions may not apply to subjects with more advanced OA. Lastly, the annual change in cartilage volume in our sample can not be generalized to other populations as half of our longitudinal study sample had a higher genetic susceptibility to OA [ 23 , 33 ]. Conclusion Knee cartilage volume and its rate of change can be accurately measured with every second 1.5 mm thick MR slice. This approach will lead to a substantial decrease in post-scan processing time and make large-scale studies of knee cartilage volume more feasible. Competing interests The author(s) declare that they have no competing interests. Authors' contributions GZ designed and carried out the study planning, data collection, analysis and interpretation of the analysis, and preparation of the manuscript. CD participated in data collection and critical revision of the manuscript. FC participated in the study planning and critical revision of the manuscript. GJ designed the study, participated in analysis and interpretation of the analysis, and critical revision of the manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC552310.xml |
546403 | The socioeconomic gradient and chronic illness and associated risk factors in Australia | Objective To examine the prevalence of major chronic diseases and their risk factors in different socioeconomic groups in the Australian population, in order to highlight the need for public policy initiatives to reduce socioeconomic inequalities in health. Methods Data were provided by the Australian Bureau of Statistics (ABS) from the 2001 National Health Survey (NHS) for selected chronic diseases and associated risk factors. Conditions selected were those, which form the National Health Priority Area (NHPA) conditions (other than injury, which has not been included in this paper, with its focus on chronic disease); plus other 'serious' chronic conditions, in line with the classification developed by Mathers; and for which sufficient cases were available for analysis by socioeconomic status. Indirectly age-standardised prevalence rates were calculated by broad age group for Australia and for five groups of socioeconomic status; rate ratios were calculated to show variations in prevalence between these groups. Results Significant socioeconomic inequalities were evident for many of the major chronic diseases; the largest was for diabetes mellitus (at ages 25 to 64 years); and for many diseases, there was also a strong, continuous socioeconomic gradient in the rates. Circulatory system diseases (in particular, hypertensive disease) and digestive system diseases also exhibited a strong differential in the 25 to 64 year age group. In the 65 years and over age group, the strongest inequalities were evident for mental and behavioural problems, diabetes (with a continuous socioeconomic gradient in rates) and respiratory system diseases. A number of risk factors for chronic diseases, namely self-reported smoking, alcohol misuse, physical inactivity and excess weight showed a striking association with socioeconomic status, in particular for people who were smokers and those who did not exercise. Conclusion This analysis shows that the prevalence of chronic disease varies across the socioeconomic gradient for a number of specific diseases, as well as for important disease risk factors. Therefore, any policy interventions to address the impact of chronic disease, at a population level, need to take into account these socioeconomic inequalities. | Background As in other developed countries, chronic diseases in Australia are major contributors to the extent of illness, disability and premature mortality in the population. They are estimated to make up the greatest proportion of the burden of disease, mental problems and injury for the population as a whole (about 80%), and for particular sub-population groups [ 1 ]. Chronic diseases are exemplified by having multifactorial aetiologies, including common disease risk factors and determinants; significant latency periods and protracted clinical courses; and are seldom cured completely [ 2 , 3 ]. Causal factors interact together at an individual and at a population level to determine the degree of disease burden and illness, and unhealthy risks can be passed on through families, communities, and populations following demographic gradients [ 4 ]. At different life stages, common risk factors and determinants include poor intra-uterine conditions; stress, violence and traumatic experiences; educational disadvantage; inadequate living environments that fail to promote healthy lifestyles; poor diet and lack of exercise; alcohol misuse and tobacco smoking [ 5 , 30 ]. Risk factors are also increasingly more prevalent in areas of low socioeconomic status and in communities characterised by low levels of educational attainment; high levels of unemployment; substantial levels of discrimination, interpersonal violence and exclusion; and poverty. There is a higher prevalence of such factors among Indigenous communities, and other socioeconomically disadvantaged Australians [ 5 , 6 ]. The inequalities in health observed across populations are many – some of them are inevitable and others, unnecessary and unfair. Those inequalities that are potentially avoidable are deemed 'inequitable' [ 7 ]. Despite significant medical advances and improved public health in recent decades, socioeconomically disadvantaged communities continue to suffer an unequal burden of illness, premature death and disability. Therefore, the study of socioeconomic inequalities in chronic diseases and conditions and in risk factors is important and necessary. This is particularly so, if we wish to develop more effective policy mechanisms for preventing and intervening earlier in the progression of chronic diseases and their associated risk factors across the diverse Australian population, and to reduce some of the existing health inequities. Our approach There have been a number of studies published in Australia on socioeconomic inequalities in mortality from various chronic diseases and conditions. The earlier ones were analysed using information on occupation recorded on the death certificate [ 8 - 10 ]. An alternative approach has been to examine variations in mortality rates by grouping residential locations according to socioeconomic criteria. A number of such studies have documented substantial variations in mortality for different age groups [ 11 - 17 ]. However, to date, there have been fewer studies that have examined socioeconomic inequalities in chronic disease prevalence in those still living (analyses of hospital admissions for chronic conditions have been published, but not of prevalence) [ 15 ]. One of the earliest was undertaken by Broadhead, who analysed data on morbidity and social status from the 1977–78 Australian Health Survey (ABS), and found that men in lower status occupations tended to suffer a higher age-standardised rate of self-reported chronic conditions and days of reduced activity; the picture for women appeared less clear [ 18 ]. Lee et al found that low income males were more likely to report mental health problems, chronic symptoms and acute symptoms than their high income counterparts [ 19 ]. Similar findings were reported in other studies, and risk factors associated with chronic diseases also were also associated with low income [ 15 , 19 - 21 ]. The work of Mathers is significant for its systematic documentation of health inequalities among working aged Australians (25 to 64 years) in the late 1980s. He examined mortality, disability, disease groups, specific diseases, self-perceived health, risk factors, health service use and use of preventive screening services, using data from the 1989–90 National Health Survey (NHS) [ 15 ]. Mathers found that there were no clear gradients of chronic, recent or minor illness with level of socioeconomic disadvantage of area, although there were some specific health status indicators (self-reported health, reduced activity, unhappiness) and certain risk factors (inactivity, smoking and alcohol use) that were reported more frequently by those in the more disadvantaged quintiles [ 15 ]. Current information on inequalities in health other than mortality is limited in Australia because of a dearth of suitable data collections. However, the release of data from the 2001 NHS provides an opportunity to examine the prevalence of self-reported chronic disease in Australia and the way in which this impacts on different socioeconomic groupings within the population. Results Information for a selection of chronic diseases is shown in Table 1 . Diseases were included on the basis of either high prevalence or their contribution to the burden of disease. Table 1 Inequality in prevalence of selected chronic diseases 1 , 2001 Age group (years) and chronic disease Rate 2 Rate ratio by quintile of socioeconomic disadvantage of area 3 First Second Third Fourth Fifth 0–14 4 Mental and behavioural problems 5 6 596 1.00 1.04 1.10 1.12 1.52** Respiratory system 21 807 1.00 1.07 1.05 1.11 0.99 Asthma 13 363 1.00 1.10 1.12 1.25* 1.12 15–24 Mental and behavioural problems 5 10 284 1.00 1.02 0.97 1.08 1.28 Respiratory system 33 373 1.00 1.04 1.12 1.09 1.00 Asthma 16 263 1.00 0.82 1.14 1.02 1.00 Bronchitis/emphysema 1 701 1.00 6 1.32 6 1.66 6 1.94 6 1.97 6 Musculoskeletal system 7 19 088 1.00 1.11 1.00 1.08 0.94 25–64 Diabetes mellitus 2 234 1.00 1.37 1.67* 1.72* 2.28*** Mental and behavioural problems 5 11 093 1.00 1.05 1.20* 1.36*** 1.67*** Circulatory system 17 491 1.00 1.04 0.97 1.15* 1.28*** Hypertensive disease 9 751 1.00 1.12 1.01 1.24* 1.54*** Respiratory system 32 964 1.00 1.00 0.99 0.99 1.01 Asthma 10 393 1.00 1.10 0.99 1.19* 1.14 Bronchitis/emphysema 3 429 1.00 0.97 1.14 1.55** 1.70*** Digestive system 8 074 1.00 1.03 1.07 1.12 1.37*** Musculoskeletal system 7 39 840 1.00 1.10* 1.16*** 1.16*** 1.22*** 65 & over Diabetes mellitus 8 981 1.00 1.13 1.14 1.52* 1.56* Mental and behavioural problems 5 7 222 1.00 1.21 1.62* 1.67* 1.56* Circulatory system 56 592 1.00 1.09 1.06 1.10 1.19* Respiratory system 31 442 1.00 1.03 0.87 0.95 1.22* Musculoskeletal system 7 63 669 1.00 1.02 1.06 1.03 1.08 1 Survey respondents can report more than one disease. 2 Rate is the number of persons per 100,000 population reporting the disease. 3 The extent of any inequality is shown by the rate ratio, which expresses the ratio of the rate in each quintile to the rate in Quintile 1 (the most advantaged areas, with a rate ratio of 1.00); rate ratios differing significantly from 1.0 are shown with * p < 0.05; ** p < 0.01; *** p < 0.001. 4 Information about these age groups were collected by proxy, using parental report. 5 Information may be based on self-diagnosis, rather than diagnosis by a health practitioner. 6 Indicates rate ratio based on estimates with a Relative Standard Error of between 25% and 50% and should be used with caution. 7 Includes diseases of the connective tissue. Source: National Health Survey, ABS 2002 The main findings are: The largest differential between those in the most well off and those in the most disadvantaged areas was for diabetes mellitus at ages 25 to 64 years, with the prevalence in the most disadvantaged areas being just over two and a quarter times (a rate ratio of 2.28) the prevalence for the least disadvantaged; there is also a strong, continuous gradient in the rates, with the rate ratios in each of the third to fifth quintiles statistically significant. There was a statistically significant differential of 67% at ages 25 to 64 years, with a strong, continuous gradient, in the prevalence of self-reported mental and behavioural problems across the socioeconomic gradient; differentials (also statistically significant) in the 0 to 14 year and 65 years and over age groups were 52% and 56%, respectively. Circulatory system diseases (in particular, hypertensive disease) and digestive system diseases also exhibit a strong differential in the 25 to 64 year age group (statistically significant differentials of 28% and 54%, respectively). In the 65 years and over age group, the strongest differentials were evident for mental and behavioural problems (a statistically significant 56%), diabetes (with a continuous gradient in rates, statistically significant in quintile3 four and five) and respiratory system disease (a statistically significant 22%). Asthma accounted for almost two thirds of the rate of reporting of respiratory system disease in the 0 to 14 year age group, almost half in the 15 to 24 year age group, and for about a third of the rate in the 25 to 64 year age group. The NHS also included data on a number of risk factors for chronic diseases, namely self-reported smoking, alcohol misuse, physical inactivity and excess weight. A number of these risk factors show a striking association with socioeconomic status, in particular for people who are smokers and those who did not exercise, with continuous gradients and highly elevated rates of statistical significance (Table 2 ). The differences in male and female rates are also of interest. It was only for underweight females, and for the risk factor of high-risk alcohol consumption by females, that the socioeconomic gradient was reversed. Table 2 Inequality in prevalence of selected health risk factors, 18–64 years, 2001 1 Health risk factors Rate 2 Rate ratio by quintile of socioeconomic disadvantage of area 3 First Second Third Fourth Fifth Current smokers - Male 30 582 1.00 1.40*** 1.55*** 1.71*** 1.95*** - Female 24 009 1.00 1.29*** 1.34*** 1.48*** 2.00*** - Persons 27 275 1.00 1.35*** 1.45*** 1.61*** 1.96*** Alcohol – High risk - Males 6 976 1.00 1.09 1.26* 1.26* 1.45*** - Females 2 127 1.00 0.59* 0.94 0.76 0.87 - Persons 4 537 1.00 0.93 1.16 1.12 1.22* Did not exercise - Males 28 772 1.00 1.20** 1.36*** 1.52*** 1.68*** - Females 28 220 1.00 1.19** 1.29*** 1.35*** 1.65*** - Persons 28 494 1.00 1.20*** 1.32*** 1.43*** 1.66*** Underweight females 12 675 1.00 0.89 0.83* 0.72*** 0.91 Overweight/obese - Males 54 701 1.00 1.09* 1.11* 1.04*** 1.00 - Females 37 004 1.00 1.09 1.21*** 1.16** 1.17** - Persons 45 798 1.00 1.09** 1.15*** 1.09** 1.06 1 Survey respondents can be shown under more than one type of risk factor. 2 Rate is the number of persons per 100,000 population estimated to be at risk from the health risk factor. 3 The extent of any inequality is shown by the rate ratio, which expresses the ratio of the rate in each quintile to the rate in Quintile 1 (the most advantaged areas, with a rate ratio of 1.00); rate ratios differing significantly from 1.0 are shown with * p < 0.05; ** p < 0.01; *** p < 0.001. Source: National Health Survey, ABS 2002 It is important to note that the inequalities reported above relate to the health of those people living in a geographic area and to the overall level of socioeconomic disadvantage of that area. Most areas will contain varying levels of individual socioeconomic disadvantage and, to the extent that the poorer health is associated with individual economic circumstances and living conditions rather than communal environment, the inequalities will understate the true differences in health status according to socioeconomic disadvantage [ 15 ]. Furthermore, there are limitations to the use of area-based measures of SES. Due to misclassification error (i.e. ascribing area-SES to individuals), estimates of difference across the quintiles will be smaller than if data on individual-level measures of SES were used [ 28 ]. Thus, chronic disease inequalities in the wider population by SES are likely to be larger than those reported in this study. In addition, the exclusion of the 'sparsely settled' areas of Australia in NHS data collection results in the omission of data from a high percentage of Indigenous people, who are the population group with the poorest health. Discussion Our analysis indicates that socioeconomic inequalities in the prevalence of chronic diseases and their concomitant risk factors are evident across the Australian population. However, the diseases with substantial disparities across the socioeconomic quintiles are different, for different stages in the life course. Although these results cannot be directly compared with those of previous studies, because of definitional and methodological differences, the recurring finding of inequalities for chronic disease morbidity and risk factor prevalence across the socioeconomic gradient remains a significant concern. The burden in the Australian population attributable to socioeconomic inequality is large, and has far-reaching implications in terms of unnecessary disability and suffering, the loss of potentially economically productive members of society, and increased costs for the health and social care systems [ 35 ]. Despite the expenditure of millions of dollars to prevent and reduce the prevalence of chronic diseases and their risk factors, these inequities have persisted. However, the situation in Australia is by no means unique, for inequalities in these diseases and their risk factors have been observed for most of the developed countries in which they have been studied [ 26 ]. What should we be doing differently? There is a growing body of knowledge that will help to provide direction for developing policies to reduce inequities across the population. The socioeconomic environment is a powerful and potentially modifiable factor, and public policy is a key instrument to improve this environment, particularly in areas such as housing, taxation and social security, work environments, urban design, pollution control, educational achievement, and early childhood development [ 34 ]. However, attention must be paid to the nature of any action that is taken, to ensure that social and economic inequalities are not increased. Some programs, by their very success, can increase inequality by widening the gap between groups in the population; for example, such programs may be more attractive to those who are already healthier, or not as effective for certain groups with poorer health, less education or more stressful lives. In one smoking cessation initiative, it was found that the prevalence of smoking decreased predominately in those adults with higher education, thus increasing the existing difference with those who were more disadvantaged [ 37 ]. While smoking prevalence in Australia has reduced considerably over the last 20 years, attributes such as lower education and occupational status, unemployment, rented housing, and living in disadvantaged areas reflect a higher probability of reporting tobacco expenditure [ 32 ]. As a result, the tax revenue from the sale of tobacco products is being disproportionately drawn from the poorest households and represents a greater proportion of their household budget [ 32 ]. It is also evident that the ways in which systems such as education and health are delivered and structured can increase existing inequality. For example, schooling can be a way of addressing inequality and also a way of reproducing it. It has been suggested that there are two goals for a social justice program in education: to work to eliminate the contribution that the education system makes to the production over time of social inequality in general; and to maximise the positive contributions that the education system makes to reducing social inequality [ 33 ]. Therefore, different approaches and mixes of policies and programs must be mounted to address inequalities. These approaches may include more precise targeting, but also greater attention to community-based dimensions of 'interdependence' between individual behaviours, key determinants, and community and institutional resources. Policy-makers who wish to address socioeconomic inequalities in health may favour one of the following approaches. Some view the impact of socioeconomic disadvantage on those groups with the poorest health in the population, such as Aboriginal people and Torres Strait Islanders, as the priority policy goal. Others identify the gap between the health of those groups at the outer ends of the socioeconomic hierarchy (those with the poorest health and those with best health), and see the narrowing of the gap as the goal. Others prefer to focus on the socioeconomic gradient in health that runs across the whole population [ 31 ]. Graham has identified that the last approach widens the policy debate in three ways [ 31 ]. Firstly, it looks for the causes of health inequality in the systematic differences in life chances and opportunities, living standards and lifestyles that are associated with people's unequal positions across the socioeconomic hierarchy, and for the pathways through which they influence health [ 31 , 36 ]. Secondly, as a result, addressing health inequalities becomes a population-wide goal that includes every citizen [ 31 ]. Thirdly, 'reducing health gradients' provides a more comprehensive policy approach: one that encompasses 'remedying disadvantages' and 'narrowing health gaps' within the broader goal of 'equalising health chances across all the socioeconomic groups' [ 31 ]. She also observes that, "improving the health of poor groups and improving their position relative to other groups are necessary elements in a strategy to reduce the socioeconomic gradient. However, neither is sufficient on its own. To reduce the socioeconomic gradient, health in other socioeconomic groups also needs to improve at a faster rate than in the highest socioeconomic group. Thus, policies to ameliorate health disadvantage, to close health gaps and to reduce health gradients need to be pursued together, and not at the expense of each other" [ 31 ]. There is also an urgent need to make health inequalities a research priority for each stage of the life course – not just to monitor the size and extent of the disparities but also to undertake research that will find preventive approaches and further policy interventions that will be effective in reducing them, and that are likely to be implemented by governments and communities. Conclusions Clearly, any moves to address the impact of chronic disease at a population level must take into account socioeconomic inequalities in prevalence. More research is needed to determine which approaches are effective and why others have failed to have the desired impact, particularly for those who are from socioeconomically disadvantaged areas. Finally, although rates are generally highest at the oldest ages, the development of risk factors for many chronic diseases occurs early in life, and thus, it is essential those health inequities are addressed right across the life course. Methods Data sources The ABS conducts the National Health Survey (NHS) on a regular basis, most recently in 2001 [ 22 ]. The NHS collects information from approximately 26,900 people from all States and Territories living in private dwellings, selected at random using a multi-stage area sample of private dwellings. The survey is undertaken across much of Australia, but excludes the 'sparsely settled' areas, which comprised less than 1% of the non-Indigenous population and 25% of the Indigenous population at the 2001 Census: a separate Australia-wide survey of the health of Indigenous people, also conducted in 2001, surveyed these sparsely settled areas. The survey includes self-reported details of health conditions (both acute and long term) and major risk factors, as well as demographic and socioeconomic information about the survey respondent. Respondents were asked if they had been told by a doctor or nurse that they had asthma, cancer, heart and circulatory conditions, and/or diabetes. These conditions, together with injuries and mental health, form the NHPAs [ 27 ]. However, for long term mental health problems, respondents were not asked whether they had been told by a doctor or nurse that they had any mental health problems; thus, the responses may be based on self-diagnosis, rather than diagnosis by a health practitioner [ 22 ]. Respondents were also asked a series of questions about other specific, non-NHPA, conditions, covering eye and sight problems, ear and hearing problems, and arthritis, rheumatism and gout. They were then shown a series of three prompt cards (two with conditions listed, while the third contained more general descriptions of condition types) and asked whether they had any of the conditions shown or conditions similar to those shown or described. In each of these cases, details were recorded for conditions reported as current at the time of the survey; respondents were also asked whether the condition had lasted, or was expected to last, for six months or more. Information was gathered directly from individuals aged 15 years and older. For children up to the age of 15 years, information was provided by proxy, from a parent or guardian. The particular conditions for which data were requested from the ABS for this analysis were: the NHPA conditions (other than injury, which has not been included in this paper, with its focus on chronic disease); plus other 'serious' chronic conditions, in line with the classification developed by Mathers [ 15 ]; and for which sufficient cases were available for analysis by five groups of socioeconomic disadvantage of area (see below for details of the way these groups were constructed). The risk factors used by the ABS were those identified for the NHPA conditions [ 27 ]. The ABS has coded conditions reported by respondents to output disease categories based on ICD-10. Conditions described as 'chronic' in this article include those long-term conditions reported in the NHS, which are commonly recognised by health practitioners as chronic diseases [ 23 ]. The risk factor for 'high risk due to alcohol' reflects the National Health and Medical Research Council's risk levels for harm in the long term from alcohol consumption [ 24 ]. The risk factors for overweight and underweight were calculated from self-reported height and weight information and grouped to reflect World Health Organization (WHO) guidelines [ 25 ]. Given the policy importance of the NHPAs, the 2001 NHS questionnaire underwent significant revisions to more precisely capture information on several of the NHPAs. Consequently, while the quality of the information on NHPAs has been improved from the 1995 NHS, the degree of comparability with previous surveys has been somewhat compromised for many of the major health conditions. Some specific conditions (e.g., diabetes) appear to be comparable between the 1989–90 and 2001 surveys, however, for most groups of conditions based on ICD chapter headings (e.g., all circulatory) the ABS advise that the combined effect of major conceptual changes as well as major classification changes between the 1989–90 and 2001 surveys would make direct comparisons very difficult. This analysis is therefore restricted to the 2001 data. Measurement of socioeconomic status The socioeconomic status (SES) of the address of residence of each survey respondent is available at the Census Collection District (CD) level and was added to the NHS file, as was the quintile of socioeconomic disadvantage of area into which that CD fell at the Census. The measure used to allocate CDs to quintiles was the 1996 Census Index of Relative Socio-Economic Disadvantage (IRSD). The IRSD is one of five Socio-Economic Indexes for Areas produced by the ABS using Principal Components Analysis. It summarises information available from variables collected in the 1996 Population Census including those related to education, occupation, and income. The variables are expressed as percentages of the relevant population. The NHS records were aggregated to the quintiles derived from the Census data, where Quintile 1 comprises the CDs with the highest IRSD scores (highest socioeconomic status, or most advantaged, areas) and Quintile 5 comprises the CDs with the lowest IRSD scores (lowest socioeconomic status, or most disadvantaged areas). Each quintile comprises approximately 20% of CDs. The ABS produced the estimates of the number of people with chronic diseases and risk factors by quintile. The method used resulted in the production of quintiles of varying sizes, ranging from 17.4% of the population in Quintile 5 (most disadvantaged areas) to 22.8% in Quintile 2 and 21.1% in Quintile 1 (most advantaged areas). This is a differential of over five percentage points between Quintile 5 and Quintile 2, or 1,027,030 fewer people in the most disadvantaged areas when compared with Quintile 2 (and 708,980 fewer in Quintile 5 than in Quintile 1). The effect of this lack of precision on the results by quintile is not known. Although, in part, the difference arises as a result of the method used (that is, that the quintiles are based on the Census population, and applied to the NHS population which has a different profile), it may also reflect different response rates to the survey from different socioeconomic groups. The NHS includes other measures that could potentially be used to measure socioeconomic status. For example, income reporting in ABS surveys is known to be incomplete, in particular for low income groups and, in the 2001 NHS, income was only collected for the respondent and their spouse. Information on education is also available, but a single measure has yet to be agreed. Problems in using education include that it is typically completed early in adulthood; it captures neither differential on-the-job training and other career investments made by individuals with similar levels of formal schooling, nor the volatility in economic status during adulthood that has been shown to have adverse implications for health [ 29 ]. The age structure of the population may also influence the indicator: an older population generally has lower education levels than a younger population due to improved access to education over time. Analysis Chronic disease and risk factor rates are expressed as rates per 100,000 population, indirectly age-standardised. The rates were also calculated by age although, given the small sample size of the NHS, only broad age groups (0–14 years, 15–24 years, 25–64 years and 65 years and over) were available. The standard population and quintile populations are the weighted 2001 survey populations from the NHS. The extent of any differential between the quintiles is shown by the rate ratio, which expresses the ratio of the rate in each quintile to the rate in Quintile 1 (the most advantaged areas, with a rate ratio of 1.00). Competing interests The author(s) declare that they have no competing interests. Authors' contributions JG conceived the study, and was responsible for its design and coordination. ST participated in the design of the study and performed the statistical analysis. DH participated in the drafting of the manuscript. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC546403.xml |
509300 | IRE1-Independent Gain Control of the Unfolded Protein Response | Nonconventional splicing of the gene encoding the Hac1p transcription activator regulates the unfolded protein response (UPR) in Saccharomyces cerevisiae . This simple on/off switch contrasts with a more complex circuitry in higher eukaryotes. Here we show that a heretofore unrecognized pathway operates in yeast to regulate the transcription of HAC1 . The resulting increase in Hac1p production, combined with the production or activation of a putative UPR modulatory factor, is necessary to qualitatively modify the cellular response in order to survive the inducing conditions. This parallel endoplasmic reticulum–to–nucleus signaling pathway thereby serves to modify the UPR-driven transcriptional program. The results suggest a surprising conservation among all eukaryotes of the ways by which the elements of the UPR signaling circuit are connected. We show that by adding an additional signaling element to the basic UPR circuit, a simple switch is transformed into a complex response. | Introduction In eukaryotes, the endoplasmic reticulum (ER) serves as the first station of the secretory pathway, through which all secreted and membrane proteins must pass. Within the ER, proteins are folded into their native structure and multisubunit protein complexes are assembled. The ER is a dynamic organelle, capable of sensing and adjusting its folding capacity in response to increased demand: when misfolded proteins accumulate in the ER, a signaling pathway, termed the unfolded protein response (UPR), is activated (reviewed in Ma and Hendershot 2001 ; Patil and Walter 2001 ; Kaufman 2002 ; Ron 2002 ). The UPR activates the expression of genes that enable the cell to adapt to and survive the stress, including those encoding ER-resident chaperones ( Lee 1987 ; Kozutsumi et al. 1988 ), key enzymes in lipid biosynthesis ( Cox et al. 1997 ), members of the ER-associated degradation (ERAD) machinery, and other components of the secretory system ( Ng et al. 2000 ; Travers et al. 2000 ; Urano et al. 2000 ). In yeast, the UPR is controlled by a binary switch imposed by a nonconventional splicing reaction that governs the production of the Hac1p transcription factor responsible for the activation of UPR target genes ( Cox et al. 1993 ; Kohno et al. 1993 ; Cox and Walter 1996 ; Mori et al. 1992 , 1996 ). In uninduced cells, direct base pairing between the 5′ untranslated region (UTR) and an intron at the 3′ end of the mRNA prevents HAC1 mRNA translation ( Chapman and Walter 1997 ; Ruegsegger et al. 2001 ). Accumulation of unfolded proteins activates the ER-resident transmembrane kinase/endoribonuclease Ire1p, which then cleaves the HAC1 mRNA at two precise splice junctions, excising the intron ( Cox et al. 1993 ; Mori et al. 1993 ; Sidrauski and Walter 1997 ). The two HAC1 exons are then joined by tRNA ligase, allowing translation of Hac1p ( Sidrauski et al. 1996 ). To date, Ire1-dependent HAC1 mRNA splicing is the only identified way by which signals from the ER lumen affect transcription in yeast. By contrast, in metazoan cells three mechanistically distinct pathways are known that operate in parallel, although their relative importance in different tissues remains to be determined (reviewed in Ma and Hendershot 2001 ). Hints that further complexity also exists in yeast comes from data presented in the accompanying paper ( Patil et al. 2004 ): these data demonstrate that Hac1p activity is modulated by interaction with Gcn4p, a transcription factor central to regulation of amino acid biosynthesis. The UPR, therefore, may integrate signals from more than one source to compute a transcriptional output appropriate for the physiological conditions of the cell. In this paper, we show that HAC1 mRNA transcription is regulated, resulting in control of Hac1p abundance. Thus the on/off switch provided by IRE1 -dependent splicing is not the only regulatory step of the UPR. This regulation responds to a bipartite signal that emanates from the ER and is communicated by an Ire1p-independent pathway. As a consequence, an alternate transcriptional program is triggered, with specific alterations to the normal UPR allowing the cell to survive. Thus, quantitative modulation of Hac1p imposes gain control on a binary switch in the UPR circuitry and, in collaboration with an additional signaling input, transforms a discrete transcriptional response into a more complex signaling function. Results Secretory Stress Boosts HAC1 mRNA Abundance To define the basic circuitry of signal transduction in the UPR, we evaluated the HAC1 mRNA processing step in a quantitative manner. To this end, we induced the UPR with either dithiothreitol (DTT) or tunicamycin (both agents that cause protein misfolding selectively in the ER) and monitored HAC1 mRNA by Northern blot analysis ( Figure 1 A). In agreement with previous results, we observed rapid and efficient splicing of HAC1 mRNA, as apparent from the conversion of unspliced HAC1 u mRNA ( u for UPR- u ninduced) to spliced HAC1 i mRNA ( i for UPR- i nduced). Quantitation of the results shows that the relative abundance of HAC1 mRNA (the sum of HAC1 u and HAC1 i mRNAs) remained unchanged over at least 12 h ( Figure 1 A; unpublished data). These data demonstrate that acute induction of unfolded proteins triggers a simple on/off switch that controls HAC1 mRNA splicing. Figure 1 ER-Distal Secretory Stress Boosts HAC1 mRNA Abundance (A) Determination of HAC1 mRNA abundance during the UPR. The UPR was induced in WT cells by addition of either 6 mM DTT (lanes 1–4) or 1 μg/ml tunicamycin (lanes 5–8) for the times indicated. Total RNA was harvested at the indicated intervals, and the relative abundance of HAC1 and ACT1 mRNAs was analyzed by Northern blot analysis (see Materials and Methods ). Splicing was calculated at the ratio of spliced (HAC1 i ) to total (HAC1 i + HAC1 u ) mRNA. (B) Determination of HAC1 mRNA abundance during ER-distal secretory stress. WT, sec12–1, sec14–3, and sec1–1 strains were grown at 23 °C and shifted to 37 °C. (C) Determination of HAC1 mRNA abundance during ER-proximal secretory stress. WT, sec14–3, sec61–101, sec62–101, and sec63–201 strains were grown at 23 °C and shifted to 37 °C. In light of these observations, we were surprised to find that blocking the secretory pathway distal to the ER resulted in a pronounced increase in HAC1 mRNA abundance. As shown in Figure 1 B, HAC1 mRNA levels increased 3- to 4-fold in mutant strains compromised at various steps in the secretory pathway when shifted to the nonpermissive temperature ( sec12–1: ER → Golgi, lanes 5–8; sec14–1: intra-Golgi, lanes 9–12; and sec1–1: Golgi → plasma membrane, lanes 13–16) ( Novick et al. 1980 ). Splicing was also induced, albeit to a lesser degree than was observed with DTT or tunicamycin treatment. The observed splicing suggests that blockages in ER-distal compartments of the secretory pathway lead to activation of Ire1p in the ER. Temperature shift alone only transiently induced HAC1 mRNA splicing and had no effect on HAC1 mRNA abundance ( Figure 1 B, lanes 1–4). To determine if any disruption of the secretory pathway had similar consequences, we blocked earlier stages of protein traffic. Mutations that blocked protein entry into the ER had no effect ( Figure 1 C: sec62–101, lanes 13–16; sec63–201, lanes 17–20) or only a mild effect ( sec61–101, lanes 9–12) on HAC1 mRNA abundance. Thus, a surveillance pathway operates to adjust HAC1 mRNA levels in response to altered conditions in the secretory pathway. In the experiments described above, we observed HAC1 mRNA induction only in sec mutants that block transport distal to the ER, not in those that block protein entry into the ER. One common consequence of blocking the secretory pathway at later stages is that proteins in transit will eventually back up into the ER ( Rose et al. 1989 ; Chang et al. 2002 ). This condition results in protein folding defects, thereby activating Ire1p, as indicated by the observed HAC1 mRNA splicing. From the data discussed above ( Figure 1 A), however, we know that an accumulation of unfolded proteins alone is insufficient to trigger an upregulation of HAC1 mRNA, suggesting that an additional inducing signal is required. HAC1 mRNA Induction Requires a Bipartite Signal To determine the nature of this second signal, we sought conditions that induce HAC1 mRNA when combined with ER protein misfolding drugs. Canvassing different conditions, we found two scenarios under which wild-type (WT) cells can be induced to upregulate HAC1 mRNA: (1) ER protein misfolding combined with a temperature shift from 23 °C to 37 °C ( Figure 2 A) and (2) ER protein misfolding combined with inositol starvation ( Figure 2 B). Intriguingly, while ER protein misfolding and inositol starvation each activated the UPR individually (as shown by the activation of HAC1 mRNA splicing; Figure 2 A, lanes 5–8; Figure 2 B, lanes 1–4 and 5–8), neither stress alone was sufficient to cause HAC1 mRNA upregulation. Similarly, the temperature shift reproducibly caused a transient UPR induction (see Figure 1 B, lanes 1–4; Figure 2 A, lanes 1–4) but by itself did not affect HAC1 mRNA levels. Only the combination of ER stress with either temperature shift ( Figure 2 A, lanes 9–12) or inositol starvation ( Figure 2 B, lanes 9–12) led to an increase in HAC1 mRNA abundance. Subjecting cells to both temperature shift and inositol deprivation had no additive effect, nor did treating cells with both DTT and tunicamycin (unpublished data). Thus, HAC1 mRNA induction requires a bipartite signal, consisting of one input provided by unfolded proteins in the ER (UP signal), and the other input provided by inositol starvation or temperature shift (I/T signal). Figure 2 HAC1 mRNA Induction Requires a Bipartite Signal and Is IRE1 -Independent (A) Determination of HAC1 mRNA abundance during ER stress and temperature shift. WT cells were grown at 23 °C and shifted to 37 °C (lanes 1–4 and 9–12) or kept constant at 30 °C (lanes 5–8). DTT was added as indicated (lanes 5–8 and 9–12). (B) Determination of HAC1 mRNA abundance during ER stress and inositol deprivation. WT cells were grown at 30 °C in synthetic medium supplemented with inositol and shifted to synthetic medium lacking inositol (lanes 1–4 and 9–12), or continuously grown in medium supplemented with inositol (lanes 5–8). Tunicamycin was added to a final concentration of 1 μg/ml as indicated (lanes 5–8 and 9–12). (C) Distinction between heat shock response and HAC1- mRNA-inducing conditions. WT (lanes 1–4 and 9–12) and HSF1 c (lanes 5–8) strains were grown at 23 °C and shifted to 37 °C (lanes 1–4 and 5–8) or continuously grown at 37 °C (lanes 9–12), and DTT added as indicated. (D) Analysis of IRE1 pathway for a role in HAC1 mRNA induction. Δ ire1 cells were grown at 23 °C and shifted to 37 °C (lanes 1–4 and 9–12) or continuously grown at 30 °C (lanes 5–8), and DTT was added as indicated (lanes 5–8 and 9–12). Note that in Δ ire1 cells, HAC1 mRNA is modestly induced in response to DTT alone (lanes 5–8). This observation is indicative of feedback regulation, whereby a block in the UPR induces the I/T signal. The heat shock response is transiently induced by shifting cells from 23 °C to 37 °C. To determine whether the heat shock response is an important component of the I/T signal, we tested whether continued growth at 37 °C or expression of a constitutively active allele of the heat shock factor Hsf1p ( Sorger 1991 ; Bulman et al. 2001 ) would substitute for the temperature shift described above. Constitutive expression of active Hsf1p ( Figure 2 C, lanes 5–8) led to upregulation of SSA1, a known target of the heat shock response ( Slater and Craig 1989 ), but did not substitute for the I/T signal for HAC1 upregulation. In contrast, continued growth at 37 °C ( Figure 2 C, lanes 9–12) allowed for modest induction of HAC1 mRNA. Thus, elevated temperature elicits effects other than heat shock, which are important for HAC1 mRNA upregulation. HAC1 Induction Is IRE1 -Independent The UP signal was experimentally induced by DTT or tunicamycin treatment of the cells. As Ire1p is a sensor of folding conditions within the ER lumen, we tested next whether Ire1p was required to transmit this signal. Surprisingly, it was not. HAC1 mRNA abundance was induced 2.6-fold in Δ ire1 cells ( Figure 2 D, lanes 9–12), similar to the 3-fold induction observed in WT cells ( Figure 2 A, lanes 9–12). These results show that a previously unrecognized Ire1p-independent surveillance mechanism must exist that monitors protein folding in the ER. HAC1 mRNA Abundance Is Regulated Transcriptionally Increase of HAC1 mRNA abundance could result from increased transcription, reduced degradation, or both. To distinguish between these possibilities, we constructed a reporter gene consisting of the HAC1 promoter driving transcription of the open reading frame encoding the green fluorescent protein (GFP) flanked by ACT1 untranslated regions (HAC1pro-GFP). The resulting heterologous GFP mRNA therefore contained no HAC1 mRNA sequences. Under conditions providing both the UP and I/T signals, the change in abundance of the GFP mRNA ( Figure 3 A, lanes 5–8) mirrored that of the endogenous HAC1 mRNA ( Figure 3 A, lanes 1–4), both in the kinetics and magnitude of the response. These data demonstrate that the observed increase in HAC1 mRNA abundance was caused by increased transcriptional activity of the HAC1 promoter. Figure 3 Activation of the HAC1 Promoter Controls Increase in HAC1 mRNA Abundance (A) Analysis of HAC1 promoter activity during bipartite stress conditions. Δ hac1 cells containing either a construct restoring HAC1 expression (lanes 1–4) or a construct expressing GFP driven by the HAC1 promoter (lanes 5–8) were grown at 23 °C and shifted to 37 °C concurrent with addition of DTT as indicated. (B) Determination of mRNA half-life during HAC1 -mRNA-inducing conditions. polII ts cells were grown at 23 °C and were shifted to 37 °C either in the absence (open symbols) or presence (filled symbols) of DTT. HAC1 mRNA abundance (squares) and ACT1 mRNA abundance (circles) are normalized to the abundance of the PolIII transcript SCR1 . To further test this notion, we compared the rate of decay of HAC1 mRNA under both HAC1 mRNA-inducing and noninducing conditions. To this end, we employed a strain bearing a temperature-sensitive allele of RNA polymerase II, which was subjected to either elevated temperature alone, or to both elevated temperature and DTT treatment. In both cases, polymerase II transcription ceased upon temperature shift, and mRNA decay was measured. As shown in Figure 3 B, the rate of decay of HAC1 mRNA was indistinguishable under the two conditions. Therefore, the increase in HAC1 mRNA abundance in response to the combination of UP and I/T signals is due solely to activation of the HAC1 promoter. HAC1 Promoter Regulation Is Required to Survive Certain Stress Conditions The results presented so far define a novel regulatory mechanism whereby cells adjust the amount of HAC1 mRNA. This mRNA is the substrate for the Ire1p-mediated splicing reaction, which in turn produces HAC1 i mRNA that is translated to produce Hac1p transcription factor. We therefore asked whether elevated levels of HAC1 mRNA led to a proportional increase in the level of Hac1p. Quantitative Western blot analysis showed that this is indeed the case: when cells were treated with DTT and concomitantly shifted to 37 °C, the levels of Hac1p increased 3-fold ( Figure 4 A, lanes 5–8), relative to the Hac1p levels observed in cells subjected to DTT treatment alone ( Figure 4 A, lanes 1–4). Therefore, the transcriptional induction of HAC1 mRNA combined with Ire1p-mediated splicing results in elevated Hac1p levels, characterizing a new physiological state. Henceforth, we refer to this state as the “Super-UPR” (S-UPR). Figure 4 HAC1 Promoter Regulation Is Required to Survive Stress (A) Determination of Hac1p levels during either ER stress alone or during both ER stress and temperature shift. WT cells were either grown at 30 °C and treated with DTT (lanes 1–4) or grown at 23 °C and simultaneously shifted to 37 °C and treated with DTT (lanes 5–8). Protein lysates were prepared, and protein levels were analyzed by Western blot analysis. The relative Hac1p/Pgk1p ratio is normalized to the DTT-treated sample (lane 4). (B) Characterization of HAC1 expression in strain used to approximate basal HAC1 expression. Cells expressing HAC1 from the endogenous promoter (lanes 1–4) or the ADH1 promoter (lanes 5–8) were grown at 30 °C in synthetic medium supplemented with inositol and shifted to synthetic medium lacking inositol simultaneous with the addition of tunicamycin. (C) Reduced viability of strains unable to express HAC1 at elevated levels. The strains described in (B) were plated in serial dilutions (left to right) on synthetic medium lacking inositol (“−ino”) and synthetic medium lacking inositol and containing tunicamycin (“−ino +TM”). To assess the physiological role of the S-UPR, we sought conditions that would allow us to directly monitor the consequences of changes in HAC1 mRNA levels under otherwise identical growth conditions. To this end, we engineered a yeast strain unable to transcriptionally upregulate HAC1 . In these cells, HAC1 mRNA expression was removed from the control of the HAC1 promoter and was instead driven by the heterologous ADH1 promoter ( ADH1pro - HAC1 ), at levels closely approximating the uninduced HAC1 state ( Figure 4 B, compare ADH1pro - HAC1 , lanes 5–8, to HAC1pro - HAC1 , lanes 1–4). Expression from the ADH1 promoter was constitutive, and the levels of HAC1 mRNA did not change significantly under the various inducing conditions described above. As expected, induction of the UPR in these strains led to efficient HAC1 mRNA splicing and Hac1p production. This strain therefore allowed us to fix the cellular Hac1p concentration to a level closely approximating the basal HAC1 expression state observed during the UPR. We next assessed whether we could identify physiological conditions under which elevated HAC1 mRNA levels were required for cell growth. Therefore, we subjected WT cells and the engineered strain described above to the combinations of stresses described in Figure 2 . Cells expressing HAC1 from the endogenous or from the ADH1 promoter grew equally well on plates lacking inositol ( Figure 4 C, left, first and third rows). This condition induces the UPR and requires the expression of at least a minimal amount of HAC1 mRNA, as Δ hac1 cells fail to grow ( Figure 4 C, left, second row). In contrast, only WT cells, which are able to upregulate HAC1 mRNA production, grew on plates lacking inositol and also containing tunicamycin. Cells expressing HAC1 mRNA only at the basal levels from the ADH1 promoter were nonviable on these plates ( Figure 4 C, right, third row). As shown previously in Figure 2 B, this combination of stresses induces the S-UPR. The data therefore reveal that regulation provided by the HAC1 promoter is necessary for cells to survive certain stress conditions that otherwise are lethal. Differential UPR Target Gene Induction by Elevated Hac1p Levels To begin to characterize the cause for increased viability, we next determined differences in UPR target gene expression resulting from either UPR or S-UPR induction. To this end, we used DNA microarray chip analysis to determine the complete mRNA profile of cells grown under UPR and S-UPR conditions. The results of this analysis are shown in Figure 5 A. Each spot represents the fold induction of a UPR target under UPR conditions (x-axis) or S-UPR conditions (y-axis) (see Materials and Methods for definition of the UPR target set used in this analysis). UPR target genes for which the S-UPR has no additional effect should undergo equal induction under both conditions, and are expected to scatter around the diagonal, indicated by the dashed line. This was the case for many UPR targets. However, induction of a substantial number of genes was skewed to the top of the graph, indicating stronger induction under S-UPR conditions than under UPR conditions. These same data are displayed in Figure 5 B to highlight and categorize these differences. In the histogram, the x-axis represents the ratio of the induction of a target gene during S-UPR and UPR conditions, and the y-axis shows the number of genes with a given ratio. We have operationally divided UPR target genes into three classes, based on their fold induction during the S-UPR compared to their fold induction during the UPR. (1) Class 1 targets ( Figure 5 , red bars) exhibit little if any difference in induction during the UPR and S-UPR (S-UPR induction / UPR induction < 2). Thus, the increased Hac1p during the S-UPR does not lead to enhanced transcription, indicating that for these genes the response is already saturated at UPR Hac1p levels. Class 1 targets include many of the known genes encoding ER lumenal chaperones (including KAR2, SCJ1, LHS1, and JEM1 ) and redox proteins (including PDI1, EUG1, and ERO1 ). (2) Class 2 targets ( Figure 5 , blue bars) are induced to a 2- to 4-fold greater extent during S-UPR than during the UPR. Transcription of these genes is therefore roughly proportional to the Hac1p levels in the cell. Class 2 targets include YIP3, involved in ER-to-Golgi transport, OPI3, encoding a phospholipid methyltransferase, and the hexose transporters HXT12, HXT15, HXT16, and HXT17 . (3) Class 3 targets ( Figure 5 , green bars) are induced by the S-UPR greater than 4-fold more than by the UPR. Class 3 contains the UPR targets DER1, involved in ER-associated degradation ( Knop et al. 1996 ; Ng et al. 2000 ; Travers et al. 2000 ), and INO1, critical for membrane biogenesis ( Hirsch and Henry 1986 ). Figure 5 Differential UPR Target Gene Induction by Elevated Hac1p Levels (A) Comparison of UPR target gene induction under either UPR or S-UPR conditions. Whole-genome mRNA expression analysis was carried out on WT cells harvested after 60 min of treatment, either grown at 30 °C and treated with 6 mM DTT (x-axis), or grown at 23 °C and simultaneously shifted to 37 °C and treated with 6 mM DTT (y-axis). Fold changes in gene expression are in reference to the untreated ( t = 0) samples. Shown are only those genes designated as targets of the UPR (see Materials and Methods ). The dashed diagonal line represents equal induction under both conditions. (B) Comparison of UPR target gene induction under either UPR or S-UPR conditions (alternate display). The data from (A) were analyzed to generate a ratio (x-axis) for each gene, dividing the induction during S-UPR-inducing conditions by the induction during UPR-inducing conditions, with target genes of similar ratio grouped together (y-axis). (C) Characterization of HAC1 expression in a strain constitutively expressing HAC1 at high levels. Cells expressing HAC1 from the endogenous promoter (WT; lanes 1 and 2), or a modified promoter constitutively expressing HAC1 at high levels ( HAC1pro HI ; lanes 3 and 4) were treated with 6 mM DTT for 60 min. Although the basal transcription of HAC1pro HI is elevated, the promoter is still capable of further induction during the S-UPR (unpublished data). (D) Determination of Hac1p level in a strain constitutively expressing HAC1 at high levels. Protein lysates were prepared from the strains described in (C), and protein levels were analyzed by Western blot analysis. The relative Hac1p/Pgk1p ratio is normalized to the WT DTT-treated ( t = 60) sample from Figure 4 A. (E) Transcriptional response of different classes of UPR targets to high levels of Hac1p. Whole-genome mRNA expression analysis was carried on HAC1pro HI and WT cells treated with 6 mM DTT and harvested after 60 min. For the genes in each of the three classes of UPR targets defined in (B), a ratio (x-axis) is calculated by dividing the fold induction in DTT-treated HAC1pro HI cells by the fold induction in DTT-treated WT cells. This ratio is plotted against the number of genes with a similar ratio (y-axis). The Class 2 target YFR026C (asterisk), which is DTT-induced approximately 10-fold more in HAC1pro HI than in WT cells, is of unknown function. a, DER1; b, INO1; c, YOR289W; d, YHR087W . (F) Transcriptional response of different classes of UPR targets to UMF. Whole-genome mRNA expression analysis was carried on ADH1pro-HAC1 cells grown at 23 °C and simultaneously shifted to 37 °C and treated with 6 mM DTT, and WT cells treated with 6 mM DTT, both harvested after 60 min. For the genes in each of the three classes of UPR targets defined in (B), a ratio (x-axis) is calculated by dividing the fold induction in ADH1pro-HAC1 cells under S-UPR-inducing conditions by the fold induction in WT cells under UPR-inducing conditions. This ratio is plotted against the number of genes with a similar ratio (y-axis). a, DER1; b, INO1; c, YOR289W; d, YHR087W . Role for a Putative UPR Modulatory Factor The increased transcriptional output under S-UPR conditions could occur for two reasons. It could be due to increased Hac1p concentrations in the cell, or it could result because an additional S-UPR-specific transcription factor is produced or activated (perhaps the same that regulates HAC1 transcription). It could also be due to a combination of these two scenarios. To distinguish among these possibilities, we determined the target gene induction profile in cells in which the HAC1 mRNA concentration was artificially elevated to a similar level as that found after S-UPR induction. We took advantage of a specific 15-bp deletion in the HAC1 promoter (HAC1pro HI ), which increases basal expression by about 3-fold, as compared to the endogenous promoter ( Figure 5 C). In cells bearing a HAC1pro HI - HAC1 gene (“ HAC1pro HI cells”), splicing of HAC1 mRNA was somewhat reduced upon UPR induction (47%, compared to 67% for WT); however, even with this reduction, HAC1pro HI cells produced approximately 2.5-fold more spliced HAC1 i mRNA than WT cells ( Figure 5 C, compare lanes 3 and 4 to lanes 1 and 2). The increased levels of HAC1 i mRNA led to a corresponding increase in Hac1p ( Figure 5 D, compare lanes 3 and 4 to lanes 1 and 2). The amount of Hac1p produced by DTT induction of HAC1pro HI cells is approximately the same as the amount of Hac1p produced during the S-UPR (compare Figure 5 D, lanes 2 and 4 with Figure 4 A, lanes 4 and 8). The ability to set HAC1 mRNA levels to S-UPR levels allowed us to compare directly UPR target gene induction with the cellular Hac1p concentration being the only variable. We induced the UPR in both WT and HAC1pro HI cells with DTT and determined the mRNA expression profiles. For each class of UPR target defined above, the expression analysis of UPR-induced WT and HAC1pro HI cells is shown in Figure 5 E. In the histograms, the x-axis shows the ratio of target gene induction during the UPR driven by a high level of Hac1p from HAC1pro HI cells compared to induction during the UPR in WT cells. The y-axis shows the number of genes at any given ratio. As expected, Class 1 targets ( Figure 5 E, top panel) did not further respond to the higher levels of Hac1p produced in HAC1pro HI cells. The majority of Class 2 and Class 3 targets ( Figure 5 E, middle and bottom panels) also did not respond to higher levels of Hac1p (ratio less than 2), indicating that only raising the Hac1p concentration in cells is not sufficient to account for their full increased induction during the S-UPR. By contrast, ten of the 32 Class 2 and Class 3 targets were significantly induced (ratio greater than 2) in cells expressing high levels of Hac1p. For the Class 3 target DER1, high levels of Hac1p were sufficient to elevate expression to S-UPR levels (compare 8-fold induction in DTT-treated HAC1pro HI cells to 9-fold induction in WT cells during the S-UPR). Otherwise, however, high levels of Hac1p did not fully reconstitute the induction seen during the S-UPR. For example, while the Class 3 gene INO1 was induced 7.5-fold more in the S-UPR than in the UPR, it was induced only 3-fold more by high levels of Hac1p, compared to normal levels. We conclude that elevated Hac1p levels are sufficient to selectively increase the induction of a few UPR targets, but that the full transcriptional program of the S-UPR predicts the production or activation of an additional transcriptional activator, which we term UPR modulatory factor (UMF). To dissect further the UMF contribution during the S-UPR, we sought conditions under which UMF activity was the only variable. To this end, we induced the S-UPR in ADH1pro - HAC1 cells, which are prevented from achieving high level Hac1p expression, and compared the mRNA expression profile against the UPR in WT cells. In this analysis, Hac1p levels were approximately equivalent in the two conditions, so variations from the normal UPR transcriptional program reflect the activity of UMF. The results are shown in Figure 5 F, with the data displayed similarly to Figure 5 E: the x-axis shows the ratio of target gene induction during the S-UPR in ADH1pro - HAC1 cells, compared to induction during the UPR in WT cells, and the y-axis shows the number of genes at any given ratio. Not surprisingly, the induction of Class 1 targets ( Figure 5 F, top panel) was unaffected: these are targets that are fully induced by even low levels of Hac1p and are not more induced during the S-UPR. Two Class 3 targets, YOR289W and YHR087W (both of unknown function) reach near WT S-UPR induction levels, without elevated levels of Hac1p; for these targets, UMF likely plays a leading role in their induction, with Hac1p having less influence. Most Class 2 and Class 3 targets ( Figure 5 F, middle and bottom panels), however, do not reach full S-UPR induction levels in the absence of elevated Hac1p levels. For example, the Class 3 target INO1 is induced roughly 25-fold in ADH1pro - HAC1 cells during S-UPR conditions; while this is roughly twice the induction observed during the UPR, it falls far short of the 75-fold S-UPR induction in WT cells. These results reinforce the in vivo requirement for high levels of Hac1p to survive S-UPR stress, demonstrated in Figure 4 C. Taken together with the data shown in Figure 5 E, we conclude that the full S-UPR transcriptional program results from a collaboration between elevated Hac1p levels and UMF, with the relative contribution from each varying among different target genes. Discussion The Circuitry of the UPR In this paper, we describe a novel ER surveillance pathway in yeast that modulates the UPR, resulting in a new physiological state that we term the S-UPR. In response to a bipartite signal transmitted from the ER by an IRE1 -independent pathway, the HAC1 promoter is activated, resulting in increased HAC1 mRNA levels that, upon splicing, yield more Hac1p. The increased Hac1p concentration, in conjunction with an additional postulated factor(s) produced or activated by the S-UPR (UMF), allows the cell to mount a modified transcriptional response to cope with the inducing stress conditions. Figure 6 shows the UPR as a circuit diagram utilizing multiple logical operations to integrate various signals. In the “classical UPR” (in red), basal transcription of HAC1 produces HAC1 u mRNA, which is translationally inactive due to the presence of the inhibitory intron. In response to unfolded proteins, Ire1p performs an on/off operation, excising the intron from HAC1 u mRNA to generate spliced HAC1 i mRNA, which is translated to produce the Hac1p transcription activator. The S-UPR provides another layer of regulation superimposed on the UPR (in blue). If ER folding stress is combined with either a shift to elevated temperature or inositol starvation, an AND gate integrates this bipartite signal and boosts HAC1 mRNA levels. In turn, this regulation causes increased Hac1p production. Together with UMF, Hac1p induces UPR target genes, with particular genes responding differentially to differences in Hac1p and UMF concentration. Thus the S-UPR can be seen as an adaptation of the classical (or basal) UPR, fine-tuning the activation of select targets to produce a response suited to the challenge faced by the cell. Figure 6 A Schematic of the Circuitry of the UPR The model depicts the circuitry of the UPR (red) and the S-UPR (blue). Transcriptional control of HAC1 is indicated by an icon representing a rheostat affording gain control of the UPR; Ire1p-dependent HAC1 u mRNA splicing is indicated by an icon representing an on/off switch. The I/T and UP signals in the S-UPR are integrated by an AND gate (semicircle, top right), i.e., both conditions must be met to propagate the S-UPR signal. The putative UMF may collaborate with Hac1p to control transcription of UPR target genes (shown) and also be involved in regulating HAC1 transcription (not shown); alternatively, different factors may be involved. The collaboration of Hac1p and UMF is indicated by the diamond-shaped icon, which integrates the information coming from both Hac1p and UMF concentration and activity. In the accompanying paper, Patil et al. (2004) describe a third signaling element, which additionally modifies the transcriptional program of the yeast UPR. The authors show that the transcriptional activator Gcn4p collaborates with Hac1p at the promoters of UPR targets, providing an additional opportunity for integration of information about the physiological state of the cell. Gcn4p is a highly regulated transcription regulator that responds to metabolic conditions, such as amino acid availability. Gcn4p is not UMF, as S-UPR induction of HAC1 proceeds normally in Δ gcn4 cells (unpublished data). A recent report from Ogawa and coworkers ( Ogawa and Mori 2004 ) demonstrates autoregulation of HAC1 expression under conditions of extreme and prolonged ER stress, mediated by Hac1p binding to its own promoter. Because the S-UPR can be triggered in Δ ire1 cells that do not produce Hac1p, autoregulation and the S-UPR are distinct pathways. The existence of multiple mechanisms of HAC1 regulation reinforces the notion that multiple cellular stimuli become integrated to fine-tune an appropriate response. Bipartite Signal Requirement for S-UPR Activation Presently, the molecular details of the pathway by which the S-UPR signal exerts transcriptional control are not known. In particular, it will be of interest to determine where in the cell the two branches of the S-UPR signal are integrated, i.e., how the AND gate is constructed and where it resides. One possibility is that this signal integration event occurs close to the source at the ER membrane. Both temperature shift and inositol starvation can equally induce the I/T signal pathway, and it is conceivable that both conditions affect ER membrane properties similarly. Inositol is an essential precursor for phosphatidylinositol, a major structural phospholipid in yeast that is required for proper functioning of the secretory system ( White et al. 1991 ; Zinser and Daum 1995 ; Greenberg and Lopes 1996 ). Previous work has demonstrated an intimate link between inositol regulation and the UPR, presumably to coordinate the concentration of ER lumenal and membrane components ( Cox et al. 1997 ). A similar sensing mechanism operates in cholesterol homeostasis, with sterol composition in ER membranes affecting the activity of SCAP, a membrane-bound regulator of SREBP intramembrane proteolysis ( Espenshade et al. 2002 ). It is likely that elevated temperatures also affect ER membrane properties, such as fluidity ( Laroche et al. 2001 ). If such a property were sensed, it would explain how the temperature effect contributing to the I/T signal is separate from the heat shock response. ER membranes distressed by either inositol deprivation or elevated temperature (the I/T signal) might then control the activity of a membrane-bound component of a signal transduction machine that also senses protein folding conditions (the UP signal) in the ER lumen. Alternatively, the AND gate might be well removed from the ER membrane, with I/T and UP signals traveling separately through the cell and meeting possibly as late as at the promoters of the affected target genes. Components that map onto either signaling pathway need to be identified and placed into the circuit to distinguish between these possibilities. The Transcriptional Output of the S-UPR The transcriptional response elicited by the S-UPR reveals different classes of UPR targets. During the S-UPR, the further activation of UPR targets is not simply proportional to the increase in Hac1p concentration; rather, we observe a multitude of complex responses. Some targets are already maximally transcribed during UPR conditions and are not induced further during the S-UPR, while other targets become significantly more induced. For some targets (a minority), elevated Hac1p concentrations are sufficient to increase transcriptional induction, while for others, S-UPR-derived UMF is also required. We find evidence for both kinds of regulation. The promoters of target genes, therefore, display differential responsiveness to Hac1p concentration and UMF activity. The production of different levels of Hac1p allowed us to isolate and directly assess the responsiveness of target genes to Hac1p concentration under otherwise identical conditions. Those target genes that undergo equivalent activation under both conditions likely have promoters that are saturated by the lower amount of Hac1p, and thus reach full activation more readily. For UPR targets at the other end of the spectrum, induction continues to increase as Hac1p levels increase; lower concentrations of Hac1p are inadequate for full stimulation of these genes, which may have lower affinity for Hac1p. Because genes respond differentially to Hac1p levels, regulation of HAC1 mRNA abundance can be used as a gene-specific gain control for target activation. This control is similar to that observed in regulation of phosphate metabolism, where the differential affinity of certain Pho4p phosphoforms for target promoters allows for the selective activation of a subset of phosphate-responsive genes ( Springer et al. 2003 ). For most target genes, however, the S-UPR further enhances the transcriptional activity even in the presence of high concentrations of Hac1p. For example, INO1 is induced over 75-fold by the S-UPR in WT cells, compared to 33-fold during the UPR in HAC1pro HI cells, while the amount of Hac1p produced in both cases is approximately the same. This added induction during the S-UPR is dependent on Hac1p, as ADH1pro - HAC1 cells treated with DTT and shifted to elevated temperature show significantly reduced induction of INO1 . The simplest interpretation of these findings is that S-UPR-induced UMF, which may or may not be identical to the transcription factor regulating HAC1 mRNA, collaborates with Hac1p to further boost transcription of these genes. The cis determinants that instruct genes to behave as Class 1, 2, or 3 targets are unknown. One attractive possibility is that target gene promoters have differential affinity for Hac1p and/or UMF. Promoters with stronger affinity for Hac1p would be maximally occupied and fully activated during a normal UPR and would not further respond to increased Hac1p levels (i.e., Class 1 targets). Promoters with lesser affinity for Hac1p would increase in occupancy, and hence transcriptional activation, as Hac1p levels rose during the S-UPR, and would possibly achieve full transcriptional activity only with the additional binding of UMF (i.e., Class 2 and 3 targets). Such a mechanism of promoter-encoded differential responsiveness to transcription factor concentration would explain the selective regulation of subsets of UPR target genes. Links with the Metazoan UPR Higher eukaryotes possess three separate pathways to sense ER stress and direct overlapping but distinct transcriptional outputs (reviewed in Ma and Hendershot 2001 ). In the first branch, Ire1p senses unfolded proteins in the ER lumen and directs the cleavage of an intron from the mRNA encoding the XBP-1 transcription factor, analogous to the splicing of HAC1 in yeast ( Yoshida et al. 2001 ; Calfon et al. 2002 ). In a second branch, the transmembrane kinase PERK phosphorylates and inactivates the eIF2-α translation initiation factor ( Harding et al. 1999 ). This attenuates global protein synthesis, but selectively increases the translation of a small number of proteins including the ATF-4 transcriptional activator. Interestingly, ATF-4 is the metazoan ortholog of Gcn4p, the yeast transcription factor demonstrated by Patil et al. (2004) to collaborate with Hac1p. Finally, in a third branch, activation of the UPR in metazoans allows for the ER-to-Golgi transit of the membrane-tethered ATF-6 protein. In the Golgi apparatus, ATF-6 undergoes proteolytic cleavage within its membrane-spanning domain, and the soluble fragment subsequently travels to the nucleus as an active transcription factor ( Haze et al. 1999 ; Ye et al. 2000 ). XBP-1, ATF-4, and ATF-6 all activate separate but overlapping transcriptional programs that enable the cell to respond to changing conditions in the ER. Notably, the XBP-1 promoter is a target of ATF-6 activation ( Yoshida et al. 2001 ), reminiscent of the circuitry described here for yeast. Conceptually, therefore, HAC1 mRNA upregulation by the S-UPR pathway in yeast takes the place of XBP-1 upregulation by the ATF-6 fragment in metazoans. Moreover, ATF-6 and XBP-1 can heterodimerize ( Lee et al. 2002 ), reminiscent of the proposed collaboration of UMF and Hac1p. Thus, intriguing parallels between yeast and metazoans in the wiring that connects the elements of the UPR signaling circuit are beginning to come to light. These findings suggest a common strategy among all eukaryotic cells for responding to challenges to the secretory system. Maintaining separate ER surveillance pathways creates the potential for cells to integrate multiple signals that, in principle, could convey precise information regarding the nature of the imbalance to afford finely tailored corrective measures. In this view, the UPR operates as a homeostatic control circuit, in which such regulation ensures that components of the secretory apparatus are produced according to need. The challenge now at hand is to decipher the logic between the UPR inducing conditions and the transcriptional output to add physiological explanations to the complex regulation of the response that we observe experimentally. Materials and Methods Yeast strains. The WT strain W303–1A, the Δ ire1 strain CS165, and the Δ hac1 strain JC408 are as described previously ( Cox et al. 1993 ; Cox and Walter 1996 ). All sec strains used in this study were provided by Robert Fuller (University of Michigan, Ann Arbor, Michigan, United States) and are otherwise genotypically identical to W303. The HSF1 c strain was a kind gift of Hillary Nelson (University of Pennsylvania, Philadelphia, Pennsylvania, United States) and contains the R222A allele of HSF1 ( Bulman et al. 2001 ) replacing the chromosomal locus in a W303 background. Strains used in the experiments described in Figure 3 A were Δ hac1 transformed with pPW598 ( HAC1pro-HAC 1 , HA -tagged HAC1 [ Cox and Walter 1996 ] under its own promoter and with native HAC1 flanking sequences, in a pRS304 background) or with pPW599 (HAC1pro-GFP, the GFP ORF, driven by the HAC1 promoter [defined as the region starting at the mapped start site of HAC1 transcription ( Ruegsegger et al. 2001 ) and extending 500 bp upstream] and flanked by 5′ UTR and 3′ UTR sequences from ACT1). Strains used in experiments described in Figure 4 were HAC1pro-HAC1 and Δ hac1 (described above) and Δ hac1 transformed with pPW600 ( ADH1pro-HAC1, HA -tagged HAC1 with 5′ and 3′ UTR HAC1 sequence subcloned into the p414 ADH expression vector [ Mumberg et al. 1995 ] and transferred to a pRS304 backbone). In Figure 5 , HAC1pro HI (pPW601) was made by subjecting HAC1pro-HAC1 to QuikChange mutagenesis (Stratagene, La Jolla, California, United States) following the manufacturer's protocol, using oligonucleotides to remove the 15 bp at coordinates −338 to −323 (+1 representing the start site of transcription). Cell culture and plates Yeast cultures were grown in YPD medium (unless otherwise specified) at the indicated temperatures to midlog phase (OD 600 ≈ 0.5). For temperature shift experiments, cultures were transferred to a preheated 37 °C water bath shaking incubator. DTT (Roche, Basel, Switzerland) was added to a final concentration of 6 mM, and tunicamycin (Boehringer Mannheim, Indianapolis, Indiana, United States) was added to a final concentration of 1 μg/ml. For experiments involving inositol deprivation in liquid medium, yeast cells were grown in liquid complete synthetic medium described by Sherman (1991) , supplemented with myo-inositol (Sigma, St. Louis, Missouri, United States) to a final concentration of 100 μg/ml. Cells were then harvested by filtration, washed three times in prewarmed complete synthetic medium lacking inositol, and then filter-transferred to a flask containing prewarmed complete synthetic medium lacking inositol. For the experiment described in Figure 4 C, yeast strains were grown in YPD to midlog phase (OD 600 ≈ 0.5), transferred to a 96-well microtiter plate, and serially 5-fold diluted in fresh YPD. Using a liquid transfer prong (“frogging”) tool (Aladin Enterprises, San Francisco, California, United States), approximately 3 μl of all serial dilutions of all strains was simultaneously transferred to complete synthetic plates lacking inositol (described above), either in the absence or presence of 0.2 μg/ml tunicamycin. After approximately 2 d of incubation at 30 °C, plates were photographed using the Epi Chemi II Darkroom GelDoc system (UVP, Upland, California, United States). RNA analysis Isolation of total RNA from yeast cells was carried out with the modified hot-phenol extraction method described in Ruegsegger et al.(2001) . For Northern blot analysis, 10 μg of total RNA was separated on a 1.5% w/v agarose gel and transferred to a Duralon-UV nylon membrane (Stratagene), which was incubated with a probe directed against the 5′ exon of HAC1 . The mRNA abundance was quantitated using a PhosporImager (Molecular Dynamics, Sunnyvale, California, United States). The membranes were then stripped with two serial washes using 0.1% SDS at 65 °C for 60 min each and incubated with a probe directed against the 3′ exon of ACT1, and mRNA abundance was again quantitated. To control for the variable strength of Northern blot probes across multiple experiments, the relative HAC1 / ACT1 mRNA abundance ratio is always normalized to the untreated ( t = 0) sample. For the detection of other mRNAs, membranes were incubated with the additional relevant probes (GFP, SSA1) concurrent with the HAC1 probe. All data shown are an average of at least two independent experiments. PolyA+ mRNA was isolated from total RNA using the PolyATract system (Promega, Madison, Wisconsin, United States) according to the manufacturer's instructions. Microarray analysis, using yeast ORF arrays printed at the University of California, San Francisco, Core Center for Genomics and Proteomics ( http://derisilab.ucsf.edu/core/ , was performed as in Carroll et al. (2001) using protocols and reagents described at http://microarrays.org/ . All array data are the average of two independent experiments. For this study, we were obliged to evaluate UPR targets differently than in Travers et al. (2000) , as we considered HAC1 -independent responses, whereas the former study specifically isolated genes induced by unfolded proteins via Hac1p ( z- score ≥ 3.6 σ). Here, UPR targets were defined as those genes that met the following three criteria in a parallel set of microarray experiments using WT (W303) and Δ hac1 (JC408) strains. First, induction (log 2 of the fold change in gene expression) in WT cells treated with DTT must be at least one standard deviation greater than the mean ([ induction WT,DTT − μ WT,DTT ]/ σ WT,DTT ≥ 1). Second, induction in WT cells treated with tunicamycin must be at least one standard deviation greater than the mean ([ induction WT,tunicamycin − μ WT,tunicamycin ]/ σ WT,tunicamycin ≥ 1). Third, induction in Δ hac1 cells treated with DTT must be at least one standard deviation less than the induction in WT cells treated with DTT (or, more awkwardly, [([( induction WT,DTT − μ WT,DTT )/ σ WT,DTT ] − [( induction Δ hac1 ,DTT − μ Δ hac1 ,DTT )/ σ Δ hac1 ,DTT ]) − μ WT,DTT − Δ hac1 ,DTT ]/ σ WT,DTT − Δ hac1 ,DTT ≥ 1). Isolation and detection of protein from yeast cells Cells were collected by filtration, frozen in liquid nitrogen, and disrupted in 150 μl of 8 M urea/1% SDS by vortexing for 5 min at 4 °C in the presence of 150 μl of silica beads. The samples were then boiled for 5 min and the lysates cleared by centrifugation at 16,200 g for 5 min at room temperature. SDS-PAGE was performed on 20 μg of protein separated on NuPAGE 10% w/v SDS-polyacrylamide gels (Invitrogen, Carlsbad, California, United States), and Western blots were visualized using SuperSignal West Dura Extended Duration ECL Substrate (Pierce Biotechnology, Rockford, Illinois, United States) according to the instructions of the manufacturer. Hac1p was detected using a polyclonal antibody raised against the carboxy terminus (see Figure 4 ) or a monoclonal antibody raised against the HA epitope and directly coupled to horseradish peroxidase (see Figure 5 ) (Molecular Probes, Eugene, Oregon, United States), and Pgk1p was detected using a commercially available polyclonal antibody (Molecular Probes). Protein abundance was quantified using the Epi Chemi II Darkroom GelDoc system (UVP). Parallel experiments using serial protein dilutions were performed to confirm that the detected protein levels were within the linear range of the system. Transcription shut-off The yeast strain JC218 ( Sidrauski et al. 1996 ; rbp1–1 ) was grown in YPD at 23 °C to OD 600 ≈ 0.5 and then shifted to a 37 °C water bath, shaking at 250 RPM. To induce the UPR, DTT was added to a final concentration of 6 mM. Cells were harvested and total RNA isolated, at 20 min intervals, as described above. Supporting Information Accession Numbers The GenBank accession numbers of the sequences discussed in this paper are Hac1p (NP_ 116622), Ire1p (NP_011946), and tRNA ligase (NP_012448). Microarray data can be accessed at the Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (NCBI) database as platform number GPL999 and sample numbers GSM16978–GSM1984. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC509300.xml |
534095 | Genomic structure and cloning of two transcript isoforms of human Sp8 | Background The Specificity proteins (Sp) are a family of transcription factors that have three highly conserved zinc-fingers located towards the carboxy-terminal that bind GC-boxes and assist in the initiation of gene transcription. Human Sp1-7 genes have been characterized. Recently, the phenotype of Sp8 null mice has been described, being tailless and having severe truncation of both fore and hind limbs. They also have malformed brains with defective closure of the anterior and posterior neuropore during brain development. Results The human Sp8 gene is a three-exon gene that maps to 7p21.3, close to the related Sp4 gene. From an osteosarcoma cell line we cloned two transcript variants that use two different first exons and have a common second exon. One clone encodes a 508-residue protein, Sp8L (isoform 1) and the other a shorter 490-residue protein, Sp8S (isoform 2). These two isoforms are conserved being found also in mice and zebrafish. Analysis of the Sp8L protein sequence reveals an amino-terminal hydrophobic Sp-motif that is disrupted in Sp8S, a buttonhead box and three C 2 H 2 zinc-fingers. Sp8 mRNA expression was detected in a wide range of tissues at a low level, with the highest levels being found in brain. Treatment of the murine pluripotent cell line C3H10T1/2 with 100 ng/mL BMP-2 induced Sp8 mRNA after 24 hours. Conclusions There is conservation of the two Sp8 protein isoforms between primates, rodents and fish, suggesting that the isoforms have differing roles in gene regulation. Sp8 may play a role in chondrogenic/osteoblastic differentiation in addition to its role in brain and limb development. | Background We are interested in understanding how Specificity proteins (Sp) govern extracellular matrix deposition during bone formation. The Sp proteins are a family of transcription factors with three zinc-fingers that bind GC-boxes and assist the further binding of the multiprotein complex TFIID promoting the initiation of gene transcription [ 1 - 4 ]. GC-boxes have the consensus sequence GGGCGGG. Sp1 and Sp3 are associated with chondrocytic differentiation by regulating the alpha1(II) procollagen gene (COL2A1), a chondrocytic marker. Sp1 has been shown to be an activator and Sp3 a repressor of COL2A1 transcription [ 5 ]. Despite the ubiquitous expression of Sp3, Sp3 null mice suffer from specific skeletal defects associated with a delay in osteoblast differentiation, having reduced ossification and impaired bone formation and show a decrease in the expression of osteocalcin, an osteoblast specific gene [ 6 ]. Sp7, also known as Osterix, is required for osteoblast differentiation since in Sp7 null mice bone formation does not occur [ 7 ]. Previously, we cloned two isoforms of human Sp7 that are generated from distinct promoters upstream of exon 1a and exon 1b [ 8 ]. We identified another human Sp gene that is closely related to Sp7 and named it Sp8. Recently, Sp8 null mice have been described [ 9 , 10 ]. Note that the mouse mBtd gene [ 10 ] is Sp8 and not a specific homologue of the Drosophila gene buttonhead (Btd) [ 11 ]. However, Btd is a member of the Sp family [ 12 ]. Sp8 null mice have a well-defined phenotype, having severe truncation of all limbs and no tail. Their malformed brains have defective closure of the anterior and posterior neuropores that leads to exencephaly and spina bifida. Sp8 is required for proper maintenance and maturation of the apical ectodermal ridge, a signalling centre, which forms at the limb bud apex and governs the outgrowth of the limb during development. Here we describe the sequence of two isoforms of human Sp8. Results In our search for additional Sp transcription factors, we performed a TblastN search of the human genome with the sequence of Sp7 [ 8 ]. The most closely related sequences were found on genomic DNA BAC clones CTA-324D18 and CTB-86D3 that were in the draft stage (GenBank accession Nos. AC005251 and AC005060 Genome Sequencing Center, Washington University, USA). This new gene was named Sp8 since it is closely related to Sp7. Human Sp8 cDNAs Utilising the sequences of the two BAC clones, PCR primers were designed to amplify probable cDNAs. Two different cDNA clones were obtained by RT-PCR from the human MG-63 osteosarcoma cell line and sequenced (Fig. 1 ). Successful amplification and sequencing required methods designed to overcome problematic GC-rich regions. These transcript variants differed at their 5' termini. Clone 1 comprises exon 1a and 2, and together they encode the long protein isoform of Sp8 (Sp8L)(Fig. 1 )(GenBank accession No. AY167047). Clone 2 comprises the untranslated exon 1b and exon 2. Exon 1b contains an in-frame stop codon and no initiation methionine codon. Together they encode the short protein isoform (Sp8S)(Fig. 1 )(GenBank accession No. AY167048). The first ATG codon of Sp8L is in excellent sequence consensus for an initiation methionine having a guanosine residue at positions -3 and +4, and that of Sp8S is in good sequence consensus having a guanosine at -3 [ 13 ]. The ORF is GC rich (70%). Another Sp8 cDNA clone was isolated from the osteoblasts of a patient with osteoporosis. This has a glycine-165 deletion; having lost one of the five sequential GGC codons (nucleotides 569–583) (Fig. 1 ). Figure 1 Human Sp8 cDNA sequences encompassing the ORF and translation. (A) Sequence encoding exon 1a encoding the start of Sp8L protein isoform. (B) Sequence encoding the untranslated exon 1b encoding the Sp8S protein isoform. (C) Sequence encoding the common exon 2. The atg/methionine start codon for the long protein isoform is shown in bold and coloured red and that of the short protein isoform coloured pink. The ORF stop codon, tga , is shown in bold and coloured red and is indicated by a red asterisk. An in-frame stop codon, tag , in the 5'UTR of exon 1b is shown in bold and coloured red. The exon/exon boundaries were determined by comparison with the sequence of genomic DNA clone CTA-324D18. The nucleotides underlined and shown in bold type indicate the location of the exon/exon boundaries. The loss of a GGC (glycine) codon found in an additional Sp8 clone is shown in green and bolded. Human Sp8 gene After the completion of the human genome the location and structure of the human Sp8 gene was identified by mapping the sequences of Sp8L and Sp8S cDNA clones. The human Sp8 gene maps to 7p21.3 between the multidrug resistance protein (P-glycoprotein) and ribosomal protein L23 genes and is separated by 0.65 Mbp from the related Sp4 gene by seven genes encoding two ribosomal proteins, four putative genes of unknown function and a gene similar to argininosuccinate synthetase. The Sp8 and Sp4 genes are linked to the homeobox gene cluster HOX A, which is 5.6 Mbp away. The human Sp8 gene spans 4.6 kb and has three exons (Fig. 2 ). Exons 1a and 1b encode different transcript variants that are likely to be initiated from distinct promoters. The first two exons are separated by a 733 bp intron 1a, followed by a shorter, 152 bp intron 1b. All splice donor/acceptor sites contained consensus GT/AG dinucleotides. The gene possesses a CpG island upstream of exon 1a (438 bp, 63% GC). Computer analysis of the immediate human Sp8 promoter preceding exon 1a contains a number of RNA polymerase II promoter elements including a TATA-box, CCAAT-box and two GC-boxes that are conserved in rodents and fish. Figure 2 Gene structure of human Sp8. The human genomic sequences corresponding to the Sp8 cDNAs are located on clones CTA-324D18 and CTB-86D3. The location of a CpG island is shown in green. The first exon encoding the start of the Sp8L transcript is labelled 1a and that of Sp8S, 1b. The ORF is indicated by closed boxes. The sequences around the initiation methionine codons (underlined) are shown with those nucleotides corresponding to Kozak consensus sequence coloured red. The location of the TGA stop codon and a potential polyadenylation signal, AATAAA are also shown. The sizes of the exons and introns, in base pairs, are indicated. Comparison of the human Sp8 protein with those of other species The ORF of Sp8L clone 1 encodes a 508-residue protein with a 50,500 Da molecular mass and an isoelectric point 9.02. The first ATG codon of clone 2 corresponds to the second ATG of clone 1, thereby omitting the first 18 residues of the ORF, encoding a 490-residue protein with a 48,674 Da molecular mass and an isoelectric point 9.10. The human Sp8 protein belongs to the Sp/Krüppel like factor (KLF) family of proteins that are characterised by three Cys2-His2 zinc-fingers [ 14 ]. The zinc-fingers are located towards the COOH terminus and are involved in binding DNA in a sequence specific manner. However, the Sp proteins are distinguished from the KLF family by an amino-terminal hydrophobic domain called the Sp-motif with the consensus sequence PLALLA and a buttonhead box (BTD) with the consensus sequence CxCP(N/Y)C prior to the zinc finger domain. Database searches identified complete Sp8 genes and protein sequences from other vertebrate species. The chimpanzee gene is located on chromosome 7 (clone RP43-37F3); the mouse gene at 12F2 (BAC clone RP23-161L22); the rat gene at 6q33 on BAC clone CH230-1K24 and the zebrafish within BAC clone CH211-180P9. A comparison of the protein sequences shows that Sp8 is well conserved through vertebrate evolution (Fig. 3 ). The human and chimpanzee proteins are very similar with one conservative substitution and an additional glycine residue. The mouse Sp8 protein [ 9 , 10 , 15 ] has 97% identity and the zebrafish 79% identity to human Sp8. An amphibian Sp8 EST from Xenopus laevis was identified by our database search (accession No. BI313193). The mammalian Sp8 proteins differ from that of the zebrafish by the insertion of two poly-alanine and a poly-glycine region in the amino-half of the protein. Figure 3 Sequence comparison of the human Sp8 protein with those from other vertebrate species. The species are man, Homo sapiens , Hs; mouse, Mus musculus , Mm; chimpanzee, Pan troglodytes , Pt; rat Rattus norvegicus , Rn and Zebrafish, Danio rerio , Dr. The location of the exon/exon boundaries are underlined in black on the protein sequences. The methionine residues at the start of translation of Sp8S are underlined in red. The positions of the Sp, BTD-box and three zinc-finger domains found in most Sp protein family members are indicated. The conserved residues are shown by (*), strongly conserved residues by (:) and weakly conserved residues by (.). Stop codons are indicated by a hash. Residues are colour coded: basic, DE, blue; acidic,_KR, pink; polar, CGHNQSTY, green and hydrophobic, AFILMPVW, red. Database searches also showed that homologues of the vertebrate Sp8 genes are found in insects, but not in nematodes, yeasts and higher plants (Fig. 4 ). The fruit fly, Drosophila melanogaster , D-Sp1 gene is a homologue of Sp8 and not, as the name implies, Sp1 nor Sp4 as previously stated [ 16 ]. The red flour beetle, Tribolium castaneum , Sp8 protein [ 17 ] has 43% identity and 80% similarity to human Sp8. However, D-Sp1 differs from the vertebrate and beetle Sp8 proteins possessing a much longer, glutamine and histidine-rich carboxy-terminus. The possession of Sp, BTD and zinc-fingers domains in a protein is characteristic of Sp proteins in general; these domains in these three Sp8 proteins are very highly conserved. Outside these domains, a motif of unknown function is located towards the amino-terminal with the sequence GKGFHPWKKS and is unique to Sp8 proteins. Figure 4 Sequence comparison of the human Sp8 protein with those from other invertebrate species; red flour beetle, Tribolium castaneum , TcSp8 and fruit fly, Drosophila melanogaster Sp8, DmD-Sp1 proteins. The locations of the exon/exon boundaries are underlined in black on the protein sequences ( GE and S ). The methionine residues at the start of translation for the short protein isoforms are underlined in red ( M ). The positions of the Sp, BTD-box and three zinc-fingers motifs are shown. Stop codons are indicated by a hash. By comparison with other Sp proteins, the Sp8 protein can be divided up into 5 domains. The amino-terminal domain A (residues 1–53) is highly conserved in Sp8 proteins from other species. Two alpha-helices, residues 3–7 and 15–27, are predicted in this domain; whereas the rest of the protein appears to lack any secondary structure except for the zinc-finger domain. Domain A is also found in the Sp7 proteins. Domain B, residues 54–167, is a low-complexity region, having poly- serine, alanine and glycine stretches that are not present in fish Sp8 proteins. Domain C, residues 168–373, is similar to the same region of Sp7 and Sp6 that has been shown to be involved in transcriptional activation [ 7 ] and contains a basic region (residues 360–371) that probably contributes to a nuclear localization signal and a BTD-box, a highly conserved region that is present in all Sp family members. The BTD-box was first described as a 10-amino-acid region in the Drosophila zygotic gene, buttonhead [ 11 ]. Domain D, residues 374–456 contains three classical zinc finger structures (residues 374–456) of the Cys2-His2 type where the conserved cysteine residues in two short beta-sheets and the histidine residues in an alpha-helix tetrahedrally co-ordinate a zinc ion [ 18 ]. Between the fingers are five-residue linker sequences with the consensus sequence TGE+x. The negatively charged carboxy-terminal domain E, residues 457–508, contains a glycine-rich region (458–475). Tissue distribution of Sp8 mRNA The expression of Sp8 in human and murine adult tissues was examined by reverse transcriptase PCR using primers located in the 3'UTR. Human Sp8 mRNA expression was detected in a wide range of tissues at a low level, being found in heart, brain, placenta, liver, pancreas, prostate, testis, ovary and colon, but was below the limit of detection in lung, skeletal muscle, kidney, spleen, thymus, small intestine and peripheral blood leukocytes (Fig. 5A ). Murine Sp8 mRNA expression, relative to the expression of beta-actin mRNA, was detected in all tissues examined, with the highest levels being found in brain and prostate and the lowest in spleen (Fig. 5B ). Figure 5 Expression of Sp8 mRNA in human and mouse adult tissues determined by RT-PCR using primers located in the 3'UTR. A . Ethidium bromide stained agarose gel of PCR products from human tissues. The lanes are: skeletal muscle, muscle; small intestine, intestine; peripheral blood leukocytes, leukocyte; negative control, -c and 100 bp ladder, m. The expression of beta-actin mRNA, a housekeeping gene, in the same samples is shown below. B . RT real time PCR of murine Sp8 mRNA expression normalised to that of beta-actin. Expression of human Sp8 mRNA in osteoblast-like cells The expression of both transcript variants mRNAs encoding Sp8L and Sp8S in osteoblast-like cells was examined by reverse transcriptase PCR using primers specific for each isoform (Fig. 6 ). The Sp8L was most abundant in the osteosarcoma cell lines HOS and MG63 and was present in adult and craniofacial osteoblasts, but was below the limit of detection in foetal osteoblasts and chondrocytes. The Sp8S was most abundant also in the osteosarcoma cell lines HOS and MG63 and was present in craniofacial and foetal osteoblasts, but was below the limit of detection in adult osteoblasts and chondrocytes. Runx2, a transcription factor important for osteoblast differentiation, and beta-actin, a housekeeping gene, were expressed in all the osteoblast-like cells. Figure 6 Expression of Sp8 mRNAs in human osteoblast-like cells by reverse transcriptase PCR. Amplicons were run on an ethidium bromide stained agarose gel. The amplicons were: Sp8 long protein isoform, Sp8L; Sp8 short protein isoform, Sp8S; Runx2/Cbfa1, Runx2 and the housekeeping gene beta-actin. The sizes of the amplicons are indicated. The lanes are: primary adult osteoblasts, AO; HOS osteosarcoma cell line, HS; MG63 osteosarcoma cell line, MG; primary craniofacial osteoblasts, CF; primary foetal osteoblasts, FO; primary adult chondrocytes, CH; negative control, (-) and 100 bp markers, m. BMP-2 regulation of Sp8 mRNA expression in the murine cell line in C3H10T1/2 BMPs are potent secreted factors that promote osteoblast differentiation during development and bone remodelling. Since Sp8 was found in human osteosarcoma cell lines and is important in skeletal development [ 9 , 10 ] we examined the regulation of Sp8 by BMP-2 in the murine pluripotent embryonic cell line, C3H10T1/2 that can be induced to form a chondrogenic/osteoblastic phenotype by treatment with BMPs [ 19 - 21 ]. Sp8 mRNA expression, normalised to that of beta-actin, was measured by RT-real time PCR and was induced by 100 ng/mL BMP-2 after one day and remained upregulated at 20 days (Fig. 7 ). Data normalised to the housekeeping gene G3PDH gave similar results (data not shown) indicating that Sp8 was upregulated and that beta-actin was not down-regulated. Figure 7 Time course of effect of 100 ng/mL BMP-2 treatment on C3H10T1/2 cells. Murine Sp8 and beta-actin mRNA expression was determined by RT real time PCR. The relative levels of expression in treated (•) and untreated cells (◆) were examined after 0, 7, 14 and 20 days treatment. Significant increases in Sp8 expression were seen at 7, 14 and 20 days (p < 0.05). Discussion In the human genome three Sp gene pairs had been described previously; being maximally separated by 3.2 Mbp, transcribed in opposite directions, being orientated 5' to 5' manner and colocalized with a specific HOX gene clusters [ 8 ]. Sp3 and Sp5 are located on chromosome 2q31.1; Sp1 and Sp7 on 12q13.13 and Sp2-Sp6 on 17q21.3-q22. A search for a partner to Sp4 located on 7q21.3-q22 revealed a putative eighth Sp gene most similar to Sp7 suggesting that Sp8, like Sp7, may play a role also in skeletal development. Consequently, we used RT-PCR to amplify Sp8 cDNAs from a human osteosarcoma cell line and isolated two transcript variants clones that encode a full-length protein (long isoform) Sp8L and an amino-terminal truncated protein (short isoform) Sp8S, with 508 and 490 residues respectively. We found a glycine-165 deletion mutation in an additional cDNA clone isolated from a patient with osteoporosis. The length of this poly-glycine region is conserved in primates and rodents, but is absent in fish. We speculate that this Sp8 mutation, and other mutations yet to be discovered, may play a role in susceptibility to osteoporosis, and since Sp8 plays an important role in neuropore closure be a candidate gene for spina bifida [ 9 ]. The sequence of our Sp8L clone is similar to that of two other human clones, but the other two clones lack 170 and 128 bp of the low-complexity, B domain (GenBank accession Nos. BAB71297 and AAH38669). These deletions may be caused by incomplete reverse transcriptase reactions occurring during cDNA library synthesis, presumably because the high GC content (81%) generates strong secondary structure in the mRNA in this region. These deletion-carrying clones are unlikely to be generated by the presence of cryptic intron(s) since no suitable donor or acceptor splice sites are present in the genomic DNA sequence. One of the deletion-carrying clones, AAH38669, possesses a polymorphism, cac>cgc, at nucleotide 1263 that results in a His>Arg mutation at residue 448. His-448 is located in the third zinc- finger and is conserved in all Sp proteins and, by homology with Sp1, is likely to contact DNA [ 18 ]. This mutated protein would be expected to be deleterious, having reduced affinity for GC-box binding sites in promoter regions. We found that in man Sp8 has two transcript variants utilising two different first exons. Sequence analyses of mouse and zebrafish genomic and EST data support this gene structure for Sp8 in vertebrates and not a gene structure with two untranslated 5' exons that has been suggested for murine Sp8 [ 10 ]. Interestingly, the drosophila D-Sp1 gene, a Sp8 homologue, also has two transcript variants, encoding long and short protein isoforms (accession No. AE003448) [ 22 ]. Both Sp8 and Sp7 have a similar exon structure, being three-exon genes with two transcript variants with different first exons [ 8 ]. The long protein isoforms use translated first exons that encodes seven residues and the short protein isoforms use untranslated first exons; overall, the proteins have 39.5% identity and 65.9% similarity. Both the Sp8S and Sp7S proteins lack an 18-residue amino-terminus thereby disrupting a hydrophobic region termed the Sp domain that is conserved in other Sp proteins. The conservation of the two protein isoforms through evolution suggests that they have differing roles. Although the function of this hydrophobic region is unknown, other zinc-finger transcription proteins often have a conserved protein-protein interaction domain at their amino-terminus (e.g. BTB or kelch domains). This indicates that the Sp motif may also be involved in a protein-protein interaction and that the short protein isoforms do not have this protein interaction domain. Amino-terminal spliced variants expressed from separate promoters are a feature of other important transcription factors that regulate skeletal development. An oestrogen regulated protein in osteoblasts, KLF10 is another member of the Sp/Krüppel-like factor family that has two amino-terminal variant isoforms generated in a similar way to those of Sp8. These isoforms are named the TGFβ-inducible early gene (TIEG1) and the early growth response gene-alpha [ 23 - 25 ]. Runx2 has two major amino-terminal isoforms that exert different functions during the process of osteoblast differentiation; the Runx2-type-I isoform is widely expressed in osteoprogenitor cells and active osteoblasts, whereas the Runx2-type-II isoform is restricted to cells lining mineralised bones [ 26 ] and BMP-2 preferentially upregulates the Runx2-type-II isoform [ 27 ]. Sp8 does not initiate BMP-mediated signalling during apical ectodermal ridge formation, but may function downstream of the BMP receptor-1a in the signalling cascade [ 9 , 10 ]. The signalling events downstream of the BMP receptor that result in tissue-specific gene expression and skeletal development have been only partially elucidated. We found that Sp8 expression was induced by 100 ng/mL BMP-2 only after 24 hours in C3H10T1/2 cells. BMP-2 has been shown to induce several other transcription factors that promote differentiation such as Runx2/Cbfa1, Sp7/osterix and ZNF450 in addition to the negative regulator Id1 [ 7 , 28 - 30 ]. Induction of these genes in C3H10T1/2 cells occurs within 4 hours, preceding that of Sp8, suggesting that Sp8 is not directly regulated by BMP signalling and that it may be induced by one or more of the BMP-early-induced genes. The Sp8 gene has a wider phylogenetic distribution than Sp7, being found in coelomates, whereas Sp7 is limited to vertebrates. In view of the close similarities between human Sp8 and Sp7 in gene structure and amino acid sequence we speculate that they evolved from an ancestral Sp8 gene during a duplication of a Sp/Hox gene cluster [ 31 ]. The Sp8 gene has retained its function in regulating appendage/limb growth [ 9 , 10 , 12 , 17 ] and that Sp7 has subsequently evolved a novel function, namely, regulating cartilage/bone formation [ 7 ]. Conclusions In humans, there are two transcript variants of Sp8 that utilise different first exons encoding long and short protein isoforms. These two isoforms are conserved being found also in the zebrafish suggesting that they have distinct functions. Sp8 is upregulated by BMP-2 in the murine pluripotent cell line C3H10T1/2 and may play a role in mesenchymal differentiation. Methods Cell culture and RNA extraction Two cell lines derived from osteosarcomas, HOS and MG-63, were cultured at 37°C in 5% CO 2 using Dulbecco's Modified Eagle's Medium, containing 4 mM L-glutamine, 4500 mg glucose/L, 1500mg bicarbonate/L (Invitrogen, UK) with the addition of 10% foetal bovine serum, 10 μM ascorbic acid, 100 IU/mL penicillin and 50 μg/ml streptomycin. Primary human osteoblasts were isolated from trabecular bone of femoral heads taken during total hip arthroplasty and cultured as previously described [ 32 ]. Primary human craniofacial osteoblasts were obtained from paediatric skull and cultured as previously described [ 33 ]. Human primary foetal osteoblasts were obtained and cultured as previously described [ 34 ]. Human primary articular chondrocytes were obtained from isolated femoral heads and cultured as previously described [ 35 ]. The murine cell line C3H10T1/2 was cultured in Dulbecco's modified Eagle's medium supplemented with 50 U/mL penicillin, 50 μg/mL streptomycin (Invitrogen) and 10% new born bovine serum (Sigma) at 37°C under 5% CO 2 in a humidified incubator. Prior to BMP-2 treatment, cells were seeded at a density of approximately 40,000 cells/cm 2 , left for 24 hours, then treated with fresh media with or without 100 ng/mL human recombinant BMP-2 (Wyeth Corporation) and the media replaced every three days. Total RNA was extracted from cells using guanidine thiocyanate and treated with DNase-I to remove any contaminating genomic DNA (SV Total RNA isolation system, Promega, UK). The concentration and purity of eluted RNA was determined spectrophotometrically and the quality of the RNA was verified by non-denaturating agarose gel electrophoresis. For Sp8 cloning, total RNA was reverse transcribed with an oligo-dT primer using ThermoScript, an AMV RNase H- reverse transcriptase at an elevated temperature of 60°C (Invitrogen, UK). Molecular cloning of human Sp8 The human Sp7 protein sequence [ 8 ] was used to search the human genome sequence and the EST database to identify closely related genes. PCR primers designed from the sequence of genomic DNA clones CTA-324D18, CTB-86D3 and EST sequence from Image clone 2721342 (accession No. AW207154) were used to clone two transcript variants spliced isoforms of the human Sp8 cDNA by PCR from MG-63 osteosarcoma cell line cDNA. The PCR primers were: Sp8L forward, ATTGTATTGCACACCTCTAAAAAAAACA; Sp8S forward, GCGTGGTGCTTGCTCCC and common reverse, GCGTCACTCTAGGCCGTTG (Helena Biosciences, France). The cDNAs were amplified by PCR with an annealing temperature of 60°C using the Advantage-GC Advantage kit (Clontech, UK) with the addition of 0.5 M GC-Melt since the DNA sequence of Sp8 is GC-rich. PCR products were excised from agarose gels stained with ethidium bromide and eluted from the agarose using a DNA extraction kit (Qiagen, UK). The PCR products were cloned into the T-A vector pCR4-TOPO (Invitrogen, The Netherlands). Transformed colonies were picked and vectors containing inserts were extracted using the Wizard Plus SV minipreps DNA purification system (Promega, UK) and sequenced in both directions using ThermoFidelase 2 (Fidelity Systems Inc., USA). Tissue and cellular distribution of human Sp8 mRNA by reverse transcriptase PCR Human cDNA was analysed for the relative expression of the Sp8, Runx2/Cbfa1 and beta-actin mRNA by PCR. Sixteen adult tissue cDNAs (BD Clontech, UK) were generated from polyA + selected RNA and reverse transcribed using an oligo-dT primer. Total RNA from cell type cDNAs were reversed transcribed using random hexamer primers using an AMV RNase H- reverse transcriptase (ThermoScript, Invitrogen, UK) at 60°C. Approximately 4.0 ng of cDNA from each tissue, and cDNA derived from 50 ng of total RNA from each cell type were amplified by PCR using Taq Gold polymerase. Tissue master mixes were divided into gene specific mixes with the addition of PCR primers to a final concentration of 200 μM. The primers were: Sp8L forward primer, GCAACTTCACTTCTAGGG GAAGA (exon 1a/ 2 ); Sp8S forward primer, TGGGGGTGCCAG GAAGAAC (exon 1b/ 2 ) and a common reverse primer, AGCTGTCCGAGAGGGAGGA (exon 2), producing a 118 and 112 bp amplicons respectively; Sp8 3'UTR, TTAGTCCGGCCGTCAATTGT and TGGTATTTAAACTACAGCCTCGTCTGA, producing a 151 bp amplicon; Runx2, AGAAGAGCCAGGCAG GTGCT (exon 6/7) and TTCGTGGGTTGGAGAAGCG (exon 7), producing a 102 bp amplicon, as measure of the sum of all Runx2 isoforms, and beta-actin, GGCCACGGCTGCTTC and GTTGGCGTACAGGTCTTTGC, producing a 208 bp amplicon. The amplification conditions were; a 10 min hot start to activate the polymerase followed by 50 cycles of 95°C for 15 sec and 60°C for 1 min. Amplification specificity was confirmed by direct sequencing of the amplicons. Expression of murine Sp8 mRNA by real time PCR Utilising a CAS-1200 robotic precision liquid handling system, PCR was carried out on a Rotor Gene 3000 (Corbett Research, Australia) using a SYBR Green I double-stranded DNA binding dye assay. Copy DNA derived from 200 ng of total RNA from each sample was amplified by PCR using Taq Gold polymerase using PCR primers to a final concentration of 50 nM. The primers were: mouse Sp8 3'UTR, CCATTCAGCTCTGGCTAGGTCTT and GATTCCCGTTCGCAGAACTC producing a 67 bp amplicon. Beta-actin mRNA was used as a control gene as previously described [ 36 ]. The amplification conditions were; a 10 min hot start to activate the polymerase followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. The number of cycles required for the fluorescence to become significantly higher than background fluorescence (termed cycle threshold [C t ]) was used as a measure of abundance. A comparative C t method was used to determine gene expression. Expression levels in each cDNA sample were normalised to the expression levels of the control gene (ΔC t ). The ratios of Sp8 mRNA/control gene RNA from each cDNA were standardised to that of the untreated cells on day 0 that was taken as 100% (ΔΔC t ). The formula E -ΔΔCt was used to calculate relative expression levels where E is the efficiency of the PCR per cycle. Statistically significant changes in gene expression were determined using the t-test on data from three replicate experiments. The amplification specificity was confirmed by melting curve analysis and agarose gel electrophoresis. Authors' contributions AJE conceived of the study, and participated in its design and coordination. MAM, JEG and AJE carried out the cell culture work, MAM and AJE carried out the expression studies and MAM and AJE carried out the gene cloning. All authors participated in writing the manuscript, read, and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC534095.xml |
554761 | Acute effects of cigarette smoking on inflammation in healthy intermittent smokers | Background Chronic smoking is the main risk factor for chronic obstructive pulmonary disease. Knowledge on the response to the initial smoke exposures might enhance the understanding of changes due to chronic smoking, since repetitive acute smoke effects may cumulate and lead to irreversible lung damage. Methods We investigated acute effects of smoking on inflammation in 16 healthy intermittent smokers in an open randomised cross-over study. We compared effects of smoking of two cigarettes on inflammatory markers in exhaled air, induced sputum, blood and urine at 0, 1, 3, 6, 12, 24, 48, 96 and 192 hours and outcomes without smoking. All sputum and blood parameters were log transformed and analysed using a linear mixed effect model. Results Significant findings were: Smoking increased exhaled carbon monoxide between 0 and 1 hour, and induced a greater decrease in blood eosinophils and sputum lymphocytes between 0 and 3 hours compared to non-smoking. Compared to non-smoking, smoking induced a greater interleukin-8 release from stimulated blood cells between 0 and 3 hours, and a greater increase in sputum lymphocytes and neutrophils between 3 and 12 hours. Conclusion We conclude that besides an increase in inflammation, as known from chronic smoking, there is also a suppressive effect of smoking two cigarettes on particular inflammatory parameters. | Background Chronic obstructive pulmonary disease (COPD) is one of the leading causes of morbidity and mortality world-wide, and its prevalence is still rising [ 1 ]. In order to develop strategies for its prevention and treatment, it is important to understand the underlying pathophysiologic mechanisms of this disease. Since chronic smoking is the main risk factor to develop COPD most studies in this field have been carried out in chronic (ex)smokers with or without COPD. It is also important to study the initial response to cigarette smoke to better understand the effects of chronic smoking, since repetitive acute smoke effects may cumulate and ultimately lead to irreversible lung damage associated with COPD. In addition, to appropriately evaluate the impact of chronic smoking, the "background" effects of acute smoking should be determined. Until now, only a few studies have investigated acute effects of smoking in humans [ 2 ]. Unfortunately, these studies investigated only a small number of time points after smoking, hence little information is available on the time course and resolution of smoking induced changes. Furthermore, all studies assessed acute effects of smoking in chronic smokers who refrained from smoking for maximally 24 hours. It is unknown whether this is sufficiently long to exclude the influence of previous smoking on the acute smoke results. Finally, no study so far investigated acute smoke effects in sputum. In the present study we investigated acute effects of smoking of two cigarettes by healthy intermittent smokers who refrained from smoking nine days before the study period. In this way, temporary effects on the airways due to chronic smoking will probably not affect the acute response to smoke. We assessed the time effects of cigarette smoking on both induction and resolution of the inflammatory response in exhaled air, induced sputum, blood and urine. We hypothesised that smoking of two cigarettes would induce an increase in inflammatory cells and markers within a limited time interval. Methods Design of the study We performed a randomised, two-period cross-over, pilot study. Subjects were randomised into smoking two cigarettes or no smoking. Subjects refrained from smoking during nine days before each study period, verified by exhaled carbon monoxide (CO < 6 ppm) and urinary cotinine (< 25 ng/ml). The time interval between the two study periods varied between 9 to 20 days. Measurements of exhaled CO, exhaled Nitric Oxide (NO), blood sampling and Forced Expiratory Volume in 1 second (FEV 1 ) were performed immediately before (baseline) and 1, 3, 6, 12, 24, 48, 96 and 192 hours after smoking and at the same time points in the no smoking period. Sputum was induced at 3, 6, 12, 24, 48, 96, 192 hours after smoking and no smoking. All subjects smoked two cigarettes from the same brand within 30 minutes and were encouraged to inhale deeply (Caballero unfiltered cigarettes, tar 12 mg, nicotine 1.0 mg, commercially obtained, no gifts). Adequacy of smoke inhalation was verified by the investigator. The working groups sputum induction from the ERS stated recently that sputum inductions should not be repeated within 48 hours to avoid carry over effects [ 3 ]. Taking this into account, we used a cross-over design (including no smoking) in this study to correct for this carry over effect. We have analysed the results of the control arm as a separate study in order to investigate the induction and resolution of the inflammatory response generated by repeated sputum inductions [ 4 ]. Subjects Sixteen healthy intermittent smokers were recruited by advertisements in the local newspaper. Intermittent smoking was defined as smoking more than one cigarette a month, but not daily, during the last 6 months. We chose to investigate intermittent smokers because they are able to refrain from smoking for a certain time period (in contrast to most current smokers) and they are used to inhale smoke (in contrast to non-smokers). Included were subjects older than 40 years, with normal lung function (prebronchodilator FEV 1 /IVC [Inspiratory Vital Capacity] > 89% of predicted for women and > 88% of predicted for men [ 5 ] and a prebronchodilator FEV 1 > 1 litre). Excluded were subjects with: 1) a history of asthma, allergic rhinitis, or allergic eczema; 2) atopy, confirmed by a positive skin prick test; 3) any current respiratory disease, symptoms of cough or sputum production; 4) a respiratory tract infection within the preceding 8 weeks or a nasal infection within the preceding 4 weeks; 5) treatment with glucocorticosteroids within the preceding 8 weeks; 6) use of aspirin, NSAIDs, paracetamol or antihistamines within the preceding 4 weeks. Subjects were asked to avoid places with high environmental tobacco smoke exposure during the study periods. The study was approved by the medical ethics committee of the University Medical Center Groningen, the Netherlands. Written informed consent was obtained from all subjects. Pulmonary function, exhaled NO and CO FEV 1 and IVC were measured according to the guidelines of the European Respiratory Society [ 5 ], using a pneumotachograph (Jaeger, Wurzberg, Germany). Exhaled NO levels were determined according to the guidelines of the American Thoracic Society [ 6 ], exhaling with a flow of 100 ml/sec against a resistance between 5 and 20 cm H 2 O, using a chemiluminescence analyser (Ecophysics CLD 700 AL). Exhaled CO levels were measured using an infrared CO analyser (UNOR 6N, Maihak AG, Hamburg, Germany) [ 7 ]. Blood analyses Blood differential cell counts were analysed automatically with a haematology flow cytometer (Coulter-STKS, Beckman Coulter, Miami, USA). Flow cytometric analysis was performed on blood cells using peridinin chlorophyll protein (PerCP) labelled anti-human leukocyte antigen (HLA)-DR, phycoerythrin (PE) labelled anti-CD11b, allophycocyanin (APC) labelled CD14 and fluorescein-isothiocyanate (FITC) labelled CD63 monoclonal antibodies (Becton Dickinson, Franklin Lakes, NJ USA). HLA-DR, CD63 and CD11b are activation markers for respectively monocytes and granulocytes. CD14 is used to discern between monocytes and granulocytes. Functional assays were performed on unstimulated and lipopolysacharide (LPS, 1 ng/ml, BioWhittaker, Walkerville, USA) stimulated blood cells, measuring tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-8 and IL-10 by ELISA (Sanquin, Amsterdam, the Netherlands). Sputum induction and processing Sputum was induced according to a modified standard technique [ 8 ], using 4.5% hypertonic saline. Whole sputum was processed within 120 minutes according to the modified method of Rutgers and colleagues [ 8 ]. The cell-free supernatant was collected and stored in aliquots at -80°C pending analysis of soluble mediators. Sputum analyses Flow cytometric analysis was performed on sputum cells using PerCP labelled anti-HLA-Dr, PE labelled anti-CD11b, APC labelled CD14 and FITC labelled CD63 monoclonal antibodies (Becton Dickinson, Franklin Lakes, NJ USA). Immunocytology was performed to quantify the percentage of inducible NO synthase (iNOS) positive macrophages. Cytospins were double stained with a monoclonal antibody against CD68 (IgG1 isotype, Dako, Glostrup, Danmark) as a marker for macrophages and rabbit polyclonal antiserum against iNOS (Transduction Laboratories, Lexington, KY, USA). The following soluble mediators were measured in sputum supernatant. NO 2 - /NO 3 - was measured using the Griess reaction, eosinophilic cationic protein (ECP) using the fluorenzyme immunoassay UniCAP ECP (Pharmacia, Uppsala, Sweden). IL-8 and Leukotriene B4 (LTB4) were measured by a commercial ELISA (IL-8: Sanquin, Amsterdam, the Netherlands, LTB4: Amersham Biosciences, UK). Matrix metalloproteinase-9 (MMP-9) was measured by gelatine zymography [ 9 ], and tissue inhibitor of metalloproteinase-1 (TIMP-1) by ELISA (R&D, Abingdon, UK). Neutrophil elastase (NE) activity was measured by chromogenic substrate assay (N-methoxysuccinyl-ala-ala-pro-val-p-nitoanilide, Sigma, UK)] [ 10 ]. Urinary measurements Before inhalation of smoke (or control), a urine portion was collected to measure urine cotinine. Cotinine was measured by gaschromatography-mass-spectrometry (Pharmacy Department, Groningen, the Netherlands). Furthermore, urine was collected over 24 hours in five consecutive fractions: 0–1 hour, 1–3 hours, 3–6 hours, 6–12 hours and 12–24 hours from all subjects to assess leukotriene E4 levels (ELISA, Amersham Biosciences, UK). Statistical analyses Since the start and duration of the acute effects of smoking of two cigarettes on our parameters were unknown, time series of all variables were plotted. Based on visual inspection of these plots the time intervals to be analysed were selected. The slopes of parameters were estimated using linear mixed effect models [ 11 ] by including the variables time (hours), smoking (yes or no) and their interaction. For the sputum parameters no baseline values were present, therefore time point 192 hours was used as baseline value. After log-transformation of all blood and sputum variables, the residuals of the models were normally distributed. All analyses were performed in S-plus 2000 (Insightful Corporation, Seattle, WA, USA). A p value <0.05 was considered statistically significant. Results Subjects Clinical characteristics of the 16 subjects are listed in table 1 . Fifteen subjects successfully refrained from smoking for nine days. One subject smoked one cigarette five days before the start of the study, but the urinary cotinine and exhaled CO levels were within the required range. The analyses are performed on data from all 16 subjects. Table 1 Subject characteristics (healthy intermittent smokers). Sex, male/female 12/4 Age, years 49 (39–71) Smoked pack years 4 (0–40) Smoked cigarettes per month 14 (1–60) FEV 1 , % predicted 119 (68–144) FEV 1 / IVC, % 77.6 (68.1–87.0) Values expressed as medians (ranges). FEV 1 : Forced Expiratory Volume in 1 second, IVC: Inspiratory Vital Capacity. Exhaled NO and CO and FEV 1 Exhaled CO increased significantly more with smoking than without between 0 and 1 hour and subsequently decreased significantly more between 1 and 12 hours (table 2 , figure 1 ). Smoking had no significant effect on exhaled NO (data not shown) or FEV 1 (table 2 ). Figure 1 Time course of smoking of two cigarettes on exhaled carbon monoxide (CO). Black circles represent the values after smoking two cigarettes and grey circles represent the values of the control period. Table 2 Linear mixed effect models: CO and FEV 1 Independent variable Time interval B 95% CI P value CO, ppm 0–1 hour 3.61 2.67–4.54 <0.0001 1–12 hours -0.29 -0.38 – -0.21 <0.0001 FEV 1 , L/sec 0–1 hour 0.06 -0.07–0.20 0.38 The time intervals of the above parameters were selected based on visual inspection of the plots. The slopes of the parameters were estimated using linear mixed effect models [11] by including the variables time (hours), smoking (yes or no) and their interaction. B: regression coefficient for the variable time (for further information see the method section). CO: carbon monoxide; FEV 1 : Forced Expiratory Volume in 1 second. Blood The number of blood eosinophils decreased more with smoking than without between 0 and 3 hours (table 3 and figure 2 ). Smoking had no significant effect on the number of other blood cells (table 3 and additional file 1 , table 1 ). IL-8 release from LPS stimulated blood cells increased more with smoking than without between 0 and 3 hours (table 3 ). Smoking had no significant effect on TNF-α, IL-10 and IL-1β release compared with no smoking ( additional file 1 , table 2 ). There was no significant difference in the expression of CD11b, CD63 and HLA-DR on CD14 high and CD14 low cells between smoking and no smoking (data not shown). Table 3 Linear mixed effect models: blood cells, IL-8 and TNF-α Independent variable Time interval B 95% CI P value Log (leucocytes, 10 9 /L) 0–12 hours 0.00 -0.01–0.01 0.44 Log (neutrophils, 10 9 /L) 0–12 hours 0.00 -0.01–0.02 0.58 Log (monocytes, 10 9 /L) 0–1 hour 0.04 -0.10–0.19 0.55 Log (eosinophils, 10 9 /L) 0– 3 hours -0.11 -0.18 – -0.03 0.01 Log (lymphocytes, 10 9 /L) 0–12 hours -0.00 -0.01–0.01 0.65 Log (IL-8, pg/ml)* 0–3 hours 0.09 0.04–0.14 0.001 Log (TNF-α, pg/ml)** 0–3 hours 0.02 -0.08–0.12 0.75 The time intervals of the above parameters were selected based on visual inspection of the plots. The slopes of the parameters were estimated using linear mixed effect models [11] by including the variables time (hours), smoking (yes or no) and their interaction. B: regression coefficient for the variable time (for further information see the method section). * Release from whole blood cells after lipopolysacharide (LPS) stimulation. ** Spontaneous release from whole blood cells IL: interleukin, TNF-α: tumor necrosis factor-α Figure 2 Time course of smoking of two cigarettes on blood eosinophils. Black circles represent the values after smoking two cigarettes and grey circles represent the values of the control period. Sputum The total number and percentage of sputum cells within the first 24 hours after smoking and no smoking are shown in table 4 . The number of neutrophils increased significantly more with smoking than without between 3 and 12 hours (table 5 , figure 3 ). The number of sputum lymphocytes decreased more with smoking than without between 0 and 3 hours (table 5 , figure 4 ). Subsequently, however, the percentage and number of sputum lymphocytes increased more with smoking than without between 3 and 12 hours (table 5 , figure 4 ). Smoking had no significant effect on the percentage and number of sputum eosinophils (table 5 , figure 5 ) and macrophages (table 5 ). Smoking had also no significant effect on the levels of inflammatory mediators in sputum ( additional file 1 , table 3 ) and the expression of CD11b, CD63 and HLA-DR on CD14 high and CD14 low cells and the number of iNOS positive macrophages (data not shown). Table 4 Inflammatory cells in sputum after smoking and no smoking. Baseline (192 hours) 3 hours 6 hours 12 hours 24 hours SMOKING Sputum cells, 10 6 /ml 1.8 (0.1–16.3) 2.0 (0.3–6.6) 2.4 (0.1–7.3) 2.4 (0.5–6.6) 2.6 (0.0–9.0) Neutrophils, % 56.9 (22.0–97.3) 56.4 (4.0–96.0) 83.2 (13.7–97.3) 77.5 (32.2–98.3) 67.3 (39.0–84.3) Macrophages, % 37.9 (2.5–74.5) 42.2 (3.7–84.8) 13.8 (2.5–68.5) 16.8 (1.7–61.7) 27.4 (14.8–57.7) Eosinophils, % 0.1 (0.0–6.2) 0.5 (0.0–5.2) 0.0 (0.0–0.3) 1.1 (0.0–8.3) 1.0 (0.0–5.2) Lymphocytes, % 1.1 (0.0–3.8) 0.4 (0.0–1.7) 1.0 (0.0–2.0) 1.2 (0.0–7.3) 0.4 (0.0–1.8) NO SMOKING Sputum cells, 10 6 /ml 2.8 (0.8–23.8) 3.1 (0.1–20.4) 2.0 (0.7–7.9) 2.1 (0.4–6.2) 2.1 (0.6–9.5) Neutrophils, % 50.9 (20.3–84.8) 58.9 (31.8–94.2) 73.2 (22.8–94.7) 83.2 (26.7–98.3) 64.5 (29.0–80.3) Macrophages, % 46.9 (15.0–77.7) 38.5 (4.2–64.0) 20.8 (4.5–71.2) 10.3 (1.7–67.8) 28.7 (16.0–66.5) Eosinophils, % 0.2 (0.0–3.2) 0.3 (0.0–1.2) 0.2 (0.0–4.2) 1.7 (0.0–15.5) 2.2 (0.5–12.5) Lymphocytes, % 0.7 (0.0–4.0) 0.9 (0.0–2.8) 0.4 (0.0–3.5) 0.2 (0.0–3.7) 0.9 (0.0–1.5) Values are expressed as medians (ranges). Table 5 Linear mixed effect models of sputum cells Independent variable Time interval B 95% CI P value Log (total cells, 10 6 /ml) 0–3 hours 0.075 -0.15–0.30 0.52 Log (neutrophils, % and 10 6 /ml) % 0–3 hours -0.17 -0.34 – -0.00 0.07 number 0–3 hours -0.22 -0.49–0.30 0.11 % 3–12 hours 0.03 -0.02–0.07 0.27 number 3–12 hours 0.13 0.04–0.22 0.007 Log (macrophages, % and 10 6 /ml) % 0–3 hours 0.12 -0.01–0.25 0.08 number 0–3 hours 0.13 -0.11–0.36 0.31 Log (eosinophils, % and 10 6 /ml) % 0–6 hours -0.22 -0.71–028 0.39 number 0–6 hours -0.12 -0.30–0.06 0.20 % 3–6 hours -0.84 -3.63–1.96 0.57 number 3–6 hours -0.31 -1.35–0.73 0.57 Log (lymphocytes, % and 10 6 /ml) % 0–3 hours -0.28 -0.59–0.02 0.08 number 0–3 hours -0.26 -0.42–-0.11 0.004 % 3–12 hours 0.19 0.06–0.31 0.006 number 3–12 hours 0.23 0.09–0.36 0.002 The time intervals of the above parameters were selected based on visual inspection of the plots. The slopes of the parameters were estimated using linear mixed effect models [11] by including the variables time (hours), smoking (yes or no) and their interaction. B: regression coefficient for the variable time (for further information see the method section). Figure 3 Time course of smoking of two cigarettes on sputum neutrophils. Black circles represent the values after smoking two cigarettes and grey circles represent the values of the control period. Figure 4 Time course of smoking of two cigarettes on sputum lymphocytes. Black circles represent the values after smoking two cigarettes and grey circles represent the values of the control period. Figure 5 Time course of smoking of two cigarettes on sputum eosinophils. Black circles represent the values after smoking two cigarettes and grey circles represent the values of the control period. Urine Smoking had no significant effect on leukotriene E4 levels in urine compared to no smoking (data not shown). Discussion In order to better understand the effects of chronic smoking, it is important to study the initial (acute) response to cigarette smoke, since repetitive acute smoke effects may cumulate and ultimately lead to irreversible damage. We therefore investigated the acute effects of cigarette smoking on both induction and resolution of the inflammatory response in healthy intermittent smokers. This study shows that smoking of two cigarettes acutely suppresses blood eosinophils. Furthermore, smoking induces a biphasic response in sputum lymphocytes, after an initial smoke-related suppression, the cells increase more with smoking than without. Finally, smoking increases sputum neutrophils and the release of IL-8 from whole blood cells. A remarkable finding in our study is that smoking of two cigarettes decreases eosinophils in blood. Three other studies have reported similar results: eosinophils decreased in blood from healthy female smokers within two hours after smoking 12 cigarettes [ 12 ], in lung tissue of rats within 6 hours after smoke exposure [ 13 ], and in lung lavage fluid of ovalbumin sensitised mice after 3 weeks smoke exposure [ 14 ]. A decrease in eosinophils may be due to a direct (apoptotic) effect by toxic substances in cigarette smoke [ 15 ], or to anti-inflammatory substances in cigarette smoke, like CO [ 16 , 17 ]. Smoking did not show a significant suppressive effect on sputum eosinophils in our study, although the figures show that sputum eosinophils are decreasing more from 3 hours onwards with smoking than without. The reason for this is probably a lack of study power, due to the lower number of successful measurements in sputum than in blood, or due to the low baseline levels of sputum eosinophils in our healthy non-atopic subjects. One has to realise that the decrease in eosinophils in blood in our study is significant but relatively small. This study is the first to report that sputum can be used to study acute smoke effects. The number of sputum neutrophils increased between 3 and 12 hours after smoking. In line with this, we demonstrated a higher release of IL-8 by LPS stimulated blood cells after smoking, which may have contributed to increased neutrophil chemotaxis. The rise in neutrophils is in line with two studies on acute effects of smoking in humans, showing increased neutrophils in bronchoalveolar lavage fluid 1 hour after smoking [ 18 ] and increased neutrophil retention in the lung during smoke exposure [ 19 ]. The fast increase in neutrophils in sputum might result from detachment of neutrophils from the pulmonary vascular endothelium (the so-called marginated pool) [ 20 ] or from recruitment from the bone marrow [ 21 , 22 ]. Smoking also shortly suppressed the number of lymphocytes in sputum. Thereafter sputum lymphocytes increased more with smoking than without. The initial decrease might result from increased adherence of lymphocytes in the lung tissue due to the fast upregulation of adhesion molecules after smoking [ 23 ] or may also be caused by the suppressive CO as mentioned in the prior paragraph [ 16 ]. The subsequent increase in sputum lymphocytes may reflect the outwash of lymphocytes from the tissue into the sputum, which can be regarded as the waste bin of lung inflammatory cells. Smoking did not affect all inflammatory markers we investigated. A few factors may contribute to this lack of response. First, the number of subjects and the number of cigarettes (n = 2) might have been too low. Second, we may have included a heterogeneous group of subjects regarding their response to cigarette smoke. We know that approximately 80% of all smokers never develop COPD. Therefore it is conceivable that a part of our healthy smokers does not respond to cigarette smoking. Third, we included subjects with a broad range in current and past smoking. Fourth, sputum may reflect only a part of the acute inflammatory changes of the airway wall [ 8 ]. It would be interesting to study the acute effects of smoking on lung tissue. Finally, CO in cigarette smoke may have dampened the inflammatory response, especially in the early phase. After continuous smoking the damaging and irritating effects may prevail, giving rise to more pronounced inflammation. Studying the acute effects of smoking in intermittent healthy smokers has both advantages and disadvantages. We choose the presented model for a number of reasons. First, intermittent smokers can refrain from smoking for three weeks in contrast to most current smokers. Second, intermittent smokers have a normal lung function (in contrast to COPD), and likely (nearly) no structural airway changes, which may affect a normal response to cigarette smoke. Third, we assumed that detecting an acute inflammatory response to cigarette smoking after an abstinence period of 9 days would be easier than detecting a response on top of chronic smoke exposure. Finally, intermittent smokers are used to inhale cigarette smoke (in contrast to non-smokers). We realise that our model has the disadvantage that the results of our study cannot easily be extrapolated to the chronic effects of smoking or COPD development. Nevertheless, when comparing the airway inflammation of our subjects with that of smoking COPD patients, both show increased levels of neutrophils, lymphocytes and IL-8 in sputum. However, in COPD patients after quitting smoking lymphocytes and neutrophils do not normalise [ 24 ], in contrast to the short-lived acute effects of smoking in this study. This suggests that not smoke but structural changes in the airways are responsible for the ongoing inflammation in COPD. Despite above limitations we think that knowledge on both the acute and chronic effects of smoking will help to better understand the mechanisms of cigarette smoke induced inflammation, which may underlie the development of COPD. Conclusion We conclude that besides an increase in inflammation, as known from chronic smoking, there is also a suppressive effect of smoking of two cigarettes on particular inflammatory parameters. Although this seems beneficial, it may disturb physiologic responses, like repair processes, in which inflammatory cells play a role. Competing interests The author(s) declare that they have no competing interests. Authors' contributions HV: Participated in the design of the study, performed the study and drafted the manuscript. DP: Conceived the study, participated in the design and co-ordination of the study and helped draft the manuscript. WT: Participated in the design of the study and helped draft the manuscript. MH: Participated in the design of the study, co-ordinated the FACS analyses and helped draft the manuscript. BW: Performed some of the laboratory analyses, participated in the design of the study and helped draft the manuscript. HB: Performed statistical analyses and helped draft the manuscript. JV: Performed statistical analyses and helped draft the manuscript. DR: Performed and co-ordinated most laboratory analyses and helped draft the manuscript. HK: Co-ordinated laboratory analyses, participated in the design of the study and helped draft the manuscript. NH: Conceived the study, participated in the design and co-ordination of the study and helped draft the manuscript. Supplementary Material Additional File 1 Table 1 . Number of blood cells (10 9 /L) after smoking and no smoking. Table 2 . Release of IL-1β, IL-10, IL-8 and TNF-α from blood cells after smoking and no smoking. Table 3 . Inflammatory mediators in sputum after smoking and no smoking. Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC554761.xml |
544572 | Colorectal cancer screening among African American church members: A qualitative and quantitative study of patient-provider communication | Background A healthcare provider's recommendation to undergo screening has been shown to be one of the strongest predictors of completing a colorectal cancer (CRC) screening test. We sought to determine the relationship between the general quality of self-rated patient-provider communication and the completion of CRC screening. Methods A formative study using qualitative data from focus groups and quantitative data from a cross-sectional survey of church members about the quality of their communication with their healthcare provider, their CRC risk knowledge, and whether they had completed CRC screening tests. Focus group participants were a convenience sample of African American church members. Participants for the survey were recruited by telephone from membership lists of 12 African American churches located in rural counties of North Carolina to participate in the WATCH (Wellness for African Americans Through Churches) Project. Results Focus Groups. Six focus groups (n = 45) were conducted prior to the baseline survey. Discussions focused on CRC knowledge, and perceived barriers/motivators to CRC screening. A theme that emerged during each groups' discussion about CRC screening was the quality of the participants' communication with their health care provider. Survey . Among the 397 participants over age 50, 31% reported CRC screening within the recommended guidelines. Participants who self-rated their communication as good were more likely to have been screened (36%) within the recommended guidelines than were participants with poor communication (17%) (OR = 2.8, 95% CI 1.2, 6.4; p = 0.013). Participants who had adequate CRC knowledge completed CRC screening at a higher rate than those with inadequate knowledge (p = 0.011). The percentage of participants with CRC screening in the recommended guidelines, stratified by communication and knowledge group were: 42% for good communication/adequate knowledge; 27% for good communication/inadequate knowledge; 29% for poor communication/adequate knowledge; and 5% for poor communication/inadequate knowledge. Conclusions Participants who rated their patient-provider communication as good were more likely to have completed CRC screening tests than those reporting poor communication. Among participants reporting good communication, knowledge about colorectal cancer was also associated with test completion. Interventions to improve patient-provider communication may be important to increase low rates of CRC screening test completion among African Americans. | Background Colorectal cancer (CRC) screening among average risk adults 50 years and older can decrease the incidence and mortality rates for CRC [ 1 - 7 ]. National policy-making expert organizations recognize and support this evidence by recommending a variety of CRC screening testing strategies [ 8 - 13 ]. However, the 2001 Behavioral Risk Factor Surveillance System (BRFSS) reported only 23.5% of adults age ≥ 50 had fecal occult blood test (FOBT) within the last year, 38.7% had sigmoidoscopy or colonoscopy within the past 5 years, and only 53.1% of adults had been screened with either test within the recommended time periods. Data specifically from North Carolina are similar: 30% have had FOBT within 1 year, 31% have had sigmoidoscopy within 5 years, and 45% have had either test within the recommended time periods [ 14 ]. A healthcare provider's recommendation to undergo screening has been shown to be one of the strongest predictors of completing a CRC screening test [ 15 - 17 ], and has also been shown to be strongly correlated with initial and repeat mammography [ 18 , 19 ], and the completion of Pap smears[ 20 ]. Such recommendations may be more likely to occur when patients and providers communicate well, but previous research has not directly explored the relationship between patients' perceptions about the quality of patient-provider communication and the use of CRC screening. We sought to examine the association between perceived communication and CRC s screening in a sample of African-American church members in rural North Carolina. Methods Data for this study were obtained as part of a larger study, the WATCH (Wellness for African Americans Through Churches) Project . The WATCH Project was a church-based colorectal cancer prevention study designed to increase fruit and vegetable consumption, reduce fat intake, increase moderate physical activity, and increase CRC screening among church members. In this study of patient-provider communication and CRC screening, we used data from two different sources: 1) focus groups of African American men and women, and 2) surveys of church members. The Institutional Review Board from the University of North Carolina at Chapel Hill approved this study. Focus groups As part of the formative data collection in early 1998, we conducted six focus groups of African American adults. There were three focus groups of African American adult men and three of African American adult women. The focus groups comprised of a convenience sample drawn from members of African American churches located in central North Carolina. The focus groups were conducted in the churches, lasted approximately 60 minutes, were tape recorded, transcribed, and reviewed for accuracy. A $15 incentive and refreshments were provided for the participants. The focus groups were led by sex-matched African American moderators and explored issues associated with colon cancer, CRC screening (barriers and motivators), nutrition, and physical activity. The information from the focus groups was used to develop the questionnaires and to help guide the intervention strategies used in the culturally appropriate colon cancer prevention program. Survey The second source of participants for the survey study was the baseline intervention sample for the larger WATCH project. The telephone survey collected information focused on general health status, nutrition, physical activity, patient-provider communication, CRC risk knowledge and screening behaviors. The baseline survey was completed, on average, in approximately 40 minutes and was administered by trained interviewers prior to the intervention. The baseline telephone surveys were conducted between October 1998 and October 1999. In this study, a history of CRC screening was defined as the self-report of undergoing a fecal occult blood test (FOBT), flexible sigmoidoscopy, or colonoscopy within the recommended time period. Participants were asked whether they had each of the CRC screening tests, and if they responded yes, the participants were asked when they had their last test. The items in the survey included a brief explanation of each screening test. Items were described as follows: FOBT, "which is stool slides"; sigmoidoscopy, "which is a tube inserted in the rectum to look at the colon and the bowel"; and colonoscopy, "which is a tube inserted in the rectum to look at the entire intestine, usually given in a hospital or specialist's office." Responses included, "less than 1 year"; "1–2 years"; "3–5 years"; or "more than 5 years." CRC screening was considered to be within the recommended time period based on an algorithm using ACS guidelines considering which test and how recent the test was performed (e.g. a person who reported having FOBT within the past year was considered within the recommended time period). Participants were considered to have been screened within the recommended time period if they had a FOBT within the preceding year, and sigmoidscopy within the preceding 5 years. Although current CRC screening guidelines recommend colonoscopy every 10 years for average-risk adults, we examined colonoscopy use in the past 5 years because of the limitations of the survey instrument. Self-report of CRC screening behavior has been demonstrated to be a reliable endpoint for intervention trials [ 22 ]. Statistical analyses Analyses of data from this study included factor analysis, analysis of variance, and logistic regression and were conducted using SPSS, version 10.1. Logistic regression analyses were performed to evaluate whether the level of perceived patient-provider communication was significantly related to CRC screening behavior in this population. Sociodemographic variables were identified as potential covariates if there was plausible theoretical or empirical evidence that the variable might be associated with the communication variable or with CRC screening. Variables that were significantly associated with communication level (p < 0.05) were retained and tested as covariates in the logistic models. Only the sex of the participant, receiving healthcare at a doctor's office versus a clinic/emergency room, and knowledge of CRC risk were significantly associated with communication and only these three covariates were entered into the initial logistic model. Variables in the model were evaluated by the Wald test and interpreted using odds ratios and confidence intervals. The overall fit of each model was evaluated using the Hosmer-Lemeshow Goodness of Fit test and by examining classification tables [ 23 ]. Results Focus groups Several major themes associated with CRC emerged from the personal experiences expressed about the medical community during the focus groups. One of the themes discussed by the participants in each focus group was patient-provider communication. Selected comments by focus group members about CRC and patient-provider communication are listed in Table 1 . This important theme that emerged during the focus group was addressed in the baseline survey by adding items to the questionnaire that specifically addressed perceived communication with health care providers (Table 2 ). Table 1 Focus groups (>50 years old): selected comments about patient-provider communication, colon cancer knowledge, and screening "My doctor never even told me that I needed a digital or colonoscopy, or the technical term. The only concern is said was prostate. Nobody said anything about colon.'' [Male] "I'm sixty years old and he's never told me to take one of these. I need to change doctors.'' [Female] "Well you know, I think of men when I think of colorectal cancer.'' [Female] "I've heard from 48 on up is a prostate exam more than anything else. I haven't heard anything about the colon. I really hadn't." [Male] Table 2 Patient-provider communication scale*. Five Items** A) I receive enough understandable information from my doctor/healthcare provider to make good decisions about my health. B) I feel rushed during visits. C) My doctor/healthcare provider involves me in decisions about my health care treatment. D) I feel uncomfortable asking my doctor for tests or information if he/she doesn't mention it. E) My doctor/healthcare provider understands my health needs. *These items were included in the baseline survey because of the importance of the patient-provider communication theme that emerged from the focus group participants. **Responses: Always, Almost always, Sometimes, Rarely, Never Items A, C, and E loaded on the same factor (α = 0.74) and these three items were used in the final measure. Survey Originally, 2480 names were obtained from the 12 church rosters located in five rural counties in North Carolina. Many members were ineligible (not 18 years old, deceased, no longer living in the state, medically incapable, phone number no longer working) or we were unable to contact them by telephone, and 239 members who declined to participate in the WATCH Project. There were 850 church members who participated in the WATCH Project and completed the baseline survey. The adjusted response rate was 66% using a calculation method, suggested by the Council of American Survey Organizations (CASRO) [ 21 ], that accounted for individuals whose eligibility and response status were unknown because program staff were never able to contact them. The participants in this study were the 397 church members who participated in the WATCH Project and were 50 years and older. The characteristics of the 397 participants are shown in Table 3 . Participants were mostly female (74%) and African American (98%). The mean age was 63 years (SD = 9.7). About half of the sample was currently married, 25% were widowed, and 14% were divorced. Thirty-seven percent had less than a high school education, 30% had a high school diploma or GED, 16% had some college or trade/beauty school, and 18% had a college degree or post-college education. Household income was answered by only 52% of the participants, and of the responders, 51% reported an income of less than $20,000. Table 3 Communication and the characteristics of the participants ≥ 50 years old* Communication* Total** n (%) Good n (%) Fair n (%) Poor n (%) F-test p-value Sex Male 103(25.9) 71(23.8) 14(28.0) 18 (42.9) F(2, 387) = 3.503 Female 293(73.8) 227(76.2) 36(72.0) 24 (57.1) p = .031 Education Less than HS 145(36.5) 103(34.6) 21(41.2) 18 (42.9) F(2, 388) = 0.409 HS/ GED 117(29.5) 91 (30.5) 14(27.5) 9 (21.4) p = .665 Trade School/ College 135(34.0) 104(34.9) 16(31.4) 15 (35.7) Income <$20,000 184(51.0) 132(48.4) 26(57.8) 23 (59.0) F(2, 354) = 1.134 $20,000–$49,999 128(35.5) 102(37.4) 12(26.7) 13 (33.3) p = .323 ≥ $50,000 49 (13.6) 39 (14.3) 7 (15.6) 3 (7.7) Marital Status Married 213(53.8) 164(55.2) 22(43.1) 24 (57.1) F(2, 387) = 1.350 p=.261 Divorced/ Widowed/ Separated 162(40.9) 116(39.1) 26 (51.0) 18 (42.9) Never married 21 (5.3) 17 (5.7) 3 (5.9) ----- Healthcare Facility*** Doctor's office 325(82.7) 250(85.0) 40(78.4) 29 (69.0) F(2, 384) = 3.605 Clinic/ER/Health Dept. 68 (17.3) 44 (15.0) 11(21.6) 13 (31.0) p = .028 Insurance**** Medicaid/Medicare 176(44.3) 137(46.0) 19 (37.3) 19 (45.2) F(2, 388) = 0.669 p = .513 No health insurance 21 (5.3) 14 (4.7) 4 (7.8) 3 (7.1) F(2, 388) = 0.566 p = .568 Employer/self-paid 219(55.2) 167(56.0) 29 (56.9) 18 (42.9) F(2, 388) = 1.344 p = .262 *Communication categories: the mean score for each category was calculated from the individual scores (continuous values) ** Numbers may not reflect the total n = 397 because of participants' refusals or missing data ***This variable is based on the questionnaire item: Where do you usually go when you need health care? ****Insurance variables were dichotomized for each category. Totals may exceed 100%, some respondents marked multiple categories (e.g., they had Medicare in addition to self-paid insurance). Factor analysis of the five communication items was performed from the baseline survey responses and two factors were identified; one with three items and the other with two items. The second factor was dropped because it had only two items and did not add reliability to the scale. The three communication items about shared decision-making and patient satisfaction demonstrated good reliability (Cronbach's alpha = 0.74) and were summed to calculate a communication score. The communication score was used to categorize the participants into three groups: good, fair, and poor communication with providers. Participants were categorized as having "good" communication if they perceived receiving enough information from their provider, being involved in medical decisions, and thinking that their provider understood their health needs almost all the time or always. Participants who rated all three items 'sometimes', 'rarely', or 'never' scored "poor" on the communication scale, and individuals who rated the items with a mix of the above listed responses were assigned to the "fair" group. In terms of quality of communication, 75% (298/397) responded positively to all 3 questions and were considered have "good" communication; 10% (42/397) responded positively to none of the 3 questions and were considered to have "poor" communication; and 13% (50/397) had fair results. Participants in the good communication group were more likely to be female (p = 0.031), and were more likely to receive their healthcare at a doctor's office versus a clinic/emergency room/health department (p = 0.028). None of the other sociodemographic factors listed in Table 3 appeared to vary significantly among communication groups. Participants categorized in the good communication group were more likely to report having been screened for CRC in the recommended time period compared to those in the poor communication group (35.9% vs. 16.7%; OR = 2.8, CI 1.2, 6.4, p = 0.013). Only 45% (175/389) of the participants reported that their providers had recommended CRC screening, and just 31% (120/389) of all participants reported being screened within the recommended time interval. Of the individuals who reported being screened, 65% (78/120) stated that their doctor had recommended CRC screening, compared with 36% (97/269) of those who did not report screening. Knowledge of CRC was assessed using seven items (Table 4 ) with a mean correct response of 3.8. If the participants answered at least four out of the seven items correctly, they were categorized as having adequate knowledge about colorectal cancer. The participants were considered to have inadequate CRC knowledge if they answered incorrectly or 'don't know' to ≥ 4 of the 7 items. Table 4 Knowledge of colorectal cancer risk factors among 397 African American participants (≥50 years old) in the WATCH Project Seven Items Correct Answer Percent* 1. A low fat and high fiber diet helps decrease colorectal cancer risk. True 70.8% 2. The risk of colorectal cancer is higher in men than women. False 13.6% 3. Physical activity decreases the risk for colorectal cancer. True 42.6% 4. Colorectal cancer risk increases after age 50. True 69.3% 5. A family history of colorectal cancer does not increase your risk. False 49.1% 6. Finding cancer early will not increase the chances of surviving it. False 65.7% 7. You only need to have a colorectal cancer screening test if you are having symptoms. False 67.5% *The percentage of participants who responded with the correct answer to each CRC knowledge item (n = 397) Knowledge about CRC was considered adequate (knowledge score > = 4) for 57% (228/397) and inadequate for 43% (197/397). Participants with adequate CRC knowledge were more likely to have completed a CRC screening test within the recommended time period compared to those with inadequate CRC knowledge (21% vs. 10%). Adequate knowledge was associated with a higher level of education (p < 0.001), a higher level of income (p < 0.001), having health insurance (p < 0.001), and having Medicare/ Medicaid as one's health insurance (p < 0.001). Multivariate analyses Results of the logistic regression analyses are shown in Table 5 . Results were similar when using communication and CRC risk knowledge as continuous exposure variables, and when using a history of CRC screening anytime in the past as the outcome variable (instead of recent screening). For ease of interpretation, we chose to present the categorical analyses and use recent screening as the outcome of interest. After adjustment for the sex of the participant and source of healthcare, quality of communication remained significantly associated with completion of a CRC test. Table 5 Factors associated with receiving CRC screening among 397 African American participants in the WATCH Project Variable OR (95% CI) p-value Sex .65 (0.39, 1.07) 0.093 Source of healthcare (M.D. office vs. Clinic/ER) 1.07 (0.58, 1.95) 0.838 CRC Knowledge (Adequate vs. Inadequate) 1.82 (1.14, 2.89) 0.011 Patient-provider communication (Good vs. Poor/Fair) 1.95 (1.29, 2.94) 0.002 CRC screening within recommended guidelines by perceived communication and knowledge is listed in Table 6 . The poor and fair communication groups were combined because of the small numbers within each category. Adequate knowledge is statistically significant for the good communication group but not for the fair/poor communication group. A test for interaction of communication and knowledge was performed for CRC within recommended guidelines and demonstrated no significant interaction. Table 6 CRC screening results by communication and knowledge CRC screening in recommended time (%) Poor and Fair communication Inadequate knowledge 15.0 (n = 40) p = 0.654 Adequate knowledge 18.5 (n = 54) Good communication Inadequate knowledge 27.4 (n = 124) p = 0.012 Adequate knowledge 41.6 (n = 173) Discussion Our study found that participants had higher rates of CRC screening when their self-rated communication with their healthcare provider was classified as perceived as more positive. In addition, we found that participants who self-rated their communication as good and who had adequate CRC knowledge completed recent CRC screening at higher rates than those with good communication and inadequate knowledge. In addition, screening rates are higher with both good communication and adequate CRC knowledge than when only one factor is present. These findings suggest that both good patient-provider communication and CRC knowledge are important for CRC screening. Because of the cross-sectional nature of our study, we cannot determine causality, and it is possible that CRC knowledge improves by going through the screening process. Improving CRC screening rates in the African American population may require strategies that address both improving physician-patient communication skills and increasing CRC knowledge. Good patient-provider communication is fundamental to a patient's perceived quality and satisfaction with their healthcare. Better communication, including the use of shared decision making, is associated with trust between patients and providers [ 24 ]. However, trusting the medical community remains a key concern for many African Americans. This is due, in part, to a long history of justifiable fear and mistrust of the medical and research communities stemming from the historical Tuskegee incident and other discriminatory practices in health care [ 25 , 26 ]. Our findings suggest that African Americans in this study generally had positive perceptions about communicating with health providers. Prospective data about communication and trust are needed to determine whether these perceptions predict greater screening compliance. The association between the lack of physician recommendations for cancer screening tests and low patient utilization of those tests has been documented in previous investigations [ 18 , 19 , 27 - 30 ]. In a recent study of rural primary care practices, discussion about CRC screening occurred in only 14% of eligible patients [ 31 ]. Additionally, previous research has documented that African American patients receive less physician recommendations for cancer screening tests and utilize cancer screening tests at lower rates than other ethnic groups [ 27 , 32 - 34 ]. In another recent study of 150 African Americans (aged 50–79), 39% reported never having a recommendation for FOBT, 60% never had a flexible sigmoidoscopy recommended, and 57% never had a colonoscopy recommended [ 35 ]. In our study, only 45% of participants stated that CRC screening had been previously recommended by their healthcare provider. The results of our study and other previously published work [ 24 , 35 - 39 ] suggest that CRC screening rates may improve by: 1) a focus on methods to improve patient-provider communication skills both for the patient and for providers; 2) addressing physician attitudes and behaviors toward recommending tests, and 3) providing patient CRC education that takes into account patients' literacy skills and preferred style of receiving information. Providing CRC risk knowledge and training to improve communication skills may be accomplished with various interventions [ 40 ]. Decision aids, systematically developed tools to provide information, increase knowledge, and encourage shared decision making, may also empower individuals to become more involved with their healthcare. Decision aids for CRC screening may be very useful because there is evidence that patients vary in their preference for how to be screened [ 36 , 37 ]. Limitations Although the results from our cross-sectional study have implications for developing colorectal cancer prevention programs, it must be recognized that prospective longitudinal studies are needed that specifically address the effect of patient-provider communication on CRC screening. Because of its cross-sectional design, we cannot infer causal relationships. In addition, since all participants were recruited from 12 churches, there may be some clustering of exposures and outcomes by physician that may affect the results; we did not measure this effect. Patient-provider communication was measured by using only three self-reported communication items, which may not fully capture the full extent of the patient-provider relationship. In addition, we do not know whether our participants had racially discordant or concordant physicians, and we did not have information regarding the physician's beliefs or practices about CRC screening or the physicians' assessment of the quality of communication with the patients. The outcome measure of having undergone CRC screening was self-reported, and these results were not validated. Finally, because of the relatively small number of participants and the use of convenience sampling, results from this study may not be generalizable. Conclusions The findings from this cross-sectional study suggest that not only do patients need to be informed about their CRC risk and the importance of screening tests but that having good communication with their healthcare provider may also important. The burden of having good communication in the patient-provider relationship is the responsibility of both individuals. Both the patient and the provider should strive to improve their communication skills so that patients who want to participate in the decision about how to be screened for colorectal cancer may do so effectively. Because culturally distinct factors may contribute to poor communication and mistrust for African-Americans, specific new strategies need to be developed [ 38 ]. Beliefs, attitudes, and concerns about cancer, prevention behaviors, and participation in medical decisions may not be the same for all races. If new prevention strategies for the African American population are developed, then there may be a chance at reducing the disparities associated with colon cancer and race. Competing interests The author(s) declare that they have no competing interests. Authors' contributions M.L.K. participated in the conception of the study, the analyses, the interpretation of the results, writing the first draft and revising the manuscript. A.S.J. participated in the conception of the study, performed the statistical analyses, participated in the interpretation of the results, and the editing of the manuscript. M.P.P. participated in the interpretation of the results and the writing of the manuscript. M.H., E.J., and V.O. assisted with the qualitative analysis and surveys. M.K.C. participated in the conception of the study, the statistical analyses, interpretation of the results, and the writing of the manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544572.xml |
554984 | Ageing and the immune system in vivo: commentary on the 16th session of British Society for Immunology Annual Congress, Harrogate, December 2004 | The problems associated with the ageing immune system coupled with possible solutions were discussed recently at the British Society for Immunology Annual Congress in Harrogate in December 2004. The session "Ageing and the Immune System in vivo" dealt in details with the immune risk phenotype and the potential methods of reversing the problems of an ageing immune system. This is a commentary on that session. | From an early age our immune system is introduced to a variety of infectious agents through contact with infected individuals. We also deliberately present our immune responses either to attenuated organisms or to components of infectious pathogens in order to provoke a response. A successful response to these organisms when first encountered provides us with immunological memory to enable our immune systems to make more rapidly responses to the same potentially infectious agents at a later date. In theory the longer we live the better our memory responses and our ability to cope with any potential pathogen seen previously. Since there are no completely sterile environments, individuals who have survived to reach a ripe old age must have combated many possible infectious agents, and have an immune system with a prodigious memory component. Only part of this is true. As we age the memory component of our immune system, as measured by the number of memory cells, does increase. Unfortunately this is not accompanied by improved immunity even to infectious agents that have been overcome earlier in life [ 1 - 4 ]. The problems associated with the ageing immune system coupled with possible solutions were discussed recently at the British Society for Immunology Annual Congress in Harrogate in December 2004. The session "Ageing and the Immune System in vivo" dealt in details with the immune risk phenotype and the potential methods of reversing the problems of an ageing immune system. The concept of the immune risk phenotype was described by Dr Anders Wikby (Jönköping University) who carried out a longitudinal immunological study in order to establish predictive factors for longevity. Dr Anders first worked on a group of normal octogenarians in the OCTA study in Sweden which later became the NONA study as these individuals entered their nineties. Within this group he was able to describe the immune risk phenotype which he characterised by a CD4:CD8 ratio of <1, poor in vitro T cell proliferation, an increase in the number of CD8 + CD28 - cells (or CD8 + CD28 - CD27 - in the very old), low numbers of B cells and the presence of CD8 T cells which were cytomegalovirus (CMV) tetramer positive. Within both the OCTO and NONA groups Dr Anders was able to show that the immune risk phenotype had some predictive input towards morbidity. This was even more apparent when cognitive impairment was included in the calculations. The effect of CMV infection on the immune system in the elderly was also discussed by Professor Paul Moss (University of Birmingham) with particular emphasis on the clonal expansion of T cells. Clonally expanded T cells are usually CD8 + and show an increased incidence with age, so far it seems that clonal expansion is not due to malignant transformation but may follow antigen stimulation. The control of CMV within the body is mainly attributed to CD8 + T cells and Professor Moss revealed that in older individuals up to 50% of CD8 + T cells may be specific for CMV as judged by tetramer staining. It is possible that this increased response to CMV may lead to the impairment of responses to other viruses in the aged. This theme of CMV infection and ageing was then continued by Professor Graham Pawelec (University of Tuebingen) who introduced the use of the KLRG-1 marker as a marker of senescence. T cells which are KLRG-1 + do not proliferate in vitro when provided with a stimulus which would induced proliferation in T cells which lack this marker. In elderly individuals fewer CMV tetramer positive CD8 + T cells secreted interferon-γ in Elispot assays when compared with similar cells from younger individuals. Furthermore between 96 and 99%of the CD8+ T cells which stained with the CMV tetramer also expressed the KLRG-1 marker. The theme of CMV infection in the elderly was further continued by Professor Arne Akbar (University College London) whose research group has recently been analyzing telomere lengths in different T cell subsets using a fluorescent staining technique. In this method telomeres are stained with a fluorescently labeled probe and the brighter the fluorescent labeling the longer the telomere. Using this technique Professor Akbar reported that CMV specific CD4+ T cells in older individuals had very short telomeres indicating their limited replicative capacity and that more of these CMV specific CD4+ T cells expressed interferon-γ. This view of a bleak future for displaying the immune risk phenotype was countered by the last two presentations which discussed different approaches to reversing the defects seen in the immune system with age. Both of these methods centered around reversing the atrophy of the thymus seen with age. Dr Jayne Sutherland, currently at the Anthony Nolan Research Institute, detailed her work carried out whilst with Professor Richard Boyd (Monash University) on reversing age associated thymic atrophy by surgical or chemical castration. Studies in mice revealed that reversal of thymic atrophy followed surgical castration and in humans chemical castration produced changes in the TREC levels which indicated improved thymic output. Functional studies showed that this treatment improved immune function in the treated aged subjects. This theme of improved immunity following reversal of thymic atrophy was carried further in work reported by Dr Richard Aspinall (Imperial College, London). He showed how therapeutic intervention with interleukin 7 and derivates could reverse atrophy of the thymus in old animals and also lead to improved immune function compared with age and sex matched controls. In addition he described work done in the Gambia which suggested a link between the development and enlargement of the thymus after birth and the level of interleukin 7 in breast milk. Two speakers from the posters presentations were also chosen to present their results. K. Wang from Professor Janet Lord's group (Birmingham) presented work on dehydroepiandrosterone, (DHEA). This steroid hormone declines with age and treatment of NK cells with DHEA increases NK cytotoxicity and further analysis shows that DHEA induces PKC-β translocation and upregulates perforin expression in these cells. D. Silva from Donald Palmers laboratory (Royal Veterinary College) presented work on the expression of neuropeptides in epithelial cells from the thymuses of different species. The meeting revealed that much needs to be done in characterizing the changes in the immune system which are associated with ageing, and members of the audience were asked whether they were interested in receiving more details about the newly formed "Differentiation and Immunosenesence Affinity Group" associated with the BSI. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC554984.xml |
554985 | Postage stamp multiple anterior capsulorhexisotomies in pediatric cataract surgery | Background Capsule related complications are common following pediatric cataract surgery. We report a new technique of multiple anterior capsulorhexisotomies after lens aspiration and intraocular lens (IOL) implantation. Methods After performing automated lens aspiration, an IOL was implanted into the capsular bag. A bent 26 gauge needle was introduced through one side port and multiple small cuts were made in one half of the circumference of the anterior capsular rim by making a radial movement of the needle tip centripetally over the margin of the anterior capsular rim. The needle was again introduced through the other side port and multiple similar cuts were made in the other half thereby creating nearly 20 – 30 cuts at the margin of the anterior capsular rim. Results The mean size of the primary capsulorhexis was 4.33 ± 0.20 mm. A uniform enlargement of the capsulorhexis could be performed in all the eyes without peripheral extension in any of the eyes. There was no damage to the posterior capsule and no scratch mark on the IOL. In one eye, the primary capsulorhexis was slightly eccentric, though it was covering the IOL optic all around. The rhexisotomies in this eye were limited to the capsular rim that was overlapping more on the IOL optic (sectoral anterior capsulorhexisotomies). Conclusion The technique of postage stamp anterior capsulorhexisotomies is a feasible technique in pediatric cataracts. | Background Pediatric cataract surgery has always been a challenge for the anterior segment surgeons and short term and long term complications have been reported in the literature [ 1 ]. Automated lens aspiration is the preferred technique employed these days for pediatric cataracts and central continuous curvilinear capsulorhexis is the method of anterior capsulotomy in this procedure. Performing a continuous curvilinear capsulorhexis (CCC) is more difficult in children than in adults because the capsular bag is more elastic. It has been reported that mechanized circular capsulectomy with a vitrector is easier to perform and safer than manual CCC in very young eyes [ 2 ]. However, most prefer using Utrata capsulorhexis forceps to complete the capsulorhexis after an initial nick in the anterior capsule with a bent 26 gauge needle. We herein, describe a new technique of postage stamp multiple capsulorhexisotomies, a modification in the primary capsulorhexis after performing lens aspiration with intraocular lens implantation in pediatric cataracts. Methods We performed this technique of anterior capsulorhexisotomies in 6 eyes of 4 children; 2 patients with bilateral developmental cataract and 2 with unilateral post-traumatic cataract. Surgery was performed in all the eyes under general anesthesia. A clear corneal 3.2 mm 3-plane tunnel was created superiorly. Two side ports were created at 10 and 2 o'clock positions using a microvitreoretinal blade (Alcon laboratories, Fort Worth, TX). Trypan blue (Vision blue, DORC, Netherlands) was injected (0.1 ml of 0.1%) into the anterior chamber under air bubble to stain the anterior capsule and was completely washed out after 15 seconds by balanced saline solution. Anterior capsulorhexis was initiated with a bent 26 gauge needle and completed by Utrata capsulorhexis forceps to create a circular, 4.0 – 4.5 mm, central capsulorhexis. The size of the capsulorhexis was measured on a television monitor according to a previously described method [ 3 ]. Hydroprocedure was performed to soften the lens matter and a complete lens aspiration was performed using Universal II (Alcon laboratories, Fort Worth, TX) phaco machine. Vacuum cleaning of the posterior capsule and the undersurface of the anterior capsular rim was performed in all the eyes. After inflating the capsular bag with 1.4% sodium hyaluronate (Healon GV; Pharmacia & Upjohn, Kalamazoo), a single piece foldable Acrysof intraocular lens (IOL) of optic size of 6.0 mm with an overall diameter of 13.0 mm (SA60AT, Alcon laboratories) was implanted in the capsular bag and the viscoelastic substance was aspirated by rock and roll technique. Sodium hyaluronate 1.4% was injected under the anterior capsular rim. A bent 26 gauge needle was introduced through one side port and multiple small cuts were made in one half of the circumference of the anterior capsular rim by making a radial movement of the needle tip centripetally over the margin of the anterior capsular rim taking care not to put any scratch mark on the IOL. The needle was again introduced through the other side port and multiple similar cuts were made in the other half thereby creating nearly 20 – 30 cuts all around circumferentially at the margin of the anterior capsular rim (Figure 1 ). The viscosurgical device was aspirated by rock and roll technique and 0.1 ml of 1% vancomycin was injected intracamerally. The anterior chamber was reformed with balanced salt solution and the corneal tunnel was hydrated. Postoperatively, patients were prescribed topical betamethasone sodium phosphate 0.1% and ciprofloxacin 0.3% QID each for 4 weeks and tropicamide 1% TID for 1 week. Results The mean age of the patients was 7.87 ± 1.60 (9, 8.5, 8.5 & 5.5 years) years and all patients were males. The mean size of the primary capsulorhexis was 4.33 ± 0.20 mm. The nicks at the margin of the anterior capsular rim could be performed successfully with a bent 26 gauge needle in all the eyes. A uniform enlargement of the capsulorhexis could be performed in all the eyes without peripheral extension in any of the eyes. No eye suffered damage to the posterior capsule. There was no scratch mark of the needle on the optic of any IOL. In one eye, the primary capsulorhexis was slightly eccentric, though it was covering the IOL optic all around. The rhexisotomies in this eye were limited to the capsular rim that was overlapping more on the IOL optic (sectoral anterior capsulorhexisotomies). Discussion Size of anterior capsulorhexis has always been a matter of debate for automated lens aspiration in pediatric cataracts. If the anterior capsulorhexis opening is small, there is risk of anterior capsular opacification, capsular contraction syndrome and phimosis of the anterior capsular opening, decentration of the IOL and capsular bag hyperdistension [ 4 , 5 ]. Despite an intact capsulorhexis, IOL decentration may still occur due to capsular contraction syndrome. If the capsulorhexis is too large, there is risk of development of adhesion between the anterior capsular rim and the posterior capsule. This can have a zipper effect on the IOL, which can result in forward popping up of the IOL and a significant change in the refractive status of the eye. The size of the anterior capsulorhexis is considered adequate when the margin of the anterior capsular rim just covers the IOL optic margin all around. However, it is very difficult to create an exactly similar size of anterior capsulorhexis in all the eyes. Intraoperative enlargement after performing lens aspiration with a previously performed smaller rhexis is possible [ 6 - 8 ]. But in pediatric eyes, it is very difficult to predict whether the enlarged rhexis opening will be of optimum size. In many of these situations, the size of the capsular opening enlarges more than optimum in one half or one quadrant. In an effort to prevent complications related to both smaller and larger anterior capsulorhexis, we performed multiple anterior capsulorhexisotomies like the configuration of a postage stamp, in which after creating a primary capsulorhexis of slightly smaller than optimum size, multiple nicks were made at the margin of the anterior capsular rim all around after lens aspiration and implantation of IOL. These nicks if made before lens aspiration can extend to the periphery towards the equator of the lens as the capsule is taut due to the presence of positive intralenticular pressure. However, after lens aspiration, the capsule becomes lax and hence these nicks do not extend to the periphery. If performed before IOL implantation, there is risk of damaging the posterior capsule and moreover, during IOL implantation, these nicks can extend to the periphery. We have observed that after curvilinear capsulorhexis, capsular opacification and fibrosis is most marked at the margin of the anterior capsular rim. We performed capsulorhexisotomies to prevent phimosis of the anterior capsular opening and capsular contraction syndrome, as these small nicks act as relaxing incisions. Since there were multiple nicks all around, the direction of the force is well distributed. The vector of the force generated at the opacified margin of curvilinear capsulorhexis is directed inwards, while after performing multiple nicks in this margin (capsulorhexisotomies), the vector is directed in both inward as well as outward directions. Therefore, there is less chance of development of capsular contraction syndrome. More over, there is negligible chance of adhesion between the anterior and posterior capsules as the anterior capsular rim rests well on the IOL optic. Conclusion The postage stamp multiple anterior capsulorhexisotomies is a feasible and safe technique after lens aspiration and IOL implantation in pediatric cataracts. Declaration of competing interest The author(s) declare that they have no competing interests. Individual contribution of authors JST designed the study and performed surgeries. RS performed the data collection and wrote the manuscript. NS followed up the patients. RBV performed the surgeries. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC554985.xml |
553998 | Direct analysis of thymic function in children with Down's syndrome | Background Down's syndrome (DS) is characterized by several immunological defects, especially regarding T cell compartment. DS is considered the best example of accelerated ageing in humans. Direct observations of the thymus have shown that in DS this organ undergoes severe histological and morphological changes. However, no data on its capacity to generate T cells are present in the literature. Here, using a new technology based upon real time PCR, we have investigated the capacity of the thymus to produce and release newly generated T lymphocytes (the so called "recent thymic emigrants", RTE) in children with DS. Methods We studied 8 children affected by DS, aged 2–7 years, compared with 8 age- and sex-matched healthy controls. Flow cytometry was used to determine different lymphocytes subsets. Real time PCR with the Taqman system was used to quantify the amount of RTE, i.e. peripheral blood lymphocytes that express the T cell receptor rearrangement excision circles (TREC). Results In comparison with control children, those with DS had a significant lower number of TREC+ peripheral blood cells. Moreover, in DS children but not in controls, a strong negative correlation between age and the levels of TREC+ cells was found. Conclusions The direct measure of thymic output indicates that the impairment of the organ results in a reduced production of newly generated T cells. This observation could suggest that cytokines able to modulate thymic function, such as interleukins, could be useful to improve the functionality of the organ and to treat the immunodeficiency present in DS subjects. | Introduction Down's Syndrome (DS) is the most common chromosomal abnormality in humans, that occurs in 1 out of every 800–1,000 births. It is an autosomal disorder resulting by triplication of chromosome 21. Many characteristics are commonly seen in DS, including some degree of intellectual impairment, that varies widely from individual to individual, heart defects, hypotonia, hyperuricemia, and the development of Alzheimer disease-type neuropathology beginning at about 40 years of age [ 1 ]. DS subjects also present alterations of the immune system which are similar to those of aged people, including increased susceptibility to bacterial and viral infection and to the onset of different types of haematological malignancies, along with a high frequency of autoantibodies. Alterations of B lymphocytes, T cell subsets, and natural killer cells, defective phagocytosis and chemotaxis of polymorphonuclear leukocytes and interleukin-2 production by activated T cells were also reported [ 2 - 7 ]. Quantitative studies of peripheral blood T lymphocytes reveal a reduction, often quite small, in the percentage and/or absolute number of T lymphocytes, although normal proportions or numbers of T and B lymphocytes in DS children have also been reported [ 8 ]. Several studies have focused their attention on the role of the thymus, and have described a variety of structural and anatomic alterations present in DS [ 9 ]. Few data, if any, exist on the direct measure of thymic functionality in terms of production of newly generated T cells. Accordingly, the aim of this study was to measure the capability of the thymus of DS children to produce new T lymphocytes, and to analyze how such capability changes with age. For this reason, we have quantified the so called "recent thymic emigrants" (RTE), that are the main contributors to the naïve T-cell pool, and are characterized by the presence in the nucleus of a circular, episomal DNA molecule called TREC ( T cell receptor R earrangement E xcision C ircles), generated during the intrathymic rearrangement of the α-chain locus of the T cell receptor (TCR). Genes for the δ-chain of the TCR are distributed within the genomic region that codifies for the α-chain, and are removed in two steps during the recombination of Vα with Jα. Thus, a thymocyte that starts to rearrange the α-chain produces the first TREC, called signal-joint (sj)-TREC, then proliferates three or four times, and finally completes the rearrangement of Vα with Jα, so producing the second TREC, called coding-joint (cj)-TREC. The removal of genes for the δ-chain from the α region does not imply their elimination, as such DNA remains into the nucleus as a circle that is not able to replicate. As a consequence, when a cell undergoes a division, TREC are passed only to one of the two daughter cells. During the following cell cycles, TREC are then diluted into the population that origins from the first cell. Several data indicated that the percentage of TREC+ cells is a marker of thymic activity, that TREC+ cells are almost present within the subset of virgin T lymphocytes, and that their number consistently declines with age [ 10 ]. Materials and Methods Subjects We could analyze blood samples from 8 children with DS (with a mean age of 4.62 years, range: 2–7 years), 7 males and 1 females. As control group, we studied 8 children with a mean age of 4.75 years (range, 2–8 years), 7 males and 1 female. All subjects were in good conditions, and had no acute or chronic disease affecting the immune system. Written informed consent was obtained by their parents, according to the Italian laws. Isolation of PBMC from blood Peripheral blood mononuclear cells (PBMC) were separated from a minimum of 8 mL freshly collected blood according to standard procedures. Analysis of the phenotype of peripheral blood lymphocytes Blood samples were stained with different monoclonal antibodies for the cytometric analysis, as described [ 11 ]. The quantification of the main subpopulations present among lymphocytes (electronically selected on the basis of an electronic gate put in the region of lymphocytes) was performed by flow cytometry using a CyFlow ML instrument (Partec, Münster, Germany), according to standard procedures [ 11 ]. DNA extraction DNA was extracted from 5 × 10 6 PBMC using the QIAamp DNA Mini Kit (QIAGEN) according to the manufacturer's protocols and stored in sterile water at -20°C until use. Quantification of percentage of sjTREC positive PBMC by a Real Time PCR approach The percentage of PBMC containing sjTREC was measured by an original method that has been recently developed (patent pending) using a Real Time PCR approach. This assay was performed by two parallel polymerase chain reactions (PCR), that quantify sjTREC or nuclear DNA (nDNA) in a given sample, carried out in two different reaction tubes, but in the same plate, in order to have similar reaction conditions. In the first reaction, we quantified sjTREC using a mix that consisted of: Supermix Biorad 1X, primers for sjTREC 500 pmol and we added 3 μl of the sample in each tube. The sjTREC primers we used were: sjTREC Dir (5'-CAC ATC CCT TTC AAC CAT GCT-3') and sjTREC Rev (5'-GCC AGC TGC AGG GTT TAG G-3'). TaqMan probe for sjTREC (5'-FAM-ACA CCT CTG GTT TTT GTA AAG GTG CCC ACT – BLACK HOLE-3') was included in the reaction mixture at the concentration of 200 nM, as a real time detector for the amplified product. One cycle of denaturation (95°C for 6 min) was performed, followed by 50 cycles of amplification (95°C for 30 sec, 58°C for 1 min 30 sec). The same approach was used to quantify nDNA, which was required to obtain the number of cells present in the sample. In this case, primers GenDir (5'-GGC TCT GTG AGG GAT ATA AAG ACA-3') and GenRev (5'-CCA ACC ACC CGA GCA ACT AAT CT-3'), designed on FasL gene sequence, present in two copies in the human genome (one in each chromosome 1), were used at the concentration of 600 and 400 nM respectively, and the TaqMan probe GenProbe (5' TexasRed – CTG TTC CGT TTC CTG CCG GTG C – BlackHole Quencher2 3') was included in the reaction mixture at the concentration of 300 nM. One cycle of denaturation (95°C for 3 min 30 sec) was performed, followed by 45 cycles of amplification (95°C for 30 sec, 60°C for 35 sec). All of the aforementioned primers and probes have been designed using the program "primer3", available on the internet address: . The length of amplified fragments is 104 pb for nDNA and 101 pb for sjTREC. PCR was performed by using an iCycler Thermal cycler (BioRad, Hercules, CA, USA), that monitors changes in the fluorescence spectrum of each reaction tube during the annealing phase. The fluorescence signal was processed using the "Real-time detection system iCycler iQ" software, that calculates the threshold and the threshold cycles of each sample. All reactions were carried out in triplicate. To optimize the precision of the assay, whose results derive from the comparison between the threshold cycles of two different real time PCR reactions, we have developed a new approach. Indeed, the regions used as template for the two amplifications ( i.e. , those of sjTREC and nDNA) were purified and cloned tail to tail in a vector (pGEM-11Z, from Promega), to have a ratio of 1:1 of the molecules used as reference. The nDNA region used has been excised as a SacI-ApaI fragment and cloned in the pGEMT easy vector (Promega), obtaining the pGEM11Z-nDNA vector. The sjTREC region used has been excised as a PstI-SacI fragment and cloned in the same vector, obtaining the pSJ vector we used. The plasmid has been purified, quantified with the spetrofotometer and linearized using SacI. Then a series of reactions has been performed for nDNA and sjTREC with serial dilution of the standard. All the times the threshold cycles of nDNA and sjTREC were the same, proving the ratio 1:1 of the two fragments. Then, serial known dilutions of this vector, amplified in triplicate, were included in each PCR run to generate a standard curve from which the relative copy number of either sjTREC or nDNA present in the unknown samples was determined. The measured values for sjTREC and nDNA were always within the range of the standard curve, whose correlation coefficient was always >0.990. The values of sjTREC and nDNA present in each sample were calculated using the mean of the threshold cycles of the three replicates. Then, the percentage of PBMC containing sjTREC in each sample was simply obtained from the ratio between the relative values of sjTREC (obtained dividing for 3 the result given by the iCycler, as we used 3 μl of DNA of the sample and 1 μl of standard to quantify sjTREC) and nDNA obtained (obtained versus the same vector), multiplied by 2 (as two copies of the nuclear gene are present in a cell). Results Peripheral blood cell phenotype in DS Down's syndrome is characterized by severe immunological alterations, which mainly affect the T cell compartment and are often regarded as signs of accelerated ageing. Numerous reports suggested that thymic retention of T cells or maturation defects might be the cause of the observed alterations in the T cell compartment. Accordingly, we first checked the presence of possible changes in the phenotype of peripheral blood lymphocytes from DS children. As shown in Figure 1 , we found that, in comparison with karyotypically normal healthy controls, DS children had less T helper cells and more cytotoxic T lymphocytes or cells expressing markers related to NK activity. This indicates that the population we analyzed can well represent the immunological situation present in DS children. Figure 1 Phenotypic analysis of peripheral blood lymphocytes in patients with Down's syndrome (DS) and healthy subjects. Data are referred to 8 individuals per group. Asterisk indicates a statistically significant difference (p < 0.05). Quantitative analysis of cells expressing TREC We then studied directly the capacity of the thymus to produce new T lymphocytes, by the analysis of TREC+ cells. Figure 2 shows a typical example of real time PCR assay for quantification of the amount of TREC per cell. It is to note that the threshold cycle, i.e. the cycle of PCR in which the fluorescent signal deriving from the amplification of the DNA becomes evident, is quite different in DS children and controls. Indeed, such cycle is much lower in control children, indicating the presence of a higher number of TREC. Figure 2 Representative example of real time PCR for the quantification of TREC in a group of 3 children with Down's syndrome (DS) and in 4 healthy controls (each measure is performed in triplicate). Note how the threshold cycle is different in the two groups. The percentage of TREC+ lymphocytes in children affected by Down's syndrome and in controls is reported in Figure 3 . A statistically significant difference was present between the two groups (p = 0.007), indicating that DS is characterized by a lower thymic output. Figure 3 DS children have less TREC+ lymphocytes than healthy controls, as shown in this box-and-whiskers graphics. The boxes extend from the 25 th percentile (x [25] ) to the 75 th percentile (x [75] ) [ i.e. , the interquartile range (IQ)]; lines inside boxes represent median values. Lines emerging from boxes ( i.e. , the whiskers) extend to the upper and lower adjacent values. The upper adjacent value is defined as the largest data point ≤x [75] +1.5xIQ, and the lower adjacent value is defined as the smallest data point ≥x [25] -1.5xIQ. Note that no outliers are present in the two groups. Correlation between age and TREC Finally, we have investigated the correlation between age and TREC levels in the two groups. As reported in Figure 4 , it is noteworthy that while control children did not display any age-related change (in the age range 2–8 years), those with DS had a significant age-related decrease in the number of TREC+ cells. This indicates that in DS, in contrast with control children, major changes in thymic output occur in the very first years of life. Figure 4 Correlation between age and TREC levels in DS (squares) and control (triangles) children. In the year range 2–8, the correlation was significant in DS but not in control children. Discussion Immunological ageing is part of a continuum of developmental processes, encompassing complex reorganizational events, compensatory mechanisms and qualitative alterations in the functionality of several systems and organs. Among those organs that undergo major changes with ageing, thymus plays a special role [ 12 ]. The thymus is a central lymphoid organ that is the primary site of T-cell maturation and development. Shortly after birth, the thymus undergoes a life long process of involution whereby the organ is replaced by adipose tissue. The result is a reduction in the number of constituent thymocytes with age, a consequent shrinking of the thymus and a decline in the output of newly generated T lymphocytes [ 13 ]. Aged peripheral T-cell pool is characterized by the accumulation of T-cell capable of limited replication [ 14 ]. Since an efficient immune response is based upon the expansion of antigen-specific clones, the consequence of a qualitative and quantitative impairment of the system is an increased susceptibility to infections or cancer. Several groups, including ours, have studied the immune system during human ageing, using different models, including the one that represents the best example of successful ageing, i.e. healthy centenarians, and conditions of accelerated ageing, such as DS [ 2 - 6 , 15 ]. In DS subjects, the proportion of T-helper cells is decreased, resulting in a decreased ratio of helper/cytotoxic cells, and we could observe this phenomenon also in the patients here described. Furthermore, peripheral blood T cells have a decreased number of cells expressing the T-cell receptor-αβ (TCRαβ) complex, elevated numbers expressing TCRγδ and a decreased proportion of CD4+, CD45RA+ naïve T cells, along with a high number of cells with NK phenotype, suggesting that the DS thymus is inefficient in the release of functionally mature T cells [ 16 ]. An age-dependence in the proliferative response to phytohemagglutinin (PHA) of DS lymphocytes, but not of lymphocytes from healthy individuals, was observed. Allogeneic mixed lymphocyte proliferative responses are decreased, as are PHA-induced interleukin-2 production and cytotoxic T-lymphocyte activity [ 7 ]. All of the above mentioned alterations have a common substrate, i.e. the dysfunction of thymus. Indeed, defects in the capacity to produce thymic hormones has been described several years ago [ 17 , 18 ]. Furthermore, anatomic studies gave further evidence that changes in the T-lymphocyte system derive from structural abnormalities of the thymus. In comparison to age-matched controls, thymuses from infants with DS from 1 day to 15 months of age have marked lymphoid depletion, with a thin cortex and poor corticomedullary demarcation. The Hassall corpuscles are increased in size and frequently cystic. The presence of lower proportions of cells bearing high levels of the TCR-αβ complex and of CD3, a signal-transducing complex for the TCR, in thymuses of children with DS and the increase in the proportions of cells with these markers with age are indicative of delayed maturation of T cells within the thymus [ 9 ]. In addition, DS thymuses contain elevated levels of IFN-γ and TNF-α mRNA expressing cells, and there is mast cell hyperplasia and overexpression of class I MHC, CD18 and ICAM-1 [ 19 ]. DS thymocytes also have a greater than normal sensitivity to inhibition of IL-4-induced proliferation by IFN-γ and TNF-α [ 20 ]. Taken together, these findings indicate an abnormal thymocyte maturation and cytokine dysregulation in the DS thymus, possibly initiated by gene dose-related increased sensitivity to IFN-γ and to overexpression of CD18 (LFA-1β). In conclusion, in this paper we show that the aforementioned alterations of the thymus result in the reduced production and output of newly generated lymphocytes, that can be directly measured by the assay we have developed and used. The quantitative analysis of TREC+ cells in the periphery is a relatively new and sensitive marker of thymic functionality, which is able to provide information on the status of the organ in different pathological conditions, and on its capability to generate new T cells [ 21 - 25 ]. A markedly reduced capacity of the thymus to produce RTE is present in the DS children we have studied. Such observation, if confirmed in a higher number of cases, could be useful to develop novel strategies to treat the immunodeficiency typical of this syndrome, based for example on the use of cytokines such as interleukin-2 or interleukin-7, which is capable of maintaining or restoring an efficient thymic output [ 26 ]. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC553998.xml |
522807 | cuticleDB: a relational database of Arthropod cuticular proteins | Background The insect exoskeleton or cuticle is a bi-partite composite of proteins and chitin that provides protective, skeletal and structural functions. Little information is available about the molecular structure of this important complex that exhibits a helicoidal architecture. Scores of sequences of cuticular proteins have been obtained from direct protein sequencing, from cDNAs, and from genomic analyses. Most of these cuticular protein sequences contain motifs found only in arthropod proteins. Description cuticleDB is a relational database containing all structural proteins of Arthropod cuticle identified to date. Many come from direct sequencing of proteins isolated from cuticle and from sequences from cDNAs that share common features with these authentic cuticular proteins. It also includes proteins from the Drosophila melanogaster and the Anopheles gambiae genomes, that have been predicted to be cuticular proteins, based on a Pfam motif (PF00379) responsible for chitin binding in Arthropod cuticle. The total number of the database entries is 445: 370 derive from insects, 60 from Crustacea and 15 from Chelicerata. The database can be accessed from our web server at . Conclusions CuticleDB was primarily designed to contain correct and full annotation of cuticular protein data. The database will be of help to future genome annotators. Users will be able to test hypotheses for the existence of known and also of yet unknown motifs in cuticular proteins. An analysis of motifs may contribute to understanding how proteins contribute to the physical properties of cuticle as well as to the precise nature of their interaction with chitin. | Background One particular family of cuticular proteins constitutes one of the largest multigene families known in insects [ 1 ]. Unrelated cuticular proteins are also numerous within a single species [ 2 , 3 ]. This diversity of cuticular proteins is extraordinary when one considers that chitin, the other principal constituent of cuticle, is a simple filamentous polymer of N-acetylglucosamine. Over 60 sequences have been obtained from proteins extracted from arthropod cuticles freed from adhering cells, primarily through the work of Svend Andersen and his colleagues in Copenhagen. An additional 9 have been extracted from cuticle and had their N-terminal sequences determined [ 2 , 3 ]. These verified cuticular protein sequences revealed motifs, unique to arthropod proteins, that have made it possible to classify sequences that came from cDNAs and genomes as cuticular proteins. In addition to sequence determination, studies of cuticular proteins have emphasized spatial distribution and expression in different developmental stages (reviewed in [ 2 , 3 ]). Consequently, a wealth of information exists. We have used and organized this information in a relational database, named cuticleDB, the first database of arthropod cuticular proteins. The current total number of entries is 445, including proteins from 6 orders of Insects, 2 orders of Crustacea, and 2 orders of Chelicerata. This first version of cuticleDB is restricted to structural proteins of the cuticle; enzymes active in sclerotizing (tanning) or digesting cuticle and proteins involved in defense and pigmentation have been omitted. The database nomenclature is based either on the names given by those who deposited the sequences or on codes assigned by genome projects. Thus, we have retained the existing names/codes for the convenience of the users. Construction and content Data collection The data collection has been basically done in two ways. First, by submitting appropriate keywords ( cuticle , exoskeletal , carapace ) to the Protein databases of Entrez and Uniprot (release 1.8) [ 4 ] we collected a number of entries, which were manually filtered. Results from the two databases were checked to eliminate duplicates. Secondly, we obtained genome data for Anopheles gambiae and Drosophila melanogaster , from Ensembl [ 5 ] and EBI, respectively. These are currently the only Arthropods with annotated genomes. We searched these genomes for a Pfam motif, PF00379, setting as cutoff the recommended gathering cutoff of the corresponding Pfam entry [ 6 ]. This motif has been shown to be responsible for chitin binding [ 7 ] and most probably adopts a precise, well-defined structure [ 8 , 9 ]. A short version of this motif was first recognized by Rebers and Riddiford in 7 cuticular proteins [ 10 ], and, as more sequences became available [ 11 ], was widely recognized. The initial consensus was 35 amino acids long, but now encompasses 68 residues as sequence similarity was recognized at its amino-terminus and the carboxy-teminus was shortened. This 68 amino acid region, named the "extended R&R consensus" is what is recognized by PF00379 ([ 2 , 3 ]and references therein). In order to ensure that our data collection is complete, we scanned all protein sequences of Uniprot (release 1.9) for PF00379. Again manual filtration was required. In addition to PF00379, other motifs have been described in cuticular proteins, some are found along with PF00379 while others define other families of cuticular proteins [ 2 , 3 ]. All recognized cuticular protein motifs were used to construct the database. The data for our database was obtained by parsing the fields Definition, Accession, GI from Version, Organism and Origin from the Entrez entries. From the Uniprot entries we used Primary accession number, Protein name, Origin of the protein, Cross-references and Sequence information. This retrieval was done with Perl scripts. Additional information, concerning temporal and local expression of the proteins or corresponding mRNAs, was drawn from literature. Implementation The data have been organised based on a relational model and is stored in a PostgreSQL database system. The user has supervisory access through our Apache web-server. The database is managed by an interferential software, written in Java, which tends to settle any web-server's query. Also, it implements a homemade computational tool that performs motif search as described below. Data retrieval The main page of cuticleDB includes the following interfaces: Introduction, Data Retrieval, User manual and Contact. On clicking the Data Retrieval icon, users are presented with the search interface of the database. The query can be done in two ways: either by searching in fields or by gathering a set of proteins (Figure 1 ). The separate fields in which the user may search are Protein name, Taxonomy, references in other databases (the user may submit Entrez GenInfo Identifier, Entrez Accession Number, Uniprot AC, Flybase ID, Ensembl code, Interpro AC or Pfam AC as a query) and the protein sequence. The protein sequence can be searched against any pattern according to the user's imagination and, therefore, hypotheses for novel motifs can be tested. This is performed by a separate, homemade tool that has been integrated in cuticleDB and which gives the user the opportunity to detect new motifs in cuticular proteins. The integration of this tool is of importance especially in a database such as this, given the significance of motifs not only in cuticular proteins, but in structural proteins in general. Users can gather all protein entries from a single species (35 species are included in cuticleDB) or all protein entries whose protein sequence contains one of a series of motifs. However, this series of motifs has been pre-selected by the constructors of the database and cannot be modified by the user. The selection criterion was the frequency of appearance of these motifs in the literature. The most commonly found motifs were searched against all protein sequences of the database and have properly been assigned to each entry. Description of an entry A typical cuticleDB entry contains the following fields: Protein Name, References to other databases (Entrez Protein Database, Uniprot, Interpro, Pfam, Flybase, Ensembl), Taxonomy, Expression Details, Protein Sequence and its Length, Database-Source of the sequence and the method by which the sequence was obtained (Figure 2 ). The field 'Expression Details' supplies the user with information about the anatomic region where each protein has been detected or the tissue where the corresponding mRNA is expressed, as well as the developmental stage in which the protein/mRNA appears. This field is usually accompanied by literature-citations. Moreover, another field named Patterns shows all patterns that have been searched for and found in the protein sequence, together with the start and end position of each. A text-box where the user can write his/her pattern is also available. If the user pattern matches the sequence, it is appended to the list of the predefined patterns. It remains there, as long as the user's session lasts. Also present are a field giving the known or predicted signal peptide and fields indicating whether the protein is putative, preliminary or fragment. Taxonomic distribution of the entries Taxonomic data are taken from Entrez. The total number of entries in cuticleDB is 445. These proteins are distributed in the three large taxa: Insecta (370 entries), Crustacea (60 entries) and Chelicerata (15 entries). The database includes entries from 6 orders of the class Insecta: Diptera (258 entries), Lepidoptera (39 entries), Orthoptera (37 entries), Hemiptera (6 entries), Coleoptera (22 entries), Dictyoptera (8 entries). The large number of proteins in Diptera is due to the inclusion of cuticular proteins from the two genomes ( D. melanogaster , A. gambiae ). The only verified cuticular proteins are those where the complete protein sequence or a unique N-terminal region was determined from a protein extracted from a cleaned cuticle or where a specific antibody reacted with proteins in cuticle or extracted from it. Finding mRNA in the epidermis is presumptive evidence that a protein is cuticular. The majority of cuticular proteins in this database were designated as cuticular proteins based on their sequence similarity to authentic cuticular proteins. Such proteins where sequence is the sole criterion for assignment are marked as "putative" in the database. Furthermore, at present, the annotation of the proteins of A. gambiae is preliminary. Many proteins are missing signal peptides, other clearly have been incorrectly assembled. Such sequences are marked as preliminary as well as putative. This database will be continuously updated at regular intervals to accommodate annotation. The distribution of the proteins in the subphylum Crustacea is 59 entries from the order Decapoda, and 1 entry from the order Sessilia, whereas the distribution in the subphylum Chelicerata is 5 entries from the order Araneae and 10 from Xiphosura. Motif distribution Apart from collecting and organizing data, this database also contains results of experimental computational work. Based on the classification of the "extended R&R" motif into two main types, RR1 and RR2 [ 12 ], which, at present, appears to correlate with their presence in proteins from soft and hard cuticles respectively, we built a Profile Hidden Markov Models for the two types. For this purpose we used the HMMER software package (Version 2.3.2) [ 13 ] utilizing its hmmbuild function. As an input to this function we used an alignment derived from 14 RR1 protein sequences from D. melanogaster for the RR1 HMM and an alignment derived from 9 RR2 protein sequences from the same species for the RR2 HMM (suitably selected from reference [ 3 ]). Both of the alignments were restricted to the area of the 'extended R&R consensus', thus they did not include the whole sequences. Subsequently, we used these Profile Hidden Markov Models as a prediction tool for classifying the cuticular proteins into two groups RR1, RR2. The prediction was in agreement with the literature as far as the known RR1 and RR2 proteins are concerned. The total number of RR1 and RR2 proteins in cuticleDB are 132 and 148, respectively. The start and end positions of the two motif-types are shown in the corresponding entry of each protein. A smaller class, RR-3, with 75 conserved residues was also identified by Andersen [ 14 ]. We have also studied the appearance of another motif: AAP(A/V). This small, hydrophobic tetrapeptide has been found to occur mainly in proteins of hard cuticles [ 2 , 3 ], where the water content is low and the sclerotization is intense. We have found that the AAP(A/V) motif occurs in 43% of the RR2 proteins, whereas only in the 12% of the RR1 proteins of cuticleDB. Utility and discussion The most severe problem of genome projects to date is that of correct annotation. So, accurate and specialized databases as cuticleDB with its description of highly conserved motifs will be of help to genome annotators. Therefore, cuticleDB can be used as a basis for annotating new cuticular proteins by similarity in future Arthropod genome projects. cuticleDB can also be utilized in molecular research as well, due to its focus on motif appearance. Cuticular proteins, as is the case with all structural proteins are marked by the presence of characteristic motifs. Some motifs are repeated within a protein sequence, whereas others appear only once. cuticleDB has been designed in such a way that the user can have a complete view of motif occurrence in the sequence of each protein entry. First, each entry shows the exact position of the most common cuticle motifs in the protein sequence. Secondly, the user is given the opportunity to search the sequence for novel motifs and therefore, test hypotheses for the existence of new patterns. Subsequently, hypotheses for possible interactions between cuticle macromolecules (either proteins with chitin or proteins with proteins) can be tested. Moreover, our results of the RR1 and RR2 predictions can be used as a guide for identifying a certain protein as coming from either soft or from hard regions of the cuticle. Most importantly, the information about the RR1 and RR2 distinction can be used for studies of cuticle's mechanical properties. As RR1 and RR2 proteins appear in soft and hard cuticles respectively, which means that the former interact with chitin more loosely than the latter, one can gain an insight in cuticle's molecular construction combining our data on the sequences of RR1 and RR2 proteins with some experimental work. Moreover, one could use the Expression Details, namely where and when each protein is expressed, when studying the differential construction of the cuticle among different developmental stages or among different regions of a single cuticle. Conclusions The goal of cuticleDB constructors was the collection of all cuticular protein sequences that have appeared to date and their detailed and correct annotation. The better the organisation of the data, the easier the work will be for researchers dealing with cuticle and structural proteins in general. cuticleDB will help them to answer questions like : 'What kind of proteins appear in hard cuticles?' 'Why do RR2 proteins interact with chitin more tightly than RR1 proteins?' 'Which motifs contribute to protein-protein interaction in the cuticle?' 'From which stage can a certain protein be extracted?' Furthermore, it is hoped that, detection of common properties of these proteins, as well as recognition of important differences that are responsible for cuticle's complexity and important functions will be facilitated by its existence. Last but not least, it is hoped that this database will be of help to genome annotators in the near future as more arthropod genomes become available. Availability and requirements cuticleDB was created and is maintained in the Department of Cell Biology and Biophysics, Faculty of Biology of the National and Kapodistrian University of Athens. It is freely available at the URL: . An e.mail biodb@biol.uoa.gr may also be used for comments, corrections and further data (sequence) submission. List of abbreviations RR1: The extended Rebers and Riddiford Consensus, type I RR2: The extended Rebers and Riddiford Consensus, type II HMM: Hidden Markov Model Authors' contributions CKM performed the data collection, and test procedures, and also participated in the design and the implementation of the database ICS carried out the design of the algorithms and the database,. implemented all the algorithms, and also created the web interface VAI supervised the data collection and the tests JHW compiled the first draft of known cuticular proteins, provided a critique of the data base during its construction SJH coordinated and supervised the whole project, suggesting the general directions and innovative features of the database All authors have read and accepted the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC522807.xml |
554991 | The chalcone butein from Rhus verniciflua Stokes inhibits clonogenic growth of human breast cancer cells co-cultured with fibroblasts | Background Butein (3,4,2',4'-tetrahydroxychalone), a plant polyphenol, is a major biologically active component of the stems of Rhus verniciflua Stokes. It has long been used as a food additive in Korea and as an herbal medicine throughout Asia. Recently, butein has been shown to suppress the functions of fibroblasts. Because fibroblasts are believed to play an important role in promoting the growth of breast cancer cells, we investigated the ability of butein to inhibit the clonogenic growth of small numbers of breast cancer cells co-cultured with fibroblasts in vitro. Methods We first measured the clonogenic growth of small numbers of the UACC-812 human breast cancer cell line co-cultured on monolayers of serum-activated, human fibroblasts in the presence of butein (2 μg/mL) or various other modulators of fibroblast function (troglitazone-1 μg/mL; GW9662-1 μM; meloxican-1 μM; and 3,4 dehydroproline-10 μg/mL). In a subsequent experiment, we measured the dose-response effect on the clonogenic growth of UACC-812 breast cancer cells by pre-incubating the fibroblasts with varying concentrations of butein (10 μg/ml-1.25 μg/mL). Finally, we measured the clonogenic growth of primary breast cancer cells obtained from 5 clinical specimens with normal fibroblasts and with fibroblasts that had been pre-treated with a fixed dose of butein (2.5 μg/mL). Results Of the five modulators of fibroblast function that we tested, butein was by far the most potent inhibitor of clonogenic growth of UACC-812 breast cancer cells co-cultured with fibroblasts. Pre-treatment of fibroblasts with concentrations of butein as low as 2.5 μg/mL nearly abolished subsequent clonogenic growth of UACC-812 breast cancer cells co-cultured with the fibroblasts. A similar dose of butein had no effect on the clonogenic growth of breast cancer cells cultured in the absence of fibroblasts. Significantly, clonogenic growth of the primary breast cancer cells was also significantly reduced or abolished when the tumor cells were co-cultured with fibroblasts that had been pre-treated with a fixed dose of butein. Conclusion We conclude that fibroblasts pre-treated with non-toxic doses of butein (a natural herbal compound) no longer support the clonogenic growth of small numbers of primary breast cancer cells seeded into co-cultures. These results suggest that interference with the interaction between fibroblasts and breast cancer cells by the natural herbal compound, butein, should be further investigated as a novel experimental approach for possibly suppressing the growth of micrometastases of breast cancer. | Background Butein (3,4,2',4'-tetrahydroxychalone-Figure 1 ), a plant polyphenol, is one of the major biologically active components of the bark and stems of Rhus verniciflua Stokes. In Far Eastern countries such as Korea, Japan, and China, the compound has been traditionally used for treatment of pain, thrombotic disease, gastritis, stomach cancer, and parasitic infections [ 1 , 2 ]. In Korea, it has also long been used as a food additive [ 2 ]. Figure 1 Chemical structure of butein. Recently, butein has been shown to possess potent activity against fibroblast function [ 3 ], possibly related to its ability to suppress differentiation of fibroblasts to myofibroblasts that are characteristically involved in wound healing [ 4 ]. Because fibroblasts and myofibroblasts are now believed to play a critical role in promoting the growth of cancer cells [ 5 , 6 ], we performed this study to determine if butein could suppress the growth of human breast cancer cells co-cultured with fibroblasts by interfering with the function of the fibroblasts. Methods Clonogenic assay The UACC-812 human breast cancer cell line (ATCC, Manassas, VA) was passaged in Leibovitz's medium supplemented with 15% fetal calf serum. Normal fibroblasts (CCD-1068SK, ATCC) obtained from the breast of a 65 year old female were passaged at 37°C in minimal essential medium (Eagle's) supplemented with 2 mM L-glutamine, Earle's balanced salt solution (1.5 grams/Liter), sodium bicarbonate, 0.1 mM non-essential amino acids, 1 mL sodium pyruvate, and 10% fetal calf serum in a 5% CO 2 atmosphere. All cell culture reagents were obtained from ATCC. Our co-culture experiments used confluent monolayers of fibroblasts that had been passaged no more than 21 days. This precaution assured that the fibroblasts were not senescent or transformed. We seeded 100 UACC-812 breast cancer cells into individual wells of a 96-well cell culture plate containing a confluent monolayer of fibroblasts growing in minimal essential growth medium supplemented as described above. At intervals of 3–4 days, fresh medium was added. After 14 days, the cells were fixed with 70% ethanol for 10 minutes prior to staining for 3 minutes with 0.1% toluidine blue. The wells were then washed with distilled water, and the numbers of colonies of tumor cells containing eight or more confluent cells were counted using inverted microscopy. Each experiment was performed in triplicate, and the means and standard deviations for each treatment and control group were then compared using a two-tailed, unpaired t-test. Co-culture with fibroblast modulators (Experiment 1) A monolayer of normal fibroblasts was seeded with 100 UACC-812 tumor cells/well containing complete culture medium (control) or medium supplemented with various modulators of fibroblast function. The fibroblast modulators that we tested included troglitazone (an activator of PPAR-γ in fibroblasts [ 7 ]; 1 μg/mL); GW9662 (an inhibitor of PPAR-γ[ 8 , 9 ]; 1 μM); butein (an inhibitor of myofibroblast differentiation; 2 μg/mL); meloxican (a COX-2 inhibitor in fibroblasts [ 10 , 11 ]; 1 μM); or 3,4 dehydroyproline (an inhibitor of collagen synthesis by fibroblasts [ 12 ]; 10 μg/mL). Because butein is relatively insoluble in aqueous solution, it was first dissolved in dimethylsulfoxide to produce a stock solution (10 mg/mL) that was then serially diluted into growth medium to produce the final desired concentrations of butein. At intervals of 3–4 days, old medium was removed and replaced with fresh medium containing the drugs. After 14 days, the numbers of colonies of 12 or more tumor cells were counted as described above. As a control to determine the effects of butein on the clonogenic growth of breast cancer cells in the absence of fibroblasts, we seeded 100 tumor cells into wells without fibroblasts but containing butein (2 μg/mL) Co-culture with fibroblasts pre-treated with butein (Experiment 2) In the next experiment, we sought to determine if pre-treatment of fibroblasts alone with various doses of butein would also inhibit clonogenic growth of breast cancer cells. A monolayer of fibroblasts in individual wells of 96-well plates was incubated for 3 days with growth medium containing serial, 2-fold dilutions of butein ranging from 10 μg/mL to 1.25 μg/mL. The adherent fibroblasts were then washed three times to remove any residual butein, and 100 UACC-812 cells were seeded per well. The co-culture was then incubated in culture medium without butein for 14 days, after which the numbers of colonies of tumor cells were counted as described above. Co-culture of primary breast cancer cells with fibroblasts (Experiment 3) This experiment was performed in order to determine if butein could also suppress the clonogenic growth of primary breast cancer cells obtained directly from clinical specimens of human breast cancer. After obtaining appropriate approval from the Institutional Review Board to perform the study in compliance with the Helsinki Declaration, we aseptically dissected small fragments of tumor tissue from five cases of invasive ductal adenocarcinoma of the breast. These specimens were submitted for routine diagnostic evaluation to the Surgical Pathology Department at UCI Medical Center (Orange, CA). The tissues were carefully minced into small pieces and then digested overnight at 37°C in collagenase II (900 U/mL; Sigma-Aldrich, St. Louis, MO) solution in cell culture medium. Epithelial cells and organoids were then isolated by differential centrifugation of the digest [ 13 ], washed, and counted. We then seeded 100 tumor cells onto monolayers of fibroblasts that had been pre-treated with butein (2.5 μg/mL) for 3 days as described in the preceding experiment. The co-cultures were then incubated in culture medium without butein for 14 days, after which the numbers of colonies of tumor cells were counted as described above. Results Clonogenic assay Co-culture of 100 UACC-812 human breast cancer cells on a monolayer of human fibroblasts without butein treatment typically yielded 50–75 distinct colonies of easily recognizable tumor cells on a background of fibroblasts (Figure 2 ). If a monolayer of fibroblasts was not used, an average of five small colonies of tumor cells was generally observed. Figure 2 Representative colony of tumor cells in clonogenic assay. Breast cancer cells co-cultured with fibroblasts typically formed clusters of confluent cells and were easily distinguished from the fibroblasts when stained with 0.1% toluidine blue. Original magnification 400×. Co-culture with fibroblast modulators The results of Experiment #1 are presented in Figure 3 . Only GW9662 and butein significantly (p < 0.01) reduced clonogenic growth compared to the control (cell culture medium alone with no drug). There was only one small colony of tumor cells visible in one of three wells containing butein. Surviving individual tumor cells were not visible in any of the wells containing butein. Notably, in the absence of fibroblasts, there was an average of 4 colonies of tumor cells regardless of the presence or absence of butein. Thus, butein had no detectable effect on the clonogenic growth of breast cancer cells grown in the absence of fibroblasts. Figure 3 Co-culture of breast cancer cells and fibroblasts with drugs that modulate fibroblasts. Compared to the Control (no drugs), butein and GW9662 almost completely eliminated clonogenic growth of the breast cancer cells. Co-culture with fibroblasts pre-treated with butein The results of Experiment #2 are presented in Figure 4 . There was no clonogenic growth when 100 UACC-812 breast cancer cells were seeded onto monolayers of fibroblasts that had been pre-treated for 3 days with butein at 10 or 5 μg/mL, and clonogenic growth was significantly reduced at butein concentrations as low as 2.5 μg/mL. Figure 4 Co-culture of breast cancer cells with fibroblasts pre-treated with butein. Pre-treatment of the fibroblasts with butein at concentrations greater than or equal to 2.5 μg/mL completely eliminated or substantially reduced clonogenic growth of co-cultured breast cancer cells, suggesting an indirect mechanism of action that interferes with the interaction between fibroblasts and breast cancer cells. Co-culture of primary breast cancer cells with fibroblasts pre-treated with butein The results of Experiment #3 using primary breast cancer cells are presented in Table 1 . For each of the 5 clinical specimens that we tested, there were significantly fewer colonies of breast cancer cells when the co-culture was performed in the presence of fibroblasts that had been pre-treated with butein. Table 1 Clonogenic growth of primary breast cancer cells co-cultured with fibroblasts pre-treated with butein (2.5 μg/mL). Clinical Sample # Mean number of colonies with butein pre-treatment (1 s.d.) Mean number of colonies without butein pre-treatment (1 s.d.) 1 0 (0) 13 (4)* 2 0 (0) 36 (9)* 3 0 (0) 6 (4) * 4 3 (2) 12 (5)* 5 7 (4) 29 (8)* *p value < 0.02 by two-tailed, unpaired t-test comparing means and s.d. of butein pre-treatment versus no pre-treatment Discussion Clonogenic growth of small numbers of breast cancer cells in vivo [ 5 , 6 ] as well as in our in vitro culture system appears to be critically dependent on the presence of fibroblasts. In this study, we have demonstrated that butein can suppress the clonogenic growth of breast cancer cells in vitro through an indirect mechanism that involves interfering with the function of co-cultured fibroblasts. Significantly, the concentration of butein that we found to be effective in these experiments (2.5 μg/ml) does not significantly reduce fibroblast viability [ 4 ] and is also non-toxic in animal experiments [ 1 , 2 ]. Moreover, this concentration of butein had no significant effect on the clonogenic growth of breast cancer cells cultured in the absence of fibroblasts. Previous studies have shown that higher concentrations of butein extract from Rhus verniciflua Stokes are directly cytotoxic in vitro to human colon adenocarcinoma cells and lymphoma cells [ 2 , 14 ]. Butein has also been shown to inhibit cell growth and induce apoptosis in murine B16 melanoma cells [ 15 ]. The mechanism of induction of apoptosis in tumor cells appears to be related to increased caspase-3 activity, decreased Bcl-2 expression, and increased Bax expression [ 16 ]. Butein has also been shown to have multiple other activities, including inhibition of epidermal growth factor receptors [ 17 ], tyrosine kinase inhibition [ 18 ], suppression of E-selectin expression [ 19 ], inhibition of tyrosinase enzymes [ 20 ], and inhibition of cyclooxygenase-2 [ 21 ]. Our results suggest that butein may also suppress the growth of tumor cells through a second, indirect mechanism that does not involve direct toxicity to the tumor cells themselves. At this time, however, the precise molecular mechanisms by which butein interferes with the interaction between fibroblasts and breast cancer cells remain undefined. Conclusion Clonogenic growth of small numbers of breast cancer cells co-cultured with fibroblasts pre-treated with butein is markedly reduced. These results suggest that butein can inhibit the growth of tumor cells through an indirect mechanism that interferes with the interaction between fibroblasts and breast cancer cells. Our results also suggest that the herbal compound butein should be further investigated as a potentially useful experimental approach for suppressing the growth of small numbers of breast cancer cells in early micrometastases. List of abbreviations PPAR-peroxisome proliferation activator receptor COX-cyclooxygenase Competing interests The authors declare that they have no competing interests. Authors' contributions MKS conceived and designed the study, performed the experiments with the primary tumor cells, analyzed the data from all experiments, and wrote this report. JT and GC performed the other experiments and reported the data. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC554991.xml |
521498 | Variable Surface Glycoprotein RoTat 1.2 PCR as a specific diagnostic tool for the detection of Trypanosoma evansi infections | Background Based on the recently sequenced gene coding for the Trypanosoma evansi ( T. evansi ) RoTat 1.2 Variable Surface Glycoprotein (VSG), a primer pair was designed targeting the DNA region lacking homology to other known VSG genes. A total of 39 different trypanosome stocks were tested using the RoTat 1.2 based Polymerase Chain Reaction (PCR). Results This PCR yielded a 205 bp product in all T. evansi and in seven out of nine T. equiperdum strains tested. This product was not detected in the DNA from T. b. brucei , T. b. gambiense , T. b. rhodesiense , T. congolense , T. vivax and T. theileri parasites. The Rotat 1.2 PCR detects as few as 10 trypanosomes per reaction with purified DNA from blood samples, i.e. 50 trypanosomes/ml. Conclusion PCR amplification of the RoTat 1.2 VSG gene is a specific marker for T. evansi strains, except T. evansi type B, and is especially useful in dyskinetoplastic strains where kDNA based markers may fail to amplify. Furthermore, our data support previous suggestions that some T. evansi stocks have been previously misclassified as T. equiperdum . | Background Surra is an animal disease occurring in Africa, Asia and Latin America, caused by Trypanosoma evansi . T. evansi belongs to the subgenus Trypanozoon , together with T. equiperdum and T. brucei . The parasite can infect different host species and is mechanically transmitted by different biting flies such as Tabanidae and Stomoxys as well as by vampire bats such as Desmodus rotondus [ 1 ]. Camels and horses are very susceptible to the infection and death can occur within weeks or months. Moreover, T. evansi infections of cattle and buffaloes usually lead to a pronounced immunosuppression resulting in an increased susceptibility to other opportunistic diseases such as Pasteurella and anthrax [ 2 ]. Diagnosis of a T. evansi infection usually starts with clinical symptoms or the detection of antibodies to T. evansi . Conclusive evidence of T. evansi infection, however, relies on detection of the parasite in the blood or tissue fluids of infected animals. Unfortunately, parasitological techniques cannot always detect ongoing infections as the level of parasitaemia is often low and fluctuating, particularly during the chronic stage of the disease [ 3 ]. As an alternative to parasitological tests, DNA detection based on PCR has been proposed. Trypanozoon specific primers have been designed previously: TBR primers which target a 177 bp repeat [ 4 ], pMUTEC primers targeting a retrotransposon [ 5 ] and ORPHON primers which target the spliced leader sequence [ 6 ]. Most of them have been tested on cattle [ 7 , 8 ], water buffaloes [ 9 ] or goats [ 10 ]. PCR tests for diagnosis of T. congolense and T. vivax infections exist as well [ 11 ]. The development of a PCR test that would be able to differentiate between the different members of the Trypanozoon subgenus still remains a challenging issue. For T. evansi infections, the only specific test available so far is based on the detection of a kinetoplast DNA sequence [ 12 , 13 ]. However, the existence of dyskinetoplastic trypanosomes such as T. evansi RoTat 5.1 [ 14 ] and E152 [ 12 ] casts doubt about the diagnostic potential of such tests to detect all infections caused by T. evansi parasites. Recently, Ventura et al . [ 15 ] developed a PCR (PCR-Te664) for the detection of T. evansi based on a Random Amplified Polymorphic DNA (RAPD) fragment. The taxon specificity of this PCR remains uncertain since it was only tested on nine T. evansi strains, one T. equiperdum , two T. b. gambiense and one T. b. rhodesiense . Following evidence that the variable epitope of RoTat 1.2 VSG is expressed by all T. evansi strains tested so far [ 16 ], and that the gene encoding RoTat 1.2 VSG is present in all T. evansi but not in T. brucei isolates [ 17 ], we designed primers derived from the sequence of this VSG cDNA. In this article we will present and discuss the results obtained with these primers and compare them to the results we obtained using the PCR-Te664. Results PCR RoTat 1.2 : taxon specificity The 39 different trypanosome stocks used in this study are listed in Table 1 [see Additional file 1 ]. They were derived from a wide range of hosts and from distinct geographical locations. In all PCR runs, RoTat 1.2 DNA was used as a positive control. As shown in Figure 1 , the RoTat 1.2 PCR yielded a 205 bp amplicon in the positive control (lane 1) as well as in all other T. evansi populations (lanes 3–8). Moreover, the same fragment was found in seven out of the nine T. equiperdum populations tested. The T. equiperdum BoTat 1.1 was PCR negative (lane 10), while the T. equiperdum OVI strain yielded a PCR product shorter than 205 bp (lane 11) probably due to mispriming. All other tested trypanosome populations, including six T. b. brucei , eight T. b. gambiense , five T. b. rhodesiense , two T. congolense , one T. vivax and one T. theileri , were negative. (lanes 18–40). As a negative control, a PCR-mix without template DNA was included (lane 2). Sequencing of the positive samples revealed that all amplicon were identical (data not shown). The weak band in OVI did not yield sufficient material to enable sequencing. Figure 1 PCR specificity results for the different Trypanosoma ( T. ) species and subspecies in this study. Lane 1 pos. control RoTat 1.2, Lane 2 neg. control, Lanes 3–8 ( T. evansi ) are, respectively, AnTat 3.1, STIB 816, Zagora I.17, Colombia, Merzouga 56, CAN 86 K; Lanes 9–17 ( T. equiperdum ) are, respectively, AnTat 4.1, BoTat 1.1, OVI, STIB 818, Alfort, Hamburg, SVP, Am. Strain, Can. Strain ; Lanes 18–23 ( T.b.brucei ) are, AnTat 1.8, AnTat 2.2, AnTat 5.5, KETRI 2494, TSW 196, STIB 348; Lanes 24–31 ( T.b.gambiense ) are, respectively, AnTat 9.1, AnTat 11.6, AnTat 22.1, NABE, SEKA, ABBA, LIGO, LiTat 1.6; Lanes 32–36 ( T.b. rhodesiense ) are STIB 884, STIB 850, AnTat 25.1/S, Etat 1.2/S, AnTat 12.1/S ; Lanes 37–38 ( T. congolense ) are IL1180, TRT 17; Lane 39 ( T. vivax ) is ILRAD 700 and Lane 40 ( T. theileri ) is MELSELE ; Lanes M 100 bp molecular marker (MBI Fermentas, Germany). PCR RoTat 1.2 : analytical sensitivity A tenfold dilution series (10 5 trypanosomes down to 1 trypanosome per 200 μl sample) of RoTat 1.2 trypanosomes in mouse blood was prepared to determine the analytical sensitivity of the PCR. As shown in figure 2 , the PCR was able to detect as few as 10 trypanosomes per PCR reaction, which corresponds with a lower detection limit of 50 trypanosomes per ml. In principle, this limit can still be lowered if a blood sample of 200 μl extracted with the QIAamp DNA mini kit is eluted in less than 200 μl. Figure 2 Analytical sensitivity of the RoTat 1.2 PCR. Lanes M 100 bp molecular marker (MBI Fermentas, Germany); lane 1: 10 5 trypanosomes, lane 2: 10 4 trypanosomes, lane 3: 10 3 trpyanosomes, lane 4: 10 2 trypanosomes, lane 5: 10 trypanosomes, lane 6: 1 trypanosome, lane 7: 0.1 trypanosome, lane 8: negative control. PCR-Te664 : taxon specificity To evaluate the RoTat 1.2 diagnostic system alongside other published methods, we compared our method to the PCR-Te664 method as published by Ventura et al. [ 15 ] using the same trypanosome stocks. The PCR-Te664 method yielded the expected amplicon in all seven T. evansi strains and in seven out of nine T. equiperdum . As with the RoTat 1.2 PCR only T. equiperdum strains OVI and BoTat 1.1 remained negative. Unexpectedly, four out of six T. b. brucei (AnTat 2.2, AnTat 5.2, TSW 196 and KETRI 2494) and two T. b. gambiense type II strains (ABBA and LIGO) tested positive in this PCR (data not shown). Discussion This study was initiated to develop a specific PCR test that would be able to distinguish T. evansi from the other members of the Trypanozoon subgenus. The study is an extension of the initial observation that the RoTat 1.2 VSG gene only is found in T. evansi and not in T. brucei strains [ 17 ]. This study mainly focused on the presence and expression of the RoTat 1.2 VSG gene in T. evansi rather than the use of this VSG in diagnosis of Salivarian trypanosomes. Previously, other research groups have used VSG genes as target sequences for PCR detection of T.b. gambiense infections (sleeping sickness). In these studies, five different primers derived from VSG genes, AnTat 11.17, LiTat 1.3, 117, 2 K and U2 were used in PCR screening of different trypanosome populations, originating from distinct geographical locations [ 18 - 20 ]. AnTat 11.17 based PCR tests were capable of distinguishing T.b. gambiense from T.b. brucei parasites from most foci of sleeping sickness in countries such as Nigeria, Cameroon, Côte d'Ivoire, R. P. Congo/Brazza. and Sudan. However, populations originating from the Moyo focus in North-west Uganda and from Cameroon were shown to be negative in AnTat 11.17 and in LiTat 1.3 (2 K) PCRs respectively. According to Bromidge et al . [ 18 ], this might be due to antigenic variation and genetic evolution of the VSG genes. On the other hand, the presence of 117 and U2 genes was shown to be a common feature among all T. brucei populations tested. In T. evansi , a similar phenomenon may occur in certain Kenyan isolates. A recent study by Ngaira et al . [ 21 ] pointed out that some T. evansi stocks in the Isiolo district in Kenya seem to lack the Rotat 1.2 VSG gene. It is believed that these stabilates belong to the T. evansi type B group. So far, this type of T. evansi has only been observed in this specific region in Kenya [ 22 , 23 ]. To our knowledge, all other T. evansi isolated elsewhere, are from the classical T. evansi type A group. Thus, we assume that, except for these few Kenyan strains belonging to the type B group, our PCR is specific for T. evansi . Compared to the PCR-Te664 presented by Ventura et al . [ 15 ], the PCR RoTat 1.2 seems to have a higher taxon specificity, since no reaction with T. b. brucei , nor with T. b. gambiense type II was observed. However, regarding T. equiperdum , both PCR test positive for the same seven T. equiperdum strains and are both negative for the BoTat 1.1 and OVI strains. Since the RAPD fragment (AF397194) shares no homology with the Rotat 1.2 VSG gene (AF317914) and is not found within the expression site of trypanosomes, both sequences can be considered as independent molecular markers. Based on the observations with both markers, it appears that on the genomic level the Botat 1.1 and the OVI strains are different from the other T. equiperdum and T. evansi strains. The observed analytical sensitivity with the RoTat 1.2 PCR is comparable to what was reported for the Te664 PCR (25 cells per reaction) [ 15 ]. The presence of a RoTat 1.2 specific DNA sequence in some T. equiperdum strains corresponds with the serological evidence that rabbits experimentally infected with these strains develop RoTat 1.2 specific lytic antibodies within 30 days post infection [ 24 ]. In contrast, rabbits infected with the BoTat 1.1 clone and the OVI strain, which are negative in the present PCR, did not produce specific antibodies to the RoTat 1.2 clone when tested in immune trypanolysis. This might be explained by the loss of the RoTat 1.2 gene in the OVI and the BoTat 1.1 strain. It is also possible that there has been a sequence drift at the sites where these primers could bind. However, we hypothesize that RoTat 1.2 VSG truly is T. evansi specific and that RoTat 1.2 PCR positive T. equiperdum strains are actually T. evansi and not T. equiperdum . Indeed, in a previous molecular characterization study using Random Amplified Polymorphic DNA (RAPD) and the Multiplex-endonuclease Genotyping Approach (MEGA) it appeared that the T. equiperdum collection is not as homogenous as previously believed and that the generally followed concept that T. equiperdum is very closely related to T. evansi and more distant from T. b. brucei , seems incorrect. From the cluster analysis on the available strains, it appeared that only two clusters can be identified: a homogeneous T. evansi / T. equiperdum cluster and a more heterogeneous T. b. brucei / T. equiperdum cluster [ 25 ]. Interestingly, all strains of that homogeneous T. evansi / T. equiperdum cluster are all PCR RoTat 1.2 VSG positive while the strains found in the more heterogeneous T. b. brucei / T. equiperdum cluster, in casu BoTat 1.1 and OVI are PCR RoTat 1.2 VSG negative. Conclusions PCR amplification of the RoTat 1.2 VSG gene is a specific marker for T. evansi strains, except T. evansi type B, and is especially useful in dyskinetoplastic strains where kDNA based markers may fail to amplify. Furthermore, our data support previous suggestions that some T. evansi stocks have been previously misclassified as T. equiperdum . Methods Trypanosome populations A total of 39 different trypanosome populations were used in this study. They belong to 39 stocks and six species, isolated from a variety of host species at distinct geographical locations (Table 1 [see Additional file 1 ]). Only three T. equiperdum strains, BoTat 1.1, OVI and STIB 818 are well documented, i.e. known origin and host. The other six are putative T. equiperdum , based on publications or on their use as reference strains in different national dourine reference laboratories [ 26 - 30 ]. Preparation of trypanosome DNA Procyclic trypanosome populations were grown in vitro in Cunningham's medium [ 31 ] and in the Kit for In Vitro Isolation (KIVI) [ 32 ]. Pure procyclic trypanosomes were obtained by repeated centrifugation (20 min., 2000 g) and sediment washes with Phosphate Glucose Sacharose buffer (PGS) (38 mM Na 2 HPO 4 .2H 2 0, 2 mM NaHPO 4 , 80 mM glucose, 100 mM sacharose, pH 8.0). Bloodstream form trypanosomes were expanded in mice and rats and were purified from the blood by di-ethyl-amino-ethyl (DEAE) chromatography [ 33 ], followed by repeated centrifugation (20 min., 2000 g) and sediment washes with Phosphate Buffered Saline Glucose (PSG) (38 mM Na 2 HPO 4 .2H 2 0, 2 mM NaHPO 4 , 80 mM glucose, 29 mM NaCl, pH 8.0). Trypanosome sediments were subsequently stored at -80°C. Twenty μl of trypanosome sediment (approximately 2.10 7 cells) were resuspended in 200 μl of Phosphate Buffered Saline (PBS) (8.1 mM Na 2 HPO 4 .2H 2 0, 1.4 mM NaHPO 4, 140 mM NaCl, pH 7.4) and the trypanosome DNA was extracted using the commercially available QIAamp DNA mini kit (Westburg, Leusden, The Netherlands), resulting in pure DNA in 200 μl of TE buffer. The typical yield of DNA extracted from a 20 μl pellet was 150 ng/μl or 30 μg total DNA. The extracts obtained were diluted 200 times in water and divided into aliquots of 2 ml in microcentrifuge tubes for storage at -20°C. For trypanosome dilution series, 180 μl of each heparinized blood sample were mixed with an equal volume of the Qiagen AS-1 storage buffer and subsequently extracted using the QIAamp DNA blood mini kit (Westburg, Leusden, The Netherlands) resulting in 200 μl of extracted DNA in Millipore water. Manipulation was performed according to the manufacturer's instructions. PCR RoTat 1.2 Primers were derived from the RoTat 1.2 VSG sequence (AF317914), recently cloned and sequenced by Urakawa et al . [ 17 ]. Using DNA sequence homology search programs to interrogate databases at TIGR (The Institute for Genomic Research) and GenBank, primer sequences were identified within the region (608–812 bp) lacking homology with any other known VSG sequence present in the databases. Primer sequences (in 5'-3' direction) and annealing temperatures were as follows: RoTat 1.2 Forward GCG GGG TGT TTA AAG CAA TA, T ann. 59°C and RoTat 1.2 Reverse ATT AGT GCT GCG TGT GTT CG, T ann. 59°C. Twenty μl of extracted DNA were mixed with 30 μl of a PCR-mix containing: 1 U Taq DNA recombinant polymerase (Promega, UK), PCR buffer (Promega, UK), 2.5 mM MgCl 2 (Promega, UK), 200 μM of each of the four dNTPs (Roche, Mannheim, Germany) and 0.8 μM of each primer (Gibco BRL, UK). All amplifications were carried out in a Biometra ® Trio-block thermocycler. Cycling conditions were as follows: denaturation for 4 min. at 94°C, followed by 40 amplification cycles of 1 min. denaturation at 94°C, 1 min. primer-template annealing at 59°C and 1 min. polymerization at 72°C. A final elongation step was carried out for 5 min. at 72°C. Twenty μl of the PCR product and ten μl of a 100 bp size marker (MBI Fermentas, Germany) were subjected to electrophoresis in a 2 % agarose gel (25 min. at 100 V). Gels were stained with ethidium bromide (0.5 μg/ml) (Sigma, USA) and analyzed on an Imagemaster Video Detection System (Pharmacia, UK). PCR Te-664 PCR on purified DNA samples was performed using primers and PCR conditions according to Ventura et al . [ 15 ]. Only the Taq DNA polymerase was purchased from another distributor, i.e. Promega (UK) instead of Gibco BRL (UK). Competing interests None declared. Authors' contributions FC carried out the molecular work and drafted the manuscript. MR and TU participated in the molecular analysis. PM, BG and PB participated in the design and co-ordination of the study. All authors read and approved the final manuscript. Supplementary Material Additional File 1 Table 1. Data on the different Trypanosoma ( T .) populations used in this study Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC521498.xml |
554760 | New York City HIV superbug: fear or fear not? | On February 11, 2005, the New York City Department of Health and Mental Hygiene announced that a city resident had recently been infected with a multi-drug resistant form of HIV and rapidly progressed to AIDS. The Health Commissioner, Thomas R. Frieden, called for increased vigilance against this new strain. Is this situation an emerging crisis or simply an unusual case report of rapid HIV progression? | On February 11, 2005, New York City (NYC) health officials announced the discovery of a "rare strain of multi-drug resistant HIV that rapidly progresses to AIDS." According to the NYC Department of Health and Mental Hygiene, a man in his mid-40s was diagnosed with HIV infection in December 2004. Shortly after his diagnosis, testing, at the Aaron Diamond AIDS Research Center in Manhattan, revealed that his virus was resistant to almost all anti-HIV therapeutics. Further, despite being infected for only 2–20 months, the man had developed AIDS. NYC Health Commissioner Thomas R. Frieden, MD, MPH, stated, "This case is a wake-up call. First, it's a wake-up call to men who have sex with men, particularly those who may use crystal methamphetamine...now, we've identified this strain of HIV that is difficult or impossible to treat and which appears to progress rapidly to AIDS." Dr. Frieden called on this community to help stop the spread of this and other drug resistant strains of HIV. He also called on NYC doctors and the public health system to improve HIV prevention counseling, to perform HIV drug resistance testing among treatment naïve, HIV + persons, and to improve anti-HIV drug adherence. At the 12 th Conference on Retroviruses and Opportunistic Infections (CROI) in Boston, Drs. David Ho and Martin Markowitz of the Aaron Diamond AIDS Research Center in Manhattan presented clinical and laboratory data regarding the NYC resident[ 1 ]. He had tested negatively for HIV-1 antibodies several times before and in May 2003. His total lymphocyte counts during these time points were repeatedly normal. Investigators believe that the NYC resident may have been infected in October 2004, when he, while on crystal methamphetamine, engaged in unprotected, receptive and insertive anal sex with multiple partners. In early November 2004, the NYC resident developed a febrile illness, and then in December 2004, he tested positive for antibodies against HIV. His personal physician, concerned over the possibility of recent acute HIV-1 infection, referred the NYC resident to Dr. Martin Markowitz. At the time of diagnosis, his CD4 + T-cell count was 80 cells/mm 3 and it has since fallen to less than 50. The NYC resident meets one criterion for the diagnosis of AIDS; his CD4 + T-cell count is less than 200 cells/mm 3 . His viral load has varied from ~100 K to 650 K/ml. The NYC resident's virus was tested and found to be resistant to all but two anti-HIV drugs, efavirenz (Sustiva ® ) and enfuvirtide (Fuzeon ® ; T20). This high degree of drug resistance existed before the NYC resident was treated with any anti-HIV compound. Is this case a harbinger of a new epidemic with this superbug or is it just an isolated, forme fruste of HIV infection? No one knows the answer to this question yet, but we do have plenty of data to suggest that the latter is the case. In people naïve to drug therapy, bone fide antiviral resistance is uncommon. A recent USA based study of treatment-naïve patients found that the prevalence of mutations associated with drug resistance was 8.8%[ 2 ]. This means 8.8% of the subjects' viruses tested positive by genotyping for 1 or more mutations associated with drug resistance. Having a single mutation associated with resistance does not necessarily make a virus drug resistant. For many drugs, HIV must contain several mutations to become resistant. This fact is true for most protease inhibitors (PIs) and for several nucleoside analogue reverse transcriptase inhibitors (NRTIs). Therefore, the overall level of drug resistance is well less than 8.8% reported in this study. Indeed, in this study, no significant resistance to protease inhibitors was seen. A similar study found the overall prevalence of drug resistance mutations was 8.3%, as also determined by genotyping[ 3 ]. However, when viruses containing these mutations were analyzed by phenotyping, only 39% demonstrated decreased reduced drug susceptibility. In other words, less than 3.5% of all isolates had phenotypic resistance. A commonly held view on why the level of drug resistance is low is that most mutations associated with drug resistance, also decrease viral fitness. Except the K103N reverse transcriptase mutation, which confers resistance to the three approved non-nucleoside inhibitors (NNRTIs), other mutations, associated with high level drug resistance, are thought to significantly decrease viral fitness. Theoretically, in the absence of drug pressure in a newly infected individual, wild-type virus is either selected for during transmission or the transmitted, resistant virus mutates back towards the fitter wild type. The current observation is that the vast majority of viruses in treatment-naïve patients are sensitive to almost all drugs. In addition to CD4, the virus, isolated from the NYC resident, uses CXCR4 or CCR5 to enter cells[ 1 ]. Such viruses, termed dual-tropic, are rarely seen in newly infected individuals. Typically, an R5 virus, which utilize CCR5, is the transmitted type. After years of infection, in approximately 50% of individuals, the viruses' tropism changes from CCR5 to CXCR4[ 4 ]. This phenotypic change is associated with an accelerated disease course. People, who are homozygous for the CCR5, 32-bp deletion, do not express functional CCR5 and have a high relative resistance to infection with HIV. In those Δ32 homozygotes, who have become infected with HIV (8 individuals reported), the disease course appears to be more rapid[ 5 ]. Most of these individuals had CD4 + T-cell counts less than 300 cells/mm 3 at the time of diagnosis. It is unclear why these individuals became infected, while the vast majority of Δ32 homozygotes remain uninfected. Possibly, these 8 individuals have some other aberration, which allows them to become infected with an X4 virus and, in turn, leads to an accelerated disease course. Perhaps, the NYC resident has a similar abnormality, which has lead to an increased rate of CD4 + T-cell depletion. Also at CROI this year, Drs. Stephen Gange and Alvarez Muňoz from Johns Hopkins University Bloomberg School of Public Health presented models of rapid HIV progression probability, based on two, large prospective cohorts[ 6 ]. These studies, the MACS and theWIHS, have been on-going for the past 21 and 11 years respectively and have collected longitudinal data on 391 seroconverters. In the pre-HAART era, the median time to AIDS was 8.3 years. Using the cohorts' data, Drs. Gange and Muňoz estimated the probability of clinical AIDS developing within 6–24 months or a low CD4 + T-cell count existing at the first visit after diagnosis (within the first 9 months of infection). Their model predicts that 7 in 10,000 patients develop clinical AIDS within 6 months of infection. This number increases to 45 in 10,000 after 12 months. Similarly, 10 in 10,000 HIV infected individuals have a CD4 + T-cell count less than 200 cells/mm 3 after 4.5–9 months of infection. Several reports of rapidly progressing HIV infection have been published. The rapid disease course of the NYC resident is rare, but hardly unique. To date, no cluster of rapid progressors has been described. All rapid progressors have been unrelated, either genetically or virologically. While multi-drug resistant (MDR) viruses may be overall less fit compared with wild-type, drug sensitive strains, MDR HIV still causes steady CD4 + T-cell depletion. Therefore, it is highly probable that the NYC resident has a genetic predisposition, which led to rapid progression, rather than a new strain of HIV-1, which is simultaneously super-aggressive and multi-drug resistant. Our experience at a large, inner city HIV clinic is in agreement with the above data. We do not see rapidly progressing, newly diagnosed individuals. We also do not see MDR HIV in our treatment-naïve patients. Review of our data does not reveal any evidence of MDR virus in persons, who have never been on therapy. To determine whether this "superbug" has spread to others, the NYC Department of Health is appropriately and aggressively investigating the sexual contacts of the NYC resident. The reason for the NYC Department of Health press release at this early point in the investigation is unclear. In the absence of a documented cluster of patients, should the entire health system react? No, we should wait for more information. I do agree that genotypic resistance testing for treatment-naïve HIV + patients is prudent, especially when the person is thought to have been infected within the past year. Of course, most patients do not know when they were infected, so we are testing each new patient. Given the availability of free, rapid testing for HIV in New Jersey, we are strongly encouraging any one with current or previous high-risk behavior to get tested and determine his/her HIV status. The best way to fight this disease is with knowledge: knowledge on one's infection status, knowledge on how to avoid becoming infected, and knowledge on how not to infect some one else. HIV is not the common cold. It is transmitted through well-described behaviors, predominantly sex, especially receptive anal intercourse, and intravenous drug use with shared needles. These behaviors can be modified to reduce or eliminate the risk of contracting HIV. Two recent studies conclude that universal testing for HIV is a cost effective way to combat this infection in the USA [ 7 - 9 ]. Outreach prevention education and widespread testing are probably more effective public health strategies than sensational press releases. Dr. Frieden's call for increased vigilance against drug resistant HIV implies that regular, old-fashioned HIV infection is not horrific enough. Any one who has seen this disease up close knows that is not the case. While we have partially effective therapies and we better understand its pathogenesis, HIV infection in this country, not as life threatening as it once was, remains quite life altering. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC554760.xml |
554990 | Sarcocystosis of chital-dhole: conditions for evolutionary stability of a predator parasite mutualism | Background For parasites with a predator-prey life cycle, the completion of the life cycle often depends on consumption of parasitized prey by the predator. In the case of such parasite species the predator and the parasite have common interests and therefore a mutualistic relationship is possible. Some evidence of a predator-parasite mutualism was reported from spotted deer or chital ( Axix axis ) as a prey species, dhole or Indian wild-dog ( Cuon alpinus ) as the predator and a protozoan ( Sarcocystis axicuonis ) as the parasite. We examine here, with the help of a model, the ecological conditions necessary for the evolution and stability of such a mutualistic relationship. A two – level game theory model was designed in which the payoff of a parasite is decided not only by alternative parasite strategies but also by alternative host strategies and vice versa. Conditions for ESS were examined. Results A tolerant predator strategy and a low or moderately virulent parasite strategy which together constitute mutualism are stable only at a high frequency of recycling of parasite and a substantial prey – capture benefit to the predator. Unlike the preliminary expectation, parasite will not evolve towards reduced virulence, but reach an optimum moderate level of virulence. Conclusion The available data on the behavioral ecology of dhole and chital suggest that they are likely to meet the stability criteria and therefore a predator-parasite mutualism can be stable in this system. The model also points out the gaps in the current data and could help directing further empirical work. | Background Preferential killing of sick and disabled prey individuals by the predator has been the focus of many ecologists working with different predator – prey systems. In a variety of prey predator systems, diseased or weaker animals are shown to be consumed in greater proportion by predators [ 1 - 5 ]. Increased susceptibility of parasitized prey to predation, or predator preference for parasitized prey is possible under a set of conditions [ 6 - 8 ]. Where the prey species is an intermediate host and the predator is the definitive host for a parasite species, the capture of prey is often an essential part of the life cycle. Therefore any mechanism that makes the prey susceptible to predation would enhance the parasite fitness. In such relationships the susceptibility induced by the parasite can be very specific towards the predator host [ 9 ]. A mutualistic relationship can be said to exist between a predator and a parasite [ 10 ] if the cost of harboring the parasite is less than the benefit of greater success in catching the prey [ 1 ]. Some evidence suggestive of predator-parasite mutualism comes from dhole or Indian wild dog ( Cuon alpinus ) and a protozoan parasite ( Sarcosystis axicuonis ) with chital or spotted deer ( Axis axis ) as the prey-host [ 1 , 11 ]. There can be a potential problem in such a mutualistic relationship. Low virulence of the parasite towards the predator host and parasite tolerance by the predator host are essential factors for the maintenance of a mutualistic relationship. However, it is possible that a virulent parasite can grow faster and invade a mild parasite population. On the other hand a parasite resistant predator can avoid the cost of parasitism but share the benefit of prey capture and therefore invade a tolerant population. Either of the events can destabilize the mutualistic relationship. It is essential therefore to examine the evolutionary stability of the mutualism. In a completely randomized distribution, a mild parasite population can be easily invaded by a virulent one and a tolerant predator can be invaded by a resistant one. Population viscosity, group selection and kin selection can alter the evolution of virulence [ 12 ]. The dhole – Sarcocystis system makes group selection and viscosity very likely factors in shaping the relationship [ 1 ]. The life cycle of the parasite is very short as compared to dhole life span. Dhole groups are stable and retain their territories over a long time. Dhole territories are large and encompass home ranges of several chital packs [ 13 , 14 ]. Therefore the protozoan harboured by a dhole pack is very likely to be recycled to the same pack. The benefit of the parasite is more likely to be gained by the same pack. The distribution of a parasite within a pack is shown to follow a consistent pattern in which only a few individuals carry most of the parasite load. On the other hand, one or two individuals in each pack are found to be parasite free. This suggests that within a pack there can be distribution of labor. [ 1 ]. A distribution of labor, in which some individuals do active hunting and some disseminate the parasite, can reduce the effective cost of carrying the parasite. Evolution can take a different route under such conditions. We examine here the effect of parasite recycling on the evolution of a predator-parasite mutualism, using a theoretical model. The model We consider two alternative strategies, namely mild and virulent, for the parasite (Table 1 ). The virulent parasite multiplies rapidly in the predator host and therefore enjoys greater success (v) and exerts a higher cost (x) on the predator host. The mild parasite exerts relatively low cost (y) on the predator host and gains a limited success (m). The predator has two alternative strategies, namely tolerant and resistant. A tolerant predator always harbors the parasite population whereas the resistant one attempts to resist or eliminate the parasite. However, since the parasite virulence mechanisms also evolve, there is a probability (p) that the parasite can infect a resistant predator. The predator gets an additional net benefit (z) from capturing a prey infected with the parasite as compared to capturing an uninfected prey. The prey can have only one viable strategy, that of becoming resistant to the parasite. The prey will not get any benefit by tolerating the parasite since it would make it more susceptible for predation. Therefore, we do not consider alternative prey strategies in the model. Table 1 Pay – off matrix for predator and parasite strategies. parasite predator Mild Virulent Tolerant Resistant parasite Mild 1-y fr(1-y)+(1-fr)(1-x) m p* m Virulent fr(1-x)+(1-fr)(1-y) (1-x) v p* v predator Tolerant -y -x z (1-fr) p *z + fr *z Resistant - p * y - p* x (1-fr)*z + fr (p* z) p* z The table differs from pay-off matrix tables for classical game theory models. The table accounts for two alternative strategies each for two different types of players namely parasite and predator. The pay-off of the parasite is not only decided by other parasites but also by the predator strategy and vice-versa. Therefore the complete pay-off of a mild parasite invading a virulent population in a tolerant host population is m * [fr(1-y)+(1-fr)(1-x)]. Others to be calculated similarly. If the parasite population consists of the mild type, they enjoy a fitness of 'm' from the tolerant host and 'p*m' from a resistant one. Since they exert a cost y on the host, there is erosion of the host resource. The host resource available to them is therefore (1-y). Similarly for a population of virulent parasites the mean fitness gain is v and the host resource available (1-x). A virulent host invading a mild population will gain a fitness of 'v' such that v > m. In the absence of recycling the host resource available to it would be (1-y). However with a frequency of recycling 'fr', the host resource would be, fr (1-x) + (1-fr)(1-y) Similarly, that for a mild parasite invading a virulent population would be, fr (1-y) + (1-fr)(1-x) If the predator population is tolerant the parasite will be harbored in large numbers and disseminated to the prey population. Since the parasitized prey is more susceptible to predation the predator gets a benefit 'z' of easy catching. A resistant population, on the other hand has a small probability 'p' of harboring the parasite. Therefore the benefit the predator gets would be 'p*z'. If a resistant predator invades a tolerant population, with the recycling factor 'fr', the benefit of prey capture would be, fr * p *z + z(1-fr). The benefit for a tolerant one invading a resistant population would be, fr * z + (1-fr)* p*z. We assume that 'v', the benefit for a virulent parasite by infecting single host is directly proportional to 'x' i.e the loss to the host from infection by virulent parasite. V = α *x similarly, m = α *y Results and discussion A mild parasite will be able to invade a virulent population if the pay-off to the mild invader is greater than that for the virulent population. When the predator population is tolerant, this condition is satisfied when, y * [fr (1-y)+(1-fr) (1-x)] > x (1-x) The condition under which a virulent invader is unable to invade a mild population is (1-y)* y > [fr *(1-x) + (1-fr)*(1-y)]*x Since x > y, for satisfying both these conditions, fr should be large, y should be moderate and x should be large. Thus selection would favour a moderate virulence in the parasite towards the predator host. Unlike our expectation, low virulence is unlikely to be stable. However, a mutualistic relation can remain if the prey capture benefit is sufficiently large. It can be easily seen that the above conditions remain unaffected even if the predator population is resistant. Considering predator strategies, a tolerant predator will be able to invade a resistant population in the presence of a mild predator if, p *z - p*y < (1-fr) p *z + fr*z - y i. e. p*fr * z - p* y < fr *z - y This condition will be satisfied if fr *z > y since p < 1. A resistant predator will be unable to invade a tolerant population since the necessary condition is z-y < (1-fr)*z + fr*(p*z) - (p * y) fr*z - y < p*fr*z - p*y This invasion is impossible if fr*z > y. If the parasite is virulent, the necessary condition would be fr*z > x. Since x is assumed to be large, a resistant predator would be stable if the prevalent parasite strategy is virulent. Thus when fr and z are large and y is small mutualism would be stable. When this condition is not satisfied the predator will evolve resistance to the parasite and the parasite will evolve greater virulence. Conclusion A large recycling frequency (fr) appears to be the only critical factor in the evolution of parasite virulence. However, the parasite is unlikely to evolve towards low virulence. There will be a moderate virulence optimum. For a net benefit to the host the cost associated with this level of virulence should be less than the benefit in terms of ease of prey capture. For the evolution and stability of the tolerant strategy in the predator a large fr as well as large z and small y are necessary. A predator-parasite mutualism therefore critically depends upon these factors, whereas it is independent of p and α. In the case of Dhole-chital – Sarcocystis system, these condition are very likely to be satisfied. Field data show that the frequency of sarcocystosis of the heart in dhole kills was approximately double that of chital dyeing of other causes [ 1 ]. This suggests a substantially large z. Dhole have large, stable and defended territories upto 80 Km 2 [ 13 ]. There is only marginal overlap between neighboring packs unlike the overlap in tigers [ 15 ]. The territory of a dhole pack encompasses the home ranges of several chital groups. The home ranges of chital groups are small and stable [ 13 , 16 ]. This can ensure a large 'fr'. Unlike wolves, which tend to defecate more on the boundaries of the territory [ 17 ], the frequently used defecation sites of dhole tend to be towards the center of their home range and close to the major hunting areas [ 18 ], further ensuring a large 'fr'. The intensity of intestinal infection is reflected in the density of sporocyst in Dhole scat. Dhole shading large numbers of sporocysts of S. axicuonis show no apparent symptoms of disease or abnormality [ 19 , 20 ]. This indicates that the virulence of S. axicuonis towards dhole is low. Further, the division of labor can substantially reduce the effective cost of carrying the parasite. Currently we are unable to quantify these parameters empirically and have no estimate of the actual cost of harboring the parasite. Therefore we are unable to state quantitatively that all the necessary conditions for mutualism are met by the system. The importance of the model is that it helps us identify the gaps in the data and thus orient future empirical work. Although we are far from having an empirical estimate of y, fr and z, the known ecology of chital and dhole suggest that fr and z could be sufficiently large. This makes the system a likely candidate for the evolution of predator parasite mutualism. Any other predator-prey system that satisfies these conditions is also likely to co-evolve with some parasite species towards a predator-parasite mutualism. Parasites of diverse taxa have evolved predator-prey life cycles and any of them could be possible candidates for a mutualism. Predator-prey-parasite systems that satisfy the following three criteria are the most likely candidates for a stable mutualistic relationship: i) parasitized prey individuals are killed with substantially greater frequency by the predator ii) pathogenicity of the parasite towards the predator host is low or moderate iii) there is a high rate of parasite recycling to the predator host. We need to look at a number of systems that could satisfy these criteria. The chital-dhole- Sarcocystis system may not be unique and many possible examples of predator-parasite mutualism may be present in nature. Authors' contributions Both authors have contributed approximately equally to the model development. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC554990.xml |
520745 | Enhancer trapping in zebrafish using the Sleeping Beauty transposon | Background Among functional elements of a metazoan gene, enhancers are particularly difficult to find and annotate. Pioneering experiments in Drosophila have demonstrated the value of enhancer "trapping" using an invertebrate to address this functional genomics problem. Results We modulated a Sleeping Beauty transposon-based transgenesis cassette to establish an enhancer trapping technique for use in a vertebrate model system, zebrafish Danio rerio . We established 9 lines of zebrafish with distinct tissue- or organ-specific GFP expression patterns from 90 founders that produced GFP-expressing progeny. We have molecularly characterized these lines and show that in each line, a specific GFP expression pattern is due to a single transposition event. Many of the insertions are into introns of zebrafish genes predicted in the current genome assembly. We have identified both previously characterized as well as novel expression patterns from this screen. For example, the ET7 line harbors a transposon insertion near the mkp3 locus and expresses GFP in the midbrain-hindbrain boundary, forebrain and the ventricle, matching a subset of the known FGF8-dependent mkp3 expression domain. The ET2 line, in contrast, expresses GFP specifically in caudal primary motoneurons due to an insertion into the poly(ADP-ribose) glycohydrolase (PARG) locus. This surprising expression pattern was confirmed using in situ hybridization techniques for the endogenous PARG mRNA, indicating the enhancer trap has replicated this unexpected and highly localized PARG expression with good fidelity. Finally, we show that it is possible to excise a Sleeping Beauty transposon from a genomic location in the zebrafish germline. Conclusions This genomics tool offers the opportunity for large-scale biological approaches combining both expression and genomic-level sequence analysis using as a template an entire vertebrate genome. | Background Human, mouse and rat genomes likely have less than 40 000 genes each [ 1 - 4 ]. This is only two to three times as many genes as in Caenorhabditis elegans and Drosophila melanogaster , and only six times as many as Saccharomyces cerevisiae [ 5 - 7 ]. The increased complexity of vertebrates therefore can not be simply accounted for by a larger gene number. A part of the increased complexity is thought to be accomplished by alternative splicing, RNA editing and the use of protein modifications to generate a variety of protein products from a single gene, but everything starts with increased complexity at the level of transcriptional regulation. While promoters are relatively simple and short in yeast, their complexity increases in multicellular organisms, making regulatory sequences ever harder to identify. In humans, enhancer elements can be located over a megabase away from the transcriptional start site [ 8 ]. Furthermore, current gene prediction programs used to annotate genomes often fail to correctly identify the 5' start site of a transcription unit, making in silico analysis of the regulatory sequences even more complex. To further complicate the matter, enhancer sequences diverge in evolution, co-evolving with their respective transcription factors, and often do not work across large evolutionary distances – worm to fly, for example [ 9 ]. This makes information from non-vertebrate model systems sometimes inapplicable to vertebrate sequences. Enhancer detection ("trapping") using insertion site context vectors was popularized as a genomics tool in Drosophila . The first fly enhancer trap vectors were based on the P element transposon and often used the transposase's own promoter fused to the beta-galactosidase reporter gene for enhancer detection [ 10 - 12 ]. Several of the enhancer trap lines were shown to express the Lac Z reporter in cells corresponding to the expression patterns of nearby genes, validating the approach [ 12 , 13 ]. In other work, promoters such as engrailed, fushi tarazu and Hsp70 were successfully developed for enhancer trapping in the fruitfly [ 14 - 16 ]. Further modifications to the system included the implementation of a bipartite system with a Gal4 transactivator [ 17 ], green fluorescent protein (GFP) [ 18 ], and even a GFP- Lac Z fusion protein [ 19 ] as reporters. In addition to the P element, other transposons such as hobo and piggyBac with insertion site preferences distinct from those of the P element have been used in Drosophila [ 20 , 21 ]. The availability of a variety of transposons, promoters and reporters for enhancer trapping in the fruitfly enabled researchers to obtain enhancer trap insertions into a considerable fraction of Drosophila genes (reviewed by [ 22 ]) and allows an investigator to choose vectors most suitable for the problem at hand. The ability to excise from a genomic location has been instrumental to the utility of P element based vectors. For mutation-causing insertions, reversion of the mutant phenotype by P element excision proves that a given insertion causes the mutation. Since the mutagenicity of Drosophila enhancer trap transposons is not significantly higher than the average 15% rate obtained with regular P elements, most insertions do not result in a mutation [ 22 ]. In these instances, the P element's ability to induce genomic deletions by "imprecise excision" can be used to obtain a mutation in the neighboring gene(s) [ 23 ]. The success of enhancer trapping in Drosophila prompted application of this approach in the mouse [ 24 - 27 ]. As was the case in Drosophila , the lac Z reporter was shown to be expressed in part of the target gene's expression domain [ 28 ]. Despite the considerable success of these early experiments, enhancer detection as an experimental approach in mouse was not explored further, giving way to different versions of gene traps (for a review, see [ 29 ]). We believe the success of enhancer trapping in Drosophila can be largely attributed to the advantages of this experimental system over mouse. In Drosophila , large numbers of transgenic organisms can be readily generated and screened for gene expression patterns. It is far less practical in the mouse. This is partly due to the availability of efficient and precise transgene delivery tools in the fruitfly: the native P element, hobo and piggyBac transposons. In contrast, early mouse experiments were carried out by non-facilitated DNA transgenesis. This approach is less efficient and prone to induce deletions and other genome rearrangements in the recipient locus, as noted in the first published mouse enhancer trap locus [ 25 , 30 ]. The compact nature of the Drosophila genome also contributed to the success of enhancer trapping, making the path from an enhancer trap insertion to the identification of the affected gene straightforward, especially once the Drosophila genome was sequenced. The zebrafish Danio rerio is a vertebrate model system that provides many of the advantages found in invertebrates. A few hundred transparent, externally developing embryos can be obtained from a single pair of fish per week. The zebrafish genome is about two-fold smaller than the mouse genome, and its sequencing and annotation are nearing completion. Finally, transposon tools for efficient and precise transgene delivery into the zebrafish genome are available. We focused our research on the Sleeping Beauty (SB) transposon system [ 31 , 32 ]. While not as efficient as the highest titer retrovirus used in zebrafish [ 33 , 34 ], the Sleeping Beauty transposon system offers advantages in expression as well as ease of construction and testing of diverse vectors that can be done using basic molecular biology tools. Furthermore, the SB system offers the possibility of transposase-induced excision out of the genome to induce local deletions or to revert possible mutant phenotypes. In this report, we investigated the potential of the SB transposon system for enhancer detection in zebrafish. Our results indicate that zebrafish enhancer trap lines with diverse GFP expression patterns can be readily generated using the SB system. Most of the obtained lines harbor a single transposon insertion event, facilitating the rapid identification of transposon insertion sites responsible for specific GFP expression patterns. We show that two enhancer trap lines exhibit GFP expression patterns matching the expression patterns of the target genes, and that both expected and novel gene expression patterns can be identified using this genomics tool. We conclude that enhancer trapping using the Sleeping Beauty transposon system is a viable experimental approach using as template a vertebrate genome. Results The Sleeping Beauty transposon can detect enhancers in cis We have previously established multiple zebrafish lines using SB transposons with ubiquitous and tissue-specific promoters driving reporter expression [ 32 ]. Surprisingly, we did not observe any dependency of the expression pattern on the genomic context of the transposon insertion. Multiple studies describing insertion-site dependent transgene expression in vertebrates have suggested that many of those events are due to the transgene falling under control of nearby enhancers [ 35 - 40 ]. For enhancer detection approaches it is imperative that the reporter gene be sensitive to neighboring transcriptional regulatory elements. At least three explanations can be put forward to explain the absence of expression patterns in our previous work in zebrafish. First, the Sleeping Beauty transposons are flanked by relatively large inverted repeats. These repeats might function as silencer elements and not allow for transcriptional regulation across them. Second, the promoter we used ( Xenopus laevis EF1α, [ 41 ]) may not be subject to transcriptional regulation by tissue-specific enhancers. Third, the expression level from the selected promoter may be too high to be effectively modulated, as enhancer traps usually contain attenuated promoters. To test these hypotheses, we decided to produce an artificial enhancer trapping event by cloning a tissue-specific promoter / enhancer just outside the inverted repeats and test for an increase in tissue-specific expression in injected embryos (Figure 1 ). We started with the transposon used in our previous work, pT2/ S1 EF1α-GM2, which contains a shortened version of the Xenopus laevis EF1α promoter driving the GM2 version of GFP in a pT2 transposon vector [ 32 ]. We took advantage of the observation that relatively few pT2/ S1 EF1α-GM2 injected embryos express GFP in the eye. We added a lens-specific Xenopus laevis γ1-Crystallin promoter [ 42 ] to the pT2/ S1 EF1α-GM2 vector, as we had previously shown that this promoter specifically expresses in the lens of injected (F0) and transgenic (F1) zebrafish [ 32 ]. Embryos injected with pT2/ S1 EF1α-GM2 or γ Cry1/pT2/ S1 EF1α-GM2 were scored for any GFP fluorescence and for eye-specific GFP fluorescence at 3 dpf (Figure 1 ). The addition of γ Cry1 to pT2/ S1 EF1α-GM2 caused a modest (two-fold) increase in injected embryos showing any GFP fluorescence. In contrast, the increase in eye-specific GFP expression was ten-fold (Figure 1 ). We concluded that at least in this assay, the Sleeping Beauty inverted repeat sequences do not block transcriptional regulation and that the EF1α promoter can be subject to transcriptional regulation from external, tissue-specific regulatory sequence elements. Figure 1 Artificial enhancer trapping with a Sleeping Beauty transposon. Comparison of GFP expression in embryos injected with pT2/ S1 EF1α or γCry1/pT2/ S1 EF1α. Plasmids are diagramed as cartoons on the top of the picture. The SB transposon's inverted repeats are shown as boxes with open triangles, and the GFP open reading frame is depicted as a grey arrow. The gamma-Crystallin promoter/enhancer is shown as a hatched box. DNA-injected embryos which survived to 3 dpf were counted and scored for GFP fluorescence anywhere in the embryo (any GFP) and for fluorescence in the eye (eye GFP), even if there was additional fluorescence elsewhere. The average percentage of embryos positive for particular GFP fluorescence in three independent experiments is shown ± standard deviation. Promoter truncations and pilot screens We next tested the hypothesis that the absence of expression patterns in our previous work was due to the fact that S1 EF1α is expressed too strongly in transgenic animals. Since most successful enhancer traps in Drosophila and mice were based on truncated or weak promoters, we decided to attenuate the S1 EF1α promoter by removing sequences upstream of Bst 1107I ( S2 EF1α) and Eco RI ( S3 EF1α) restriction enzyme sites (Figure 2 ). We then co-injected the corresponding transposon constructs with SB10 transposase mRNA to assess germline transmission, expression and enhancer trapping rates. In pilot experiments, progeny from over 20 fish were screened with each construct. While overall germline transmission and expression rates were comparable (Figure 2 ), there was a difference in the expression patterns of the two constructs in transgenic embryos. Most of the transgenic animals generated with pT2/ S3 EF1α-GM2 exhibited weak GFP expression, and we could not detect any expression patterns (data not shown). In contrast, when pT2/ S2 EF1α-GM2 was used, most of the GFP-positive fish exhibited fairly strong, ubiquitous expression. Closer analysis indicated that many of these "ubiquitous" expression patterns were rather unique, with GFP expression often noted to be particularly strong in some tissues, consistent with a tissue-specific expression pattern superimposed on a ubiquitous expression pattern (data not shown). In most cases, GFP expression segregated as a single integration event, indicating that a tissue-specific expression pattern was not likely being masked by a ubiquitous expression pattern from a different insertion event. A similar phenomenon has been observed with Drosophila enhancer traps [ 11 ]. We speculate that in those instances, the GFP expression cassette may have fallen under the control of multiple enhancers – some tissue-specific, some ubiquitous. Alternatively, the ubiquitous expression may stem from the ubiquitous activity of the EF1α promoter used in the screen, with tissue-specific enhancers only elevating the expression levels in certain tissues, but not restricting it. We did not consider such expression patterns valuable and did not establish any fish lines with such GFP expression. Importantly, one of the founders in the pT2/ S2 EF1α-GM2 pilot screen yielded three kinds of GFP expression in its progeny. Some were ubiquitously GFP positive, some showed a hatching gland-specific expression profile, and some exhibited both. When three F1 fish with both ubiquitous and hatching gland specific expression were raised and outcrossed, the two expression patterns exhibited independent segregation: 24% of the embryos were GFP negative, 26% expressed GFP ubiquitously, 25% had hatching gland-specific GFP expression, and 25% had both hatching gland-specific and ubiquitous expression (n = 245). Independent segregation indicates that the two transposon insertions causing the two expression patterns are unlinked. Two independent integration events were confirmed by Southern hybridization and inverse PCR (data not shown). The hatching gland-specific GFP expressing embryos were used to establish our first enhancer trap line, ET1 (Figure 3A ). We concluded from our pilot screens that pT2/ S2 EF1α-GM2 demonstrated the desired properties for potential use in enhancer trapping studies. Figure 2 EF1α promoter truncations and endogenous enhancer trap screening. A diagram of the S1 EF1α promoter [32, 41]. Restriction enzyme sites are shown on top as single letters. S is Sph I, N is Nhe I, B is Bst 1107I and R is Eco RI. G/C, G and C rich box. Sp1, Sp1-like site. TATA, TATA box. Numbering below is relative to the first T of the TATA box. The table below the diagram shows the results of the pilot and scale-up (*) screens. Transgenesis and expression rates are shown, non-expressing transposon insertions were not scored. Transgenesis and expression rate from scale-up screen ( # ) is an underestimate since many founders were screened by incross and crosses from doubly transgenic founders were scored as a single transmission event (see text). Figure 3 Enhancer trap lines exhibit a variety of unique GFP expression patterns. (A). Lateral view of GFP expression in Enhancer Trap line 1 (ET1) at 38 hours post fertilization (hpf). (B) ET3 at 5–6 somite stage. (C) ET3 at 36 hpf. (D) ET4 at 26 hpf. (E) ET5 at 30 hpf. (F) ET5 at 48 hpf. (G) ET6 at 26 hpf. (H) ET7 at 32 hpf. (I) Ventral view of ET7 at 5 dpf. (J) Lateral view of ET8 at 26 hpf. (K) Dorsal view of ET9 at 28 hpf. (L) Lateral view of ET9 at 30 hpf. In all panels, anterior is to the left. See text for details. Germline excision of a Sleeping Beauty transposon insertion We have previously demonstrated the excision of a Sleeping Beauty transposon from the genome in somatic tissues of transposase-injected zebrafish embryos [ 43 ]. We tested if such an excision event could be inherited by examining transposon excision in the germline. Embryos homozygous for the ET1 insertion were injected with SB10 transposase mRNA, and while some were used for a somatic excision assay the rest were raised to test for germline transmission of an excision event. A PCR reaction on genomic DNA from transposase-injected embryos with primers flanking ET1 insertion point produced two bands. A large band corresponded in size to the transposon insertion allele, and a small band corresponded to a transposon-less allele (data not shown). Both cannonical Sleeping Beauty transposon footprints (ATGTCAT and ATGACAT, [ 44 , 45 ]) were obtained upon cloning and sequencing of the smaller band, indicating a transposase-mediated excision and DNA repair. 26 fish were screened for germline transmission (see Materials and Methods), and one was shown to transmit the expected excision footprint. We conclude that the Sleeping Beauty transposon can be excised from a genomic location in the zebrafish germline. pT2/ S2 EF1α-GM2 scale-up screening: 10% of GFP-expressing integrations yield tissue-specific patterns One tissue-specific expression pattern was recovered from our pilot screen. We sought to recover more patterns and to test if enhancer detection in zebrafish is amenable to scale-up. To that end, we co-injected 3248 zebrafish embryos with the pT2/ S2 EF1α-GM2 and SB10 transposase mRNA mix. 2102 embryos survived to day 3 for scoring, of which 848 were mosaic GFP positive and were selected to be raised. 330 survived to adulthood and were screened for germline transmission of GFP expression, primarily by sibling incrossing. This approach provided a lower estimate of the transgenesis and expression rate because it does not distinguish instances were both parents are transgenic. In this screen, at least 80 of the founder fish produced GFP-expressing progeny resulting in a minimum estimate of a 24% transgenesis rate. The actual transgenesis rate is closer to 30% because most of the fish were screened by incross, and if a pair produced GFP-expressing progeny, only one parent was counted as a transmitter. Eight of the GFP-expressing fish displayed distinct GFP expression patterns (Figure 3 ). Together with the pilot screen, 9 tissue-specific expression lines were obtained from 90 transgenic founder fish (10%) using the pT2/ S2 EF1α-GM2 transposon. Recovered expression patterns label a diverse array of tissues during embryogenesis GFP expression in ET1 can be first observed in the polster region at 7–8 somite stage (not shown). The expression is very pronounced between 20 and 40 hours post-fertilization (hpf), when it marks the hatching gland (Figure 3A ). Expression disappears as the hatching gland is resorbed. Line ET3 represents a pattern with the earliest onset of expression. Anterior localization of GFP in the diencephalon is detected by 5–6 somite stage in this line (Figure 3B ). Extremely bright anterior expression persists in the ventral diencephalon (Figure 3C ) and by 6 days post-fertilizations (dpf) is restricted slightly more posterior in the midline. The onset of expression for ET4 is 18 hpf with a bilateral expression pattern in cranial sensory ganglia that remains strong until 2 dpf and is undetectable by 5 dpf. This anterior expression in ET4 seems to label the lateral line ganglia both anterior and posterior to the otic vesicle (Figure 3D ), however, no expression is detected in the lateral line in the trunk. In ET5 a single bilateral patch of strong GFP expression in the hyoid arch is observed by 24 hpf (Figure 3E ), that by 48 hpf marks a more anterior location in the embryo (Figure 3F ). Expression in this line is greatly diminished by 3 dpf and is undetectable by 5 dpf. Strong GFP expression is observed in ET6 by 26 hpf as a bilateral expression pattern consisting of two distinct patches in a subset of cranial sensory ganglia/placodes (Figure 3G ). The expression weakens by 2 dpf and is undetectable by 3 dpf. GFP expression in ET7 begins weakly in the midbrain-hindbrain boundary (MHB) at 12–14 somites with the most pronounced expression in the anterior side of the MHB detected by 26 hpf (Figure 3H ). Robust expression in the heart is first detected at around 32 hpf and remains ventricle-specific through 5 dpf (Figure 3I ), even though expression in the brain is no longer restricted to the MHB. GFP expression in ET8 is already localized by 10–12 somites and remains strong in the telencephalon, and posterior side of the MHB through 26 hpf (Figure 3J ). By 3 dpf the localized anterior expression is undetectable over autofluorescence, however, caudal expression appears to be enhanced in the dorsal neural tube. The onset of expression in ET9 occurs around 22 hpf and is difficult to detect by 2 dpf. At 28–30 hpf (Figure 3K,3L ), three distinct expression domains are apparent in the telencephalon, diencephalon and hindbrain of ET9. The ET2 line expresses GFP specifically in the motoneurons We analyzed the ET2 line in detail because of the highly specific expression of GFP in these fish. GFP expression was first observed at the 16 somite stage, when 2 bilateral cells in the spinal cord of the 10 anterior somites become GFP positive (Figure 4 ). At later stages, multiple cells per somite become GFP positive, either due to continued expression of GFP mRNA or due to segregation of GFP to daughter cells. GFP expression follows the wave of somitogenesis, with the posterior-most somites lagging in GFP expression. At about 24 hours, ventrally-projecting axons become visible by GFP fluorescence. Later yet a pattern of nodes appears along the axons (Figure 4 ). Based on the position of neuronal cell body and the axonal trajectory, we conclude that caudal primary motoneurons express GFP in this line [ 46 ]. To our knowledge, this is the first gene to be specifically expressed only in this subpopulation of motoneurons. We therefore sought to identify the locus tagged by this transposon insertion. Figure 4 The ET2 transgenic fish line expresses GFP in caudal primary motoneurons. GFP expression in ET2 was visualized in motoneurons using a bandpass GFP filter set at various stages of embryonic development. In all panels anterior is to the left. (A) The onset of GFP expression in ET2 line at 16 somite stage. (B) 26 somite stage. (C) 24 hpf. (D, E) 36 hpf. Axonal trajectories are visible at 24 and 36 hpf. Southern analysis indicated that there is a single transposon insertion in this line, and it is linked to GFP expression (Figure 5 ). Inverse PCR identified a transposase-mediated insertion into a TA dinucleotide at position 256083 on contig ctg9701 (zebrafish genome assembly Zv3). The insertion occurred into a Genescan-predicted gene. Further analysis indicated that the Genescan-predicted gene actually consists of parts of at least two different genes, myoferlin and poly(ADP-ribose) glycohydrolase (PARG). The insertion located in the PARG part of the predicted transcript, 649 nucleotides from an exon just upstream of the PARG catalytic site. To confirm that the transposon insertion into the PARG gene induced GFP expression in primary motoneurons, we prepared genomic DNA from both GFP positive and GFP negative embryos from an independent outcross, and we conducted a PCR with NeuroIns-F1 and NeuroIns-R1 primers specific to the flanking sequences. In GFP negative embryos, a 0.5 kb band corresponding to wild type locus is noted. In GFP positive embryos, the same band is seen in addition to a larger band of approximately 2.4 kb, corresponding to a locus with transposon insertion (Figure 5 ). Since the inverse PCR and confirming PCR was performed on DNA from different batches of embryos, we can exclude the possibility of DNA contamination or fish husbandry error and conclude that the enhancer trap transposon insertion into the PARG gene causes GFP expression in caudal primary motoneurons. Figure 5 Identification of the transposition event in the ET2 line. (A) The pT2/ S2 EF1α transposon insertion into zebrafish genome is shown; restriction enzyme sites and primers used for molecular analysis are indicated. Transposon IR/DR's are shown as solid boxes with open triangles, and the GFP open reading frame is shown as a grey arrow. Genomic DNA is shown as a dotted line. N is Nsi I, E is Eco RV. (B) Southern blot on ET2 line outcross embryos. DNA from GFP positive (lanes 1 and 2) and GFP negative (lanes 3 and 4) embryos was digested with Nsi I (lanes 1 and 3) or Eco RV (lanes 2 and 4) and probed with a GFP-specific probe. (C) Linkage of the transposon insertion event to GFP expression. Primers flanking the transposon insertion event (arrows) were used to conduct PCR on DNA from GFP positive (lane 2) and GFP negative (lane 3) embryos from an ET2 outcross different from the one used in (B). Lane 1, λ Eco 47III Marker (Fermentas Inc). GFP expression in ET2 line matches the expression of the endogenous PARG gene Poly(ADP-ribosyl)ation is a protein modification that is extensively studied at the biochemical level and is associated with changes in DNA replication, recombination, repair and transcription [ 47 ], for a review, see [ 48 ]. Recently poly(ADP-ribosyl)ation was demonstrated to have a role in long term memory in the sea slug Aplysia [ 49 ]. Most organisms have multiple genes for poly (ADP-ribose) polymerases but only a single known gene for poly (ADP-ribose) glycohydrolase [ 48 ]. PARG activity is noted to be expressed in many cell lines, among them neuronal [ 50 - 53 ], but the tissue specificity of PARG expression during embryogenesis has not been reported for any organism. To test if the pT2/ S2 EF1α-GM2 enhancer trap recapitulates the expression pattern of the endogenous PARG gene, we conducted whole mount in situ hybridization on ET2 outcross embryos to compare PARG and GFP reporter expression (Figure 6 ). In situ hybridization visualizes axonal cell bodies, the position of which appears indistinguishable with both PARG and GFP probes. We therefore conclude that GFP mRNA expression in this enhancer trap line faithfully recapitulates the expression of the zebrafish PARG gene during embryogenesis. Figure 6 GFP expression in ET2 line embryos is indistinguishable from endogenous PARG gene expression. 23 hpf embryos collected from a heterozygous outcross were photographed for GFP fluorescence and sibling embryos were fixed for in situ hybridization. (A) In situ hybridization with PARG antisense probe. (B) In situ with GFP antisense probe. (C) Visualization of GFP expression in living embryos using a bandpass GFP filter set. (D) The same embryo as in (C) photographed using a bandpass GFP filter set with a low level of bright field illumination to visualize GFP expression in relative position to the somites. Molecular analysis of other enhancer trap lines identifies target genes We characterized insertion events in other enhancer trap lines. GFP positive F1 or F2 fish were outcrossed, and embryos were sorted into GFP positive and GFP negative pools. Genomic DNA was prepared from each pool, and Southern analysis was conducted to assess transposon copy number and linkage to the GFP expression pattern. In all lines except ET1 (see above), a single GFP expression-linked transposon insertion event was detected by Southern hybridization. We then conducted inverse PCR analysis on the DNA from GFP positive embryos to identify the insertion locus. For verification, DNA from embryos from an independent outcross was prepared and PCR was run with primers flanking the insertion site to link GFP expression and transposon insertion at a particular locus. Verified enhancer trap loci are presented in Table 1 . Notably, seven of the insertions have occurred into introns of Genescan-predicted genes. Four of the tagged genes show significant homologies to previously identified genes: PARG (see above), MAPK upstream kinase-binding inhibitory protein (MBIP) [ 54 ], a member of cytochrome P450 superfamily and Nidogen [ 55 ]. The other three tagged predicted genes do not have significant homologies to previously identified genes. In the two lines which have insertions into intergenic regions, transposons have integrated less than 25 kb from the nearest predicted transcript. Table 1 pT2/ S2 EF1α-GM2 transposon insertion events in analyzed enhancer trap lines. Trap line Sequence Insertion location Predicted gene ET1 ATTGTCCtTAGTG TA TGTGTTTGTGTGA Chr. 4 none ET2 CAAAAAGACTATA TA TAGGAGGCTTCAA ctg9701 PARG ET3 AACGCTTACCATG TA TGTTAATAAATGT Chr. 17 MBIP ET4 TATATCAAAATTA TA TATATGAACGTAT Chr. 6 Cyt. P450 ET5 GTACATAcACATG TA CAAATCaACATTA ctg13471 novel ET6 ATTTTAAACAAAC TA AGTtGAACATTAC ctg13605 Nidogen ET7 ATCACAGAGCATC TA GCTTGGATGTGCT ctg12155 novel/ mkp3 ET8 TATACAACAAACT TA TCTAACGTGCAAT Chr. 2 none ET9 TATTTAATATATA TA TTATATTATATTA Chr. 19 novel Left column, line designation. Sequence column, the genomic sequence the transposon has inserted into. The target TA dinucleotide is highlighted in bold. The sequences flanking the left inverted repeat are to the left of the target TA, and sequences flanking the right inverted repeat are to the right of the target TA. Lowercase indicates mismatches between an actual sequence read and the current zebrafish genome sequence (Assembly ZV3). Both left and right transposon/genomic DNA junctions were sequenced for ET1, ET2, ET3, ET5 and ET7. Only the left junction was read for ET8, and only the right junction was read for ET4, ET6 and ET9. Insertion location column, predicted insertion chromosome or contig (Zv3 assembly of the zebrafish genome). Predicted gene column, the gene into which the transposon has inserted as annotated in the zebrafish genome assembly Zv3. Novel indicates no significant homologies. Gene name indicates significant homology to denoted genes. The predicted integration into an intron of the PARG locus for line ET2 was experimentally confirmed (see text). For ET7, a comparison of the observed expression pattern in this line with that of a known nearby gene ( mkp3 ) indicates this insertion has most likely trapped an enhancer for this gene (Fig. 7, see text). Actual sequence reads which were used to determine the genomic location of the transposon insertions were longer than shown in this table and are available upon request. ET7 line has a transposon insertion near mkp3 locus and matches mkp3 expression pattern The transposon insertion in ET7 line has occurred into a predicted novel gene (Table 1 ). Closer investigation of the target locus revealed the presence of a previously characterized zebrafish mkp3 gene within 30 kb of the insertion site. Our attempts to amplify the predicted novel candidate gene from maternal and post-somitogenesis zebrafish cDNA libraries using 2 different primer pairs failed, while mkp3 cDNA was readily amplified in parallel PCR reactions (data not shown). This suggests that the novel target gene may be a false prediction by Genescan. The mkp3 gene encodes a dual specificity phosphatase which was cloned as a member of fgf8 synexpression group and is a negative feedback regulator of FGF8 signaling. mkp3 is expressed in the midbrain-hindbrain boundary, forebrain, tailbud, branchial arches, developing ear, pectoral fin buds and other tissues [ 56 , 57 ]. GFP expression in ET7 line closely mimics mkp3 mRNA expression pattern in 23 hour embryo (Figure 7 ). The only significant difference is that GFP expression is stronger in somites and not as bright in the tailbud, even though the tailbud expression becomes brighter at later stages of development (data not shown). We did not observe GFP expression in the pectoral fin buds, even though we reproduced mkp3 expression in the fin buds just after after 24 hpf by in situ (data not shown, [ 56 , 57 ]. An intriguing possibility is that mkp3 expression in pectoral fin buds may be controlled by a different enhancer, one we do not detect in this transgenic line. Additionally, ET7 expresses GFP in the heart after 24 hpf, and the expression clearly localizes to the ventricle at 5 dpf (Figure 3I ). Expression of mkp3 in the heart after 24 hpf was not reported, and we did not conduct in situ hybridization on late pharyngula stage embryos to test for it. However, fgf8 is expressed in the ventricle of the zebrafish heart at 36 hpf [ 58 ]. Taken together, this suggests that GFP expression in ET7 line mimics a subset of the complete expression pattern of the zebrafish mkp3 gene. Figure 7 GFP expression in ET7 line matches mkp3 mRNA expression. (A) GFP fluorescence photograph of an ET7 embryo at 23 hpf. (B) In situ hybridization on 23 hpf wild type embryo using mkp3 antisense RNA probe. Discussion In this paper, we describe the first use of enhancer trapping, or enhancer detection, as an experimental approach in zebrafish. We show that Sleeping Beauty transposons can trap enhancers by testing an artificial enhancer trapping event in vivo . This approach is likely to also be useful in the construction and testing of other trap vectors: gene (5' exon) and polyA (3' exon) and other related constructs. We then constructed two further truncations to the S1 EF1α promoter in the transgenesis cassette [ 32 ] and found one to be particularly suitable for enhancer trapping. Ten percent (9 of 90) of GFP-expressing transgenic fish generated lines with unique GFP expression patterns. All reagents described in this paper, including the enhancer trap fish strains, are readily available upon request . Many of the obtained enhancer trap lines express GFP in the nervous system. This was previously observed with both mouse and Drosophila enhancer trap vectors and was speculated to stem from the transcriptional complexity of neural tissue [ 11 , 28 ]. Several of our lines also exhibit some level of GFP expression in the eye. At least two explanations can be put forward to explain this observation. First, many genes are expressed in the developing eye. Thus, the eye expression that we see may reflect expression of the tagged genes in the eye. Second, optical properties of the tissues in the eye may permit detection of GFP expression that is lower that what would be required for detection in other tissues. The ET2 line harbors a transposon insertion into the zebrafish gene for poly(ADP-ribose) glycohydrolase (PARG). We demonstrate that both PARG and GFP in ET2 line are expressed in caudal primary motoneurons of 23 hour old embryos. Thus, GFP expression in the ET2 line mimics that of an endogenous gene (PARG), indicating that transgene expression is under control of an endogenous enhancer. A very intriguing question is what the actual trapped enhancer sequence is, how far away from the genomic enhancer the trap can insert and still detect it, and weather artificial enhancer trap approach (Figure 1 ) can be used to answer these questions. The ET7 line has a transposon insertion into a predicted novel gene 30 kb downstream of the zebrafish mkp3 locus. GFP expression in that line closely resembles part of the mkp3 expression domain, suggesting that the enhancer trap transposon in that line is under control of a subset of mkp3 enhancer elements. Zebrafish enhancer trap lines will be valuable in future developmental genetics studies, be it classical mutagenesis or morpholino "knockdown" screening [ 59 ]. GFP expression can be used as a sensitive marker for certain tissue or cell types. For example, the ET1 line expresses GFP in the hatching gland. The expression of the hgg1 gene is specific to the polster and hatching gland depends on nodal signaling and is absent in one-eyed-pinhead mutants [ 60 ]. We phenocopied the one-eyed-pinhead mutation by morpholino injection in ET1 homozygotes and observed a complete loss of hatching gland-specific GFP expression (data not shown). While the ET1 line expresses GFP in an organ that can be readily observed using regular light microscopy techniques, other lines visualize tissues that are not nearly as easily morphologically accessible. In particular, the ET2 line visualizes the position of primary motoneuron cell bodies and axonal trajectory. Morpholinos against known genes or new members of the zebrafish secretome [ 61 ] can be screened for effects on neuronal cell body position or axonal pathfinding in the developing embryos by injection into ET2 line embryos. The ET7 line may provide a fluorescent readout of FGF8 signaling, thus facilitating the identification of genes involved in that signaling pathway. A further utility offered by the transposon system is the possibility to revert a mutant phenotype or to generate localized deletions by transposon excision [ 23 ]. We successfully excised the transposon in the germline of the ET1 line, resulting in the expected transposon footprint. It has been shown that excision of the Sleeping Beauty transposon from a plasmid results in local deletions with fairly high frequency which is dependent on the cell type or tissue used [ 45 ]. Furthermore, the frequency of imprecise excision of Sleeping Beauty transposons significantly increases in cells with a compromised DNA repair pathway [ 62 , 63 ]. It remains to be determined how frequently the excision of a Sleeping Beauty transposon from a genomic location in zebrafish germ line is accompanied by a deletion of flanking genomic DNA, and it should be possible to compromise the embryo's DNA repair machinery to induce such deletions at a high frequency. Our experiments indicate that enhancer detection using Sleeping Beauty transposons is an easily scalable and efficient experimental technique in zebrafish. Obtaining fish with different GFP expression patterns is not the rate limiting step in this process. Preliminary molecular analysis of the insertion site is also straightforward using inverse PCR techniques. Identification of candidate genes should benefit from the progress in zebrafish genome sequencing and annotation. The main bottleneck step is the detailed biological analysis of GFP and the corresponding candidate gene expression profile. In Drosophila , the generation of transposase-expressing lines of flies made enhancer detection and P-element mutagenesis in general a mainstream approach. Even without a similar gain in efficiency, transposase expressing fish lines would make enhancer trapping as well as related gene- and poly(A)-trap methodologies even more accessible for high-throughput functional analysis of the vertebrate genome. Methods Plasmid construction pT2/ S1 EF1α-GFP (pDB358) was previously published [ 32 ]. To make γ Cry/pT2/ S1 EF1α-GFP (pDB375), a Bam HI- Hind III fragment from Cry1-GFP3 [ 42 ] containing part of the X. laevis γ-Cry1 promoter was cloned into the Ecl136II site of pDB358. To produce pT2/ S2 EF1α-GFP (pDB371), a part of the EF1α promoter was deleted from pDB358 by ligation of the Bst 1107I- Nco I and Nhe I- Nco I fragments of pDB358. Similarly, the Eco RI- Nco I and Nhe I- Nco I fragments of pDB358 were ligated to produce pT2/ S3 EF1α-GFP (pDB372). Inverse PCR, PCR and primer sequences For inverse PCR experiments, zebrafish genomic DNA was digested and ligated as described [ 64 ]. 1 and 2.5 microliters of the ligation reaction were used for the first PCR reaction with RP1/LP1 or RP1/GFP-R1 primers in total volume of 25 μl. 1 μl of the first PCR reaction was used as a template for the second (nested) PCR reaction with primer pairs RP2/LP2 or RP2/GFP-R2, respectively. Expand Hi Fi PCR system (Roche) was used for all PCR reactions. A MJ Research PTC-100 PCR machine was used for PCR with the following program : 92°C 4 min., 92°C 10 sec., 60°C 30 sec., 68°C 6 min., 30 cycles. Starting at cycle 11, 20 sec. per cycle was added to the extension time. The same PCR reaction with an annealing temperature 55°C was used for amplification with primers flanking transposon insertion sites, and for amplification of partial PARG cDNA from a maternal cDNA library. Primer sequences are: LP1 GTGTCATGCACAAAGTAGATGTCC [ 32 ]; LP2 ACTGACTTGCCAAAACTATTGTTTG; nRP1 CTAGGATTAAATGTCAGGAATTGTG; RP2 GTGAGTTTAAATGTATTTGGCTAAG; GFP-R1 TTCGGGCATGGCACTCTTG; GFP-R2 TATGATCTGGGTATCTCGCAA; NeuroC1-F1 CGTAAAGATGCCTTGTTCAGAA; NeuroC1-R1 ATTCCGTGACTCTCCTGAAATA; NeuroIns-F1 GGCTTGCATACATGACTAATG; NeuroIns-R1 GAAGACTGAAGTCCTCAAACT; HG1-1 ACATTGAGCCACTAAGCATTG; HG-2 TGTGTGCACTTAAGGGGCGA. Mkp3-F1 AGTGTTGCATTCTCCAGGATA; Mkp3-R1 TGACACAGAACTTCCCTGAAC; EF1a-F2 TTCCTGCAGGTCGACTCT; GFP-R0 GTGTAATCCCAGCAGCTG. Information about other primers is available from the authors upon request. In situ hybridization A partial sequence for the zebrafish poly(ADP-ribose) glycohydrolase cDNA was amplified using primers neuroC1-F1 and neuroC1-R1 and cloned using a Topo TA cloning kit (Invitrogen) to make pDB376. To make antisense RNA probe, pDB376 was digested with Spe I and transcribed with T7 RNA polymerase (Promega) and DIG labeling kit (Roche). GFP probe was made by amplifying GFP with 46 base pairs of EF1α promoter from pT2/ S1 EF1α-GM2 using primers EF1a-F2 and GFP-R0, and cloning it into Topo TA cloning kit resulting in pSS100. pSS100 was linearized with Spe I and transcribed with T7 RNA polymerase using DIG labeling kit. To make mkp3 antisense probe, mkp3 cDNA was amplified from maternal cDNA library with primers Mkp3-F1 and Mkp3-R1 and cloned into Topo TA cloning kit to produce pDB528. The plasmid was linearized with Spe I and transcribed with T7 polymerase using DIG labeling kit. Screening for germline transmission of Sleeping Beauty transposons Embryos injected with SB10 transposase mRNA and transposon DNA mix were raised as described [ 32 , 64 ]. In pilot screens, adult fish were outcrossed to brass for ease of husbandry. All collected embryos were screened for GFP expression at 1 day post fertilization (dpf) and 3 dpf. We set an arbitrary 200 embryo cutoff for screening, meaning that if less that 200 embryos were obtained from a founder, an additional cross was set up and to obtain additional embryos for screening. Analysis of transgenesis data from pilot screens indicated that 10% of transgenic lines would have been missed if cutoff was set at 100 embryos, and this less stringent coverage protocol was used in scale up screen. Also, we decided to limit screening to 1 dpf since none of the transgenics would have been missed in the pilot screens without the 3 dpf screening. Transposon excision in the germline Homozygous ET1 embryos were injected with SB10 transposase mRNA, raised and screened for loss of hatching gland specific GFP expression, or for a change in the GFP expression pattern. Twenty six fish were screened (R0, for Remobilization), and 2 gave GFP negative embryos, with an additional 2 giving ubiquitously GFP positive embryos, suggesting that germline remobilization events may have occurred in as many as 15% of transposase injected embryos. Ubiquitous GFP positive embryos (one in each of the two R0) did not survive. Of the two R0's that gave GFP negative embryos, one gave mosaic hatching gland expression in the next generation. PCR with transposon specific and flanking primers did not show any changes in the locus. The second R0 produced 19 embryos that were GFP negative from the total of 671 embryos obtainted. An R1 adult was outcrossed, embryo DNA was prepared, and PCR with primers HG1-1 and HG1-2 was conducted. The resulting PCR fragment was cloned using PCR 4 Topo cloning kit (Invitrogen). Plasmids were sequenced using M13 Forward primer, and one clone with a transposon footprint was identified. To confirm that it was not due to PCR contamination, a second clutch of embryos was obtained, the procedure was repeated, and the same footprint was obtained (data not shown). Authors' Contributions The experiments described in this paper were planned, conducted and analyzed as a joint effort between the authors. In particular, DB, AD, SH and ZW contributed to fish screening, line establishment and to scientific descriptions of these lines, DB and SS to molecular analysis, AD and DB to GFP expression and in situ analysis. DB designed and built the transposons used in this study and was responsible for drafting the manuscript for publication. SE conceived and supervised the study and edited the manuscript. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC520745.xml |
522813 | Localization of ABCG5 and ABCG8 proteins in human liver, gall bladder and intestine | Background The molecular mechanisms that regulate the entry of dietary sterols into the body and their removal via hepatobiliary secretion are now beginning to be defined. These processes are specifically disrupted in the rare autosomal recessive disease, Sitosterolemia (MIM 210250). Mutations in either, but not both, of two genes ABCG5 or ABCG8 , comprising the STSL locus, are now known to cause this disease and their protein products are proposed to function as heterodimers. Under normal circumstances cholesterol, but not non-cholesterol sterols, is preferentially absorbed from the diet. Additionally, any small amounts of non-cholesterol sterols that are absorbed are rapidly taken up by the liver and preferentially excreted into bile. Based upon the defects in sitosterolemia, ABCG5 and ABCG8 serve specifically to exclude non-cholesterol sterol entry at the intestinal level and are involved in sterol excretion at the hepatobiliary level. Methods Here we report the biochemical and immuno-localization of ABCG5 and ABCG8 in human liver, gallbladder and intestine using cell fractionation and immunohistochemical analyses. Results We raised peptide antibodies against ABCG5 and ABCG8 proteins. Using human liver samples, cell fractionation studies showed both proteins are found in membrane fractions, but they did not co-localize with caveolin-rafts, ER, Golgi or mitochondrial markers. Although their distribution in the sub-fractions was similar, they were not completely contiguous. Immunohistochemical analyses showed that while both proteins were readily detectable in the liver, ABCG5 was found predominately lining canalicular membranes, whereas ABCG8 was found in association with bile duct epithelia. At the cellular level, ABCG5 appeared to be apically expressed, whereas ABCG8 had a more diffuse expression pattern. Both ABCG5 and ABCG8 appeared to localize apically as shown by co-localization with MRP2. The distribution patterns of ABCG5 and ABCG8 in the gallbladder were very similar to each other. In the small intestine both ABCG5 and ABCG8 appear to line the brush border. However, at the level of the enterocyte, the cellular distribution patterns of ABCG5 and ABCG8 differed, such that ABCG5 was more diffuse, but ABCG8 was principally apical. Using standard deglycosylation methods, ABCG5 and ABCG8 do not appear to be glycosylated, suggesting a difference between human and mouse proteins. Conclusion We report the distribution patterns of ABCG5 and ABCG8 in human tissues. Cell fractionation studies showed that both proteins co-fractionated in general, but could also be found independent of each other. As predicted, they are expressed apically in both intestine and liver, although their intracellular expression patterns are not completely congruent. These studies support the concept of heterodimerization of ABCG5 and ABCG8, but also support the notion that these proteins may have an independent function. | Background The gastrointestinal tract is the initial barrier to dietary constituents and is important in regulating nutrient entry, as well as keeping non-nutrients out. Additionally, the hepatobiliary system acts as an additional filter to rapidly excrete such non-nutrients into bile, thus keeping the net retention of these potential toxins low. Mammals have evolved many mechanisms in the gastrointestinal tract to select out usable dietary constituents from those that maybe potentially toxic to the body. It is apparent that the ATP-binding cassette proteins (ABC proteins/transporters) are the machinery that mediate the ATP-dependent transport of a wide variety of substrates that range from xenobiotics to peptide fragments [ 1 ]. A subset of these ABC transporters, located in the canalicular membranes of mammalian liver, play key roles in bile formation and detoxification [ 1 - 3 ]. One of these processes involves the regulation of sterol entry and excretion. Whole body cholesterol homeostasis is a tightly regulated process, involving dietary absorption, de novo synthesis and hepatobiliary secretion. Sitosterolemia, a rare autosomal recessive disorder of sterol metabolism results in the disruption of dietary sterol entry and hepatobiliary sterol secretion [ 4 - 6 ]. Under normal circumstances, our diets contain equal amounts of plant sterols and cholesterol, but the plant sterols are specifically excluded from our bodies and only regulated amounts of cholesterol are retained. In Sitosterolemia, this exclusion is defective resulting in the retention of non-cholesterol sterols. Mutations in either, but not both, of two ABC transporters, ABCG5 and ABCG8, encoded by a single locus, STSL , are known to cause this disease [ 7 - 9 ]. Based upon the genetics, as well as in vitro and in vivo data, these 'half-transporters' are proposed to function as obligate heterodimers. In vitro experiments have shown that both proteins are needed to be co-expressed for apical expression and that these may function as mutual chaperones in the ER for maturation [ 10 , 11 ]. In vivo experiments in mice have not been consistent. Using the Abcg5/Abcg8 double knockout mice, Graf et al has shown that by inoculating them with adenoviral constructs for Abcg5 and Abcg8 that both are required for expression of both proteins [ 11 ]. Additionally, Plosch et al and our group have constructed mouse models deficient in either Abcg5 [ 12 ] or Abcg8 [ 13 ] that show different biliary physiology than that of the Abcg5/Abcg8 double knockout mice. This suggests these proteins may have independent function(s) in addition to their function as heterodimers. However, to date, no reports of characterization and localization of the human proteins have been reported. In this report, we examined the location of these two proteins using cellular fraction and immunohistochemical analyses of human liver, gallbladder and small intestine. We found a general concordance of co-expression of both proteins, but we also noted that ABCG5 and ABCG8 could be found in plasma membranes, as well as in intracellular membrane locations independent of each other. Additionally, deglycosylation of human liver membranes with peptidyl N-glycosidases did not alter the mobility of the proteins after SDS-PAGE, suggesting that these proteins may not be glycosylated in human liver. This differential localization suggests that perhaps ABCG5 and ABCG8 may have functions independent of each other, as well as functioning as heterodimers. Methods Tissue aquistion. Human liver donated for transplantation, but deemed unsuitable for transplantation on inspection by the transplant service (usually based upon a 'fatty' appearance) was obtained in accordance with IRB approval. As soon as the liver was deemed unsuitable (typically less than 10 h following harvesting) pieces were either snap-frozen in liquid nitrogen and stored at -80°C or liquid nitrogen until use, or placed in ice-cold 2-methylbutane and stored in liquid nitrogen. Samples from more than nine different donors were used in these studies. Additionally, human gallbladders and segments of proximal small intestine were obtained from patients under going either laproscopic cholecystectomy or pancreatoduodenectomy (Whipple procedure). These tissues were directly taken from the operating room in normal saline on ice to be processed directly for frozen sectioning. Antibodies Anti-membrin, anti-transferrin, and anti-calnexin antibodies were obtained from Stressgen (Victoria, BC Canada), anti-caveolin antibody from BD Bioscience (San Diego, CA, USA), anti-MDR1 that also detects MDR2/3 (C219) from Centocor Inc. (Malvern, PA, USA), anti-MRP2 (cMOAT) from Chemicon International (Temecula, CA, USA) and secondary antibodies were purchased from Jackson Immuno research (West Grove, PA, USA). Polyclonal rabbit anti-sera to human ABCG5 and ABCG8 peptides were generated in-house, using a 20-peptide immunogen from human ABCG5 (576–587, FQKYCSEILVVNEFYGNFTC, GenBank Accession number NP_071881) and a 22-peptide sequence from human ABCG8 (608–629, SRRTYKMPLGNLTIAVSGDKIL, GenBank Accession number NP_071882). The anti-sera were further purified using peptide affinity columns and stored at a concentration of 0.8 mg/ml in Immuno Pure Binding Buffer (Pierce, Rockford, IL, USA). For peptide blocking experiments the peptides were dissolved in DMSO (final concentration 30%), incubated with the corresponding peptide for 1 1/2 hours at 37°C then used for immunoblotting as described below. Membrane protein preparation Crude total membrane isolation was carried out with minor modifications as previously described [ 14 ]. All the procedures were carried out at 4°C. Three grams of human liver were homogenized in homogenization buffer (5 mM Tris pH7.5, 250 mM sucrose, 1 mM PMSF, 20 μg/μl of leupeptin and 1 μg/μl of aprotinin) by applying 10 strokes with a dounce homogenizer. The homogenate was centrifuged at 1000 g for 10 minutes, the pellet containing any undisrupted cells and nuclear debris were re-homogenized with one-half the initial volume of homogenization buffer, centrifuged at 1000 g for 10 minutes and this process was repeated once more. Supernatants were pooled and subjected to centrifugation at 100,000 g for 40 minutes. The resulting pellet, deemed the crude membrane fraction, was used as the starting material for Western blotting and fractionation experiments. Nycodenz gradient fractionation Human liver crude total membrane proteins were re-suspended in 30% Nycodenz solution (Nycodenz in 5 mM Tris-HCl pH 7.5 and 1 mM EDTA). This suspension was loaded on top of a 40% Nycodenz solution cushion in an ultracentrifuge tube, overlaid by consecutive 23%, 20%, 15% and 10% Nycodenz solutions and subjected to centrifugation at 39,000 rpm for 16 hours at 4°C in a SW41 rotor (Beckman Instrument, Palo Alto, CA). After centrifugation, 800–1000 μl fractions were sequentially removed from the top, combined with two volumes of homogenization buffer (see above) and centrifuged at 39,000 rpm at 4°C for 40 minutes to remove the Nycodenz. The resulting pellets were re-suspended in buffer (25 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% Triton X-100 and 0.1% SDS, 1 mM PMSF, 20 μg/μl of leupeptin and 1 μg/μl of aprotinin), the protein content determined by the method of Lowry and fractions analysed by SDS-PAGE. Equal amounts of protein (25 μg) per lane were loaded. Sucrose gradient fractionation The procedure for membrane fractionation was essentially as described for the Nycodenz fractionation, except for the homogenization buffer used (25 mM Tris pH6.8, 150 mM NaCl, 1 mM PMSF, 20 μg/μl of leupeptin and 1 μg/μl of aprotinin). The sucrose density gradient fractionation was modified as previously described [ 15 - 17 ]. Human liver crude membrane proteins were re-suspended in 1% Triton X100 buffer (25 mM Tris-HCl pH6.5, 150 mM NaCl, 1% Triton X100, 1 mM PMSF, 20 μg/μl of leupeptin and 1 μg/μl of aprotinin), adjusted to a final sucrose concentration of 40% and overlaid with a 15–30% linear sucrose gradient. The samples were subjected to centrifugation at 39,000 rpm for 16 hours at 4°C in a SW41 rotor (Beckman Instrument, Palo Alto, CA, USA) and fractions collected from the top as described above. The proteins from fractions 1–4 from top of the tube were precipitated with acetone because these fractions did not contain sufficient protein for direct analysis. After protein concentrations were determined, equal amounts of proteins (20 μg) from each fraction were resolved by SDS-PAGE. Immunoblotting Proteins resolved by SDS-PAGE were transferred onto nitrocellulose membranes. Membranes were then blocked for 1 hour in 5% dry milk in PBS-T (Phosphate Buffered saline and 0.1% Tween 20) and incubated with primary antibody against either ABCG5 or ABCG8 in 5% milk in PBS-T overnight at 4°C. Blots were washed three times for 5 minutes in TBS-T (Tris Buffered Saline/0.1% Tween-20) with 150 mM NaCl, incubated with goat-anti-rabbit conjugated HRP antibodies (1:10000 dilution), washed for three times 5 minutes and developed with Western Lightning ® Chemiluminescence Reagent Plus (Perkin Elmer Life Sciences, Inc. Boston, MA, USA). Immunohistochemical analysis and microscopy Snap-frozen liver, gallbladder and intestine tissues were used to cut 8 μm thick frozen sections, air-dried for 30 minutes onto glass slides and kept at -80°C until used. The slides were stained with hematoxylin, rinsed with PBS three times, fixed for 10 minutes with cooled methanol at -20°C and rinsed with PBS three times. The slides were treated with blocking solution (10% donkey serum in 0.1 M glycine/PBS) for 30 minutes at room temperature and incubated with primary antibody overnight at 4°C. The slides were washed with PBS and incubated with secondary antibody (goat-anti-rabbit conjugated with Cy3™ or rhodamine or FITC) for 20–30 minutes at room temperature, rinsed with PBS three times and examined under an Olympus BX-5 confocal microscope with Fluoview. Results Identification of ABCG5 and ABCG8 in crude total membrane preparations of human liver Peptide antibodies were raised against human ABCG5 and human ABCG8 and affinity purified prior to use (see Methods). The immunogen peptides used for the antibodies were selected since they were sequences that lay outside of the predicted transmembrane domains and based upon antigenicity.Western blotting experiments (Figure 1A ) showed that both anti-ABCG5 and anti-ABCG8 antibodies bound to ~75 kDa proteins in human liver crude membranes. Pre-immune sera did not detect the ~75 kDa expected bands. Pre-incubation of the immune antibodies with the peptides against which they were raised abolished specific binding (Figure 1B ). For anti-ABCG5 5 μg of peptide was needed to block 1 μg of antibody and for anti-ABCG8 12.5 μg of peptide was needed to block 1 μg of antibody. Interestingly, a ~60 kDa band was detected using the anti-ABCG8 antibody whose signal is abolished when incubated with the peptide from which the antibody was raised (Figure 1B and 1C , arrow indicated band). The significance of this is unclear at present. These antibodies were also tested against mouse and rat liver membrane preparations and no significant cross-reactivity was detected except for faint bands seen with anti-ABCG5 in mouse liver samples (Figure 1A , tracks 3 and 4). No other bands were detected above the 150 kDa marker in all western blots. Interestingly, while these proteins are predicted to be N-glycosylated [ 8 , 9 ], only single bands in the appropriate molecular weight range were detected and no higher molecular bands were observed. To investigate whether these proteins are glycosylated, crude membrane fractions were digested with EndoH, PNGase F and examined for alterations in gel migration by SDS-PAGE (Figure 1C ). Although the mobility of a known glycoprotein, transferrin, was increased in the same fractions following deglycosylation, there was no change in the mobility of ABCG5 or ABCG8 (Figure 1C ). Figure 1 Immunoblotting analyses of ABCG5 and ABCG8 in human liver. Panel A shows the immunodetection of ABCG5 (tracks 1–6), ABCG8 (tracks 7–12) in membrane preparations from human liver (HL, tracks 1, 2, 7, 8) mouse liver (ML, tracks 3, 4, 9, 10) or rat liver (RL, tracks 5, 6, 11, 12). The anti-ABCG5 peptide antibody detected a faint mouse band, but no other specific binding was identified. Although the pre-immune sera detected bands in the rodent tissue samples, none were detected in human liver (tracks 13–16, MB, mouse brain). Specificity was further shown by pre-incubation of the antibodies with the peptides they were raised against (panel B). In the presence of the specific peptides, the 75 kDa bands are not detected in human liver microsomes. Panel C shows the results of deglycosylation of human total liver microsomes, probed with anti-ABCG5 (left hand panel), anti-ABCG8 (middle panel) or anti-transferrin (right hand panel). Aliquots from the same incubation were separated for all three western blots. Although ABCG5 and ABCG8 do not appear to have their SDS-PAGE mobility's altered by either EndoH or PNGase F treatment, that of transferrin in the same samples is clearly effected (see Text for discussion). Newly synthesized (sensitivity to EndoH), as well as mature forms of transferrin (resistant to EndoH, but fully sensitive to PNGase F) are present in the liver membrane preparations. Localization of ABCG5 and ABCG8 by Nycodenz and Sucrose gradient fractionation of human liver Crude total membrane proteins from human liver were fractionated by Nycodenz gradient centrifugation and examined for localization markers by western blot analyses (Figure 2A ). After Nycodenz gradient centrifugation, ABCG5 (fractions 9–10) and ABCG8 (fractions 6–11) were found to have a broad range of distribution and appeared to be distributed in a pattern similar to calnexin (an ER membrane marker, fractions 5–10), Cytochrome C (a mitochondrial marker, fractions 4–9), transferrin (a plasma membrane marker, fractions 1–11), caveolin (fractions 6–10) and MDR1 (an apical membrane marker, fractions 4–10). ABCG5 and ABCG8 did not co-localize with cis-Golgi (Figure 2A ) markers. Figure 2 Subcellular localization of ABCG5 and ABCG8 in human liver. Panel A shows the Nycodenz gradient fractionation and panel B the Triton X-100/sucrose gradient fractionation. A representative result from each of these is shown. F1–F12 represents serial fractions collected from the top of the tube. Proteins from each of these fractions were separated by SGS-PAGE, western blotted and probed for the proteins as indicated on the figure. Calnexin (ER marker), membrin (Golgi marker) and caveolin (raft marker) did not co-localize with ABCG5/ABCG8 when both biochemical fractionation patterns are compared. However, MDR1 (apical membrane marker) and transferrin (plasma membrane marker) showed some consistency with ABCG5/ABCG8 co-localizations (tracks F9–10, panel A and F11, panel B). To examine whether ABCG5 and ABCG8 were associated with membrane rafts, total membrane proteins from human liver were solubilized with ice-cold 1% Triton X-100 detergent and fractionated by sucrose density gradient centrifugation (Figure 2B ). Fractionation resulted in two Triton X-100 insoluble complexes, as judged by the clarity of the gradient fractions. The first was found in the low-density range (15–20% sucrose, fractions 2–6, Figure 2B ) and the second in the high-density range (40% sucrose, fractions 10–12, Figure 2B ). Caveolin-rich fractions localized to the low-density range (Figure 2B , fractions 2–6). However, ABCG5 and ABCG8 were detected in the high-density fractions, F10–12, along with transferrin (fractions 5–12) and MDR1 (fractions 9–11). Calnexin or cytochrome C, under these conditions did not co-localize with either ABCG5 or ABCG8. Note that the majority of the ABCG5 and ABCG8 were present in the densest fractions, F11–12. Under these conditions, membrin, a cis-Golgi membrane marker, was also detected in fraction 12. Thus, ABCG5 and ABCG8 have significant overlap with each other suggesting co-localization. However, these proteins did not seem to co-localize with any specific membrane marker except transferrin, when two different methods of fractionation were utilized. Immunohistochemical localization of ABCG5 and ABCG8 in human liver It has been shown previously that ABCG5 and ABCG8 are expressed only in the liver and intestine [ 8 , 9 ]. To further characterize the location of ABCG5 and ABCG8 in human liver, immunohistochemical analyses were performed on frozen serial sections of human liver. Pre-immune sera were used as negative controls. The distribution of the two proteins appeared to be divergent not only histologically, but also at the cellular level. From a histological point of view, ABCG5 was detected along sinusoidal tracts (Figure 3A , upper panel) whereas ABCG8 was detected within the cells lining the bile ducts (Figure 3B , upper panel). At higher magnification, ABCG5 was detected along bile canaliculi and at the cellular level appeared to be an apically expressed (Figure 3A , lower panel). However, the distribution of ABCG8 at the cellular level appeared more diffuse consistent with plasma membrane expression and perhaps in intracellular membranes (Figure 3B , lower panel). Expression of ABCG5 within intracellular vesicular compartments could not be excluded by the techniques employed. To confirm an apical location of ABCG5 and ABCG8 immunohistochemical co-localization studies were carried out using an antibody against a known apical transporter (MRP2) in the liver. As shown in Figure 4 , ABCG5 and ABCG8 have significant overlap in expression with MRP2 (panels A and B, respectively). Figure 3 Immunolocalization of ABCG5 and ABCG8 in human liver sections. Panel A shows the staining pattern of ABCG5 and panel B that for ABCG8. The pre-immune controls for both antibodies are as marked and shown in the top right hand corners of each panel. The top panels of each section are at low magnification (bar is 50 μm) and the bottom panels at high magnification (10 μm). The images for ABCG5 and ABCG8 were visualised with red and green colors respectively using Adobe Photoshop (Adobe, Cupertino, CA). The left panels show hematoxylin stained phase contrast images and the middle panels show the fluorescence images after immune serum staining. The bottom right panel of each section shows the merged images of phase contrast and the fluorescence signals. ABCG5 was readily detectable in canalicular cells and at higher magnification seemed to be apical in expression (panel A). On the other hand, ABCG8 was more readily detectable in cells lining the bile ducts (panel B, top panels), as well as in canalicular cells; although its cellular expression appeared more diffuse (see Text for discussion). Figure 4 Co-localization of ABCG5 and ABCG8 with MRP2 by immunohistochemistry. Liver sections were simultaneously incubated with either the combination of anti-ABCG5/MRP2 or anti-ABCG8/MRP2. Panel A shows staining patterns of MRP2 and ABCG5 independently (upper portion) then merged together (lower right) with what appears to be similar over-lapping expression patterns. Likewise MRP2 and ABCG8 expression patterns appear to over lap as well as shown in panel B. Bar is 20 μm. Immunohistochemical localization of ABCG5 and ABCG8 in human gall bladder Serial sections of human gall bladder were incubated with each antibody and pre-immune serum was used as a negative control. Both ABCG5 and ABCG8 were detected in the epithelium of gall bladder mucosa (Figure 5A and 5B ). At higher magnification, the cellular distribution of the signals detected was similar and showed a diffuse cytoplasmic distribution (Figure 5A and 5B ). Figure 5 Immunolocalization of ABCG5 and ABCG8 in human gall bladder sections. Gall bladder surgical samples were stained for ABCG5 (panel A, red staining) or ABCG8 (panel B, green staining). The layout is as indicated for figure 3. Both ABCG5 and ABCG8 appeared to be confined to the epithelial cells lining the lumen and no significant staining in any of the deeper cell layers was detected. At the cellular level, both proteins seem to be diffusely expressed within the cell and at plasma membrane. A strict apical expression was not observed for either of these proteins. Immunohistochemical localization of ABCG5 and ABCG8 in human small intestine Both ABCG5 and ABCG8 were detected in the apical surfaces of the enterocytes in biopsy samples of the small intestine (Figure 6A and 6B , upper panels). However, at higher magnification the cellular distribution of ABCG5 appeared more diffuse (Figure 6A , lower panel) whereas ABCG8 was expressed apically (Figure 6B , lower panel). This divergent pattern was seen in all of the serial sections analyzed. Figure 6 Immunolocalization of ABCG5 and ABCG8 in human intestinal sections. Human intestinal surgical samples were stained with ABCG5 (panel A, red staining) or ABCG8 (panel B, green staining). The layout is as indicated for figure 3. At the cellular level, ABCG5 was expressed in a more diffuse pattern (panel A, see bottom right hand panel). In contrast, ABCG8 was expressed in the apical surfaces of the enterocytes (panel B, bottom right hand panel). Both proteins seemed to be expressed only in the enterocytes lining the villi and no significant expression was detected in any of the other cell layers. Discussion In this study, we report the localization of ABCG5 and ABCG8 in human liver, gall bladder and intestine. Our studies showed that these proteins are highly specific in the cells they are expressed. In the liver, expression is seen in cells lining the hepatobiliary tracts, both hepatocytes and ductal cells. In the intestine, robust expression was seen only in the villus enterocyte layers. In the gall bladder, expression was confined to the epithelial cells lining the lumen. However, some differences in the distribution of ABCG5 and ABCG8 within these tissues were apparent. In the liver, ABCG8 was highly expressed in the hepatocytes lining the bile ducts, whereas ABCG5 was more robustly expressed in hepatocytes lining the cannaliculae. In fractionation studies, using two different methods of separation, the distribution of ABCG5 and ABCG8 was compatible with both proteins potentially acting as heterodimers. However, we also noted that there were fractions where only one of these proteins, but not the other was detected. This could be an artefact, with one antibody being a better reagent, or that this pattern could truly reflect that each of these proteins can also exist independently, perhaps as homodimers. Overall, the distribution patterns of ABCG5 and ABCG8 in these cellular fractionations were similar to that observed with the plasma membrane marker transferrin and apical membrane marker MDR1. Additionally, immunohistochemical analyses show that ABCG5 and ABCG8 are apically expressed in the liver, gall bladder and intestine. In the liver ABCG5 and ABCG8 also appear to co-localize with the known apical protein MRP2. This confirms previous data from Graf et al , using in vitro expression in WIF-B cells [ 10 ], and support the contention that ABCG5 and ABCG8 are plasma membrane proteins. These data would suggest that expression of these proteins might not be wholly dependent upon mutual co-expression, as has been reported for the mouse and in in vitro studies [ 10 , 11 ]. However, one note of caution should be expressed. All of the liver samples analyzed were obtained because they were unsuitable for transplantation. Most livers were considered to be 'fatty' livers. While these livers were not effectively diseased, that fact that they had fatty infiltrates may have influenced the normal expression of these two half-transports. Thus, confirmation in normal human liver samples will be needed, though this may not be feasible. With that reservation in mind, our data do have important implications for sterol trafficking in humans. Firstly, the relatively robust and highly specific expression of ABCG5 and ABCG8 in gall bladder epithelium confirms the important role of this organ in regulating biliary secretion. In addition to the production of bile by the liver, the gall bladder may be able to further regulate the sterol content of bile, via ABCG5/ABCG8 activity. A similar pattern of expression has been reported in canine gall bladder epithelial cell culture and these data confirm these findings in human gall bladder [ 18 ]. Secondly, the differences in expression patterns of ABCG5 and ABCG8 in liver, gall bladder and intestine, though subtle, seem to indicate the several important possibilities. It is not clear which organ is of paramount importance in the human in regulating non-cholesterol sterol retention. And it is not clear if ABCG5 and ABCG8 play a significant role in determining cholesterol entry at the intestinal level, though they seem to be implicated strongly as determining sterol excretion at the level of the hepato-biliary system. In mice deficient for Abcg5/Abcg8 or Abcg5, cholesterol absorption rates were not dramatically affected [ 12 , 19 , 20 ]. Hepatobiliary excretion of all sterols was significantly (but not completely) reduced. In mice singly deficient for Abcg5 or Abcg8, some differences have been reported [ 12 , 13 ]. In Abcg5 KO mice, following LXR activation, sterol excretion in bile was comparable to wild-type mice, though it is not clear if this also restored plant sterol excretion. In contrast, although cholesterol absorption studies for Abcg8 KO mouse have not been reported, biliary excretion of cholesterol was dramatically reduced and no stimulation of excretion was observed after forced bile acid infusions. While all of these data have been reported using different protocols (LXR activation, bile acid infusions, or static gall bladder puncture), these data suggest that there exist other pathways for sterol trafficking in both liver and intestine. At present it would be speculative to assume that these 'other' pathways involve ABCG5 or ABCG8 as homodimers, but this possibility is supported by the circumstantial evidence of separate patterns of expression of human ABCG5 and ABCG8 in these tissues. Finally, although rodent sterolins are glycosylated and in vitro glycosylation is readily demonstrable, human ABCG5 and ABCG8 did not appear to be glycosylated as judged by deglycosylation-migration assays. It is possible that this technique is insensitive and these proteins are glycosylated. Alternatively, it is possible that the antibodies we have raised only react with unglycosylated forms and thus fail to detect the glycosylated forms. With respect to the first issue, deglycosylation-migration has been demonstrated to detect mouse glycosylated proteins and since these proteins are highly conserved, this possibility seems remote. With respect to the second possibility, if our antibodies were exclusively detecting unglycosylated (and presumably immature forms), the apical patterns of expression of these proteins in both the liver and intestine would seem to suggest that these proteins seem to traffic to these specialized membranes normally. In the absence of an independent method, and the lack of a direct assay for function, whether these proteins form an active heterodimer can not be resolved at present. Conclusion In summary, we report the first immunolocalization of ABCG5 and ABCG8 in human liver, gall bladder and intestine. Our data show that these proteins are located in membranes and can have an apical expression in all of these tissues. Biochemical, as well as immunolocalization studies show that while both proteins co-localize in general, they can also seem to have expression patterns that may be independent of each other. Competing interests None declared. Authors' contributions ELK and MHL performed the experiments, KDC and DBA provided the liver and surgical samples respectively SBP was responsible for supervision, data analyses and obtaining funding for these experiments. ELK and SBP wrote the paper. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC522813.xml |
553999 | Reduction of motion artifact in pulse oximetry by smoothed pseudo Wigner-Ville distribution | Background The pulse oximeter, a medical device capable of measuring blood oxygen saturation (SpO2), has been shown to be a valuable device for monitoring patients in critical conditions. In order to incorporate the technique into a wearable device which can be used in ambulatory settings, the influence of motion artifacts on the estimated SpO2 must be reduced. This study investigates the use of the smoothed psuedo Wigner-Ville distribution (SPWVD) for the reduction of motion artifacts affecting pulse oximetry. Methods The SPWVD approach is compared with two techniques currently used in this field, i.e. the weighted moving average (WMA) and the fast Fourier transform (FFT) approaches. SpO2 and pulse rate were estimated from a photoplethysmographic (PPG) signal recorded when subject is in a resting position as well as in the act of performing four types of motions: horizontal and vertical movements of the hand, and bending and pressing motions of the finger. For each condition, 24 sets of PPG signals collected from 6 subjects, each of 30 seconds, were studied with reference to the PPG signal recorded simultaneously from the subject's other hand, which was stationary at all times. Results and Discussion The SPWVD approach shows significant improvement (p < 0.05), as compared to traditional approaches, when subjects bend their finger or press their finger against the sensor. In addition, the SPWVD approach also reduces the mean absolute pulse rate error significantly (p < 0.05) from 16.4 bpm and 11.2 bpm for the WMA and FFT approaches, respectively, to 5.62 bpm. Conclusion The results suggested that the SPWVD approach could potentially be used to reduce motion artifact on wearable pulse oximeters. | Introduction Wearable medical devices are capable of continuously monitoring an individual's vital signs in real time. These devices are particularly important to the world's increasingly aging population, whose health conditions have to be assessed regularly or monitored continuously. The devices can warn individuals of symptoms of deterioration, e.g. alerting them when their blood pressure is increasing to a level above a predetermined threshold. The devices can also automatically notify emergency services in critical situations. In order to make wearable devices practical, a series of technical problems have to be solved. For example, these devices need to be miniature in size, must possess a user-friendly interface and be efficient in power consumption. Most importantly, these devices need to have a low failure rate and must report minimal false alarms. In other words, these devices are required to provide an accurate estimate of the monitored vital sign under normal daily life situations. This leads to the important topic on the reduction of motion artifacts [ 1 - 4 ]. In this paper, the smoothed pseudo Wigner-Ville distribution (SPWVD) is investigated as a novel motion artifacts resistant approach for estimating one of the most important vital signs – the blood oxygen saturation level (SpO2). The paper is organized as follows. Section 2 reviews the techniques commonly used for attenuating motion artifacts in pulse oximetry. Section 3 discusses the basic theory for SpO2 computation and the techniques used in this study for reducing motion artifacts. Section 4 compares the performance of two time-frequency techniques, i.e. the short-time Fourier transform (STFT) and the SPWVD. Section 5 presents the protocol and the results of an experiment to assess motion artifact reduction in real data. Section 6 discusses the performance of the SPWVD approach as compared to the traditional time domain and spectral methods. Lastly, the major findings of this paper are summarized in section 7. Background SpO2 is commonly monitored by a pulse oximeter, which has been widely adopted as a standard measure during anesthesia, neonatal care and post-operative recovery [ 5 , 6 ]. Pulse oximeters currently available on the market normally perform remarkably well when the monitored subject is in the resting position. However, their reliability is significantly reduced when the subject moves, even when movements are only involuntary, such as shivering [ 1 - 4 , 7 ]. Therefore, the reduction of motion artifacts is of particular concern in the development of pulse oximeters to be applied in ambulatory, pediatric and trauma settings, as well as for implementing them into wearable devices for personal home healthcare [ 8 ]. A number of attempts have been made in the past decade to improve the accuracy of pulse oximeters when subjects move. Typical methods can be generally classified into three categories: (1) based on an independent measure of motion; (2) based on a model of the ideal signal or the noise; and (3) based on features recognized from the corrupted signal. For techniques based on an independent measure of motion, one or more transducers (e.g. piezo or optical sensors) are employed to record the user's motion. By assuming that the artifact is a linear addition to the pulsatile photoplethysmographic (PPG) signal, the original signal can be reconstructed from the corrupted signal [ 9 - 11 ]. This hypothesis is however often doubted when inspecting PPG signals under typical artifact-producing forces [ 12 ]. This observation drives researchers to develop more realistic models for the PPG signal or the artifact. A recently proposed PPG artifact reduction methodology was based on the inversion of a nonlinear physical artifact model and could significantly reduce the effect of changes of probe coupling [ 8 , 12 ]. However, model-based techniques suffer inherently from the specificity of the model design and are unable to cope with all aspects of real-life scenarios. On the other hand, techniques based on feature recognition are free of the generic problem of model designs. Instead, these techniques often utilize some predetermined criteria to separate regions of corrupted and uncorrupted PPG signal and estimate the desired parameters from the uncorrupted portion of it. For example, Swedlow et al . calculated the derivative of a signal and identified a portion of it as a motion artifact whenever the ratio of adjacent positive and negative peaks of the derivative is below a threshold [ 13 ]. J.E. Scharf et al . evaluated the use of spectral analysis to separate the cardiac physiologic components from the recorded PPG signal that is contaminated by motion artifact for SpO2 estimation [ 14 - 16 ]. The above methodologies employ techniques in the time domain or frequency domain. However, due to the nonstationary nature of PPG signals, the use of time-frequency analysis appears to be extremely attractive. Dowla et al . proposed using a neural network together with a wavelet transform (WT) to estimate SpO2 in the presence of a motion artifact, and found out that this technique performs better than conventional algorithm that detects peaks and troughs of the PPG signal for estimating SpO2 levels [ 17 ]. In their method, a neural network was trained to identify the motion level, which was then fed into a second neural network together with the amplitude ratios at different scales of WT of the PPG signal to estimate SpO2 levels. It has been pointed out by another researcher [ 16 ] that using WT for SpO2 computation requires careful analysis and additional testing. WT does not result in a spectrum where the amplitude of a unique cardiac frequency can be directly obtained for SpO2 estimation. On the other hand, although such a unique component is available on the spectrum obtained from fast Fourier transform (FFT), the time-frequency resolution of FFT or STFT is relatively low when compared to other time-frequency techniques such as the Wigner-Ville distribution. The goal of this study is to investigate the use of SPWVD, a high resolution time-frequency transformation where the amplitude of a unique cardiac frequency is apparent, for the estimation of SpO2 levels. Methods Basic theory The traditional algorithms for estimating SpO2 detect peaks and troughs of the PPG signal in the time domain. Based on the Beer-Lambert law, which relates the optical path length and effective absorbance to the intensity of transmitted light, the relationship between intensity of transmitted light and SpO2 is commonly described as: I ( λ , t ) = I 0 ( λ ) exp[(- s ε HbO 2 ( λ ) + (1 - s ) ε Hb ( λ ))· c · d ( t )], (1) where, ε HbO 2 and ε Hb are the extinction coefficients of oxygenated and de-oxygenated hemoglobin, and s , c , and d represent SpO2, total concentration of hemoglobin and the optical path length respectively. By using two light sources – red and infrared lights – and calculating a normalized ratio of the AC component to the DC component for each light source, SpO2 can be computed from the ratio of ratios R , i.e. the normalized ratio of the red to the infrared transmitted light intensity. That is, In practice, SpO2 can be obtained from equation (3) directly or by an empirical equation that relates SpO2 and R . In this study, SpO2 is estimated directly from equation (3). SpO2 computation by weighted moving average (WMA) By calculating the ratio of the AC components and the ratio of the DC components of the two light sources, SpO2 can be obtained from every single pulse of a PPG signal. To stabilize the reading, the weighted moving average (WMA) is often used [ 18 ]. Typical averaging methods, e.g. the median averaging and standard arithmetic averaging, are applied to every several samples or samples in every several-second intervals. In this study, the SpO2 obtained by the WMA approach was the average of SpO2 samples in an 8-second period. Overlap processing was performed at 1-second interval. The 8-second period is selected in order to satisfy clinical requirements [ 15 , 16 ]. SpO2 computation by fast Fourier transform (FFT) Based on the hypothesis that cardiac rate can be estimated more easily by spectral analysis than time domain analysis, techniques in the frequency domain have been widely investigated as alternatives in pulse oximetry. For example, the FFT and discrete cosine transform (DCT) were proposed for estimating SpO2 [ 14 - 16 ]. These techniques calculate the spectrogram of the PPG signal in a fixed time period and select the strongest spectral line in the cardiac frequency band as the AC component. The cardiac frequency band is usually predetermined by certain thresholds or obtained from an independent pulse rate estimator, e.g. by applying electrocardiography in parallel. In this study, FFT is applied to every 8-second PPG signal at 1-second interval. The cardiac frequency band is predetermined as 0.8–2 Hz, i.e. corresponding to 48–120 bpm. SpO2 computation by SPWVD The Wigner-Ville distribution (WVD) of a signal x ( t ) is given as: where x ( t ) and x *( t ) are the time series of the signal and its complex conjugate respectively. The problem of the WVD is the so-called cross-term interference, which appears as frequencies that lie between the frequencies of any two strong components. In order to suppress cross-term interference, the smoothed pseudo WVD is often used: The two windowing operations h and g are equivalent to smoothing the WVD in the frequency and time domain respectively. Selection of the window is a compromise between the joint time-frequency resolution and the level of cross-term interference. Common choices of window include the rectangular and Kaiser windows [ 19 - 21 ]. In our experiment, we chose the Hamming window as both the time and frequency smoothing windows, g ( t ) and h ( τ ). The maximum magnitude within the cardiac frequency band of the SPWVD in each second was used for SpO2 computation. Since SPWVD represents energy distribution, the square root of the magnitude was used for calculating the ratio of ratios R and SpO2. Moreover, as the rate of change of SpO2 is relatively slow, SpO2 that changed by more than 2% per second was considered to be physiologically impossible, and was rejected from the calculation [ 22 ]. Simulation: STFT spectrogram versus SPWVD The performance of SpO2 computation by the SPWVD approach was evaluated in a simulation using PPG signals collected from a volunteer under modest random motions. For comparison, SpO2 was also estimated from the spectrogram obtained by STFT. The noise-mixing-composition (NMC) method was applied to mimic the clinical situation and to induce a range of signal-to-noise ratios (SNR) [ 23 ]. To synthesize a noise-contaminated signal, an artifact noise that has been verified to be similar to the real noise was added to an undisturbed basis signal. The synthesized signal and SNR are formulated as: with the basis episode S and artifact episode N and a mix parameter ε , which was adjusted to achieve the desired SNR. 512 samples of the PPG signal were recorded at 20 Hz from a volunteer. The signal was contaminated with in-band noise but with the cardiac frequency recognizable. The artifact noise was extracted by a filtering technique in the frequency domain. The spectrogram of the PPG signal was calculated using an FFT-based algorithm. Coefficients of the spectrum within 0.5 Hz of the primary cardiac frequency or its harmonics were set to zeros. The inverse FFT of the modified spectrum allowed us to obtain a pure artifact noise. Typical spectra of the contaminated signal (solid) and the resulting pure artifact noise (starred) are shown in Figure 1 . Figure 1 Typical spectra of contaminated signal and extracted artifact. Spectrum of the contaminated signal (solid) is obtained by FFT. Coefficients of the spectrum within 0.5 Hz of the primary cardiac frequency or its harmonics were set to zeros to result in the spectrum of pure artifact noise (starred). A set of synthesized signals with different SNR values was obtained by changing the value of the mix parameter ε in equation (6). The undisturbed signal was also estimated from equation (6) by setting ε = 0. SpO2 were estimated from the set of synthesized signals in an 8-second period at 1-second interval by using the STFT and SPWVD approaches. The mean SpO2 error during the complete 25.6 seconds is shown in Figure 2 as a function of SNR. Figure 2 The mean SpO2 error obtained by the STFT and SPWVD approaches at different levels of SNR. The SPWVD approach outperforms the STFT-based technique for low SNR. Figure 2 suggests that the two approaches lead to similar results for high SNR values (e.g. SNR>-5 dB). However, the SPWVD method outperforms the STFT-based technique for low SNR (e.g. SNR<-5 dB). Also, it is observed in this simulation that the errors are randomly positive or negative for high SNR values, but is mostly positive for low SNR values, i.e. the approaches consistently overestimate the SpO2 level. When the SNR value decreases, energy in the side-bands of the noise artifact that overlapped with the cardiac frequency components increases. And therefore, the resultant SpO2 level approaches a value that would have been estimated from the pure noise artifact, which differed by 7.2% from the actual SpO2 level for this specific trial. It should be noted that the pulse rate was predetermined in the simulation, which helped both approaches to determine the cardiac frequency band more accurately. In practical situations, the electrocardiogram can be recorded simultaneously and used as a reliable pulse rate estimator. As for the computational cost, the SPWVD approach can be implemented efficiently by making use of its symmetry properties, and thus, it can reduce the computational cost to a quarter of that of the STFT technique [ 20 ]. Experiment and Results Experimental protocol The purpose of this experiment is to compare the performance of three different methods (WMA, FFT, and SPWVD) in estimating SpO2 on subjects when they are (a) in a resting position and (b) in motion. Six healthy subjects participated in the study. Four kinds of motions have been investigated: horizontal movement (M1) and vertical movement (M2) of the hand, as well as the bending motion (M3) and pressing motion (M4) of the finger. These motions were selected because they are some of the common movements attributable to the motion artifact in pulse oximetry [ 7 , 24 ]. Subjects were asked to perform all four movements, 4 times each, and each time for a duration of 30 seconds. When performing each movement, subjects were asked to move their right hand, or the index finger of their right hand, for a magnitude of 2–5 cm at a frequency of 0.5–4 Hz, while keeping their left hand stationary. Signals were recorded simultaneously from the index fingers of both hands. Throughout the analysis, SpO2 or pulse rate estimated from the left hand, which was stationary at all times, was used as the reference. The reference estimates were obtained by the WMA method. The collected signals were separated into an AC and a DC component. The AC component was filtered out by a 4th order Butterworth band-pass filter with cut-off frequencies at 0.5 Hz and 20 Hz. The ratio of the DC components was computed directly in the time domain and the same value was used for the three different approaches, i.e. the WMA, FFT and SPWVD approach. On the other hand, a different ratio of the AC components was computed using each of the three approaches. To evaluate the performance of the different approaches, the SpO2 bias and precision, the pulse rate error, the dropout rate and the SpO2 performance index (PI) were calculated. The bias and precision are defined as the mean and standard deviation of the difference between reference and estimated SpO2 respectively. The pulse rate error is the difference between reference and estimated pulse rate. The dropout rate and SpO2 PI are evaluation parameters adopted from previous work by S.J. Barker [ 24 ]. The dropout rate is the percentage of time during which the technique fails to give a SpO2 reading, and SpO2 PI is the percentage of time during which the SpO2 level was within 7% of the reference reading. Results Table 1 shows the composite values from all the experiments when subjects were in a resting position and in motion. As indicated in Table 1 , all three approaches can achieve 100% SpO2 PI, 0.0% dropout rate and less than 3 bpm mean absolute pulse rate error in this experiment with a limited dataset. Table 1 Performance statistics of the different approaches. The bias, precision and performance index (PI) of SpO2, as well as the mean absolute pulse rate error and dropout rate, are used to evaluate the performance of the WMA, FFT and SPWVD approaches when subjects are in a resting position and in motion. State Approach SpO2 bias (%) SpO2 precision (%) SpO2 PI (%) Mean absolute pulse rate error (bpm) Dropout rate (%) Resting WMA 0.19 0.34 100 1.25 0.0 FFT 0.24 0.53 100 2.51 0.0 SPWVD 0.21 0.41 100 1.35 0.0 Motion WMA -1.31 3.58 81 16.4 4.6 FFT -1.42 3.18 83 11.2 0.0 SPWVD -1.07 2.42 91 5.62 0.0 However, the SPWVD approach shows significant improvement in both SpO2 and pulse rate estimation as compared to the WMA and FFT approaches when subjects were in motion. SpO2 estimated from the SPWVD, WMA and FFT approaches differed from the reference by -1.07 ± 2.42%, -1.31 ± 3.58% and -1.42 ± 3.18%, respectively. The mean absolute pulse rate error is reduced significantly (p < 0.05) from 16.4 bpm and 11.2 bpm for the WMA and FFT approaches, respectively, to 5.62 bpm for the SPWVD approach. The SpO2 PI also has the highest SpO2 PI (91%). Both the SPWVD and FFT approaches achieve 0.0% dropout rate. The WMA approach sometimes failed to give a reading during bending or pressing motions (dropout rate = 4.6%), which would lead to instrument "dropout" or "freeze" in clinical situations. Figure 3 shows the distribution of SpO2 bias and pulse rate error of the three approaches. As shown in Figure 3(a) , the SpO2 errors obtained by the SPWVD approach have a higher incidence (72%) in the main error band (-3%, 3%), which is the range of bias commonly accepted by most pulse oximeter manufacturers, as compared to that obtained by the WMA (55%) and FFT (56%) approaches. Figure 3 The distributions of (a) SpO2 bias and (b) pulse rate error obtained by the WMA, FFT and SPWVD approaches For the estimation of pulse rate, 90% of the pulse rate error falls in the error band (-10 bpm, 10 bpm) when the SPWVD approach is used (see Figure 3(b) ). When compared to the WMA and FFT approaches, where only 36% and 40% of the error fall in this error band respectively, the SPWVD significantly outperforms the other two approaches. Figure 4 shows the SpO2 output bias and precision under conditions with different kinds of motions: horizontal and vertical movements of the hand, as well as bending and pressing motions of the finger. It can be seen that the estimation of SpO2 by the SPWVD approach improved significantly (p < 0.05) as compared to the WMA and FFT approaches when subjects bend their finger or press their finger against the sensor. The three approaches show no significant differences (p > 0.05) when subjects move their hand horizontally or vertically. Figure 4 SpO2 (a) bias and (b) precision when subjects performed different types of motions: horizontal movement and vertical movement of the hand, as well as the bending motion and pressing motion of the finger. Figure 5 gives the error distribution of SpO2, obtained by the SPWVD approach, when subjects were in different types of motions. It is found that the bending (M3) and pressing motions (M4) of the finger have a relatively broader error distribution than the horizontal and vertical movements of the hand (M1 and M2). It can also be seen that the error distribution of M2 is slightly more concentrated than that of the M1. Figure 5 The error distributions of SpO2, obtained by the SPWVD approach, when subjects performed different types of motions: horizontal movement and vertical movement of the hand, as well as the bending motion and pressing motion of the finger. Discussion Spectral analysis is useful for separating motion artifact and cardiac physiologic spectra [ 14 - 16 ]. However, these techniques will not be applicable to spectra that contain frequency bands close to each other. Moreover, since both the motion and cardiac frequency are nonstationary in nature, simply using techniques in the frequency domain would not be able to separate them when one of the spectra varies within the fixed time window (i.e. an 8-second period in this study). Therefore, a time-frequency representation of the corrupted signal would be useful. The SPWVD approach is proposed for the reduction of motion artifacts because it can suppress cross-term interference while maintaining a good time-frequency concentration [ 19 ]. In addition, the approach utilizes the fact that SPWVD is an energy distribution and directly calculates the magnitude of the AC component from the spectrum. The approach solves the problem of WT, where a unique value for the cardiac frequency may not always be available [ 14 - 16 ]. Moreover, the approach does not require a large amount of samples for training, as the back-propagation neural network approach proposed in [ 17 ]. Standard parameters used to evaluate the performance of the techniques in pulse oximetry have been adopted in this study. S.J. Barker [ 24 ] evaluated 20 commercial pulse oximeters on 70 subjects, where data were recorded on each subject for 6 minutes during normal situation and 3 minutes during a hypoxemic episode. A motorized motion table was used to induce rubbing or tapping motions of the finger, with amplitude of ± 2 cm and frequency either fixed at 3 Hz or randomly varied between 1–4 Hz. As compared to the performance of some of the commercial products evaluated in [ 24 ], which have SpO2 bias in the range of 0.4–12%, SpO2 precision in the range of 2–6%, and SpO2 PI in the range of 27–94%, the proposed SPWVD approach reports comparable performance. The four motions investigated in this study are some of the common movements associated with motion artifacts affecting pulse oximetry [ 24 ]. By studying the effect of each component on the estimated SpO2, one would have a clearer picture of what kind of motion induces the largest error on SpO2 estimation. In future studies, it would be interesting to develop a model that specifically deals with one type of motion. As suggested by Figure 4 and Figure 5 , bending the finger (M3) or pressing the finger against the sensor (M4) induces a larger error on SpO2 estimation than horizontal or vertical movements of the hand (M1 or M2). In fact, this is consistent with the clinical findings discussed in [ 7 ], which suggested that bending and/or pressing the finger may cause the irregular compression of the vascular bed between the emitter and detector of pulse oximeter sensor, and thus inducing higher errors in the estimated SpO2. A potential solution would be to place multiple sensors around or along the finger so that the ratio of the light intensity received or a pressure reading could be an indication of the degree of bending, pressure exerted or even the level of distortion made on the peripheral blood vascular bed. Compared with the WMA and FFT approaches, the SPWVD approach showed a significant improvement (p < 0.05) in pulse rate estimation when subjects were in motion. Although such a significant improvement is not found in the estimation of SpO2, this is attributed to the fact that erroneous SpO2 estimates above the 100% upper bound were always rejected. It is hypothesized that when patients with SpO2 much lower than 100% are recruited as subjects for evaluating the different approaches, the performance of each approach will be more notably different from each other. However, this hypothesis remains to be proven in a clinical study involving a significantly large patient population. Conclusion Estimation of SpO2 by a time-frequency representation, the SPWVD, has been investigated in this study. The approach has been tested on four kinds of motions that are found in common movements associated with motion artifacts in pulse oximetry [ 7 , 24 ], i.e. the horizontal movement and vertical movement of the hand, as well as the bending motion and pressing motion of the finger. When compared with the WMA and FFT techniques, the SPWVD approach shows significant improvement (p < 0.05) when subjects bend their finger or press their finger against the sensor. When subjects were in motion, SpO2 levels estimated from the SPWVD, WMA and FFT approaches differed from the reference by -1.07 ± 2.42%, -1.31 ± 3.58% and -1.42 ± 3.18% respectively. The SPWVD approach achieves 0.0% dropout rate and 91% SpO2 PI when subjects were in motion. For the estimation of pulse rate, the SPWVD approach results in a mean absolute pulse rate error of 5.62 bpm, as compared to 16.4 bpm and 11.2 bpm by the WMA and FFT approaches respectively. The results of the study suggested that the SPWVD approach could potentially be used to improve the performance of wearable pulse oximeters by reducing the influence of motion artifacts, in particular when subjects bend their finger or press it against the sensor. Competing interests The author(s) declare that they have no competing interests. Authors' contributions YSY designed and carried out the experiment, analyzed and interpreted the data, and drafted the manuscript. CCYP helped to analyze and interpret the data, and assisted in drafting the manuscript. YTZ conceived of the study, and participated in its design and coordination and helped to finalize the manuscript. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC553999.xml |
522812 | Early visual evoked potentials are modulated by eye position in humans induced by whole body rotations | Background To reach and grasp an object in space on the basis of its image cast on the retina requires different coordinate transformations that take into account gaze and limb positioning. Eye position in the orbit influences the image's conversion from retinotopic (eye-centered) coordinates to an egocentric frame necessary for guiding action. Neuroimaging studies have revealed eye position-dependent activity in extrastriate visual, parietal and frontal areas that is along the visuo-motor pathway. At the earliest vision stage, the role of the primary visual area (V1) in this process remains unclear. We used an experimental design based on pattern-onset visual evoked potentials (VEP) recordings to study the effect of eye position on V1 activity in humans. Results We showed that the amplitude of the initial C1 component of VEP, acknowledged to originate in V1, was modulated by the eye position. We also established that putative spontaneous small saccades related to eccentric fixation, as well as retinal disparity cannot explain the effects of changing C1 amplitude of VEP in the present study. Conclusions The present modulation of the early component of VEP suggests an eye position-dependent activity of the human primary visual area. Our findings also evidence that cortical processes combine information about the position of the stimulus on the retinae with information about the location of the eyes in their orbit as early as the stage of primary visual area. | Background In humans, goal-directed movements to an object in space are improved by foveal vision, namely by gaze anchoring on the object [ 1 ]. Since motor and visual information are encoded in different reference frames, accurate reaching and grasping movements in space require ongoing registration and coordinate transformation of visual percepts with gaze and limb positioning. One essential transformation is to convert the retinal image from eye-centered coordinates in a target location with respect to the head by taking into account the position of the eyes in the orbit [ 2 ]. Single-cell recordings in monkeys revealed that the neural substrate of the visual-to-motor coordinate transformations is a change in the visual or motor response properties according to gaze position in extrastriate visual, parietal, oculomotor, premotor and motor areas [for review see [ 3 , 4 ]]. Functional magnetic resonance imaging (fMRI) studies have localized human homologues of such monkey areas and have showed that eye position signals modulate activity of extrastriate visual areas [ 5 ] and the parieto-frontal network related to hand-arm movements [ 6 , 7 ]. Less is known at the earliest stage of visual processing, namely in the primary visual area (V1). Previous electrophysiological [ 8 - 10 ] and modeling [ 11 ] studies described eye position dependent activity in V1 neurons to a lesser extent than that reported in parietal and premotor cortex but consistent with the idea that both retinal and eye position signals may also converge at early vision stage. However, the direct influence of eye position on V1-related activity in humans has not been investigated. Consequently, we used a specific experimental design based on pattern-onset visual evoked potentials (VEP) recordings to study the effect of eye position exclusively on V1 activity in humans. Thus, we focused our investigation on the first major VEP component C1, obtained using pattern-onset stimulation, because its distribution over the scalp and its retinotopic properties indicate an origin from the calcarine fissure that is V1. This issue has been the conclusion of all previous studies over the past 10 years regarding the cortical visual areas that generate the early components of pattern-onset VEP [for review see [ 12 ]]. The most recent reports even demonstrated in individual subjects a close anatomical correspondence between modelled dipoles for the C1 component and sites of activation in the calcarine fissure obtained in fMRI in response to the same visual stimuli [ 12 - 14 ]. The purpose of the present report was therefore to investigate the eye position-dependent activity of V1 in humans by characterizing the early C1 component of VEP and testing its properties at different eye positions. Results and discussion The C1 component reverses classically in polarity for upper vs. lower visual field stimulation [for review see [ 12 ]]. Consequently, we first characterized the C1 component over the 20 subjects by observing its polarity inversion for a central eye position. Figure 1B illustrates the representative polarity inversion of the C1 component on the grand averaged of VEP over the 20 subjects and in response to stimuli in the upper and lower quadrants of the right visual fields. Equivalent VEP and C1 polarity inversion were obtained in response to stimuli in the upper and lower quadrants of the left visual fields. The polarity inversion of the C1 components on the grand average VEPs in response to upper and lower hemifield stimuli were described previously for occipito-parietal sites using a 10–20 system montage with 62 scalp sites (see box included in Figure 1 ). It is noteworthy that a similar polarity inversion was measured in our study using only two occipital intermediate sites (IN3 and IN4) of the modified 10–20-system montage (see Methods for details). Once the C1 component was characterized for both upper and lower quadrant visual fields, the effect of eye position on C1 amplitude was measured only for both left and right lower quadrants visual field that is for the most salient C1 component which we observed. Mean peak latencies of C1 was calculated in each subject in response to both left and right lower visual quadrants and for both IN3 and IN4 recording sites. They ranged between 94.8 ms and 98.0 ms that are consistent with the C1 latency range previously observed in numerous studies [ 12 , 15 ]. For each subject, the C1 amplitude for five different eye positions (0°, 10° and 20° both left- and rightward) was then measured at these mean latencies for each lower quadrant and each recording site and, with respect to a 80 ms pre-stimulus baseline. We observed that checkerboard presented in the lower visual field elicited VEPs with modulated amplitudes of the C1 component in respect of eye position. Grand averaged VEPs over the 20 subjects in response to flashed stimuli in the right lower quadrant of the visual field for three different eye positions (0° and 20° both left- and rightward) are shown in Figure 2A . It is noteworthy that the eye position-dependent modulation of C1 amplitude was observed at the IN3 recording site for a 20° rightward eye position (Figure 2A , red trace) and at the IN4 recording site with a 20° leftward eye position (Figure 2A , green trace). In other words, the eye position effect was observed at the occipital recording site contralateral to the direction of the eye deviation. Equivalent VEPs but reverse eye-position effects were obtained in response to left field stimuli. A one-way Friedman repeated measures analysis of variance was conducted for each recording site and for each visual stimulation revealing a significant main effect of eye position on the amplitude of the C1 component. Complete data obtained at both recording sites, for each lower visual field are shown in Figure 2B . We chose to represent these data using relative amplitude measures that are amplitude for deviated eye position subtracted by the amplitude for central eye position. It means that amplitude zero corresponds to the central eye position situation. A post-hoc Dunnett's test (p < 0.05) using 0° as reference showed that the amplitude of the C1 component measured contralaterally to the deviated eye position was significantly different from 0° except for one situation (Figure 2B , white histograms). Conversely, the amplitude of the C1 component recorded ipsilaterally to the deviated eye position was not significantly different from 0° excepted for one situation (Figure 2B , grey histograms). Note that no parametric effect was observed for the C1 amplitude between 10° and 20° of eye eccentricity (Student's t-test, p < 0.05). The overall result of the present study revealed that the amplitude of first major component of VEP elicited by checkerboard (C1) is modulated by the eye position. The previous data linking the C1 component to a striate cortex generator, namely V1 activity (see introduction), led us to suggest that eye position influences the earliest cortical stage of visual processing. One may first suggest that the present results may be explained by the difficulty of maintaining eccentric fixation, which may have altered the pattern of fixational eye movements, such as microsaccades [fast, conjugate jerks, smaller than 1/3°, see [ 16 ]]. Recent studies have shown that microsaccades modulated neural activity in V1 [ 17 , 18 ], but to our knowledge no study has examined the effect of maintaining eccentric fixation on the occurrence of microsaccades. In the present study, we were able to track spontaneous saccadic eye movements superior to 1°. A one-way Friedman repeated measures analysis of variance for each visual stimulation revealed no significant main effect of eye position on the amplitude (p = 0.35) and on the frequency (p = 0.45) of saccades superior to 1°. We cannot rule out quantitatively a possible role for microsaccades in the eye position-dependent V1 activity. Regardless of such putative effects, however, one may argue that in case of an increase of the number of microsaccades related to eccentric fixation, a similar effect in terms of magnitude would be observed for both left- and rightward deviation. This was not the case in the present study. The question also arises if the effects of changing C1 amplitude may be due to oculomotor signals and/or retinal disparity that is to the difference in the position of the visual stimulus on each retina related to relative monitor distance. The absence of any parametric effect for the C1 amplitude between 10° and 20° of eye eccentricity may indicate that eye position effects are not due to the difference in the horizontal retinal disparity, but one may argue that a putative relationship between the horizontal disparity and the C1 amplitude is not linear. We evaluated the difference of both horizontal and vertical disparity between the different eye's deviations in our study (see Methods for details and Figure 3 ). Figure 3B shows that the difference between the magnitude of disparity for the central eye position and each deviated eye condition depends on both eye deviation and the distance of the visual stimulus from the fixation point. Interestingly, such a difference in terms of horizontal and vertical disparity does not depend on the stimulated quadrant visual field in the range of the present checkerboard's width. In other words, the magnitude of the relative disparity for a given distance from the fixation point and for a given deviated eye condition is similar for both left and right visual field stimulation. Since we observed that the C1 amplitude varied inversely in function of the hemifield stimulation (Figure 2 ), the eye-position effect observed on the C1 amplitude cannot be simply related to a horizontal and/or vertical disparity effect. Both direct and indirect arguments also suggest that variations in the C1 amplitude are not due to attention. Firstly, the subjects were instructed to keep firmly visual attention on the fixation point suggesting that the potential degree of attention required fixating binocularly the red fixation dot was similar across the different deviated eye positions. Secondly, numerous recent studies gave impetus to an emerging view that V1 activity may be modulated by attention through delayed feedback signals (160–260 ms) from extrastriate and/or oculomotor structures while the initial C1 sensory response (50–90 ms) was not modulated by attention [ 13 - 15 , 19 - 22 ]. Our present findings, together with those aforementioned, allow considering that both eye position and attention-related signals may affect the early stage of visual processing in different manner. The former may comes from extraocular muscle afferents and/or corollary discharges while the later is considered as a late top-down process [ 14 , 20 ]. Finally, Trotter et al. (1999) have shown that an eye position signal (extraretinal signal) is involved in the neural modulation process dealing with the eye position-dependent visual response observed in area V1 of behaving monkeys. Oculomotor signals coming from extraocular muscle afferents and/or corollary discharges are considered as the substrate of such an eye position signal and have been previously described in V1 [ 23 , 24 ]. Conclusions The present results and previous works obtained from neural recordings in monkeys indicate that changes in eye position can modify response properties in V1 that is at the earliest cortical stage of visual processing. Among the visuo-motor processing allowing accurate reaching and grasping movements in space on the basis of the image seen by the retina, the primary visual area may be therefore one of the first cortical relay to convert the image in eye-centered coordinates into a target location by taking into account the position of the eyes in the orbit. It endorses some recent arguments pointing out that V1 could no longer be considered only in relation to the pattern of light falling on the retina but appears to be a cortical area in which contextual influences take place too [ 9 ]. Methods Subjects Twenty right-handed healthy volunteers with normal visual acuity (range age 19–29 years, 9 males) participated in VEP recordings after they provided their written informed consent. The study was approved by the Basse-Normandie ethics committee (Caen, France). Experimental design The subjects were seated on a swivel armchair with their head stabilized with a headrest. An experimenter slowly rotated the armchair and locked it in one of the five different angles: 0°, 10° and 20° both left- and rightward from the center of the monitor leading to five different eye positions (Figure 1A ). The vertical swivel axis passed through the base of the nose between both eyes. Such a passive vertical movement of rotation of the whole body stimulated only the horizontal semicircular canals of the vestibular system with a time constant around 15 sec [ 25 ]. The VEP recording was started at least 1 minute after the rotational movement in order to stay away from the influence of the passive whole body rotational movement of the subjects. The subjects had to fixate binocularly a red fixation dot continuously visible in the center of the display as stimulus was flashed in 1 of the 4 quadrants of the visual field. Upper quadrants were used only in the case of central eye position in order to observe the inversion of polarity of the C1 component that allowed its detection on VEP recordings. The stimulus consisted of a black and white rectangular checkerboard (12 × 9° of visual angle, 0.6 cycle.deg -1 of spatial frequency) flashed against a black background (ISI = 500–1000 ms) and delivered by a visual stimulator (Nicolet, Madison, USA). The subjects were instructed to fixate continuously the fixation dot and to keep their attention on it, for each quadrant of the visual field so that the projection of the stimulus on the retina was equivalent whatever was the deviated eye position. The edges of the monitor and the space up to 1 meter around the monitor were masked with an opaque black sheet preventing any cue perception in the room except the flashing checkerboard. VEP recordings With respect to the purpose of the present study, we recorded VEP from the scalp using the two occipital intermediate sites (IN3 and IN4) of the modified 10–20-system montage [ 26 ] with both left and right mastoids serving as reference. The VEP from each site was recorded (Vicking, Nicolet, Madison, USA) at a sampling rate of 2500 Hz (0.1–100 Hz of band-pass filter with a 50 Hz notch filter). Prior to averaging, artifact rejection was performed to discard epochs with eye blinks. A total of 200 non-rejected epochs was averaged for each recording. Both horizontal and vertical eye movements were also monitored in each subject and during all VEP recordings, with EOG electrodes placed around the orbit of the right eye. The EOG system had a resolution superior to 1° of visual angle. All EOG records were analyzed by computer, using a dedicated software (SAMO) [ 27 ] which detects saccadic components and quantifies the amplitude and frequency of the spontaneous saccadic eye movements. Measurement of the disparity We evaluated the difference of disparity between the different eye's deviations (Figure 3 ). The difference in retinal angles (β - α) defines the magnitude of disparity, classically designated (η) [ 28 ]. Such a disparity depends on the position of the head's subject relative to the screen, designated (θ), on the interocular distance (a), and on the distance (d) between the middle of the cyclopean axis (c) and the fixation point (e1). We evaluated the magnitude of the disparity for any point (e2) of the checkerboard (Figure 3A ). Using the dot product, for the left eye, α can be expressed as following: a1e1.a1e2 = |a1e1| |a1e2| cos(α) (1) Expressing of the coordinates of the vectors a1e1 and a1e2 in the canonical reference ( x,y,z ) centred on the cyclopean axis (c): a1e1.a1e2 = a1e1 x a1e2 x + a1e1 y a1e2 y + a1e1 z a1e2 z (2) and combining both (1) and (2) led to: cos(α) = (a/2 cos(θ))(e2 x +a/2 cos(θ)) + (d +a/2sin(θ)) 2 / |a1e1| |a1e2| (3) Similarly, for the right eye, we obtained: cos (β) = (-a/2cos(θ))(e2 x -a/2cos(θ)) + (d -a/2sin(θ)) 2 / |a1e1| |a1e2| (4) Applying the numeric values defined in our study (d = 1 m, a = 8 cm), and combining both (3) and (4) allows to estimate the magnitude of the disparity (η = β-α) in function of the distance from the fixation point for each given eye's deviation. Therefore, we plotted the magnitude of disparity with each distance (e) and for each eye's deviation [-20°, -10°, 10°, 20°] normalized by the disparity calculated for the central eye position (θ = 0°, Figure 3B ). List of abbreviations used C1 : early component of visual evoked potential fMRI : functional magnetic resonance imaging IN3 and IN4 : occipital intermediate recording sites ISI : inter-stimulus interval V1 : primary visual area VEP : visual evoked potential Authors' contributions All authors designed the study. FA and OE carried out the experiments and generated the method for data analysis. OE performed the statistical analysis and generated the method for estimation of the magnitude of disparity. FA and LP wrote the first versions of the manuscript. All authors read, discussed and approved the final version of the manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC522812.xml |
509315 | Extinction, Slime, and Bottoms | Is biodiversity decreasing or is it being replaced by a splendid profusion of microbes -- and does this matter? | There is an old Chinese curse: ‘May you live in interesting times.’ According to those who know about such things, we live in a momentous time, the time of the Sixth Mass Extinction! But most of us do not feel at all cursed. Because, in fact, the Sixth is quite different to the previous Big Five—no-one would notice this one if we were not repeatedly reminded of it by ecologists. Previous mass extinctions were not so bashful, so discreet. The fossil record reveals the disappearance of pollen during previous ones, replaced by an abundance of fungus spores, telling us of a world of devastated forests rotting away. The earliest sediments after the mass extinction that did away with the dinosaurs are barren of fossils: so it is not just that species were going extinct, conditions for life itself were bad. Not only did species diversity drop, the abundance of life did as well. But conditions for life itself have never been better than today. In the history of the planet, there has never been anything as productive of life as a wheat field in Kansas. It may not have a large diversity of species, but that is a different matter. In fact, one of the reasons for the ongoing loss of plant diversity from grasslands is the very reason the wheat field is so productive—fertilisation. We are pouring nitrogen fertiliser into the environment and, through the wellstudied ‘paradox of enrichment’, this reduces species diversity while increasing actual biomass. Now, there is no question that if current trends of habitat alteration and climate change continue then we will ultimately lose large numbers of species—diversity will drop—but this does not necessarily translate into a loss of abundance of life, and that is a big difference between now and previous mass extinctions. Looking at specific groups of organisms tells the same story. So, for example, many island bird species are threatened, like the kagu of New Caledonia, but British seabird populations, like puffins, are booming. Worldwide amphibian diversity is threatened, but cane toads are a pest in Australia. Introduced species pose a threat to diversity—the ‘McDonald-isation’ of nature—precisely because they achieve enormous abundances. Actually, all six mass extinctions may have one very important thing in common: from the point of view of the vast bulk of life on the planet they are probably not mass extinctions at all. By any criterion—number of individuals or total biomass—the vast majority of life on earth is invisible—microbial. So, for example, at least 10% of the living biomass on earth consists of bacteria living deep in the oceans' sediments: it would take more than an asteroid impact to disturb them. And microbial life is extraordinarily robust: microbes can be found living happily in pressurised water hotter than your boiling kettle, in concentrated acid, and in rock, and their spores can survive for years in the rigours of outer space. In talks and lectures, the renowned oceanographer and paleontologist Jeremy Jackson paints a vivid picture of what is currently happening to coastal ecosystems, talking about a wall of slime emanating from populated areas and growing outwards inexorably towards the open oceans, replacing beloved ecologies like coral reef systems. What he means is that the visible life that we find attractive and useful—pretty fish, turtles, and so on—is being replaced by microbes in splendid profusion. It is taken completely for granted that this is disastrous. From a utilitarian point of view indeed it is disastrous, since we like eating fish and turtles, and don't like snorkling in slime. But from the point of view of life per se, again things have never been better. Life is so abundant that in some places all the oxygen in the water is completely used up. These are called ‘dead zones’, but they are no more ‘dead’ than the Dead Sea, which is actually teeming with life—just not fish. But, nonetheless, we consider what is occurring to be a disaster not just from a utilitarian point of view, but at some deeper level giving us an emotional reaction to the word ‘slime’—somehow it is just plain wrong. But this reflects nothing other than our evolutionary origins. Evolution has programmed us to be positively interested in plants and animals, our food, and to be repelled by slimes and oozes, teeming with potentially harmful microbes. These emotional responses colour our view of ecology, for example, in a way that has no parallel in other sciences: physicists do not just study particles that they find pretty. No ecologist wants to study the rich ecosystem that each of us carries around inside our gut, because evolution has programmed our brains to find bottom-related matters disgusting. I think it likely that naturalists from a different planet, silicon entities evolved under very different circumstances, would find tropical forests uninteresting (mainly primary producers with some herbivory and mutualisms) and animal guts fascinating, with their complex metabolic networks in which each node is manned by different species with wildly varying means of energy production. Our guts should be an ecological scientist's dream come true, ecological theatres that are replicated billions of times, which operate on a fast time scale and are easy to get to! (If aliens are ecologists, this would explain why they always ‘probe’ their abductees.) Many natural experiments are going on all the time as antibiotics and probiotics are administered and people find all sorts of different ways, voluntary or otherwise, to establish migration links between their gut ecologies. Microbiologists are increasingly interested in our guts from an ecological point of view but, unlike ecologists, they are used to faeces from their work in sewage plants. Two thousand years ago the Roman senator Cicero noted the creation of barren desert-like land in North Africa after the forests were felled for their timber, providing the earliest record of an ecosystem ‘service’ provided by forests—the stabilisation of soils. Other services provided by biodiversity readily come to mind, like pollination and carrion clean-up, and there may be many more. But perhaps the clearest example of an ecosystem service provided by biodiversity comes from our gut. Throughout our history, until very recently, we all had worms. In rich countries we have quite happily eradicated them from our inner ecosystems with none of the handwringing we expend on rhinos. But it is increasingly believed that the loss of worms from our internal ecology is responsible for the upsurge in inflammatory bowel disorders such as Crohn's disease and colitis. In fact, there are clinical trials underway in the United States testing the efficacy of worms as treatment for these diseases. The mechanism is clear: worms trigger one arm of the immune system which down-regulates another, inflammatory arm. Our immune system has evolved to expect a certain constellation of species in our gut: in that context, worms provide an ecosystem service of balancing the immune system. Our perception of our impact on the planet as equivalent to a mass extinction simply reflects the evolutionary prism through which we view life. Of course, we may yet live up to our own publicity and pull off something apocalyptic like a runaway greenhouse that sterilises the Earth. But it is at least as likely that the microbial world, resentful at being either ignored or exterminated, will come up with something to consign us to a footnote in the history of life when it is ultimately written by the silicon entities. The Spanish flu, SARS, and HIV have just been early experiments. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC509315.xml |
532397 | The History of the Intron—Balancing Gains and Losses | null | In the 25 years since they were first discovered, introns have puzzled molecular biologists because of their uncertain function and mysterious origin. Introns are non-coding DNA sequences that reside inside a gene, splitting it into discrete units called exons. The resulting disruption of coding sequence continuity would wreak havoc in protein assembly if eukaryotic cells did not dispose of introns in messenger RNAs—the intermediates in the decoding of gene sequences to produce protein chains—in a now well-described process known as splicing. At first glance, introns may seem like pesky parasites for which eukaryotes have cleverly evolved bypass mechanisms. But introns may also benefit their hosts. Evolutionary advantages of introns include the possibility to create new genes by cutting and pasting exons from existing genes or to diversify the protein output of a single gene by splicing the exons together in different ways. Thus, balancing intron gains and losses clearly has important evolutionary implications for a host. Yet different organisms strike that balance differently. The budding yeast Saccharomyces cerevisiae averages less than one intron per gene, whereas mammalian genes routinely have 10 or more. Whether these differences reflect different propensities for gaining or losing introns is the subject of ongoing debates. Organisms with low intron density display a bias for insertions at the beginning (5′ end) rather than the end (3′ end) of genes. A popular hypothesis is that in these organisms, genes lose their introns through a process that rewrites genomic DNA using as template the messenger RNAs purged of intron sequences. This process might preferentially remove 3′ introns because it relies on an enzyme called reverse transcriptase that can be primed to read RNAs starting at their 3′ end. The hypothesis has gained experimental support in yeast. It also presents the advantage of potentially explaining intron paucity and 5′ position bias in one stroke. In a new study, Cydney Nielsen and her colleagues present evidence that challenges this model. They address intron dynamics with a genome-wide survey of intron distribution among four Ascomycete fungi with recently completed genome sequences. The four fungi ( Neurospora crassa , Magnoporthe grisea , Fusarium gramineum , and Aspergillus nidulans ) form an evolutionary tree with branching points estimated at 200, 230, and 330 million years ago. While they diverged from yeast some 500 million years ago, they share with yeast a low intron density (one to two introns per gene) and a 5′ position bias. The authors' approach is to tally intron gains and losses during the evolution of these four species and then plot their positions along the genes' length. They identify 3,450 gene regions that are clearly conserved in all four species and harbor an intron in at least one of them. To distinguish intron gains from losses, they rely on a simple parsimony principle, which they refine with additional probability analyses. In brief, an intron present in only one species counts as a gain; an intron absent from one species but present in its closest relative and in a cousin counts as a loss. Nielsen and colleagues record between 150 and 350 intron losses in each lineage. Surprisingly, losses do not occur preferentially at the genes' 3′ end. The authors conclude that while a 3′ reverse transcriptase-based mechanism might be a factor, it cannot be the sole reason for the introns' 5′ bias. The other surprising result is that intron gains occur at almost the same rate as losses in all lineages. Intron gains therefore play an important role in the evolution of even intron-poor genomes. Clearly, intron distribution in fungi owes to forces more complex than simple 3′ intron elimination, forces that the authors propose may also shape evolution of other eukaryotic genomes. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC532397.xml |
534094 | A randomised controlled trial to determine the effect on response of including a lottery incentive in health surveys [ISRCTN32203485] | Background Postal questionnaires are an economical and simple method of data collection for research purposes but are subject to non-response bias. Several studies have explored the effect of monetary and non-monetary incentives on response. Recent meta-analyses conclude that financial incentives are an effective way of increasing response rates. However, large surveys rarely have the resources to reward individual participants. Three previous papers report on the effectiveness of lottery incentives with contradictory results. This study aimed to determine the effect of including a lottery-style incentive on response rates to a postal health survey. Methods Randomised controlled trial. Setting: North and West Birmingham. 8,645 patients aged 18 or over randomly selected from registers of eight general practices (family physician practices). Intervention: Inclusion of a flyer and letter with a health questionnaire informing patients that returned questionnaires would be entered into a lottery-style draw for £100 of gift vouchers. Control: Health questionnaire accompanied only by standard letter of explanation. Main outcome measures: Response rate and completion rate to questionnaire. Results 5,209 individuals responded with identical rates in both groups (62.1%). Practice, patient age, sex and Townsend score (a postcode based deprivation measure) were identified as predictive of response, with higher response related to older age, being female and living in an area with a lower Townsend score (less deprived). Conclusion This RCT, using a large community based sample, found that the offer of entry into a lottery style draw for £100 of High Street vouchers has no effect on response rates to a postal health questionnaire. | Background Self-completed postal questionnaires are an economical and simple method of data collection for both research studies and audit activity. They are cheaper than telephone or personal interviews [ 1 ] and may be particularly useful in medical research as the response rate to sensitive questions may be greater than from other methods of data collection [ 2 ]. However, questionnaire based studies are subject to non-response bias [ 3 ]. If there are differential response rates from certain groups in the sample population, then generalisability of results to the target population is questionable. The lower the response rate, the more open to criticism the conclusions drawn from the research will be. High response rates are important not only because they reduce the risk of non-response bias but also because they increase the precision of parameter estimates [ 4 ]. Ensuring a high response rate to initial mailings has the additional benefit of reducing the costs associated with re-mailing or other methods of follow-up such as telephone interviews. Strategies for increasing response rates can be summarised as falling into five broad categories; the covering letter (personalisation, use of an appeal etc), incentives (cash or other reward), contact (pre-notification and follow-up), mailing (return envelopes, type of outgoing postage etc) and questionnaire (length, format, colour etc) [ 5 ]. Pre-notification, follow-up, university sponsorship, cash incentives, first class postage, freepost return and questionnaire colour have all been shown to increase response rates [ 4 ]. A Cochrane review [ 3 , 6 ] reported similar findings, with monetary incentives and recorded delivery having the greatest effect on response (response rates doubled) and a range of other strategies significantly increasing the odds of response. However, individual studies demonstrated wide variability, were conducted across a variety of disciplines (psychology, medicine, business etc) and in a variety of countries; their generalisability is, therefore, questionable. An individual's decision to reply to a questionnaire is bound up in the content of the questionnaire, its relevance to the individual and who is perceived as benefiting from completion [ 3 ]. Questionnaires asking about personal experience may be treated differently to those asking for opinion, and information requested by a retail company may be perceived as less important than that sought by a charity, academic institution or health provider. Much of the research relating to the effectiveness of specific interventions on response rates in general population surveys was conducted before 1990 and from a market or sociological perspective. More recent work has been conducted on the effects of incentives on increasing response rates in populations of medical professionals. A lottery style incentive (prize of a weekend break) was reported to generate a small increase in response rates from General Practitioners (GP, Family physicians) in Quebec [ 7 ] and cash incentives also increased response rates in a GP population [ 8 ]. It is not clear, however, whether results from these professional groups can be generalised to general population surveys. Studies sampling members of the public tend to confirm the more general findings of the Cochrane review [ 6 ], which included surveys of professionals as well as the public, that financial incentives have a positive effect on response and even a small financial incentive ($1 – $2) increased response rates to a health related survey [ 9 , 10 ]. However general population surveys often require large samples and the cost of offering even a small individual monetary reward may be prohibitive. The Cochrane Review [ 6 ] identified 45 trials (44,708 participants) of non-monetary incentives (e.g. key ring, offer of entry into a lottery, offer of study results) and found that the odds of response were approximately a fifth higher when using such incentives. However, their meta-analysis did not discriminate between the type of non-monetary incentive, the population (i.e. professional versus public) or the subject matter (i.e. health versus non-health). Of the nine studies identified that utilised some form of lottery incentive [ 6 ], only one demonstrated a significant benefit [ 11 ]. The remaining eight studies had non-significant results although five of these suggested a trend towards an improved response [ 12 - 16 ], two tended to suggest a detrimental effect [ 17 , 18 ] and one showed no effect [ 19 ]. Evidence on the use of lottery style incentives is therefore conflicting. One of the studies that showed no overall benefit in including a lottery incentive [ 13 ] demonstrated a significant difference in favour of the lottery group (24.4% versus 18.5%) with respect to responses to a first mailing. Although this benefit was not maintained after a reminder, such a result could reduce the cost of undertaking surveys. Three of the five [ 13 , 15 , 16 , 18 , 19 ] evaluations of a lottery incentive using a general population sample, indicated small but non-significant increases in response. The three studies that demonstrated a tendency towards increased response rates had offered entry into a draw for a restaurant meal to the value of $100 (1985) [ 15 ], $50–$200 (1988) [ 16 ], and a range of prizes including a television, $100 and a trip to Las Vegas (1989) [ 13 ]. There is, therefore, some evidence to support the use of lottery incentives in increasing questionnaire response rates. However, studies focussing on health related issues are limited to only two that used a general population sample [ 15 , 19 ] and three that sampled patients [ 12 , 14 , 17 ]. The wide range of incentives used, study settings and subjects mean that it is difficult for medical researchers to determine the applicability of this evidence to community based surveys in the UK. Determining the community prevalence of disease is important for needs assessment, service planning and determining the potential economic implications of new treatments. Studies aiming to precisely estimate the prevalence of disease usually require large samples and often use postal questionnaires. A large community mailing to determine the prevalence of Irritable Bowel Syndrome (IBS) [ 20 ] provided an opportunity to embed a randomised controlled trial to investigate the effect of including an incentive on response and completion rates. Medical research rarely, unless sponsored by the pharmaceutical industry, has the capacity to pay a cash incentive for all returned questionnaires. This study therefore aimed to determine the effect of including a low-cost lottery style incentive (returned questionnaires being entered into a draw for a prize) on the response and completion rate of a health questionnaire survey. A secondary objective was to combine data from this study with that from previously published studies based in a health environment [ 15 , 19 ] using patient/general population responders to provide a more precise estimate of effect. Methods Location This study was undertaken in Birmingham, a large city in the West Midlands region of the United Kingdom. Ethical approval was obtained from North Birmingham and West Birmingham Research Ethics Committees prior to commencement of the study. Participants Eight general practices in the North and West Birmingham areas were recruited to participate in a health survey designed to determine the prevalence of IBS in the community [ 20 ]. All practices in these areas were invited to participate in this survey and practices were randomly selected from those practices who had expressed an interest in participation, after stratification for deprivation scores. To provide a sample with representation from all socio-economic groups, practices were selected from each of the four quartiles of the Townsend scores. Townsend scores are calculated from small area statistics collected during the decennial census (most recent in 2001) and provide an indicator of deprivation. Practices were allocated a Townsend score based on their location (postcode). All patients aged 18 and over were eligible for inclusion. The only exclusion criteria were patients for whom the general practitioner indicated that mailing would be inappropriate; this typically included patients known to be terminally ill or patients unable to complete the questionnaire e.g. those with severe learning disability. Questionnaires returned by the postal service as 'not at this address' were removed from the sampling frame (denominator) and response rates calculated as the number returned divided by this denominator. Trial size Estimates were based on the prevalence study within which this trial was embedded and indicated that 8,000 patients should be mailed. This provided 4,000 in each arm of the RCT, sufficient to demonstrate a 4% difference in response rates with 90% power at the 5% significance level, assuming a 60% response rate. Randomisation To minimise contamination (two or more individuals in a household receiving a questionnaire, but not all receiving the lottery incentive) only one person per domestic address was selected. Practice registers were utilised to generate randomly ordered lists of addresses and then one individual aged 18 or over was randomly selected from each address until the required number of patients had been identified. The prevalence survey, within which this trial was embedded, aimed to recruit a stratified random sample, comprising 1,150 patients from each participating practice to ensure sufficient cases were included from each of the Townsend quartiles. Where the practice was unable to generate 1,150 patients (due to smaller list sizes) the maximum number of patients available was included. Randomisation was performed on a 50:50 basis within each practice, to control for practice effects, using a computerised simple random number sequence. Intervention All patients selected received a health questionnaire. Questionnaires had three sections; section one requested personal and demographic details in addition to details of personal and family medical history; section two was the SF12 [ 21 ], a validated generic quality of life measure; section three was a self completed questionnaire version of the ROME II criteria [ 22 ] (to confirm diagnosis of IBS). A covering letter, sent in the joint names of the University and the relevant general practice, explained that the practice was participating in a research project to find out about the number of patients affected by certain conditions and the ways in which ill health affects people's quality of life. All patients received a reply paid envelope (addressed to the research team at the University of Birmingham) with the questionnaire and were informed that return of a blank questionnaire would indicate the wish not to be involved and they would not be contacted further. The intervention group received an identical questionnaire to the control group. The covering letter was also identical apart from the addition of a paragraph explaining that all returned questionnaires would be entered into a draw for a prize of £100 of "High Street shopping vouchers". This letter stressed that entry into the draw was dependent on return rather than completion of a questionnaire, as this was deemed to be more ethically responsible. In addition to the letter, a flyer printed on brightly coloured paper (yellow) was included for intervention patients, highlighting the fact that returned questionnaires would be entered into a draw. All non-responders were re-mailed after 4 weeks. Again, the intervention group received the additional paragraph in the covering letter and an additional flyer. All follow-up mailings included a copy of the questionnaire and a reply paid return envelope. Data handlers were not blinded to the intervention status of responders, but this was not considered to be a source of bias as response rates were the primary outcome. Mailings took place in the period January to July 2001. Mailings were conducted by practice and within each practice, control and intervention patients were mailed on the same day. Outcomes The principal outcome was the overall response rate. Response rates to initial and follow-up mailings, and numbers of blank responses were also compared between groups. Analysis Analysis was on an intention to treat basis. Response rates were compared between the two arms of the trial using chi-squared tests. Predictors of response were identified by logistic regression using backward elimination. Variables entered into the starting model included randomisation arm, age, sex, Townsend score derived from patient postcodes, practice and all two-way interactions with the randomisation arm. Two recent systematic reviews of strategies to influence response rate were identified [ 3 , 4 ] and from these, all studies of non-professional groups using lottery-style incentives in a health environment were identified and full papers obtained. Data from these was combined with those of this study and a meta-analysis undertaken (Rev Man software). Results Eight thousand six hundred and forty five patients were included in the trial; 4,325 were randomised to the lottery arm and 4,320 to the control group and a 50:50 split was maintained within each practice. The trial profile is shown in Figure 1 . Two hundred and sixty questionnaires were returned by the Royal Mail as 'not at address' and were not included in the analysis. The proportion of these returned questionnaires were comparable between trial arms (121 (2.8%) in the intervention arm and 139 (3.2%) in the control arm). Baseline characteristics of patients were similar between the two randomised groups (Table 1 ). Four thousand and twelve individuals responded to the initial mailing; 1996 (47.5%) in the intervention arm and 2017 (48.2%) in the control group (χ 2 = 0.5, p = 0.48). A further 1197 replied to the reminder mailing; 616 in the intervention arm and 581 in the control arm giving an overall response of 2612/4204 (62.1%) in the intervention arm and 2598/4181 (62.1%) in the control arm (χ 2 = 0, p = 0.99). The numbers of questionnaires returned blank was similar for both groups, 197 (4.7%) in the intervention and 217 (5.2%) in the control arm (χ 2 = 1.1, p = 0.29). No objections to the use of an incentive or requests for exclusion from the prize draw were made to the research team or participating general practices. Response rates varied by practice from 37% to 77% (Table 2 ). Practice, patient age, patient sex and Townsend score were identified as predictive of response, with response rates increasing with age ((OR per 1 year change in age) (95% CI) = 1.02 (1.016, 1.024)) (i.e. if the odds of responding at age 40 are 1.0 (1:1), then odds of responding at age 45 are 1.1 (1:1.1)). Lower response rates were associated with being male (OR = 0.53 (0.48, 0.58)) and living in an area with a higher Townsend score (more deprived) (OR per 1 point change in deprivation score= 0.93 (0.92, 0.95) (Table 3 ). The small number of practices, and limited number of explanatory variables associated with each practice, meant that whilst practice was predictive of response it was not possible to explore the practice characteristics related to this. Randomisation arm and its interaction with other terms were not identified as significant predictors of response rate. This indicates that practice, patient age, sex and Townsend score affected response rates whereas the lottery incentive did not. Given the lack of difference between groups, no economic evaluation was conducted. The lottery incentive presented additional costs in terms of production and inclusion of flyers (<£0.01 per individual) and provision of the pre-specified prize (£100 overall), and there was therefore a net disbenefit to the use of the incentive. Two previously published relevant studies were identified [ 15 , 19 ]. Results of the meta-analysis are provided in Figure 2 . Discussion This RCT using a large community sample suggests that using a lottery draw style incentive for £100 of High Street vouchers has no effect on overall response rates to a postal questionnaire asking about health and medical history. Whilst it is possible that the locality of the study (Birmingham, UK) is not typical, the range of practices and Townsend scores of the sample, suggest this is unlikely. It is also possible that a larger incentive or a cash prize, rather than vouchers, may influence response rates. However the value of the incentive offered was comparable to, or greater than, previous studies. The only published study which has shown a significant benefit of a prize draw was of professionals [ 11 ] and we, therefore, believe that the findings of this large trial are likely to be generalisable to health related surveys of the general public. There was no difference between the groups in the numbers of questionnaires returned blank indicating that the incentive did not just encourage people to return incomplete questionnaires. There was also no difference in the numbers replying to the first mailing. Had the incentive increased numbers replying to the initial mailing, even if it failed to increase response overall, it may have proved cost-effective in terms of reducing the costs of follow-up mailings. Our results indicate that previous research suggesting increased response to initial mailings [ 13 ] is not generalisable. The questionnaire used in this RCT was designed to determine disease prevalence. It is possible that such mailings over-inflate prevalence estimates, as individuals responding may be encouraged to do so because of their vested interest as an individual experiencing disease. Inclusion of an incentive not related to disease status may encourage response from a wider range of the population and may provide more accurate estimates of prevalence. However, this study does not suggest that future health surveys would gain any response benefit from the inclusion of a prize draw incentive. Indeed the addition of the data from this study to that of the two previously published studies based in a health environment [ 15 , 19 ] using patient/general population responders provides further evidence of a lack of effect (Figure 2 ). Future work in this area may best be conducted using qualitative methodologies to explore factors related to response in community surveys. Conclusions A lottery draw style incentive for £100 of High Street vouchers does not affect response rates to a postal health survey, when used in a general population sample. On the basis of this large RCT we would not recommend utilising similar incentives in general population health research. Competing interests The author(s) declare that they have no competing interests. Authors' Contributions SW and LR conceived the idea and designed the study. PB contributed to the study design and collected and validated the data. AR completed the analyses. LR wrote the first draft of the manuscript, all other authors made significant contributions to the writing of the manuscript and approved the final version. SW is the guarantor. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC534094.xml |
521499 | Use of satellite imagery in constructing a household GIS database for health studies in Karachi, Pakistan | Background Household-level geographic information systems (GIS) database are usually constructed using the geographic positioning system (GPS). In some research settings, GPS receivers may fail to capture accurate readings due to structural barriers such as tall buildings. We faced this problem when constructing a household GIS database for research sites in Karachi, Pakistan because the sites are comprised of congested groups of multi-storied building and narrow lanes. In order to overcome this problem, we used high resolution satellite imagery (IKONOS) to extract relevant geographic information. Results The use of IKONOS satellite imagery allowed us to construct an accurate household GIS database, which included the size and orientation of the houses. The GIS database was then merged with health data, and spatial analysis of health was possible. Conclusions The methodological issues introduced in this paper provide solutions to the technical barriers in constructing household GIS database in a heavily populated urban setting. | Introduction Geographic data are increasingly being employed in health studies [ 1 ]. By studying disease patterns in space, we can understand the relationships between socioecological exposure and illness [ 2 , 3 ]. Such understanding may help the formulation of need based healthcare systems and health intervention programs. Geographic methods provide a wide spectrum of geographic scales from local to global for analyzing health and health-related data. Regional variation in disease incidence be attributed to regional or global differences in ecological or socio-environmental phenomena [ 4 ]. Local-level geographic variation of disease obtained from fine resolution geographic data can provide clues about the spatial variability [ 5 ], and may pinpoint areas where health interventions are needed. One way to facilitate the measurement of local variation in health outcomes is to create household-level geographic information systems (GIS) database. Household locations can be captured by using GPS (global positioning system) receivers [ 6 , 7 ]. However precise geographic data on households are an absolute requirement for critical examination of local variation of the disease and its association with the environment [ 8 ]. A large variety of GPS receivers are available in the market and different GPS receivers provide different levels of accuracy. A low cost receiver can capture data with an accuracy of 5 to 10 meters provided that they are configured properly and the satellites have good alignment at the time the data are collected [ 9 ]. The alignment of the GPS satellite constellation at a particular time can be measured using GPS trip planning software. It is essential that the GPS receiver has a clear "view" of at least four GPS satellites which can be obstructed by large structures such as buildings or mountains. In congested urban settings, collecting household locations in narrow lanes using the GPS can be challenging. Faced by such challenges we explored satellite imagery in order to acquire household GIS data in urban slums in Karachi, Pakistan. This paper describes the methods used to construct the household GIS database and the technical barriers one might encounter during the construction of a database. The household geographic information systems project Geographic studies have been considered as one of the research disciplines of large Vi (antigen) typhoid vaccine effectiveness trials as well as typhoid disease burden studies [ 10 ]. The Vi typhoid vaccine provides a comparable degree of protection to the whole-cell type but with less severe side effects. Only one dose is required for a course of vaccination. The studies are part of the Diseases of the Most Impoverished (DOMI) program, a multi-country, multi-disciplinary health research program aimed to accelerate the development and introduction of a new generation vaccines against cholera, typhoid fever, and shigellosis in several Asian countries. The program involves a number of parallel activities including epidemiological studies, social science studies, and vaccine technology transfer. The local collaborator of the household GIS project in Karachi is the Pediatric Department of the Aga Khan University Hospital, Karachi, Pakistan. Technical support for the project was provided by Techno-Consult International, Karachi, Pakistan. The aim of the project was to construct a spatial database that includes household locations, study area boundary with administrative units, and other geographic features such as hospitals/clinics, schools, mosques, roads, lanes, and water bodies. The project area included four Karachi slums including Sultanabad, Hijrat Colony, Rehri Goth, and Sherpao Colony (Figure 1 ). Sultanabad and Hijrat Colony are adjacent areas near the port of Karachi, and Rehri Goth and Sherpao Colony are about two kilometers apart located 20 kilometers South East of Karachi. In 2002, a census was conducted in the four slums to enumerate the study population. The list of households and their addresses were obtained from the population database. Figure 1 The study sites in Karachi, Pakistan. The geographic position of the four study sites along with other geographic characteristics of Karachi are shown in the map. The base map A commercially available map of Karachi was used as the base map for this GIS project. The base map was georeferenced with four identifiable landmarks using handheld GPS receivers with accuracy of approximately five meters. This accuracy was considered sufficient to identify the study areas, to order satellite imagery, and to conduct subsequent ground surveys. After georeferencing the map, the main geographic features such as roads, hospitals/healthcare centres, and other city landmarks were digitized and incorporated into the baseline GIS database (Figure 1 ). The satellite imagery Satellite imagery is available at different spatial, temporal, and spectral resolutions [ 11 ]. Different sensors capture images of the earth surface in different spectral resolutions, which allow different surface features to be differentiated. At the time of the project, the highest resolution commercially available satellite imagery was the one-meter panchromatic from the IKONOS (Space Imaging, Inc) satellite. We acquired a panchromatic IKONOS image (Figure 2 ) for two study slums, Sultanabad and Hijrat. We found that the one-meter panchromatic imagery was not helpful for separating lanes from building and mud roofs from open ground. We therefore acquired the four-meter multispectral IKONOS imagery for the two other study slums Rehri Goth and Sherpao Colony where many houses are made of mud. The result was more appropriate for our purposes even with the loss of spatial resolution. Figure 2 IKONOS image of Sultanabad, Karachi, Pakistan. The landscape of the Sultanabad study area obtained from IKONOS satellite imagery. Image processing and georeferencing The satellite images were enhanced using an image processing software package (ERDAS Imagine, Atlanta, USA) to facilitate the digitization of house parcel boundaries. High precision, dual frequency GPS units (Trimble 4000 ssi) were used to capture data at several identifiable points on the images to be used as ground control points (GCPs). To transfer images into a GIS database, it must be geometrically rectified to a known coordinate system on the basis of a number of GCPs [ 12 ]. Most of the GCPs were selected from the periphery of the study area so that possible errors would converge towards middle of the area. After locating GCPs on the satellite image and identifying them on the ground, GPS readings were obtained at centimeter level accuracy. The GPS data were collected in the WGS-84 (World Geodatic Systems-84) datum in the latitude/longitude system and were subsequently transformed into the Universal Transverse Mercator (UTM) Zone 42-North system. The GCP coordinates within the UTM projection were then integrated with the satellite images using the ERDAS Imagine software for georeferencing. The resultant root mean square (RMS) errors were approximately two meters, which was considered sufficiently accurate for the purpose of constructing the GIS database. Digitization of house parcels After georeferencing the images were resampled, converted into JPEG files, and were imported in AutoCAD Version 14 (Cadopolis.com Inc., Canada). These processes allowed the parcel boundaries to be delineated through heads-up digitizing. The resampled satellite imagery was inserted as backdrop in AutoCAD. The image was aligned to correct its scale, translation and rotation by using two GCPs located at the corners of the image. After the alignment of the raster images in AutoCAD the house parcels were digitized. One image was used on multiple workstations to digitize different portions of the image, which were subsequently merged to form a complete area map of house parcels (Figure 3 ). Figure 3 House parcels added to the satellite image, Sultanabad, Karachi, Pakistan. The house parcels drawn using AutoCAD superimposed on to the IKONOS satellite imagery. Ground survey The georeferenced image and the household parcel maps were used during a ground survey. The survey team consisted of civil engineers skilled in drawing household parcels and other field staff who conducted the census survey. The ground survey included verification of the size and orientation of household structures, resketching of the structures where needed, and locating specific household parcels so that they could be given their unique address identification number (ID) which was created during census survey (Figure 4 ). The address ID of each household was marked on the walls or doors at the time of census survey. Each time, the ground survey was started from a known location on the image, the household address ID was verified, and the IDs were marked on the hard copy map. Incorrect sizes or orientations of the digitized house parcels were also corrected during the ground survey. Figure 4 The address ID is used to link house parcels to population database, Sultanabad, Karachi. The unique census ID (address ID) of the household assigned to each parcel is shown inside the house parcel (the map in right side). The ID is used to link household census and disease surveillance data. After completing the ground surveys, the maps were updated using AutoCAD, and the address IDs were added to household parcels in the database. Finally, the household parcel AutoCAD files were imported as polygons into the ArcGIS software package (ESRI Inc., USA). The process included several checks for missing households, duplicate address IDs, and misplacement of address IDs and data were corrected when an error was found. The corrected data were validated by randomly selecting several household parcels (about 2%) from different zones of the study area and verifying their position on ground. At this stage, we observed no discrepancies in the data between ground verification and the satellite based maps suggesting that the household level GIS database is highly accurate. Implementation of the health GIS (HGIS) study The HGIS database is composed of spatial and non-spatial components (Figure 5 ). The spatial component consists of geographic features of households, roads, rivers, prominent places (e.g., hospitals), schools, and administrative boundaries. Each type of geographic feature was drawn in a single map layer. For instance, although both hospitals and schools were spatially referenced by points, we created two map layers for these two types of geographic features. The non-spatial component of the database consists of household-level data that include household socioeconomic status and individual-level data such as vaccination and disease history. The database relationships for patients with target diseases are shown in Figure 5 . Figure 5 Entity relationship of the health study GIS database, Karachi, Pakistan. Inter-relationship between spatial and non-spatial database and intra-relationship of the database tables are shown here for the GIS-based health study research project. The descriptions of the entity relationship are descried in the texts. Entity relationships between data tables are shown as lines, and logical relationships between entities have parentheses around them (Figure 5 ). In the logical relationship (1,N), "1" indicates each entity should be linked to an entity on the other end, and "N" indicates multiple entities can be linked to an entity at the other end. Similarly, the "0" in (0,1) indicates not all entities will be linked to an entity at the other end. The "1" of the relationship indicates not more than one entity will link to an entity at the other end. For example, the relationship of "member" towards "patient" is shown as (0,1). Here "0" indicates not all records in "member" are to be linked in "patient," and "1" indicates not more than one record of the "member" can be linked to a record in "patient". Similarly, "1" in the relationship (1,N) of "patient" towards "member" indicates all records in "patient" should be linked to "member', and "N" indicates multiple records in "patient" can be linked to a record in the "member". Conclusion In this paper, we have outlined methodological issues involved in the construction of a household GIS database using satellite-based technology in a situation where the GPS was not appropriate. To our knowledge, this approach has never been reported, but may offer greater value in constructing household GIS databases compared to that based on GPS. Our household GIS offers size and orientation of individual houses in dense urban environment. Such database can be instrumental in health and disease studies because they facilitate the integration of socioecological and environmental factors that may influence health. Future health studies may benefit by using satellite-based technology to construct household GIS databases. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC521499.xml |
546402 | The Australian Health Care Agreements 2003–2008 | The Australian Health Care Agreements for the five years 1 July 2003 to 30 June 2008 were signed in August 2003 after vituperative debate and intransigence from the Commonwealth that vitiated the negotiation process. The new Agreements, which were not as generous as the Agreements they replaced, increase accountability on the States, requiring States to match increases in Commonwealth funding, and de-emphasise the prospects for further reform in Commonwealth-State relations during the course of the Agreements. This paper describes the new Australian Health Care Agreements and the process which led to them. | Introduction The most significant Australian health policy event for 2003 was the signing of the five-year Australian Health Care Agreements. The Agreements were preceded by an ultimatum to the States and Territories from the Commonwealth indicating that there would be no changes from the offer on the table. This led to bitter political recriminations, but the Agreements were eventually signed. In fact important preparations for Agreement renewal occurred in April 2002 with the Commonwealth and State Ministers, in a display of remarkable amity and accord, endorsing a new approach to the Agreements that: • Commonwealth/State relations in the health arena should focus on the provision of optimal care and health outcomes, regardless of jurisdictional boundaries. • It is in the best interests of all Australians for the Commonwealth, Stats and Territories to work co-operatively to improve the health and wellbeing of the community and the way in which health services are provided; • The 2003–08 Agreements should contain the principles, objectives and proposed health outcomes designed to achieve those objections. The Ministers also agreed to establish nine reference groups to address key issues in health reform which would feed into the Agreement "negotiation" process [ 1 ]. The reference groups addressed interaction between hospital funding and private health insurance; improving rural health; interface between aged and acute care; continuum between preventative, primary, chronic and acute models of care; improving indigenous health; improving mental health; information technology, research and e-health; quality and safety; and collaboration on workforce, training and education. The reference groups were co-chaired by senior Government officials and non-Government clinical experts involving participants from the bureaucracies, people who work in health agencies, and consumers. The reference groups created great expectations amongst the neophyte health policy contributors who believed the rhetoric of the Commonwealth about being prepared to consider wide ranging changes to the health sector. Seasoned commentators also called for reform [ 2 - 4 ]. Although extensive reports were produced by the Reference Groups and delivered to the Health Ministers the reports had no discernible impact on the 2003–2008 Agreements [ 5 ]. On 23 April 2003 the Commonwealth produced a non-negotiable offer with severe penalty clauses if States refused to sign by the Commonwealth's arbitrary deadline of 31 August 2003. An Australian Health Reform Alliance was formed to put pressure on the Commonwealth to respond to the reference group reports and to attempt to ensure that the 2003–2008 Agreements did not waste yet another opportunity to improve the efficiency, equity and quality of the health system. The Alliance's National Health Summit, which met at Old Parliament House, presented its final communiqué to non-Government politicians following a march up the hill to New Parliament House [ 6 ]. The Commonwealth deadlines remained and there was no change to the Agreement content. The Commonwealth's confrontationalist stance effectively destroyed relationships between the Minister for Health and Ageing, Senator Kay Patterson, and her State colleagues, and she was replaced as Health Minister by Tony Abbott MP in the Ministerial reshuffle of October 2003. The content of the Agreements There have been five Health Care Agreements since Medicare was introduced in 1984. The emphasis, orientation and priorities of these Agreements have changed over time (see Table 1 ). Table 1 Key elements of Commonwealth-state hospital funding agreements Agreement Political Objective Key Principles 1984–88 : Labor (Medicare Compensation Agreement) Introducing Medicare Compensation for cost increases and revenue losses 1988–93 : Labor (Medicare Agreement) Consolidating Medicare Growth and reform of public provision Incentives for system reform Penalties for lower public:private bed day shares and excess private medical service use 1993–98 : Labor (Medicare Agreement) Entrenching Medicare Expansion of public provision Reward for relatively higher levels of public provision and for increasing public provision relative to other states Post 1996, accountability for negotiated outcomes 1998–2003 : Coalition (Australian Health Care Agreement) Continuing with Medicare Increased Commonwealth funding with increased accountability for states Increased accountability on states for activity level changes Increased clarity of Commonwealth responsibility if health insurance levels change 2003–08 : Coalition (Australian Health Care Agreement) Continuing with Medicare Slowed Commonwealth funding growth Increased accountability for states Improved reporting, including of state spending Requirement on states at least to match Commonwealth funding increases Source: [11] The most significant elements of the 2003–08 Agreements are: • a base grant which is increased for weighted population increases, a further 1.7 per cent increase for utilisation drift, and indexation for wage movements • a withheld amount of 4 per cent of the grant paid on compliance with reporting schedules and funding growth matching requirements • a capital funding scheme to facilitate improvements in services involved in the transition from hospital to home ('Pathways Home Program') • funding for palliative care, mental health, and safety and quality initiatives. The most contentious difference between the 1998–2003 and 2003–2008 Agreements related to the indexation provisions (see Table 2 for significant areas of difference between the two Agreements). Table 2 Comparison of provisions of 1989–2003 and 2003–08 Australian Health Care Agreements Agreement Provision 1998–1998 Agreement 2003–2008 Agreement Indexation 2.1% above weighted population growth applied to 83% of the grant 1.7% above weighted population growth applied to 71% of the grant State matching Nil State "commits to increase its own source funding for public hospital services such that the cumulative rate of growth will at least match the cumulative rate of growth of Commonwealth funding" (Clause 11) Scope and level of services (State) "continues to provide services to public patients at an indicative public patient weighted separation rate of XX" (Clause 22) "The range of services available to public patients should be no less than was available on 1 July 1998" (Clause 7(a)) Reform The Commonwealth and Victoria recognise the need for service delivery reform and ongoing exploration of additional initiatives under a measure and share model. Victoria will work with the Commonwealth in evaluating the outcomes from the Co-ordinated Care Trials to provide information to guide future directions for the reform of health service delivery. The Commonwealth and Victoria will consider proposals which move funding for specific services between Commonwealth and State funded programs on the basis that each proposal meets the following criteria: • the proposal must be consistent with accepted evidence based best practice care models; • there should be a sound basis for believing that the reform will lead to improved patient outcomes and/or more cost effective care; • the impact of the proposal should be measurable in terms of change in services delivered and costs to the health system as a whole and to each party to this Agreement; • if the proposal is expected to lead to net savings, these should be shared equitably between the Commonwealth and Victoria; • the proposal should have potential to be replicated, be on a scale such that extension can be realistically tested and be evaluated in terms of such extension; and • the proposal must preserve eligible persons' current access to Medicare Benefits Schedule services or their equivalent. Reform proposals may result in the cashing out of State funded programs and/or Commonwealth funded programs, including the Medicare Benefits Schedule and Pharmaceutical Benefits Scheme. Victoria and the Commonwealth are committed to working with other States to progress the reform agenda agreed by Commonwealth and State Ministers for Health on 27 September 2002. The Commonwealth considers that for its part, such reform can taken place within existing funding parameters. In line with clause 18, the specific areas of national co-operation to deliver reform include: (a) improving the interface between hospitals and primary and aged care services; (b) achieving continuity between primary, community, acute, sub-acute, transition and aged care, whilst promoting consumer choice and improved responsiveness. Initial priorities for a stronger continuum of care approach will be cancer care and mental health services; and (c) exploring setting up a single national system for pharmaceuticals across all settings. Each of the predecessor Agreements provided indexation formulae to account for growth and ageing of the population. The 1998–2003 Agreements also recognised that there was further "utilisation drift", that is increases in utilisation were occurring in the hospital sector over and above that which can be explained by population growth and ageing. This utilisation drift was in part the result of new technologies that allowed for treatments for conditions for which there was previously no hospital treatment. Utilisation also increased because of shifts in treatment from general practitioners' rooms and other ambulatory settings to same day hospital admission. The 1998–2003 Agreements provided an escalation factor of 2.1% per annum over and above the growth caused by the increase in the rate of population for key elements of the grant. The 2003–08 Agreements reduced the utilisation drift factor to 1.7% and narrowed the applicable components of the grant, saving the Commonwealth Government about $1 billion from that provided for in the Forward Estimates. This reduction in growth provision was vociferously opposed by States and also by clinicians who were experiencing significant financial pressures on hospitals as a result of State Government funding constraints. The 2003–08 Agreements also addressed an ongoing concern of Commonwealth Governments (both Labor and Coalition): its perception that when the Commonwealth increased expenditure on hospital services, this often had no discernible impact on hospitals as State Governments withdrew funding concomitantly. As Deeble points out, the reality is more complex, but the evidence is that an increased Commonwealth share is associated with growth in spending [ 7 ]. The new Agreements provided that the States were required to increase their funding of hospitals at the same rate as the Commonwealth increases, otherwise the increases available to the State would not be paid. These stronger reporting frameworks built on the trend from the previous agreements and responded to a critical Auditor-General's report that concluded that the Commonwealth did not have all the performance information required to administer the Commonwealth funding allocated under the agreements [ 8 ]. The 'negotiation' processes Why were the processes so acrimonious and what shaped the Agreement outcome? To some extent the shape of the 2003–08 Agreement negotiations was inevitable. The political context, where all state and territory governments were of the opposite political colour from the Commonwealth, meant that harmonious negotiations were probably never seriously in contemplation. Commonwealth governments of both persuasions have tightened up accountability on states with successive Agreements and so tighter control was also probably inevitable. Two important political choices did exacerbate the tensions and inflamed the processes. First, the 2003–08 framework was more parsimonious than predecessor Agreements. As mentioned above, this represented a saving to the Commonwealth of about $1 billion on the Forward Estimates. A contemporary political issue was the decline in bulk billing. The Commonwealth's response to this involved an injection of around $1 billion. The link between the two policy debates within the Health portfolio is clear. Cabinet probably judged the political costs of finding a $1 billion saving from the states as low, as state premiers complaining about Commonwealth cuts and meanness is a regular part of the political landscape. Further, States would probably have criticised the Commonwealth position regardless of the offer made. The second choice that shaped the process was the Commonwealth's intransigence after the drafts were released. The Commonwealth's position here may have been based on a recognition that, eventually, all the states would have to sign the Agreements as they were politically committed to Medicare and free access at point of admission to hospitals, and that the states could not afford to suffer the cash flow consequences announced by the Commonwealth if the Agreements weren't signed by their deadline. The Prime Minister probably took a strong hand in this decision and left no room for his Health Minister to manoeuvre. The Minister's failure to attend meetings exacerbated an already difficult situation. A positive of the process was the extensive involvement of practitioners in the lead-up to the draft Agreements through the Reference Groups. Commonwealth-state negotiations had hitherto been an arcane process involving bureaucratic insiders. This widening of participation was welcomed by those involved and has set a precedent for future negotiations. Prospects for Reform The 2003–08 Agreements commit the Commonwealth and states to work towards reform in a number of areas including the interface between hospitals, primary care, and aged care; continuity of care particularly in cancer care and mental health services; and continued work on pharmaceuticals reform. A subtle shift from the predecessor Agreement model is the more sceptical and parsimonious approach to the potential for health care reform. Despite the aspirations implicit in establishing the nine reference groups, the language of the 2003–08 Agreements reflects a much more hard-nosed approach to reform with a strong emphasis on efficiency. This approach is most clearly articulated in Clause 18: "The Commonwealth considers that for its part such reforms can take place within existing funding parameters". Although predecessor Agreements also made provision for reform to Commonwealth/State relations, the progress in designing and implementing reform has not lived up to expectations. The most important shift that occurred during the course of the 1998–2003 Agreements was the rationalisation of hospital provision of outpatient pharmacy services, a long overdue response to a significant frictional issue in Commonwealth/States relations [ 9 , 10 ]. It is unclear whether the dynamic, facilitatory aspects of the 2003–08 Agreements will lead to any reform, especially given the acrimonious exchanges prior to signature of the Agreements. However it is important to note that, with a Federal election due at the end of 2004, there is a possibility that a Labor Government will be administering the remainder of the 2003–08 Australian Health Care Agreement. A new Government may be more committed to reforming and strengthening Medicare. However, despite the fact that a Labor would then hold political office throughout Australia, at all levels, this would not necessarily presage a more laissez faire attitude by a federal government to its state politically-allied counterparts. A Commonwealth Labor government would be just as keen as its Liberal predecessor to ensure that states are held accountable for maintaining spending and access. Competing interests The author(s) declare that they have no competing interests. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC546402.xml |
552311 | Resorbable screws versus pins for optimal transplant fixation (SPOT) in anterior cruciate ligament replacement with autologous hamstring grafts: rationale and design of a randomized, controlled, patient and investigator blinded trial [ISRCTN17384369] | Background Ruptures of the anterior cruciate ligament (ACL) are common injuries to the knee joint. Arthroscopic ACL replacement by autologous tendon grafts has established itself as a standard of care. Data from both experimental and observational studies suggest that surgical reconstruction does not fully restore knee stability. Persisting anterior laxity may lead to recurrent episodes of giving-way and cartilage damage. This might at least in part depend on the method of graft fixation in the bony tunnels. Whereas resorbable screws are easy to handle, pins may better preserve graft tension. The objective of this study is to determine whether pinning of ACL grafts reduces residual anterior laxity six months after surgery as compared to screw fixation. Design/ Methods SPOT is a randomised, controlled, patient and investigator blinded trial conducted at a single academic institution. Eligible patients are scheduled to arthroscopic ACL repair with triple-stranded hamstring grafts, conducted by a single, experienced surgeon. Intraoperatively, subjects willing to engage in this study will be randomised to transplant tethering with either resorbable screws or resorbable pins. No other changes apply to locally established treatment protocols. Patients and clinical investigators will remain blinded to the assigned fixation method until the six-month follow-up examination. The primary outcome is the side-to-side (repaired to healthy knee) difference in anterior translation as measured by the KT-1000 arthrometer at a defined load (89 N) six months after surgery. A sample size of 54 patients will yield a power of 80% to detect a difference of 1.0 mm ± standard deviation 1.2 mm at a two-sided alpha of 5% with a t-test for independent samples. Secondary outcomes (generic and disease-specific measures of quality of life, magnetic resonance imaging morphology of transplants and devices) will be handled in an exploratory fashion. Conclusion SPOT aims at showing a reduction in anterior knee laxity after fixing ACL grafts by pins compared to screws. | Background Anterior cruciate ligament (ACL) rupture belongs to the most common musculoskeletal injuries in the western world. In the United States, 100,000 new cases occur each year, with 10% of all injuries leading to occupational disability [ 1 ]. In Germany, the prevalence of torn ACL among subjects between 20 and 35 years averages 0.4%. In the general population, the yearly incidence of ACL rupture reaches 32/100,000, but peaks to 70/100,000 in athletes [ 2 ]. An ACL deficient knee is at risk of developing secondary damage to the cartilage and is liable to undergo progressive intra-articular worsening. Roughly half of all acute ACL disruptions attend meniscus damage. Arthroscopic reconstruction surgery by autologous grafting emerged as the therapy of choice. Predictions call for 175,000 ACL replacements performed yearly in the United States. In Germany, around 50,000 patients undergo ACL repair each year. Selecting the ideal graft remains an issue of debate. Randomized controlled trials suggest a lower degree of persistent laxity with bone-patellar-tendon-bone (BPTB) comparing with two-, three- or four-bundle hamstring (that is, semitendinosus and gracilis tendon) transplants (HT). However, the biomechanical advantage does not frame higher patient satisfaction, or differences in scoring after long-term follow-up [ 3 ]. In contrast, harvesting BPTB grafts often produces notable donor site morbidity, and refractory kneeling pain [ 4 , 5 ]. Available data from randomized, quasi-randomized and uncontrolled trials signal a weighted mean difference of 2.28 mm (95% confidence interval [CI] 1.83 – 2.73 mm) in anterior laxity between the injured and healthy knee with HT reconstruction (see Figure 1 ) [ 3 , 6 - 16 ]. Figure 1 Persisting instability following ACL repair with HT autografts (KT-1000 measurements). Individual study results were weighted by their inverse variance to derive a common point estimate with 95% confidence interval (diamond). Many features contribute to an unsatisfactory or failed ACL replacement, for example, imprecise tunnel positioning, the presence of degenerative changes, or the onset of arthrofibrosis. The choice of tibial graft fixation affects later stability. The intact ACL has a tensile strength around 2200 N. To avoid loosening, the graft must be fixed under firm traction (around 40 N), with the knee in a smoothly flexed position. A common way to anchor the tibial end of the graft is by titanium or biodegradable interference screws, for example, the BioCryl ® (DePuy Mitek) screw that contains both resorbable poly-L-lactid and osteoconductive tricalcium phosphate (see Figure 2 ). Figure 2 Appearance of the BioCryl ® screws (left, courtesy of DePuy Mitek), and their positioning (right). However, in a recent biomechanical study, extracortical fixation devices like the EndoButton ® (Smith & Nephew) or RigidFix ® (DePuy Mitek) provided better strength than did the interference screws [ 17 ]. The possible advantage of RigidFix ® over other tethering methods is a splicing of strands, tightening the contact between the tendon surface and the bony tunnel over the entire graft circumference (see Figure 3 ). Figure 3 Left: Positioning of RigidFix ® pins. Right: splicing of graft bundles leading to close adherence to the surrounding bone. Methods/ Design Objectives The present study aims at comparing later laxity in subjects undergoing arthroscopic anterior cruciate ligament replacement with either RigidFix ® pinning or BioCryl ® screwing of HT grafts. Both implants are CE approved, and were introduced to ACL-repair in Germany in 2002. We have secondary objectives in imagining graft incorporation by MRI-scanning, functional results, residual pain, resumption of occupational and leisure activity, and quality of life by generic and disease-specific questionnaires. Primary endpoint We pose the primary hypothesis that the RigidFix ® system preserves graft tension gained during surgery, and leads to lower KT-1000 arthrometer side-to-side differences than the BioCryl ® screw after six months of follow-up. Specifically, we will test the hypothesis that RigidFix ® decreases the average difference gained with interference screws by 1.0 ± standard deviation 1.2 mm. The investigators consider this difference clinically sound, important, and measurable by KT-1000 arthrometer testing. Twenty-four subjects a treatment arm will allow for detecting this difference with an 80% chance at a two-sided alpha-level of 5%. Assuming a drop-out rate of 10%, 54 patients will be enrolled in this study. Secondary endpoints As secondary endpoints, we consider functional outcomes by means of the Lysholm scale, the Tegner score, and the International Knee Documentation Committee evaluation form (IKDC) in its German translation, 2000 revision [ 18 ]. Besides disease-specific items, this questionnaire also contains the Short-Form 36 (SF-36) generic health assessment tool. The noted instruments have proven reliability, validity, and responsiveness for use in clinical research. Confirmatory testing will apply for the primary endpoint only. All secondary endpoints will be addressed in an exploratory fashion. Design SPOT is a patient and investigator blinded, randomised controlled trial conducted at a single academic institution. Randomisation is carried out in the operating theatre shortly before transplant fixation, with random codes drawn from sealed envelopes. We use block-randomisation with five subjects a block following a computer-generated random list [ 19 ]. Inclusion criteria Men and women (providing that they are not pregnant) being at least 18 years old are recruited to this trial. Subjects may engage in this study if they - faced a first one-sided total or subtotal rupture of the anterior cruciate ligament , proven either by arthroscopy or MRI-scanning - had met an acute knee distorsion event likely to have caused the index injury at least six weeks before scheduled repair - have been physically examined in the ambulatory of the study hospital before assigning an admission date, and were screened and considered suitable to enter this trial by one of the investigators Also, patients must be able to give voluntary written informed consent, and to comply with the post operative follow up regime Exclusion criteria We exclude patients - with related lower limb fractures - with active infection affecting the limb subject to needed treatment - who have previously took part in this investigation or who are taking part in another clinical investigation - with contraindications for MRI-scanning (that is, large indwelling orthopaedic implants made of metals others than titanium, or pacemakers) Ethical considerations This protocol and all accompanying documents were approved by the local Institutional Review Board (IRB). According to IRB recommendations and requirements, information leaflets explicitly note that "a benefit from participation in this trial cannot be guaranteed." We will stress the principle of randomisation as "treatment assignment by chance, without the possibility of the investigator, other health care professionals involved in this study, or the patient influencing the choice of treatment." We also tell patients that, as long as they keep agreement in participation, they will not know their assigned treatment until the six-month follow-up visit. We will notify the IRB of any significant changes to the protocol. Also, we will notify the IRB within ten working days of the discovery of any severe adverse events which occur during this investigation. Confidentiality of subject data will always be maintained. Subject anonymity will be guaranteed and all documentation about a subject (including radiographs) will be kept in secure location. This investigation strictly adheres to the relevant articles of the Declaration of Helsinki as adopted by the 18 th World Medical Assembly in 1964 and its later revisions, as well as to principles of GCP, developed within the Expert Working Group (Efficacy) of the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). Surgery and rehabilitation All devices and instrumentations used in this clinical investigation bear the CE mark. They belong to the regular implants used for ACL repair at the study hospital since 2002. Except different graft anchoring, similar treatments apply to patients in both study groups. All participants undergo internationally accepted surgically procedures by a single surgeon (D.C.) with extensive experience in ACL repair using both the BioCryl ® screw and RigidFix ® cross pins. Also, postoperative care and rehabilitation programs do not differ from those employed outside a clinical trial. All repairs are carried out under general anaesthesia, with the patient in a supine position. Perioperative antibiotic prophylaxis comprises 2 g of cefotiam. Patients receive 40 mg of enoxaparin daily for prophylaxis of thromboembolic events until full weight bearing. The knee joint is accessed through two to three standard portals. Meniscal injuries are addressed with partial resection or repair. Hamstring tendons are harvested via a small incision over the insertion of the pes anserinus at the anterior medial tibia by a closed tendon stripper, and prepared as triple-stranded grafts. Tibial and femoral tunnels are drilled to approximate graft thickness (usually 8 to 9 mm) with the use of a guiding wire. Grafts are fixed with the knee in 30° flexion to achieve firm tension. Postoperatively, the knee is stabilized for three days by a zero-degree splint. Afterwards, flexion is limited to 90° by a Secutec ® orthosis for six weeks. Patients are allowed partial weight bearing with walking crutches. Subjects are prescribed intense physical therapy for motion exercise, and to strengthen thigh muscles. Normally, full range of motion and weight bearing is achieved until week 12 after surgery. Patients in the experimental group have their grafts secured by tibial and femoral RigidFix ® pinning. Patient in the control group receive tibial and femoral BioCryl ® screws. Baseline assessment Each subject considered eligible for entry into this investigation has the following information and procedures recorded at the pre-investigational examination: - Demographic details including date of birth and gender - Medical history, coexisting diseases, and accompanying medication - Physical examination, including circumferential measurement of both legs at defined landmarks, Lachman and pivot shift tests, KT-1000 arthrometer objectifying of instability, one-legged hop test, Lysholm, Tegner, and IKDC scores, knee and kneeling pain measured by visual analogue scales - Radiographic examination, including a conventional roentgenogram of the injured knee in anteroposterior and lateral projection, and a preoperative MRI scan according to local standards Intraoperative assessments During surgery, we record procedure details in the electronic CRF. We assess the duration of surgery (from cut to skin closure), and operating theatre time (from induction of anaesthesia to arrival at the recovery room). A clinical knee examination is performed and documented with the subjects under general anaesthesia and relaxation. Arthroscopy findings (accompanying injuries to or degenerative changes of the cruciate or collateral ligaments, menisci, or cartilage) are recorded by video and/ or hard copy images. We document eventual blood loss, and any other adverse event occurring during surgery. The responsible surgeon judges the handling of implants and his overall satisfaction with the intraoperative result using five-point Likert scales. We detail any problems or complications on an Adverse Events form. Follow-up assessments Patients are appointed outpatient visits as part of the clinical investigation at 3, 6, and 12 months postoperatively. For study purposes, except quality of life measurements, patients do not undergo any diagnostic or other procedure not belonging to the common repertoire of assessments carried out after ACL repair. Specifically, we avoid invasive procedures, blood sampling, or imaging tests exposing subjects to radiation or contrast agents. The investigators consider the possible burden caused by extra clinical tests negligible. We assess the following items at the scheduled visits: Three months postoperatively - Physical examination, including circumferential measurement of both legs at defined landmarks, Lachman and pivot shift tests, KT-1000 arthrometer objectifying of instability, one-legged hop test, Lysholm, Tegner, and IKDC scores, knee and kneeling pain measured by visual analogue scales - Any complications or complaints raised by the patient, GP/ ambulatory surgeon, or Clinical Investigators - Resumption of work - MRI scan according to local standards Six months postoperatively - Physical examination, including circumferential measurement of both legs at defined landmarks, Lachman and pivot shift tests, KT-1000 arthrometer objectifying of instability, one-legged hop test, Lysholm, Tegner, and IKDC scores, knee and kneeling pain measured by visual analogue scales - Any complications or complaints raised by the patient, GP/ ambulatory surgeon, or Clinical Investigators - Resumption of work and leisure activities - MRI scan according to local standards Patients and Investigators responsible for follow-up examinations may learn about the assigned treatment after completing the six-month CRF. Twelve months postoperatively - Physical examination, including circumferential measurement of both legs at defined landmarks, Lachman and pivot shift tests, KT-1000 arthrometer objectifying of instability, one-legged hop test, Lysholm, Tegner, and IKDC scores, knee and kneeling pain measured by visual analogue scales - Any complications or complaints raised by the patient, GP/ ambulatory surgeon, or Clinical Investigators - Resumption of work and leisure activities - MRI scan according to local standards MRI studies Radiological evaluation comprises - Tunnel widening - Impingement - Transplant sufficiency, morphology of transitional areas and tendon-bone-interfaces - Degree of degradation of screws and pins - Degree of inflammation (synovitis) and effusion - Presence or progression of arthrofibrosis - Changes in cartilage and meniscal morphology Physical examination Before physical examination, both knees are prepared by applying opaque dressings to hide scars and to blind the examining doctors for the managed side. Examination is performed independently by two of three board-certified surgeons (D.S., K.B., V.T.) who had not operated on any of the patients in the study. The responsible surgeon (D.C.) conducts a third clinical examination after completion of the case report forms. We document his examination findings separately and consider them as the diagnostic reference standard. We assess both interobserver agreement by kappa statistics and the accuracy of measurements taken by independent observers comparing with those of the responsible surgeon. KT-1000 arthrometer testing Objective translation measurements comprise a defined load (89 N). We measure the anterior translation of the injured and healthy side (in mm), as well as the difference between both knees. Safety assessment and reporting of adverse event We define an adverse event as 'any undesirable clinical instance in a subject whether it is considered treatment related or not'. In addition, an adverse device effect, undesirable side effect, is defined as 'a device related adverse event'. A record of all adverse events, including details of the nature, onset, duration, severity, relationship to the device, relationship to the operative procedure and outcome, will be made on the relevant section of the subject's CRF. The subject will be questioned about any adverse event at each later follow-up assessment visit. An adverse event or an adverse device effect may be mild, moderate or severe and are usually unexpected. A severe adverse event or adverse device effect is defined as any experience that - is fatal or life-threatening - is permanently disabling - needs or prolongs in-patient hospitalization because of a potential disability, danger to life or forces an intervention All severe adverse events or adverse effects which occur during this investigation must be and will be reported immediately by telephone or facsimile to Bundesinstitut für Arzneimittel und Medizinprodukte, Kurt-Georg-Kiesinger-Allee 3, 53175 Bonn, Phone: ++49 30 228 207 30, Fax: ++49 30 228 207 5207 Data management For data collection, we set up a Microsoft XP Professional Access ® Database, run on a mobile computer separately from the hospital documentation and the intranet. Data collection and storage comply with orders fixed by the data safety board of the Unfallkrankenhaus Berlin, and follow German laws for data safeguard and protection ( Bundesgesetz über den Schutz personenbezogener Daten [Datenschutzgesetz 2000 – DSG 2000], 17. August 1999, BGBl. I Nr. 1999/165). We ensure data storage for five years. For study documentation, we assign patients an identification number. Electronic sources do not contain names or addresses of participants. Linking lists are stored in a study folder with copies of adverse events forms. Since this study runs at a single centre, we do not appoint an external monitor for data handling and management. We regularly (at least twice a month) check datasets for consistency, completeness and plausibility. Statistical analysis We will conduct all analyses following the intention to treat principle (that is, patients will be evaluated as randomized). We will express measurements as means, medians or proportions with their proper distribution indices (that is, standard deviations, ranges, and interquartile ranges). In case of skewed distributions, normalizing will be achieved by logarithmic transformation, where necessary. As pointed out earlier, we will address only the primary hypothesis in a confirmatory fashion, whereas all other results will be evaluated in a plain exploratory intent. We will employ the student's t-test for independent samples to test for the difference in anterior laxity between both fixation methods at six months of follow-up. For secondary endpoint analysis, we will calculate cross tables, 95% confidence intervals for normally or binomially distributed data, and differences in means, proportions, and ratios. In case of obvious benefits or harms with either device in a certain subgroup of patients, we will use stratified analyses (for example, according to Mantel-Haenszel). Where statistically and/ or clinically sound, we will consider linear and logistic regression analyses, or more sophisticated regression models for correlated data (for example, generalized estimating equations). We will, however, respect the small sample size of this study, and limit statistical analyses to the necessary minimum. In case of missing data, we will use both a last observation carried forward approach, and imputation methods by regression or semi-Bayesian modelling. Separate analyses will be performed for raw and modelled data. Discussion After ACL repair, most patients rarely recognize slightly weakened anterior knee stability in everyday life. However, subjects with a high recreational and sporting activity and physically strenuous professions often suffer from recurrent events of giving-way, especially on hastened movements. This poses a high risk for secondary knee injury. Of note, muscular training cannot compensate for residual laxity, outbalancing the anticipated benefit from surgical repair. Thus, attempts to optimize the surgical technique may be valuable. Currently, surgeons performing ACL reconstruction use screws, pins, buttons, and cramps for graft fixation because of individual preference, or institutional orders. The latter are chiefly driven by cost considerations. For example, the purchase price of a RigidFix ® tray is ten times higher than that of BioCryl ® screws. Obviously, the more expensive implant must prove a distinct clinical advantage over the common one to justify its further use. Unfortunately, there is lack of robust evidence on the effectiveness of either fixation method beyond laboratory and animal experiments. Although conceptually impressive, there is no comparative study that proved a clinically measurable advantage of RigidFix ® over screws. The investigators consider the equipoise principle fulfilled, since it is unclear whether screw or pins lead to better long-term stability, or show any measurable differences at all. We hope that the results from this pragmatic study can clarify this issue. Competing interests MW, DS and AE have worked as independent scientific consultants for DePuy ® International, and received project-related funding that does not apply to this work . No support in any form was provided, or will be provided by third parties to set up the trial protocol, or to conduct this study. This study aims at investigating biomechanical principles, not certain implants. The members of the SPOT Group have no financial or non-financial competing interests in this work. Authors' contributions DS drafted the manuscript. KB and GM edited the manuscript. DC is in charge of all surgical procedures. DS is responsible for statistical analyses. All authors participated in the design of this study, and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC552311.xml |
544573 | The Angiotensin Converting Enzyme Insertion/Deletion polymorphism is not associated with an increased risk of death or bronchopulmonary dysplasia in ventilated very low birth weight infants | Background The ACE gene contains a polymorphism consisting of either the presence (insertion, I) or absence (deletion, D) of a 287 bp alu repeat in intron 16. The D allele is associated with increased ACE activity in both tissue and plasma. The DD genotype is associated with risk of developing ARDS and mortality. The frequency of the D allele is higher in patients with pulmonary fibrosis, sarcoidosis and berylliosis. The role of this polymorphism has not been studied in the development of BPD in the premature newborn. Methods ACE I/D genotype was determined in 245 (194 African-American, 47 Caucasian and 4 Hispanic) mechanically ventilated infants weighing less than 1250 grams at birth and compared to outcome (death and/or development of BPD). Results The incidence of the D allele in the study population was 0.58. Eighty-eight (35.9%) infants were homozygous DD, 107 (43.7%) were heterozygous ID and 50 (20.4%) were homozygous II. There were no significant differences between genotype groups with respect to ethnic origin, birth weight, gestation, or gender. There was no effect of the ACE I/D polymorphism on mortality or development of BPD (O 2 on 28 days or 36 weeks PCA). Secondary outcomes (intraventricular hemorrhage and periventricular leukomalacia) similarly were not influenced by the ACE ID polymorphism. Conclusions The ACE I/D polymorphism does not significantly influence the development of BPD in ventilated infants less than 1250 grams. | Background Prematurely born infants who require mechanical ventilation (MV) frequently develop chronic lung disease known as bronchopulmonary dysplasia (BPD). Oxidant injury, mechanical disruption of the lung, inflammation and subsequent failure of lung development are considered the major mechanisms in the pathogenesis of BPD. The development of BPD is characterized by an initial acute inflammatory component followed by variable degrees of lung fibrosis and failure of alveolar septation, both of which ultimately impair the development of the immature lung [ 1 - 5 ]. The inflammatory component, which consists interstitial and alveolar edema, hyaline membrane formation, epithelial cell necrosis and influx of activated neutrophils, is similar (albeit not identical) to that seen in other forms of acute lung injury (ALl). There is increasing evidence to support the role for the activation of the renin-angiotensin system (RAS) system during ALI. In ARDS there is an increase in bronchoalveolar lavage and a concomitant decrease in circulating angiotensin converting enzyme (ACE) activity[ 6 , 7 ]. This increase in local ACE activity may influence the course of acute lung injury by its effects on vascular permeability, epithelial cell survival and fibroblast activity [ 8 - 12 ]. Angiotensin-II (AT-II) concentrations are increased in patients with ARDS, consistent with activation of the RAS with ALI[ 13 ]. Inhibition of AT-II with type 1 angiotensin receptor antagonists delayed the onset of ARDS and inhibited neutrophil influx in to the lung in experimental models[ 14 ]. The role of RAS activation in the lungs of premature infants with respiratory distress syndrome and its evolution into BPD has not been determined. There is significant variation in circulating ACE activity among individuals, which may be genetically determined. The human ACE gene is located on chromosome 17 (17q23) and contains a polymorphism consisting of either the presence (insertion, I) or absence (deletion, D) of a 287 bp alu repeat in intron 16[ 15 ]. The deletion is associated with increased ACE activity in both tissue and in the circulation and accounts for about 47% of the intra-individual variation in plasma ACE activity in Caucasians[ 15 , 16 ]. A role for genetic variation in ACE activity in both acute and chronic lung disease has recently been suggested [ 17 - 21 ]. Higher intrinsic ACE activity (DD genotype) is associated with an increased risk of developing ARDS and other lung diseases [ 17 - 21 ]. The potential role of this polymorphism to influence risk of developing chronic lung disease has not been studied in the premature newborn. Because of the relationship between inflammation and activation of the RAS and the association with lung fibrosis and adverse outcome in ARDS, the ACE l/D polymorphism may modify risk for the development of chronic lung disease or death in mechanically ventilated VLBW infants. The purpose of this study was to determine if there is a relationship between the ACE I/D polymorphism and respiratory outcomes of death or the development of BPD, in mechanically ventilated very low birth weight (VLBW) infants. Methods Genomic DNA used for this case controlled study was extracted from archival tracheal aspirate (TA) pellets (223 patients) or blood (22 patients) collected prospectively as part of an ongoing study of genetic factors in the development of complications of prematurity. The TAs that were used as a source of genomic DNA were originally collected as part of long term longitudinal studies examining cytokine concentrations and the development of CLD[ 22 , 23 ]. Infants were included in this study if they fulfilled the following inclusion criteria: birth weight less than 1250 grams, mechanical ventilation (MV) during the first week of life, complete clinical data on pulmonary outcome and a genomic DNA sample that could be used for genotyping. Infants were excluded if complete data on pulmonary outcome was not available or suitable DNA was not available. The study was approved by the Institutional Review Board for Human Research at Louisiana State University Health Sciences Center in Shreveport. Cultures for genital mycoplasmas were performed on samples collected in the first few days of life. Clinical and outcome data were abstracted from the clinical record and included information on respiratory outcome, survival and development of complications of prematurity. Laboratory methods DNA isolation Isolation of total DNA from blood or TA pellets was performed using the QIAmp DNA Mini kits™ (Qiagen Incorporated, Chatsworth, CA). Briefly, TA pellets were suspended in 200 μl of sterile phosphate buffered saline by vigorous vortexing, then digested in proteinase K and applied to silica gel spin columns. Columns were washed in the manufacturer's supplied buffers and the total DNA was eluted in 200 μl elution buffer. Blood (200 μl) was extracted similarly to the TA pellets. ACE I/D genotyping ACE I/D polymorphism was performed by microplate PCR method as described previously[ 24 ]. (Primers used in the assay: 5'-CTG GAG ACC ACT CCC ATC CTT TCT-3' AND 5'-GAT GTG GCC ATC ACA TTC GTC AGA T-3'. In 10 microliter PCR volume, the following components were added with the final concentration of MgCl 2 1.5 mM, KCl 50 mM, 5%DMSO, Triton X-100 0.1%, 200 micromolar each of dNTPs, 10 pmol each primer, 1–2 U of Taq polymerase, and 1 microliter of DNA. DNA was denatured at 95°C for 5–10 min, followed by 30 cycles of denaturation at 94° for 60 sec, annealing at 67 for 60 sec, and extension at 72° for 2 minutes, with a final extension at 72° for 7 min. PCR products were separated on 2% agarose gel containing 0.5-microgram/ml ethidium bromide. After gel electrophoresis the bands were visualized by UV-transillumination. The PCR product is a 190 bp fragment in the absence of the insertion (D genotype) and a 490 bp fragment in the presence of the insertion (I genotype). Data analysis Several definitions of BPD have been used through the years. Initially BPD was defined as oxygen dependency at 28 days of age[ 25 , 26 ]. More recently, the use of oxygen dependency at 36 weeks postconceptional age (PCA) has been proposed as a more suitable definition of BPD.[ 27 ] Both definitions of BPD predict long term respiratory abnormalities. Data analysis consisted of comparing the frequencies of the ACE I/D genotypes in infants with and without the outcome of interest (supplemental oxygen administration at 28 days or 36 weeks PCA, death or BPD /death before 36 weeks PCA) by Chi Square. All statistical analysis was performed using the SPSS for Windows version 12.0 (SPSS Inc., Chicago, IL). The Student t-test was used to assess normally distributed variables. The Wilcox Rank Sum test was used for analysis of factors that were not normally distributed. A probability value of less than 0.05 was considered statistically significant. The data are presented as mean ± standard error of the mean (SEM). Results Two hundred and forty five (245) patients had complete clinical information on respiratory outcome and genomic DNA available for genotyping. Mean gestational age and birth weight of the study population was 26.4 ± 0.1 weeks and 869 ± 12 grams respectively. One hundred and ninety-four (79%) infants were African-American, 47 (19 %) were Caucasian and 4 (2%) were Hispanic. Male: female ratio was 147:98. All patients required MV at birth and 228 (93%) infants were treated with exogenous surfactant therapy (Survanta ® , Ross Products Division, Abbott Laboratories, Columbus, OH). Tracheal aspirate cultures obtained during the first few days of life grew Ureaplasma urealyticum (Uu) on at least one occasion from 74 of 217 infants cultured (34%) and Mycoplasma hominis (Mh) from 29 (13%). One hundred and fifty one (67%) infants were oxygen dependent at 28 days and 55 (25%) were oxygen dependent at 36 weeks PCA. There were 39 (16%) patients who died (from all causes) during their initial hospitalization (24 before 28 days of age and 15 after 28 days). All 243 infants were genotyped for the ACE I/D polymorphism. The frequency of the D allele in the study population was 0.58. The frequency of the D allele was similar between African-American (0.57) and Caucasian infants (0.59) (p = 0.664). Fifty (20.4%) infants were homozygous II, 107 (43.7 %) were heterozygous ID and 88 (35.9%) were homozygous DD. Baseline clinical characteristics of birth weight, gestational age, race, gender, TA isolation of Mh, and the need for surfactant replacement were not different between genotype groups (Table 1 ). Isolation of Ureaplasma urealyticum from the trachea was more frequent in Caucasian infants who had the ACE II genotype (II 55%, ID 9%, DD 22%; p = 0.046). Uu isolation frequencies were similar between genotype groups in African-American infants. Baseline clinical characteristics for Caucasian and African-American infants separately can be found in the online supplement. Table 1 Baseline Clinical Characteristics ACE Genotype II (n = 50) ID (n = 107) DD (n = 88) P value Birth Weight 874 ± 29 876 ± 18 858 ± 20 0.793 Gestation 26.4 ± 0.3 26.5 ± 0.2 26.2 ± 0.2 0.386 Race (Caucasian) 11 (22) 17(16) 19 (22) 0.664 Gender (Males) 28 (56) 59 (55) 59 (67) 0.166 Uu isolated from TA a 18/46 (39) 29/88 (33) 27/83 (33) 0.719 Mh isolated from TA a 4/46 (9) 14/88 (16) 11/83 (13) 0.507 Surfactant Replacement 48 (96) 99 (93) 81 (92) 0.651 Data are presented as Mean ± Standard error of Mean. Numbers in parenthesis represent percentages. a Not all infants had TA cultures performed for Uu and Mh Uu Ureaplasma urealyticum Mh Mycoplasma hominis TA Tracheal Aspirate Clinical characteristics of surviving infants who were or were not oxygen dependent at days are shown in Table 2 . Infants who were oxygen dependent at 28 days were less mature, of lower birthweight, more likely to have TA isolation of Uu or Mh, and more likely to have received surfactant replacement therapy than those weaned from oxygen by 28 days of age. Ethnic groups and gender were not different between outcome groups. Infants who died or who were oxygen dependent at 36 weeks PCA (BPD) were similarly less mature, of lower birthweight, and were more likely to have received surfactant replacement therapy than surviving infants without BPD (Table 3 ). Isolation of either Uu or Mh from the TA had no influence on this outcome. Table 2 Comparison of Infants Oxygen Dependent at 28 days No Oxygen at 28 days (n = 75) Oxygen at 28 days (n = 151) P value Gestation (weeks) 27.4 ± 0.1 26.0 ± 0.1 <0.001 Birth Weight (grams) 996 ± 18 826 ± 14 <0.001 Race (Caucasian) 20 (27) 26 (17) 0.103 Gender (Males) 39 (52) 95 (63) 0.102 Surfactant therapy 75 (87) 144 (95) 0.020 Ureaplasma isolated a 16/67 (24) 55/134 (41) 0.016 Mycoplasma isolated a 4/67 (6) 23/134 (17) 0.028 IVH b 8/74 (11) 55/151 (36) <0.001 IVH = Grade 3 b 4/74 (5) 35/151 (23) <0.001 PVL b 2/74 (3) 15/151 (10) 0.054 Data are presented as Mean ± Standard error of Mean. Numbers in parenthesis represent percentages. a Not all infants had TA cultures performed for Uu and Mh b Not all infants had cranial US evaluations Uu Ureaplasma urealyticum Mh Mycoplasma hominis IVH Intraventricular Hemorrhage PVL Periventricular leukomalacia TA Tracheal Aspirate Table 3 Clinical Characteristics of Infants who Died or Developed BPD Survival without BPD (n = 162) Death or BPD (n = 83) P value Gestation (weeks) 26.6 ± 0.1 25.8 ± 0.2 <0.001 Birth Weight (grams) 925 ± 14 760 ± 17 <0.001 Race (Caucasian) 33 (20) 14 (17) 0.737 Gender (Males) 96 (60) 50 (60) 0.926 Surfactant therapy 147 (91) 81 (98) 0.046 Ureaplasma isolated a 49/144 (34) 25/73 (34) 0.974 Mycoplasma isolated a 20/144 (14) 9/64 (12) 0.750 IVH b 31/161 (19) 41/80 (52) <0.001 IVH = Grade 3 b 15/161 (9) 31/80 (39) <0.001 PVL b 8/161 (5) 10/80 (13) 0.033 Data are presented as Mean ± Standard error of Mean. Numbers in parenthesis represent percentages. a Not all infants had TA cultures performed for Uu and Mh b Not all infants had cranial US evaluations Uu Ureaplasma urealyticum Mh Mycoplasma hominis IVH Intraventricular Hemorrhage PVL Periventricular leukomalacia TA Tracheal Aspirate Because the ACE I/D polymorphism may have different functional effects on plasma and tissue ACE activities in different ethnic groups, we analyzed the effects of ACE on the incidence of BPD separately for Caucasian and African-American infants. (The effects of the ACE I/D polymorphism on the incidence of BPD and other outcomes for the combined group can be found in the online data supplement). Table 4 shows the effects of ACE genotype on outcomes in ventilated Caucasian infants less than 1250 grams. There was no significant effect of ACE genotype on mortality, oxygen dependency at either 28 days or 36 weeks PCA or the combined outcome of death or BPD. The incidence of periventricular leukomalacia (PVL) was significantly higher in Caucasian infants with the II genotype. The incidence and severity of intraventricular hemorrhage (IVH) was not affected by ACE genotype. Table 4 Effect of ACE genotype on Outcomes in Caucasian Infants ACE Genotype II (n = 11) ID (n = 17) DD (n = 19) P value Oxygen at 28 days 6/11 (55) 10/15 (67) 9/18 (47) 0.620 Oxygen at 36 weeks PCA 3/10 (30) 4/15 (27) 3/18 (17) 0.674 Death <28 days 0 (0) 1 (16) 1 (5) 0.724 Death or Oxygen at 36 weeks 4 (36) 6 (35) 4 (21) 0.558 Death ≥ 28 days 3 (27) 1 (6) 0 (0) 0.039 IVH a 2/10 (20) 6/17 (35) 1/19 (5) 0.076 IVH ≥ Grade 3 a 2/10 (20) 3/17 (18) 1/19 (5) 0.415 PVL a 3/10 (30) 0/17 (0) 1/19 (5) 0.022 Numbers in parenthesis represent percentages. a Not all infants had cranial US evaluations PCA Postconceptional age IVH Intraventricular Hemorrhage PVL Periventricular leukomalacia The effects of ACE genotype on outcome in ventilated African-American infants less than 1250 grams are shown in Table 5 . There was no significant effect of ACE genotype on mortality, oxygen dependency at either 28 days or 36 weeks PCA or the combined outcome of death or BPD. The incidence of PVL was not affected by ACE genotype in African-American infants. Similar to that observed in Caucasian infants, there was no apparent effect of the ACE I/D polymorphism on the incidence and severity of IVH in African American infants. Table 5 Effect of ACE genotype on Outcomes in African-American Infants ACE Genotype II (n = 39) ID (n = 88) DD (n = 66) P value Oxygen at 28 days 24/36 (67) 57/82 (70) 43/59 (73) 0.805 Oxygen at 36 weeks PCA 7/34 (21) 22/79 (28) 16/58 (28) 0.698 Death <28 days 3 (8) 6 (7) 8 (12) 0.517 Death or Oxygen at 36 weeks 12 (31) 31 (35) 25 (37) 0.792 Death ≥ 28 days 3 (8) 7 (9) 4 (7) 0.925 IVH a 14/38 (37) 23/86 (27) 25/67 (37) 0.311 IVH ≥ Grade 3 a 7/38 (18) 15/86 (17) 17/67 (25) 0.438 PVL a 1/38 (3) 5/86 (6) 8 (12) 0.145 Numbers in parenthesis represent percentages. a Not all infants had cranial US evaluations PCA Postconceptional age IVH Intraventricular Hemorrhage PVL Periventricular leukomalacia Since birth weight and gestation are the primary determinants of adverse outcome, analysis was repeated in both infants greater than 750 grams and infants ≤ 750 grams. There was no significant effect of ACE genotype when this subgroup analysis was performed (data not shown). Discussion The RAS is activated during lung injury and plays a role in several pathological processes. In addition to the pulmonary endothelium, respiratory epithelium also possesses significant ACE activity. Fas-induced alveolar epithelial cell apoptosis is dependent on local AT-II production and interaction with it receptor.[ 10 , 11 ] Further, AT-II is mitogenic for lung fibroblasts and aberrant AT-II production has been linked with some forms of pulmonary fibrosis[ 12 , 18 , 28 - 30 ]. Inhibition of AT-II with type 1 angiotensin receptor antagonists delayed the onset of ARDS and inhibited neutrophil influx into the lung in experimental models[ 14 ]. In adults, there is an increase in bronchoalveolar lavage ACE activity and AT-II during ALI, however the contribution of activation of the RAS to neonatal lung injury has received little study[ 6 , 13 ]. The frequency of the D allele in our study population was not different than reported in our local population or for other groups[ 16 , 24 , 31 ]. The ACE D allele is common with a frequency approximately 50–60% in Caucasians and 60–65% in African-Americans. This suggests that the D allele is neither a risk factor nor a protective factor for either premature birth or the need for mechanical ventilation. We did not study early cardiovascular adaptation or record measures of early illness severity, as did Harding et al[ 32 ] In that study the DD genotype was associated with a worse cardiovascular adaptation. In our study, the infants are much more immature, of lower gestational age and almost universally had hyaline membrane disease. In our study population, the ACE ID polymorphism was not associated with an altered risk for the development of BPD, oxygen dependency at 28 days, death (early, late or total mortality) or composite outcomes (BPD or death) suggesting that factors other than genetic variation at this locus contribute to these outcomes. This is in contrast to that seen in adults with ARDS or other lung disorders such as pulmonary fibrosis, sarcoidosis and berylliosis where the D allele is associated with disease [ 18 - 21 ]. Indeed genetic factors may not have a large influence in a disease (BPD) where maturity at birth has such an overwhelming influence. Our previous study in this population failed to find any association of the tumor necrosis factor-α-308 G/A, MCP-1-2518 A/G or transforming growth factor-β 1 +915 G/C SNPs on the development of BPD[ 33 ] Additionally we have found that the IL-10-1082 G/A SNP has no effect on the incidence of BPD (Yanamandra et al, Pediatric Pulmonology, in press). The finding of increased incidence of PVL in Caucasian infants with the II genotype may be due more to the greatly increased incidence of Ureaplasma urealyticum colonization in these infants. Chorioamnionitis, which is highly correlated with isolation of Uu, is a risk factor for PVL.[ 34 ] Because of the retrospective nature of this study we were not able to systematically review placental pathology. Our study has several limitations. The ACE I/D polymorphism has only been demonstrated to be functional in Caucasians.[ 16 ] Data linking this polymorphism and ACE activity in other ethnic groups are either lacking or suggests the polymorphism is non functional (African-Americans)[ 31 ]. If this were true, then one would expect to find no effect of the ACE I/D polymorphism on outcomes in African-American infants. The numbers of Caucasian infants in this study are few and an effect of the ACE I/D polymorphism on the incidence of chronic lung disease or other outcomes may not be detected. Thus it may be relevant to examine the contribution of the ACE I/D polymorphism to chronic lung disease in a larger cohort of Caucasian infants. The observations of this retrospective case controlled study are also limited somewhat by selection bias. Because only infants who were mechanically ventilated were included, the true impact of this polymorphism on the complications of prematurity may be underestimated. Since this was a retrospective study using stored material no attempt was made to correlate phenotype with genotype (ie TA ACE measurements). Lung injury and the subsequent maladaptive repair process that leads to the development of BPD is complex with a great many factors that interplay to determine outcome. It is very likely that other genetic play a role in determining the risk of poor outcome. Polymorphisms in cytokines, their receptors, bacterial pattern recognition molecules, surfactant proteins, and heme oxygenase-1 are known to alter the course of other pulmonary diseases and should be examined as to their potential role in the development of BPD in the premature infant. Conclusions The ACE I/D polymorphism does not significantly influence the development of BPD in ventilated infants less than 1250 grams. Competing interests The author(s) declare that they have no competing interests. Authors' contributions RJB conceived and organized the study, prepared the manuscript and performed the statistical analyses. KY performed the genotyping, and assisted with manuscript preparation. JL oversaw collection tracheal aspirates, assisted with recruitment of subjects into the original studies, and assisted with editing of the manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Supplementary Material Additional File 1 Contains Tables showing baseline clinical characteristics of African American and Caucasian infants separately by Genotype group Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544573.xml |
522806 | A combinational feature selection and ensemble neural network method for classification of gene expression data | Background Microarray experiments are becoming a powerful tool for clinical diagnosis, as they have the potential to discover gene expression patterns that are characteristic for a particular disease. To date, this problem has received most attention in the context of cancer research, especially in tumor classification. Various feature selection methods and classifier design strategies also have been generally used and compared. However, most published articles on tumor classification have applied a certain technique to a certain dataset, and recently several researchers compared these techniques based on several public datasets. But, it has been verified that differently selected features reflect different aspects of the dataset and some selected features can obtain better solutions on some certain problems. At the same time, faced with a large amount of microarray data with little knowledge, it is difficult to find the intrinsic characteristics using traditional methods. In this paper, we attempt to introduce a combinational feature selection method in conjunction with ensemble neural networks to generally improve the accuracy and robustness of sample classification. Results We validate our new method on several recent publicly available datasets both with predictive accuracy of testing samples and through cross validation. Compared with the best performance of other current methods, remarkably improved results can be obtained using our new strategy on a wide range of different datasets. Conclusions Thus, we conclude that our methods can obtain more information in microarray data to get more accurate classification and also can help to extract the latent marker genes of the diseases for better diagnosis and treatment. | Background With the successful completion of the Human Genome Project (HGP), we are entering the post genomic era. Facing mass amounts of data, traditional biological experiments and data analysis techniques encounter great challenges. In this situation, cDNA microarrays and high-density oligonucleotide chips are novel biotechnologies as global (genome-wide or system-wide) experimental approaches that are effectively used in systematical analysis of large-scale genome data. In recent years, with its ability to measure simultaneously the activities and interactions of thousands of genes, microarray promises new insights into the mechanisms of living systems and is attracting more and more interest for solving scientific problems and in industrial applications. Meanwhile, further biological and medical research also promoted the development and application of microarray. Typical issues addressed by microarray experiments include two main aspects: finding co-regulated genes for classification based on different cell-type [ 1 ], stage-specific [ 2 , 3 ], disease-related [ 4 - 6 ], or treatment-related [ 6 - 8 ] patterns of gene expression and understanding gene regulatory networks by analyzing functional roles of genes in cellular processes [ 9 , 10 ]. Here we focus on the former, especially on tumor classification using gene expression data, which is a hot topic in recent years and has received general attention by many biological and medical researchers [ 11 - 19 ]. A reliable and precise classification of tumors based on gene expression data may lead to a more complete understanding of molecular variations among tumors, and hence, to better diagnosis and treatment strategies. Microarray experiments usually generate large datasets with expression values for thousands of genes (2000~20 000) but not more than a few dozens samples (20~80). Thus, very accurate classification of tissue samples in such high-dimensional problems is difficult, but often crucial, for successful diagnosis and treatment. Several comprehensively comparative and improved methods have been proposed recently [ 20 - 22 ]. In this paper, we introduce a combinational feature selection method using ensemble neural networks to remarkably improve the accuracy and robustness of sample classification. In recent years, several researchers have used ensemble neural networks for tumor classification based on gene expression data [ 12 , 23 ]. Khan et al. [ 12 ] used neural networks to classify 4 subcategories of small round blue-cell tumors. By using 3750 networks generated by three fold cross-validation 1250 times and using the list of 96 most influential genes as the inputs, they reported very excellent results based on their dataset. Also O'Neill and Song [ 23 ] used neural networks to analyze lymphoma microarray data and can predict the long-term survival of individual patients with 100% accuracy based on the datasets published by Alizadeh et al [ 18 ]. Both of them are very good work in microarray data analysis using neural networks. In this paper our motivation lies in that by combining various feature selection mechanisms we can avail of more information of samples for classification and by using ensemble neural networks we can more effectively combine these features and improve the stability and robustness of answers. So the most important distinctions between our work and these above two citations are that by using combinational feature selection we can penetrate various different profiles of the samples and can avail of more information for classification, and also these neural networks can work in a parallel way unlike those two papers. In the same time, unlike their work based on some certain dataset, we can get improved, at least comparable results on a wide range of different datasets. In the following section, we provide detailed illustration and comparison of our new method. Results The general framework and implementation of our method The flowchart of our method can be seen in Figure 1 . When we obtain the microarray raw data based on a certain classification problem, first we need to preprocess them in order to be beneficial for further analysis. Broadly defined, pre-processing includes the planning and design of experiments, the acquisition and preprocessing of images, data transformation, data inspection, and data filtering. In this paper we avail of the publicly available datasets in , so we simplify this step and only use all datasets exactly as we found them in their transformed data. Due to the characteristic of small sample numbers in microarray data, in order to improve the accuracy, robustness and generalization of issue classification, we apply bootstrap mechanism to resample 100 iterations. During each iteration, we input the resample training data into three cooperative and competitive neural networks, and then by averaging their decisions, the neural network set can output their discrimination. From Figure 2 , we can clearly understand the architecture of these three neural networks. After obtaining the transformed resampling data, we extract and select features respectively based on ranksum test, PCA, clustering and t test. Ranksum test (also named Wilcoxon/Mann-Whitney test) is a nonparametric test, which does not take values into account and only calculates their scores purely based on rank information. We chose the top-ranked 30 genes identified as differentially expressed between the two types of tissues according to the ranksum test with the highest confidence (here using training data) as the first network input. At the same time, we used PCA to extract the principle components of all genes and used the top 15 principle components as the features to input another neural network. Also, we used Jaeger's "Masked out Clustering" ideas to group all the genes into 50 clusters and then used a t test to obtain the top 30 significant genes. Here we assume that each cluster can belong to the same pathway, genes which are co-expressed or are coming from the same chromosome. In this way, we can prefilter the gene set and drop genes that are very similar or highly correlated; that is, we can select the more significant genes for our discrimination as the third network input. More information about feature selection can be found in the methods section later. Based on these above three kinds of features we selected as the input, we construct and train three neural networks. Here we adopt simple one-hidden-layer feed-forward networks, which have 10 hidden units and one output unit for binary classification problem. As for multi-class problems, we can accordingly change the number of output units. Because each of these three networks adopts different feature selection mechanism as inputs, these inputs respectively reflect different aspects of samples, that is, different feature space in discriminative problems. We believe that this strategy of feature selection for issue classification reflects more profiles of different classes and will be able to obtain more accurate solution. Actually each of three networks is just like an expert holding a different judgment mechanism. Through averaging the confidences of three experts' answers, we can get the answer of this expert system. In this way, we not only can get the confidence of each expert, also we can judge the weight of each type of features in the answer. Finally, through competitive neural networks the robustness of this problem will be improved greatly. After completing the 100 iterations, we can get 100 individual answers about the problem. In this situation, how to combine these answers into one more precise result is still a problem. Here, we simply use majority voting to combine the result and then give the ultimate solution about this classification problem. As noted above, here we adopt the soft-voting mechanism, that is, we can combine the confidence of each net. All the implementations of our framework were written in Matlab, using the hardware platform of a PC running 2.4 GHz. Datasets illustration In this section, simple illustrations of the datasets we used in this paper for exploring the performance of our classification are given. The datasets in our paper have been downloaded from the following website: . We adopted their transformed data format for further research. All datasets we used can be reduced to three categories: binary class with testing samples, binary class without testing samples and multiple class problem. Here we classify samples into binary class with testing samples and without testing samples just according to the reference authors for each dataset. One important reason is that in this way we can easily compare our result with others based on the same training and testing sets. These datasets are shown in Table 1 . We use the three datasets below as the example of the first category, for which performance of our classification can be tested using the error ratio of testing samples. ALL-AML leukemia The training dataset consists of 38 bone marrow samples (27 ALL and 11 AML), with 7129 probes from 6817 human genes. Also, 34 samples testing data is provided, with 20 ALL and 14 AML. More information and raw data can be found in Golub et al. [ 11 ]. Lung cancer The dataset can be reduced to the problem of classification between malignant pleural mesothelioma (MPM) and adenocarcinoma (ADCA) of the lung. The training set contains 32 tissue samples, which consists of 16 MPM and 16 ADCA and the testing samples are constitutive of 15 MPM and 134 ADCA. Each sample is described by 12533 genes. More information about this dataset can be found in Gordon et al. [ 17 ]. Prostate cancer For the prostate cancer dataset, detailed explanation and raw data is available in Singh et al. [ 5 ]. This dataset consists of 102 training vs. 34 testing (Tumor versus Normal classification) samples. The training set contains 52 prostate tumor samples and 50 normal samples with around 12600 genes and the independent test sets consist of 25 tumor and 9 normal samples. Another three recently popular datasets have been used as the representative of the second category. Using these kinds of datasets, we apply cross-validation to validate our classification performance. Types of diffuse large B-cell lymphoma This dataset is used for discriminating distinct types of diffuse large B-cell lymphoma (DLBCL) using gene expression data. There are 47 samples, 24 of them are from "germinal canter B-like" group while the rest 23 are form "activated B-like" group and each sample can be described by 4026 genes. More detailed explanation can be found in Alizadeh et al. [ 18 ]. Ovarian cancer The goal of this significant experiment is to identify proteomic patterns in serum that distinguish ovarian cancer from non-cancer. The proteomic spectra were generated by mass spectroscopy and the dataset provided here is 6-19-02, which includes 91 controls (Normal) and 162 ovarian cancers with 15154 molecular mass / charge (M/Z) identities. Here we use the transformed normalization data in . More information can be found in Petricoin et al. [ 6 ]. Colon tumor The dataset Contains 62 samples collected from colon-cancer patients. Among them, 40 tumor biopsies are from tumors (labelled as "negative") and 22 normal (labelled as "positive") biopsies are from healthy parts of the colons of the same patients. Two thousand out of around 6500 genes were selected based on the confidence in the measured expression levels. Raw data and more information can be found in Alon et al. [ 14 ]. Finally, we can generalize our method from binary class to multi-class problems. In this paper, we evaluate the performance using the dataset below. MLL_leukemia This dataset contains training data consisting of 57 leukemia samples (20 ALL, 17 MLL and 20 AML) and testing data consisting of 4 ALL, 3 MLL and 8 AML samples. We adopted the transformed data from . More information can be seen in Armstrong et al. [ 15 ]. Our results First we primarily focus on the binary class problem. Because most of problems can be reduced to binary class problems, such as diseased vs. normal, survival vs. lethal, two opposite subtypes of some diseases and so on. Finally we generalize our classifier to multi-class application. In this paper, we evaluate the performance of different classification methods using predictive accuracy, which can be defined as: Here, TN 1, TN 2,…, TNn respectively denote the correct classification numbers of the samples belonging to a corresponding class; totalnum represents total sample numbers. The results of binary classification with testing samples For the first category of the datasets, we evaluate the performance of our classifiers using predictive accuracy of testing samples compared with the best performance of the current available methods. In this paper we use bagging to resample just as Tan and Gilbert [ 24 ], and we also compared our results to those using their bagged decision trees. In Table 2 , we described the recognition rate of our methods compared with the best classifiers of our knowledge for each certain dataset and bagged decision trees proposed by Tan and Gilbert [ 24 ]. From this table, it is clear that our results are remarkably better than others based on these several datasets. For the most popularly used AML-ALL leukemia dataset, to our knowledge, the best classifiers of this dataset can be found in [ 25 - 27 ], which can predict the results with 97.1% accuracy. However, we designed the classifiers using our methods based on 38 training samples, 0 error number of 34 testing samples can be obtained from our classifier. In the same way, we trained our ensemble of neural networks using 32 training sets of lung cancer and then predicted the 149 separate testing sets still with 0 error number. And three (1:2) testing error numbers can be reached using methods by Li et al. [ 28 ], which is the best performance corresponding to this dataset of our knowledge. For the third prostate cancer dataset, after training the classifier using 102 training sets, only one wrong classification can happen using our ensemble neural networks to predict the 34 separate testing samples. We did not find a more accurate classification result except for the bagged decision trees in [ 24 ] based on this dataset, so here we think that is the best result. In this sense, a great improvement in predictive accuracy can be obtained by using our method. In order to further validate the effectiveness of predictive accuracy, we also performed the leave-one-out cross-validation (LOOCV) respectively only on the above three training samples. We also obtained the 100% accuracy both on the AML-ALL leukemia dataset and the lung cancer dataset, which are the same results as using individual testing samples. At the same time, 96.08% accuracy can be got based on the prostate cancer dataset, which is a little lower than using individual testing samples. For the purpose of comparison, we also list these results in Table 2 . Thus, we conclude that our performance evaluation is credible. From the results of the above three testing datasets, we can also see that many different classifiers obtain the best results when they concern some certain dataset, but there is still no general best strategy for tumor classification problems based on a wide range of different datasets. Furthermore, from Figure 3 , it is clear that our method is superior to the traditional bagging decision trees. Thus, we conclude that by using our method a more general accuracy improvement can be achieved for tumor classification. The results of binary classification without testing samples Without separate testing samples, we cannot evaluate the performance of our classifiers with the predictive accuracy of testing samples in the same way as above. Many performance evaluation methods have been proposed, of which various cross validations are most popularly used, such as 3-fold cross validation, 10-fold cross validation, leave-one-out cross-validation (LOOCV), and others. Here, we used the leave-one-out cross validation (LOOCV) to evaluate the performance of ours based on these available datasets. For further comparison with recent published methods based on the same datasets, we also perform 10-fold cross validation just as they used in their research. In Table 3 , we list the predictive accuracy of our methods using 10-fold cross validation and LOOCV respectively and the corresponding results of other methods based on the same dataset and the same evaluation mechanism. These comparisons based on data in Table 3 are shown in Figure 4 . In the first data column of Table 3 , we show our predictive accuracy 97.87% and 95.74% by LOOCV and 10-fold cross validation respectively. But unfortunately, we did not find the corresponding result based on this dataset. Cho et al. [ 25 ] artificially divide the dataset into 22 training samples and 25 test samples, and their best classification result is 96%. For the purpose of comparison, we also use the same strategy as Cho et al.'s [ 25 ] and in Figure 4 we can see that 98% predictive accuracy obtained by our method is a little better than theirs. O'Neill and Song [ 23 ] used neural network to get very good result based on the lymphoma dataset. But here the dataset we used is based on different subset and we can't compare our result with theirs. For the ovarian dataset we used, Liu et al. [ 21 ] reported that 100% predictive accuracy can be obtained running 10-fold cross validation on all 253 samples under an all-χ 2 feature selection heuristic and support vector machine (SVM). In our method, only 75 features in total were used, and 99.21% and 98.82% accuracy was obtained respectively by LOOCV and 10-fold cross validation. Thus, we think that our method is comparable to theirs to some extent. For the colon tumor dataset, we found that 85.48% predictive accuracy is the best classification result obtained in Dettling et al. [ 27 ], where they used various boosting algorithms and adopted leave-one-out cross validation (LOOCV). As shown on Figure 4 , compared with our predictive accuracy by LOOCV, a significant accuracy improvement was obtained by using our method. Our result by using 10-fold cross validation also is shown in Table 3 . From the results of the above three datasets, we can see that our method is better, or at least comparable to current other best methods. Also, we need to note that these named best methods can get the best results based on certain datasets but may get worse results based on other datasets (Here we omitted the concrete comparisons on wide range datasets; correlated information can be found in our references); however, general performance improvement can be obtained using our method. The result of multi-class problem Finally, we generalize our method from binary class to multi-class problems. After minor adjustments to the corresponding parameters of our framework, we obtained 100% classification result of the above-mentioned multi-class dataset – MLL_Leukemia dataset, that is, 4 ALL, 3 MLL, 8 AML of 15 test data can be predicted correctly. Similarly, 100% accuracy also was obtained by Li et al. [ 28 ]. In this way, we conclude that our method also is fit to multi-class problems, and the classification result is comparable to other methods. Discussion In this paper, we introduce a combinational feature selection and ensemble neural network method for the classification of gene expression data. On a wide range of recently published datasets, our method performs better, or is at least comparable to, the current best methods of our knowledge. As a further test, we randomly selected genes of the same amount as the feature instead of any of the three individual selected features in our research and then used the ensemble neural networks based on these features for classification again. The apparently worse discrimination power can be seen in this strategy. Moreover, we also used the output of a unitary network based on all the same features as the ultimate classification result and the result was also worse than ours. Thus, we believe such remarkable performance improvements of our method are due to the fact that our combinational feature selection mechanism induced more useful information for discrimination, and the ensemble neural network framework improved the stability, robustness and generalization of learning. We performed simple majority voting mechanism to combine the individual networks produced by bagging and got a more accurate solution. The advantage of the ensemble is to reduce the variance, or instability of the neural network, and avoid the error surface of neural network training being trapped into local minima. The ensemble model tends to cancel the noise part as it varies among the ensemble members, and it tends to retain the fitting to the regularities of the data. In this paper, our ensemble neural network model has 100 members; However, further research is needed to determine how many members working together can reach the best performance. In this paper, we focused on classification problem, so we didn't give a detailed analysis about how the importance of each different gene we select and the interaction between them influenced the diseases, which is a very important issue for application and will be researched in our future work. Note that the only drawback of our approach is the problem of increasing computational complexity and the fact that it consumes a little more time than others. However, considering the lost caused by wrong prognosis or diagnosis of disease, we believe that the remarkable improvement in corresponding accuracy deserve these costs. Conclusions By aggregating various information and ensemble neural networks, we reached a more accurate classification decision based on several datasets. We think that making full use of all available information will more clearly elucidate the latent mechanisms of many diseases. For example, we can combine various imaging techniques, such as CT, MRI, PET and others, which can detect the change of phenotype for the corresponding disease, with microarray data for further research. In this way, we can recognize the nature of various life phenomena both from macro and micro viewpoints. Also, we can retrieve the information of genes that are used in microarray, such as gene functions and gene locations. In this way, we can make use of prior knowledge combined with the microarray data for further research. Methods Feature selection Feature selection is one of the most important issues in classification, which is a transformation process of observations in order to obtain the best pathway for getting to the optimal solution. At the same time, it can reduce the complexity of the data to make it more comprehensible. It is particularly relevant for microarray datasets with thousands of features because it has been reported that many diseases, especially tumors, have never been caused by a single gene mutation but are the result of a series of gene changes. Such genes are highly relevant to the studied phenomena of diseases. On the other hand, the expression levels of many other genes may be irrelevant to the distinction between tissue classes. We can say that the extraction and selection of features determine the ultimate performance of classifiers. Both for cost and for biological insight, making full use of the most informative genes and finding small feature sets with high classification accuracy are very essential. At the same time, highly informative genes that are part of known biochemical pathways give insights into the processes that underlie the differences between classes, and those of unknown function suggest new research directions. Some classifiers, such as trees, perform automatic feature selection and are relatively insensitive to the variable selection scheme, but most classifiers need to perform feature selection first. So far, various feature selection schemes have been used in microarray data analysis, such as the most popular method of selecting the top-ranked genes based on various different scores (Euclidean distance, correlation coefficient, mutual information, signal to noise ratio) [ 9 , 11 , 22 , 25 , 29 ]. These feature selection methods gain better results on certain datasets, and the selective informative genes are the marker genes providing more useful information for further diagnosis and treatment. However, a problem with the above approaches is that they tend to select more correlated features so as not to provide more useful information for the purpose of classification. Li et al. [ 28 ] conclude that sometimes low-ranked genes are found to be necessary for classifiers to achieve perfect accuracy. It is conceivable that these useful low-ranked genes might have some relations with some important biological pathways and might have a vital influence on some diseases. Just selecting top-ranked genes will inevitably lose essential information. In order to compensate for this shortcoming, Jaeger et al. [ 22 ] proposed an improved gene selection for classification of microarrays. They demonstrated that the traditionally selected genes based on top-ranked scores are usually highly correlated, and they solved that problem through retrieving groups of similar genes first and then applying test-statistic to finally select genes of interest from these groups. In this way, the selected genes can correspond with some biological insights and might give out more accurate prediction about disease. The difficulty of this method lies in determining how many clusters and how many genes might directly correspond to the pathway on certain problems. Also, many researchers get the first several principle components by using PCA or SVD as the selected features, which captures most variation between samples and to some extent can obtain better results [ 12 , 30 - 33 ]. However, principle components cannot provide comprehensible rules to help elucidate the scheme of the related disease because it can be due to noise as well as true difference in expression and we do not know how many genes to pick. Just as we alleged above, in such a high-dimension space, finding accurate and significant features (genes) is very essential for classification, for cost savings and for biological insights. However, it has been verified that differently selected features reflect different aspects of the dataset and some selected features can obtain better solutions on certain problems. This is because one feature selection mechanism corresponds to one different artificial hypothesis, but which hypothesis is most near to the true hypothesis on a special problem is unknown to us. Here, we propose combining the above mentioned several feature selection methods to reflect different profiles of samples in order to obtain more useful information for classification and to produce a good approximation to true hypothesis by averaging the different hypotheses. In this paper, we select features using wilcxon's ranksum test [ 34 ] to get the top-ranked genes, use PCA [ 31 ] to obtain the principle components as the feature, and use Jaeger's clustering method to group the whole genes into different clusters, and then select the top-ranked genes by t-test [ 34 ] scores from these groups. After picking these features from gene expression data, how to make full use of these features for further accurate classification is still a problem. Detailed illustration of our strategy is given below. Ensemble neural networks An Artificial Neural Network (ANN) is an information-processing paradigm that is modeled on biological nervous system, which is composed of a large number of highly interconnected processing elements (neurons) working in unison to solve specific problems. In fact, since the basic model was proposed, various improved algorithms and theories have already been successfully applied in many fields. Because neural networks are best at identifying patterns or trends in a large amount of data with little theory, they are well suited for prediction or forecasting needs. That's just the case for microarray data. However, instability and little intrinsic knowledge of neural networks are obstacles to its further generalized application in some specific problems. Here we ensemble multiple networks in an attempt to solve this problem to some extent. Since multi-net systems were introduced by Sharkey in 1996 [ 35 ], the combination of a number of neural networks has been widely applied in many fields. Because combining the outputs of several neural networks into an aggregate output often gives improved accuracy over any individual output, the objective of this kind of ensemble module is to solve problems that are difficult for a single neural network, and is to combine the individual outputs to achieve better generalization. Remarkable advantages of the ensemble compared with a unitary network have been demonstrated previously [ 36 , 37 ], one advantage of which is that it can to some extent ease the obstacles mentioned above and can improve the stability of neural network decisions. Generally speaking, in neural networks ensemble, two problems need to be resolved: how to generate the individual network and how to combine them together. Using bootstrap or boosting resample mechanism to obtain the individual network is the most popular method to solve the first problem. Bootstrap is the most popular resample mechanisam of sampling with replacement, therefore some observations are duplicated and some are omitted. Boosting means to boost a "weak" learning algorithm into a "strong" learning algorithm. Their differences are that using bootstrap can resample uniformly and can get the individual network immediately, but boosting weights every sample in each iteration and must generate the individuals in sequence. Several recent published papers claimed that adaboost-the basic boosting algorithm is not fit to microarray data analysis [ 20 , 26 ], and some improved boosting have been made to increase the accuracy to some extent [ 26 , 27 ]. With no exception, they perform boosting in conjunction with decision trees. Here, we perform our resample mechanism using bagging, that is, bootstrap aggregating, which has been shown to work well in the presence of noise [ 24 ]. Due to the noisy fact of microarray data, here we use bagging to resample. As a further validation, we also used adaboost instead of bagging to resample in order to construct the individual network and the result is worse than bagging networks. As to the second problem, many ensemble mechanisms have been researched in recent years; for example, improvements in performance can result from training the individual networks to be decorrelated with each other [ 38 ] with respect to their errors. In this paper, we only adopt the majority voting, the basic ensemble method, to obtain the ultimate output result. Note that here we use the soft-voting mechanism, that is, the confidence of each net output is applied as voting value, rather than unit or zero. Considering the complexity and noise of microarray data, it's very difficult to get a perfect solution using a unitary neural network based on some certain selected features. At the same time, accurately extracting and selecting the most informative genes is also very difficult using individual available methods. Thus, in this paper we attempt to combine multiple modes of information available from gene expression data using neural networks ensemble to get a better solution. We presented a cooperative and competitive neural network system that each of nets has the same architecture and topology, and each can respectively learn to classify a set of patterns based on partial information of the patterns and then by combining their classification results we can get a more precise result. The detailed framework of combing various features and neural networks ensemble and its implementation methods are discussed below. Authors' contributions BL carried out the design of the method and performed the related analysis. QC participated in discussions of algorithms and manuscript preparation. TJ and SM instructed the whole study. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC522806.xml |
449902 | No Rest for the Weary: Migrating Songbirds Keep Their Wits without Sleep | null | Every spring and fall, billions of songbirds fly thousands of miles between their summer breeding grounds in North America and their wintering grounds in the more hospitable climes of southern California, Mexico, and Central and South America. While some birds fly during the day, most, including the white-crowned sparrow, fly under cover of night. Many aspects of this remarkable voyage remain obscure, especially if, and how, nocturnal migrators get any sleep at night. White-crowned sparrow ( Zonotrichia leucophrys gambelii ) A tracking study of the Swainson's thrush found that the roughly seven-inch birds flew up to seven hours straight on six of seven nights, racking up over 930 miles. While the study didn't track their daytime behavior, the birds' migratory pace—as well as the increased activity required to sustain migrations—suggests little time for sleep. Yet field observations indicate that presumably sleep-deprived fliers appear no worse for wear, foraging, navigating, and avoiding predators with aplomb. Researchers are left trying to reconcile this observation with the vast body of evidence linking sleep deprivation to impaired neurobehavioral and physiological function. How do songbirds cope with so little sleep? Do they take power naps? Have they taken “sleep walking” to new heights? Or have they managed to selectively short-circuit the adverse effects of sleep deprivation during migratory stints? To investigate these questions, Ruth Benca and colleagues studied cognitive and sleep behaviors in captive white-crowned sparrows over the course of a year. The sparrows fly nearly 2,700 miles twice a year between their Alaska and southern California homes. In laboratory cages, the birds' migratory instincts manifest as increased restlessness at night during the migratory season, with lots of hopping around and wing flapping. Niels Rattenborg et al. characterized the birds' activity levels with motion-detection measurements and video recordings, and placed sensors on their brains to monitor their seasonal sleep patterns. The brain recordings showed a marked seasonal difference in both the amount and type of sleep during a 24-hour period: migrating birds spent roughly two-thirds less time sleeping than nonmigratory birds and fell into REM sleep (the dream stage of sleep, marked by rapid eye movements) much sooner. Birds displaying active migratory behavior appeared completely awake during such activity. Cognitive tests—birds performed a task that involved pecking a key in exchange for seed—revealed that birds in the nonmigrating state suffered cognitive deficits when sleep-deprived but displayed an “unprecedented” ability to maintain cognitive function in the face of ongoing sleep loss in the migratory state. These results suggest that wild songbirds drastically reduce sleep time during migration, though Benca and colleagues concede it's impossible to know for sure without recording the birds in action. And it is unclear what molecular mechanisms jumpstart the migratory mindset. Such an ability to temporarily circumvent the need for sleep, however, could prove useful for humans in situations that demand continuous performance. Some studies link migration with increased neuroendocrine activity, which is in turn associated with sleep disruption, accelerated timing of REM cycles, and mood disorders in humans. “Like migrating sparrows,” the authors note, “both depressed and manic patients show reduced latency to REM sleep, loss of slow-wave sleep, and reduced amounts of total sleep.” Given the parallels between migratory behaviors and bipolar illness, it's possible that similar mechanisms may be involved in both. Whatever the mechanism, the unprecedented imperviousness of migrating songbirds to sleep deprivation, the authors conclude, clearly warrants further testing. But it also raises interesting questions about the role of sleep, which recent studies suggest is required to incorporate novel perceptions into the brain's memory banks. If this is true, how do songbirds consolidate memories of migratory events with so little sleep? Understanding the mechanisms that power the sleepless flight of songbirds promises to unravel one of the longstanding mysteries of their improbable journey. It may also shed light on the origins of sleep-related seasonal disorders and the much-debated role of sleep itself. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC449902.xml |
554774 | Methodological bias in cluster randomised trials | Background Cluster randomised trials can be susceptible to a range of methodological problems. These problems are not commonly recognised by many researchers. In this paper we discuss the issues that can lead to bias in cluster trials. Methods We used a sample of cluster randomised trials from a recent review and from a systematic review of hip protectors. We compared the mean age of participants between intervention groups in a sample of 'good' cluster trials with a sample of potentially biased trials. We also compared the effect sizes, in a funnel plot, between hip protector trials that used individual randomisation compared with those that used cluster randomisation. Results There is a tendency for cluster trials, with evidence methodological biases, to also show an age imbalance between treatment groups. In a funnel plot we show that all cluster trials show a large positive effect of hip protectors whilst individually randomised trials show a range of positive and negative effects, suggesting that cluster trials may be producing a biased estimate of effect. Conclusion Methodological biases in the design and execution of cluster randomised trials is frequent. Some of these biases associated with the use of cluster designs can be avoided through careful attention to the design of cluster trials. Firstly, if possible, individual allocation should be used. Secondly, if cluster allocation is required, then ideally participants should be identified before random allocation of the clusters. Third, if prior identification is not possible, then an independent recruiter should be used to recruit participants. | Background The randomised controlled trial (RCT) has a number of important features that make it the 'gold-standard' evaluation method. One of the most important aspects of random allocation is that it eliminates selection bias. Randomisation ensures that the two or more groups formed are similar, except for chance differences, in all aspects. Nevertheless unless trials are undertaken in a rigorous manner biases can be introduced that negate the effect of random allocation. Indeed, a poorly conducted RCT can be worse than a good observational study as the latter is interpreted in the light of possible confounding whereas the results of an RCT might be uncritically accepted. Random allocation can take place either at the level of the individual level or at a higher group or cluster level. In a cluster randomised trial groups of people are allocated to receive an intervention or not. In some areas of evaluation (e.g., education) the cluster is the natural method of allocation. For example, a trial among school children may well randomise by class or by school rather than by individual child. Allocation by cluster may be preferable for a number of reasons. There may be practical reasons: for instance, teaching a novel curriculum will be easier to use existing classes than form new ones through randomisation. There may be contamination issues. Individuals allocated to a control treatment may inadvertently receive some aspects of the intervention if they are in proximity to the treated group. Allocation by cluster has some important statistical issues that have been addressed 65 years ago in the educational trial literature [ 1 ] and subsequently widely in medical statistics [ 2 ]. In brief, analysis of cluster trials needs to take into account the clustered nature of the data otherwise the risk of a Type I error (i.e., erroneously concluding there was a statistically significant difference) increases. However, more seriously in our view is the potential of cluster trials producing a biased estimate of treatment effect. Randomisation should eliminate selection bias. Selection bias can be reintroduced within any trial if there is high loss to follow-up or failure to use intention to treat analysis. In cluster trials selection bias can also be introduced through participant recruitment. Because cluster trials often recruit their participants after the clusters have been randomly allocated this can lead to selection effects[ 3 , 4 ]. There are a number of potential reasons for this. Foreknowledge of allocation If the person recruiting participants has both knowledge of the clinical characteristics of the participants and of the allocation schedule biased recruitment can occur. Subversion, within individually randomised trials, can occur by recruiting participants with poor prognostic characteristics so that they are more likely to enter the 'unfavoured' group [ 5 , 6 ]. Evidence for the biasing effects of allocation foreknowledge has been shown on treatment effect sizes [ 7 , 8 ]. Consequently a rigorously designed individually randomised trial ought to conceal the allocation schedule from the people who are recruiting participants. Cluster randomised trials often do not, or cannot, conceal treatment allocation. For example, a trial was undertaken to reduce violence among children randomised by school [ 9 ]. After allocation the children were recruited into the study and the intervention was delivered. The allocation could not be concealed from the teachers researchers or children. This has two potentially unfortunate consequences. Awareness of the allocation can lead to biased recruitment in cluster trials [ 10 ]. Alternatively, or in addition, participants can differentially refuse consent to participate in the trial and this could be another source of selection bias. For example, in a cluster trial evaluating the use of advanced end of life directives among residents of nursing homes there was a differential in participant rates of 83% among people in the intervention homes compared with 92% in the control arm [ 11 ]. Such differential participation rate can lead to selection bias. Treatment effects on recruitment Recruitment of participants with different clinical characteristics is not necessarily a sign of subversion it could be simply a consequence of the cluster level intervention. For example, in an evaluation of an educational package for the treatment of back pain primary care physicians were trained in 'evidence based' management of back pain [ 12 ]. This training was associated with an increased recruitment rate among practices allocated to training compared with no training. Because training involved recognition and diagnosis of back pain with hindsight we should not be unsurprised that differential recruitment would occur in this instance. As well as having more potential, than individually randomised trials, for the introduction of selection bias. Cluster randomised trials also may be at more risk of dilution bias. Because consent for treatment is often not obtained until after randomisation more participants, than in an individually randomised study, may refuse treatment and this will consequently dilute any treatment effects. For example, Kendrick and colleagues in a cluster randomised trial to prevent accidental injuries among young children found that only 75% of the group allocated to the experimental group actually received the intervention [ 13 ]. Whatever the underlying reasons for differences in recruitment the consequences are potentially the same: selection bias has been introduced. Figure 1 shows the potential sources of bias that can occur after cluster randomisation. With the introduction of selection bias trial results are unreliable. In this paper we examine some evidence for this phenomenon and make recommendations on how to design this problem out of future cluster trials. Figure 1 Sources of bias in cluster trials Evidence for recruitment bias among individual trials A recent review identified a sample of 36 cluster randomised trials from three major general medical journals, between 1997 and 2002 [ 4 ]. This review identified all cluster randomised trials published in three major medical journals over a period of five years. In this review 15 of the trials could have experienced bias in their recruitment of participants. Of these 15 trials seven showed some evidence in the published papers of consenting differential numbers of participants or excluding participants in a selective fashion. One of the remaining 8 trials, whilst having no evidence of bias in the original published paper was later subsequently found to have experienced recruitment bias [ 10 ]. Therefore, 25% of cluster trials published in major clinical journals suffered potential selection bias. On the other hand a review of 152 cluster trials undertaken in primary care found that only 8 (5%) were found were the authors reported differential recruitment [ 14 ]. However, unlike Puffer and colleagues each trial was not carefully scrutinised to ascertain whether or not there was a problem of biased recruitment (Eldridge, personal communication). Although Puffer and colleagues noted that some trials had significant differences in recruitment and retention rates between groups they did not investigate whether or not this had an impact on important treatment covariates. To assess whether observed differences in recruitment could have had an effect on important predictors of outcome we examined the age differences between treatment groups. We chose age for two reasons: first, it is a commonly reported baseline characteristic and, second, is the most likely common confounder across different disease groups. Nevertheless, we acknowledge that biased recruitment may not manifest itself in terms of age differences [ 12 ]. From the 36 trials we identified 14 that reported, either directly or indirectly, the mean age and standard deviation of the treatment groups. Of the 14 trials that were included nine stated they had taken clustering into account in their sample size calculation, two had not and the remaining three it was not clear whether they had adjusted their sample size. We then grouped the 14 trials according to whether Puffer et al considered there was evidence for differential recruitment. Eight out of the 14 trials had been regarded as potentially biased. In Figure 2 we plot the standardised mean age differences between treatment groups (ie., age difference divided by the pooled within group standard deviation). Negative age differences were all converted to positive differences as we were uninterested in the direction of the bias. Figure 2 Standardised mean differences of patient age As can be seen in Figure 2 the age difference between treatment groups tend to be larger in the potentially biased group. The mean age difference was greater than 10% of their standard deviation in 3 out of the 8 potentially biased trials. The pooled standardised mean difference in the biased group was also twice as large as that in the non-biased group. A test for the difference using a meta-regression resulted in a non-significant p-value of 0.15; therefore, the difference observed in this instance was not conclusive. Age imbalances for any single trial could be due to chance as it is more difficult to achieve balance in cluster randomised trials compared with individually randomised trials due to the smaller number of allocated units. On the other hand, cluster trials, like individually randomised trials should be balanced across all cluster trials if there was no bias present. In Figure 3 we plot the standardised mean age differences by whether or not the trial showed a statistically significant effect. The significance was determined as p-value < 0.05. In all 14 trials the analysis took the clustering effect into account through various methods The figure suggests that significant trial results were associated more often with potentially biased recruitment with larger baseline differences in age, even though a formal test for interaction did not show a statistical significance (p-value 0.3) as this was not adequately powered. Figure 3 Standardised mean difference by bias group and treatment significance Evidence for bias from a systematic review Cluster randomised trials often answer different questions to individually randomised trials or cannot use individual allocation. Therefore, it is difficult to make a direct comparison between individual and cluster randomised trials in terms of the likely differences in effect sizes within the same subject area. However, within the area of hip protection for fracture prevention there are trials using both individual and cluster allocation. The most recent Cochrane review of hip protectors has identified 13 RCTs of hip protectors with hip fracture outcomes [ 15 ]. In addition, there is a large individually randomised trial that has not yet been included in the review (i.e., 14 in total) [ 16 ]. In figure 4 the effect sizes from these trials are plotted against their sample size. The sample sizes were adjusted for the design effect for the cluster trials and therefore the sample sizes for these trials are the effective sample sizes (sample size divided by the design effect). One out of the five cluster randomised trials reported their design effect, from which we estimated an intra-cluster correlation coefficient (ICC) and applied this to the other four studies, as they are all similar trials, to calculate a correction factor. As the figure shows the resulting funnel plot indicates little evidence of effect from individually randomised trials. In contrast, all of the cluster trials show a substantial benefit when using the cluster design. This suggestion of bias could be as a result of publication bias. On the other hand, there are a number of alternative explanations. First, the cluster trials might have been undertaken in a different setting than the individually randomised trials and this might account for the observed differences in effect. Second, the intervention (hip protection) may work better using a clustered design. Third, there might be treatment contamination in the individually randomised trials: biasing the treatment effect towards the null. Fourth, the observed differences may be due to poor implementation of cluster trial methodology, which biases the results of those trials towards the positive. Figure 4 Funnel plot of individually and cluster randomised trials There is a tendency for the cluster trials to be largely undertaken among residents of nursing homes compared with individually randomised trials – although one of the largest individually randomised trial was in a nursing home setting. It is possible that compliance might have been better in a nursing home setting and this could account for the difference in effect. However, the compliance rates were not that different from those trials using individually randomisation. The second reason that hip protection might work in a clustered design might be that the intervention is delivered as a 'package' of care and their use alerts the clinical staff responsible (e.g., nursing home staff) to the dangers of falls and this encourages other anti-fracture interventions. It is possible that in the individually randomised trials the control group could have been 'contaminated' by accessing hip protection by, for example, buying the product themselves. In a large individually randomised trial we undertook of hip protectors [ 16 ] some participants in the control arm did purchase hip protectors; however, the prevalence of this was very low. An alternative explanation of the difference was poor implementation of the cluster trial methodology: including selective recruitment, differential loss to follow-up and failure to use intention to treat analysis. For example, the largest cluster randomised trial had a 30% difference in the population that were included in the trial after random allocation [ 17 ]. Preventing biased recruitment In order for cluster randomised trials to provide unbiased evidence for treatments we must design out any sources of recruitment bias. In this section we will consider design suggestions that should minimise this threat. Use individual allocation Often cluster randomisation is used to overcome the perceived threat of contamination between the treatment groups. Although in many instances this threat is real in some cases there may be little contamination. Indeed, even if there are quite high contamination rates (e.g., 20%) it may still be more efficient in sample size terms to randomise more patients in an individual trial and accept a diluted effect size [ 3 ]. Therefore, one solution to avoiding biased recruitment is to avoid using cluster trial methods if at all possible. Prior identification of participants In some instances it may be possible to identify participants before cluster allocation. For example, if we consider a school based evaluation of a health promotion curriculum. Children within schools or intact classes can be identified before the cluster allocation. Children and their parents can be asked to participate in the study and are presented with the alternatives under consideration. Once consent has been obtained to take part in the study then the schools or classes are randomised to the different curricula. Independent recruitment Evaluation of an intervention for incidence disease cases means prior identification is not possible. For example, consider a trial of educating primary care physicians for the treatment of acute shoulder pain. Because the condition has an incident nature it is necessary to recruit participants in a prospective manner. Should the primary care physician undertake this then selection bias is likely to ensue. Therefore, to reduce this possibility an 'independent' person needs to recruit participants. Consider a recent example of such an approach. In a trial of educating GPs for the identification and treatment of depression in primary care trial participants were recruited by practice receptionists. Because the receptionists from both intervention and control practices had been exposed to the same amount of trial training then the potential for selection bias is reduced, although never eliminated [ 18 ]. Discussion The use of cluster randomised trials has significantly increased in medical research in recent years [ 19 ]. Despite Lindquist outlining an appropriate approach to the analysis of cluster trials in 1940 [ 1 ] – many fail to undertake the analysis taking the clustering effect into account. Consequently the attention of many medical statisticians has been directed at the appropriate analysis and sample size issues with less attention to more serious problems with the design of cluster trials. Whilst inappropriate analysis will give misleading precision (i.e., smaller confidence intervals and lower p values) it will rarely change the point estimate of a treatment effect. In contrast, bias can give a misleading effect size estimate. Cluster randomised trials are potentially more susceptible to some forms of bias than individually randomised trials. Biased recruitment can be a problem in some cluster randomised trials. One symptom of biased recruitment is differential recruitment rates. However, one trial noted significant selection bias even when there were similar recruitment rates [ 10 ]. Therefore, even when recruitment rates appear similar between treatment arms selection bias can be introduced. We have examined the issue of bias in cluster trials by comparing a sample of trials against similar individually randomised studies from a review of hip protectors. This suggested a difference in effect size that was dependent upon the type of study design. However, the sample size was small and there are alternative explanations to the apparent differences in effect sizes: not least the explanation of chance. We have also looked at baseline differences in ages of people in cluster trials that appeared to be free of bias with those that seem to have had bias introduced due to poor methodological application of design. There was a difference in age imbalance, which was suggestive of an interaction with statistical significance of trial results although a formal statistical test failed to show a significance of the difference. Our sample size was relatively small and we could have missed a statistically significant difference through lack of statistical power. Nevertheless, this paper does raise concerns about the design of cluster trials and signals that such trials should be used with caution. If there are important confounding variables, stratification, matching or regression models for clustered data are required. Studies with evidence of biased recruitment might try methods of analysis that allow for observed confounding. For example, if there were imbalances in patient or cluster level covariates between the randomised groups multi-level or hierarchical models explicitly model the treatment effect adjusting for the confounding, provided that there is a fairly large number of clusters. However, even the most sophisticated statistical analysis cannot adjust for the unmeasured or unknown confounder, which is one of the main reasons we undertake random allocation. Therefore, it is crucial that we avoid the introduction of bias into our cluster designs. Future cluster randomised trials should endeavour to either identify participants before randomisation or use an independent person, preferably blind to allocation, to recruit participants. Furthermore, cluster randomised trials ought to be undertaken such that loss to follow-up is similar between groups and intention to treat is always used. Competing interests The author(s) declare that they have no competing interests. Authors' contributions SH undertook the statistical analysis and wrote sections about statistical methods and implications. SP wrote the first draft and contributed to the original data collection. DT had the original idea of the paper and undertook revisions to the original draft. SW contributed to the original review and collected additional data for the paper. All authors contributed to commenting on drafts of the manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC554774.xml |
546416 | A case study of type 2 diabetes self-management | Background It has been established that careful diabetes self-management is essential in avoiding chronic complications that compromise health. Disciplined diet control and regular exercise are the keys for the type 2 diabetes self-management. An ability to maintain one's blood glucose at a relatively flat level, not fluctuating wildly with meals and hypoglycemic medical intervention, would be the goal for self-management. Hemoglobin A1c (HbA1c or simply A1c) is a measure of a long-term blood plasma glucose average, a reliable index to reflect one's diabetic condition. A simple regimen that could reduce the elevated A1c levels without altering much of type 2 diabetic patients' daily routine denotes a successful self-management strategy. Methods A relatively simple model that relates the food impact on blood glucose excursions for type 2 diabetes was studied. Meal is treated as a bolus injection of glucose. Medical intervention of hypoglycaemic drug or injection, if any, is lumped with secreted insulin as a damping factor. Lunch was used for test meals. The recovery period of a blood glucose excursion returning to the pre-prandial level, the maximal reach, and the area under the excursion curve were used to characterize one's ability to regulate glucose metabolism. A case study is presented here to illustrate the possibility of devising an individual-based self-management regimen. Results Results of the lunch study for a type 2 diabetic subject indicate that the recovery time of the post-prandial blood glucose level can be adjusted to 4 hours, which is comparable to the typical time interval for non-diabetics: 3 to 4 hours. A moderate lifestyle adjustment of light supper coupled with morning swimming of 20 laps in a 25 m pool for 40 minutes enabled the subject to reduce his A1c level from 6.7 to 6.0 in six months and to maintain this level for the subsequent six months. Conclusions The preliminary result of this case study is encouraging. An individual life-style adjustment can be structured from the extracted characteristics of the post-prandial blood glucose excursions. Additional studies are certainly required to draw general applicable guidelines for lifestyle adjustments of type 2 diabetic patients. | Background It is well established that diabetes can lead to acute and chronic complications, compromising the health and quality of life. Results from various studies [ 1 ] have demonstrated that improved control of blood glucose in type 2 diabetes reduces related complications. Type 2 diabetes results from the metabolic problem that is related to certain tissue resistance to insulin action and to the inability of the pancreas to appropriately regulate the quantity of insulin for glucose metabolism. These metabolic abnormalities lead to the many complications of diabetes. Type 2 diabetes historically occurs predominantly in adults aged 40 and over. A recent trend, however, indicates that children and adolescents of minority ethnic groups, especially in African Americans and American Indians, are increasingly susceptible to type 2 diabetes [ 2 ]. With the prevalence of type 2 diabetes and its associated risk for serious complications, issues related to proactive self-management become an urgent concern. Dietary management is frequently referred as the cornerstone, or the initial step, in treating of type 2 diabetes mellitus. Foods containing carbohydrates play an important role in the diet. The glycemic Index (GI) ranks foods according to their post-prandial glycemic responses. The GI was introduced more than twenty years ago and has been widely adopted in diabetes management in Australia, New Zealand, Canada, the United Kingdoms, and France [ 3 ]. The World Health Organization states that it is important to consider the GI in constructing a healthful diet because low GI foods help control blood sugar levels by producing minimal fluctuations in blood glucose [ 4 ]. For diabetic patients, choosing low GI foods is particularly important because consumption of high GI foods often results in far more exaggerated glycemic responses, creating a need for drug or insulin therapy [ 3 , 5 ]. Most published GI lists are for single food items only. A GI is a numerical measure of how a carbohydrate would increase one's blood glucose level over a period of two (for normal) or three hours (for diabetic patients) after eating [ 6 , 7 ]. The area of elevated blood glucose level from the baseline (the pre-prandial measure) is expressed as a percent of the area for the same amount of a reference carbohydrate such as a pure glucose or a white bread (usually 50 g) [ 8 , 9 ]. To plan a complete meal using the weighted mean [ 6 ] for various food items is not only tedious, but also impractical. Diet exchange lists are usually recommended for diabetic patients to use in formulating a sensible meal plan. However, an exchange list is not always convenient to use. Moreover, there is a lack of ethnic diet exchange lists. For a member of an ethnic minority to follow a diet exchange list, he or she must prepare his or her own meal away from the rest of the family. Nutall and Chasuk [ 10 ] have stressed that dietary recommendations for type 2 diabetes should be flexible and highly individualized, yet most of the prepared meal programs and exchange-list diets for diabetes have not had individualization in mind nor are they designed for ethnic minorities. When diet alone cannot effectively control the type 2 diabetic conditions, medical interventions, such as insulin injections or dispensing hypoglycaemic pills, are usually the next step of managing type 2 diabetes mellitus. Medical interventions notoriously exacerbate the fluctuation of blood glucose excursions. Even with the smallest dosage of hypoglycaemic drug (5 mg glucotrol or glyburide) once in the morning, the subject of this study still experienced frequent acute hypoglycaemias. Besides, his A1c levels hovered around 6.5 levels for many years following his physician's advice of taking 5 mg glucotrol per day. It became obvious that a properly designed drug dispensing regimen was needed to avoid hypoglycaemic bouts and effectively reduce A1c levels. Fasting blood glucose measurements are not consistent indicators, fluctuating widely from a low of 70 mg/dL to a high of 200 mg/dL (with most frequent range lay between 90 to 150 mg/dL) that were experienced by this type 2 diabetic subject prior to the model-based lifestyle adjustment. Initially, the subject tried to adjust lifestyle based on fasting glucose measurements, but it was not successful. His A1c measurements crept from 6.3 to 6.7 in a year. As glucose binds irreversibly to haemoglobin molecules within red blood cells, the amount of glucose that is bound to haemoglobin is directly tied to the concentration of glucose in the blood. The average life span of erythrocytes is about 120 days [ 11 ], measuring the amount of glucose bound to haemoglobin – by the A1c measurement – can provide an estimate of average blood sugar level during the 3 to 4 months period. It is obvious that A1c is a more reliable indicator than fasting glucose measurements for an effective blood glucose control self-management. It has been established that exercise can effectively alleviate diabetic conditions. Although no rigorous investigation has been performed here, nor is the focus of this current study, a forty-minute exercise of swimming, or weight lifting, or jogging, or any combination of these, prior to a meal or 3 to 4 hours after a meal, can significantly depress the volunteer's post-prandial blood glucose levels. However, it is impractical to substitute hypoglycemic pills with a multiple daily exercise schedule. A sensible lifestyle adjustment is required to manage the diabetic conditions without altering much of daily routines. Post-prandial blood glucose excursions (time series) for type 2 diabetes vary widely depending on the variety and the amount of food consumed. It also depends on long and short term physical conditions (exercise routines and stress levels such as insomnia) to a lesser scale. The recovery periods of blood glucose excursions returning to the pre-prandial level (or baseline) for diabetics are generally longer than those for non-diabetics. Although a simple glucose-insulin interaction compartmental model exists [ 12 ], not all the model parameters are readily interpretable. In addition, no case study is given to illustrate its potential applications. Compartmental models can provide first-order approximations that may be sufficient for specific goals. Simple models may not duplicate real phenomena but may reveal enough clues for which alternative approaches or experimental designs may come to light. A biophysically-based model of impulse-force-generated heavily damped oscillatory system is used here to capture the post-prandial blood glucose characteristics of type 2 diabetes. The model follows the general approach of glucose-insulin interaction model (bolus injection of glucose) with a few modifications, for which parameters can readily be interpreted and a case study is presented for exploring its potential applications. Rather than using single food items for their published GI values, or its cumbersome weighted mean of multiple ingredients in a meal, normally consumed lunch for the subject was used for the test meal. Based on the preliminary results obtained from the model, a moderate lifestyle adjustment was devised for the subject: swimming 20 laps for 40 minutes in a 25 m pool in the morning and dispensing 1/4 of 5 mg glyburide 1/2 to 1 hour before lunch and dinner – that enables him to reduce 10% of his A1c level in six months and maintain the desirable lower level for the subsequent six months. Methods The subject is a mid-sixty healthy male of 180 lbs with 5'10" frame, leading a productive professional life. He has been diagnosed with type 2 diabetes for more than 30 years. Initially, he was on diet regimen for nearly twenty years and then was instructed by his physician to dispense 5 mg glucotrol once every morning. He experienced frequent acute hypoglycemia that led him to discuss a possible self-managed regimen with his family physician. Lunch was chosen as the test meal for having sufficient time to take post-prandial measurements. The test meals were 15 sets of lunches that consisted either (1) 10 to 12 oz of steamed rice, stir-fried vegetables with 4 oz canned tuna (or steamed cod), or (2) 10 to 12 oz spaghetti with 6 medium sized meat balls (from Sam's family package). Five sets of data each were collected from: (i) without taking hypoglycemic pills before test meals; (ii) 1/4 size of 5 mg glyburide pills were dispensed pre-prandially right before the meal and (iii) 1/4 size of 5 mg glyburide pills were dispensed pre-prandially an hour before the test meals. One pre- and 8 to 12 post-prandial blood glucose measurements were taken at 30-minute intervals starting at the beginning of a meal (meal is usually consumed in 15 minutes): (i) for 6 hours, (ii) for 5 hours, and (iii) for 4 hours. In addition, for case (iii) two reference measurements were taken with one right before dispensing the pill and one an hour after completion of the 8 post prandial measurements, i.e ., at hour 5, for a total of 11 readings. The purpose of the first set of measurements was to establish the baseline for this diabetic subject: the recovery period of post-prandial blood glucose excursion without medication. The second and the third sets of the trials were designed to quantitatively measure the hypoglycemic drug effects and the most optimal time frame to administer the pills. Raw data were averaged and the corresponding standard deviations were also calculated for 5 replicates at given times. The averaged data were then used for modeling analysis. Model formulation The post-prandial blood glucose excursion can be considered as a hormone regulated resilient system. The food intake is treated as a bolus injection of glucose, and thus the impulse force f ( t ); effects of exercises and hypoglycemic medication are lumped as the damping factor, β . The differential equation of such an oscillatory system, that is used to describe post-prandial blood glucose excursions, can be found in many physics texts: where x represents blood glucose level over the baseline at time t , ω 0 is the system natural frequency [ 12 ]. The pre-prandial blood glucose levels are generally fluctuating with relatively insignificant magnitudes thus can be approximated as a flat level. If the impulse force f ( t ) takes the form of the Dirac delta function, F δ ( t -0) with F being a food intake dependent parameter, the solution of Eq. (1) is where is the frequency of the system. Equation (2) is a three parameter model: F , ω and β . Implications of these three parameters not only could reveal distinctive characteristics between diabetic and non-diabetic individuals but also provide guidelines to adjust one's lifestyle. Parametric estimation For a given blood glucose excursion, data was taken every 30 minute interval from the time a meal was initially consumed, from which the excursion peak ( MR ), x max , and the corresponding time τ to reach MR can both be estimated. Setting dx / dt = 0 in Eq. (2), the time τ can be expressed as: Substituting Eq. (3) into Eq.(2), we have The area under an excursion curve, AUC , can also be obtained: where T = 2 π / ω is the period of oscillation. The reason for setting the upper integral limit to T /2 is because the damping factor β effectively depresses the glucose excursion levels x near zero for t > T /2, i.e ., it ripples about pre-prandial level. The time T /2 is therefore defined as the recovery period ( RP ). For type 2 diabetic patients who are not in a properly structured regimen, the recovery periods are often longer than 5 hours, by which time the next meal arrives and induces another blood glucose upswing. Equations (3) – (5) can be used to estimate the three parameters, F , ω and β , from the measurable quantities of τ , x max , and AUC . The procedure is briefly described below: 1. Assign T as twice the roughly estimated recovery period in hours, which can be obtained from the raw data and thus ω = 2 π / T . 2. The damping factor β can be estimated from Eq. (3): , and thus . 3. The estimation of food intake-dependent impulse force F can be obtained from Eq. (4): . 4. Fine tune these three parameters by using MATLAB function fminsearch to minimize [ AUC data - AUC ( F , β , ω )] 2 , where AUC data is calculated from the averaged data points by the trapezoidal rule and AUC ( F , β , ω ) is calculated from Eq. (5). 5. These three parameters can further be fine-tuned by fminsearch (sum of squared errors between the averaged data points and the model predicted values). Two MATLAB user defined functions: GlucoseModel (for No pill and Pill at meal) and GlucoseModel1 (for Pill one hour prior) to estimate these model parameters and calculating the relevant diabetic characteristic measures: τ , x max , AUC are listed in the Additional files 1 and 2 , respectively. Results Table 1 lists the fine-tuned values of model parameters: F , ω , β , and those characteristic parameters: RP , τ , x max , and AUC , the latter three are calculated from Eqs. (3) to (5). Also included in Table 1 are the fitting statistics R 2 values that indicate how well model curves fit the data. Table 1 Model and characteristic parameters for the post-prandial blood glucose excursion Parameters No pill 1/4 pill at time 0 1/4 pill at time -1 F (mg/dL/hr) 47.1 73.8 59.3 ω (hr -1 ) 0.46 0.67 0.84 β (hr -1 ) 0.35 0.56 0.44 τ (hr) 2.60 1.76 1.56 RP (hr): π / ω 6.77 4.71 3.72 x max (mg/dL) 59.8 62.5 49.4 AUC (mg-hr/dL) 248 179 118 R 2 0.92 0.99 0.97 The parametric value of F is the result of food impact, or the rate of glucose being absorbed into the blood stream. The interpretation of F is rather difficult as the liver acts as a storage compartment for glucose [ 12 ]. Liver regulates blood plasma glucose levels; if it is too high, the excess will be stored in the liver, and the reverse process will take place if the plasma glucose is too low. Although all three model parameters: F , ω , and β are more or less influenced by the liver function, the impact on F deems more pronounced as it has a direct impact on the glucose levels in the blood stream. As the function of the liver is not included in the current model, the estimated F values can only be loosely inferred as a function of insulin level, F increases as hypoglycemic drug depresses the blood glucose levels that in turn increases the absorption rate of glucose into the blood stream as in the case of 1/4 pill taken right before the meal. When the drug is taken an hour before the meal, the liver may have sufficient time to regulate blood glucose levels that additional glucose absorption becomes less intensive. The increases of ω and β along with the intake of hypoglycemic drug are expected, which renders favorable characteristic parameters of τ , RP and AUC , all of these are decreasing with the moderate level of medication. The characteristic parameter x max has significantly depressed for the 1/4 size glyburide taken one hour before the meal while in the other two cases x max are roughly the same. This implies that the hypoglycemic drug has a net delay effect. Moderate hypoglycemic medication can enhance the liver function to regulate blood glucose levels, alleviating its fluctuation intensities. Interestingly, many ratios of characteristic parameters are roughly equal to constants for all three cases, which indicates that characteristic parameters are not mutually independent. Table 2 gives ratios of various combinations of characteristic parameters. The ratios of τ x max / AUC and are extremely attractive as both τ and x max can be estimated with fewer number of post-prandial measurements that one may use τ and x max to estimate more interpretable characteristic parameters of AUC and PR . Table 2 Ratio of characteristic parameters for the post-prandial blood glucose excursion Characteristic ratio No pill 1/4 pill at time 0 1/4 pill at time -1 τ x max / AUC 0.627 0.614 0.653 2.97 2.96 2.95 τ / RP 0.384 0.374 0.419 AUC / RP 36.6 38.0 31.7 AUC /( RP x max ) 0.612 0.608 0.642 No pill trial Parametric values for no-pill trial reveal that glucose absorption rate is generally slower (low F value) in comparison with the other two cases. The exceedingly long RP of nearly 7 hours is undesirable: as it implies that the next meal time arrives before the blood glucose level could return to the baseline, i.e ., an elevated blood glucose level would be sustained for a prolonged period of time. The high RP and AUC are unmistakably the characteristics for type 2 diabetes. Figure 1 compares the model and the data with the corresponding standard deviation bars. Model curves are extended for an additional hour beyond the last data point (and in all the figures herewith) to denote the trend of blood glucose excursion. Figure 1 Post-prandial glucose excursion: no pill trial 1/4 of 5 mg glyburide taken right before the meal The blood glucose characteristics are significantly improved with a 1/4 size of 5 mg glyburide taken right before lunch. Increased ω and β values translate to significantly lower RP and AUC with virtually unchanged x max . Although the mean RP is less than 5 hours, it is still a bit too long in comparison with the non-diabetics [ 12 ] (~ 4 hours). A higher F value than the one for no-pill trial may partly due to the liver intervention. Figure 2 compares the model and the data. From the figure one can tell that hypoglycemic drug has an effective delayed effect of about two hours as the rising portion of the model is almost identical to the one for no-pill trial with both x max are about 60, which may be the result of liver function that with initial stimulation of hypoglycemic drug, liver may also release glucose. As the hypoglycemic drug effect persists, the liver ceases to interfere. Figure 2 Post-prandial glucose excursion: 1/4 pill right before the meal 1/4 of 5 mg glyburide taken an hour before the meal From the personal experience of the participating subject, the hypoglycemia usually occurs 3 to 4 hours after taking the pill. The trial described in the previous section also reveals that no significant hypoglycemic drug effect is detected in the initial two hours. In order to learn the drug impact on an empty stomach, an additional glucose measurement was made prior to taking the hypoglycemic pill at -1 hour. Another measurement was also taken an hour after the blood glucose excursion returned to the baseline ( i.e ., at hour 5). This is meant to check if the blood glucose would remain near the baseline level. The drop of blood glucose levels between -1 and 0 hours are roughly 10 mg/dL, which can be contributed to the mild liver intervention. No net hypoglycemic drug effect is taking place before the meal as evidenced from the initial rise of the blood excursion curve as shown in Fig. 3 (in comparison with Fig. 2 ), where only data between hour 0 and hour 4 were used to generate the model curve. Indeed, all parametric values are improved significantly: both PR and x max are decreased by 20% and their combination that reflected in AUC dropped nearly 35% in comparison to those for pill taken at meal trial as shown in Table 1 . The food impact parameter F decreased a little from the one for pill at meal trial, which may indicate an hour after dispensing the pill, a quasi-equilibrium state has been reached among the liver function, hypoglycemic drug effects, and the bolus injection of glucose. The system frequency ω increased for more than 25%, which gives a shorter RP that compares favorably with non-diabetics. The drop of damping factor β may be the result of low F , as both τ and x max are already significantly reduced that further strengthening of β becomes unnecessary. The hour 5 measurements confirm that although the model curve shows a decreasing trend, upon returning to the base level the blood glucose excursions practically stabilizes. In addition, the volunteer patient did not experience any hypoglycemia even two to three hours after the final post-prandial measurement. Figure 3 Post-prandial glucose excursion: 1/4 pill an hour before the meal Discussion This simple impulse-forced model provides a means to shape a self-management regimen for the type 2 diabetic subject: a moderate meal coupled with minimal amount of medical intervention has effectively modulated the blood glucose excursion by reducing its recovery periods and fluctuation amplitudes. Based on the model, the type 2 diabetic subject was able to adjust a lifestyle that include (a) 40 minute swimming in a 25 m pool in the morning, (b) a fruit of mid-size apple or its equivalent and a cup of coffee with cream for breakfast without taking hypoglycaemic pill, (c) moderate lunch with 1/4 size of 5 mg glyburide taken 1/2 to 1 hour before the meal, (d) moderate early dinner, 4 hours prior to bed time, with 1/4 size of 5 mg glyburide taken 1/2 to 1 hour before the meal, (e) snack a mid-size banana, or a small bag (3.5 oz) of peanuts, or 6 crackers when needed in between meals. With this regimen, he was able to reduce his A1c level from 6.7 to 6.0 in 6 months and maintained at this level for the subsequent 6 months. Moreover, he has not had any hypoglycaemic bouts ever since he particitipated in this study more than two years ago. Elevated blood glucose excursions during the night would boost the A1c levels. To keep a low average fluctuation of blood glucose excursion amplitudes, the evening meal is crucial. In order to avoid hypoglycaemia during the sleep, an early dinner is advised. The subject has been able to keep post-prandial blood glucose levels within 200 mg/dL with the mean fasting reading of 90 ± 20 mg/dL. Occasionally he consumes a can of beer or sugar free deserts. Although no rigorous study has been performed, a forty-minute exercise of swimming, or weight lifting, or jogging, or any combination of these is roughly equivalent to the effect of 1/4 size of 5 mg glyburide. Nonetheless, it is impractical to exercise more than once a day, thus the subject takes 2.5 mg of hypoglycemic pill a day instead. His physician originally prescribed him to take one 5 mg hypoglycemic pill daily. That was more that 10 years ago. The regimen did not work very well as he experienced hypoglycaemic bouts often. This model-based regimen not only reduced A1c level but entirely eliminated hypoglycaemic symptoms. In addition, one fasting blood glucose measurement in the morning is sufficient for him to maintain a healthy daily routine of exercise, consuming meals/snacks and leading a productive life with mental and physical activities. Conclusions Lifestyle adjustments are the best regimens for many chronicle ailments such as diabetes, hypertension, high cholesterol levels, etc . Although this model-based self-management regimen for the type 2 diabetic subject is only a case study, it certainly provides a general guideline for an applicable life-style adjustment. Currently not all the model parameters are entirely clear, additional data are required to draw a meaningful general conclusion. A pilot project of testing this regimen on six type 2 diabetic patients in a regional nursing home is proposed for the next phase of study. If future studies support that the ratios of τ x max / AUC and are approximately constants, the combination of τ and x max can then be used to estimate AUC and PR with fewer number of post-prandial measurements. This would be much more convenient to characterize a type 2 diabetic subject than using AUC and PR . Although derived characteristic parameters: RP and AUC (to a lesser degree, τ and x max ), carry clear meaning that can be used to characterize type 2 diabetic subjects from non-diabetics, the implications of model parameters, F , ω and β are not as translucent. With additional data, one may be able to draw plausible conclusions about (a) how F is influenced by food intakes, drug (delaying) effects, and liver (regulatory) functions; and (b) how ω and β behave, whether they are independent of F and of each other, or all three somewhat mutually dependent. Better understanding of these parameters would definitely enhance the self-management for type 2 diabetes. This model-based lifestyle adjustment has another advantage: it can be used to manage each individual needs. Nutall and Chasuk [ 10 ] have stressed that dietary recommendation for type 2 diabetes should be flexible and highly individualized; most of prepared meal programs and exchange-list diets for diabetes have not had individualization in mind nor are they designed for ethnic minorities. Once we have a comprehensive understanding of these parameters, it is possible to tailor individual lifestyle adjustment accordingly. For those individuals who are interested in self-managing the type 2 diabetes, the general advice is: avoiding big meals, may snack moderately between meals, eat an early dinner – about 4 hours before bedtime, and exercise regularly. If one is interested in "normal" meal effects on one's post-prandial blood glucose excursion, taking a pre-prandial blood glucose measurement prior to a typical lunch and 8 to 10 post-prandial measurements at half-hour intervals for 5 or more replicates and follow the procedure described here to obtain these characteristic parameters RP , τ , x max , and AUC . Applying a small dosage of medical intervention prior to a meal can keep the blood glucose at a relatively flat level and depress the overnight blood glucose excursion; however, this practice needs the approval from one's family physician and is not recommended here. Authors' contributions Sole authorship: data collection/analysis, model building, parameter estimation/interpretation, and the design of life-style adjustment regimen for the participating subject. Supplementary Material Additional File 1 MATLAB user defined function: GlucoseModel (for No pill and Pill at meal) to estimate model parameters: F , β , ω and to calculate the relevant diabetic characteristic measures: τ , x max , AUC . Click here for file Additional File 2 MATLAB user defined function: GlucoseModel1 (for Pill one-hour prior) to estimate model parameters: F , β , ω and to calculate the relevant diabetic characteristic measures: τ , x max , AUC . Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC546416.xml |
552305 | International outbreak of Salmonella Oranienburg due to German chocolate | Background This report describes a large international chocolate-associated Salmonella outbreak originating from Germany. Methods We conducted epidemiologic investigations including a case-control study, and food safety investigations. Salmonella ( S .) Oranienburg isolates were subtyped by the use of pulsed-field gel electrophoresis (PFGE). Results From 1 October 2001 through 24 March 2002, an estimated excess of 439 S . Oranienburg notifications was registered in Germany. Simultaneously, an increase in S . Oranienburg infections was noted in other European countries in the Enter-net surveillance network. In a multistate matched case-control study in Germany, daily consumption of chocolate (matched odds ratio [MOR]: 4.8; 95% confidence interval [CI]: 1.3–26.5), having shopped at a large chain of discount grocery stores (MOR: 4.2; CI: 1.2–23.0), and consumption of chocolate purchased there (MOR: 5.0; CI: 1.1–47.0) were associated with illness. Subsequently, two brands from the same company, one exclusively produced for that chain, tested positive for S . Oranienburg. In two other European countries and in Canada chocolate from company A was ascertained that also contained S . Oranienburg. Isolates from humans and from chocolates had indistinguishable PFGE profiles. No source or point of contamination was identified. Epidemiological identification of chocolate as a vehicle of infections required two months, and was facilitated by proxy measures. Conclusions Despite the use of improved production technologies, the chocolate industry continues to carry a small risk of manufacturing Salmonella -containing products. Particularly in diffuse outbreak-settings, clear associations with surrogates of exposure should suffice to trigger public health action. Networks such as Enter-net have become invaluable for facilitating rapid and appropriate management of international outbreaks. | Background Non-typhoidal Salmonella spp. is a substantive cause of human gastroenteritis in many parts of the world [ 1 ]. In Germany, non-typhoidal salmonellosis remains the most frequently reported infectious disease. For example, in 2001, the 77,185 Salmonella reports (incidence:94/100,000) received at the federal level by the Robert Koch-Institut (RKI) accounted for 31% of all notifications for the 54 notifiable conditions [ 2 ]. Salmonella enterica subspecies enterica serotype Enteritidis ( S . Enteritidis) is the predominating serotype followed by S . Typhimurium. They represented 65% and 23% of the reported cases of non-typhoidal salmonelloses with known serotype in 2001. Thus, the remaining ~250 serotypes reported to the RKI in that year, including S. Oranienburg, accounted for only 12%. In mid-October 2001, the National Reference Center for Salmonella and Other Enteric Pathogens (NRC) in Hamburg noted an unusual increase in the number of S . Oranienburg isolates received in October. At that time, no increase was noticeable in the national database for statutorily reportable infectious diseases; 50 S . Oranienburg notifications (median: 1 per week) had been registered for 2001. On November 19, the NRC in Wernigerode informed the RKI that it had received a S . Oranienburg isolate in September. The isolate was submitted by a private laboratory for serotyping and had come with the additional source information "confectionery sample". Upon inquiry, a large German chocolate manufacturer (company A), which produced a broad variety of chocolates and products made thereof, called the RKI on November 27, and confirmed that it had sent in the confectionery sample. According to company A, the positive sample originated from an in-house control of a chocolate product and the pertaining batch, due to be exported to the United States, was completely destroyed and not distributed. Notwithstanding, the number of statutory S . Oranienburg notifications had sharply increased and continued to rise. This report describes the epidemiologic, food safety, and microbiological investigations of this outbreak. Methods Epidemiologic investigation Descriptive epidemiology A standard exploratory questionnaire was distributed on 20 November 2001 via state health departments to all local health departments to aid the collection of data on food and environmental exposure from cases. In addition, local health departments were asked to immediately interview patients with newly reported S . Oranienburg infections about chocolate consumption in the seven days before disease onset, and also to send any remaining chocolate to a food safety laboratory. International case-finding A request was distributed to participants of the Enter-net surveillance network [ 6 ] on December 10, to see if other countries were affected or had relevant information. Case-control study On December 3, while exploration of patients were ongoing and results inconclusive, S . Oranienburg isolates from patients and from the in-house chocolate control were found to be indistinguishable by pulsed-field gel electrophoresis (PFGE). On the same day, a multistate case-control study was initiated and coordinated by RKI to test the hypothesis that at least one product from company A was associated with S . Oranienburg-infections. As we were denied a product list from company A, we resorted to the company's web-site and included in our food history evaluation all the products listed there. Some products from company A, e.g., bars of chocolate (brand A), were exclusively sold at a large chain of discount grocery stores (chain X). We found that the majority of chocolates sold at chain X were produced by company A. Therefore, for the analysis we constructed a variable for chocolate(s) purchased at chain X ("chain-X-chocolate") as a proxy for chocolate-products from company A because most patients could remember the flavor of the purchased chocolate, but seldom the brand name. The hypothesis-testing questionnaire collected data on the consumption of chocolates, and some other foods, particularly those previously associated with outbreaks of S . Oranienburg in other countries [ 3 - 5 ]. Food-consumption history was evaluated for two different time periods, i.e., for the seven days prior to onset of symptoms in the case-patients and for the seven days before the interview. In addition, case-patients were asked about clinical symptoms, duration of illness and hospitalization. We defined a case-patient as a person with gastroenteritis starting after 1 October 2001 who had been reported with a S . Oranienburg infection to a public health department before December 6. Case-patients were excluded from the analysis, if they could have been secondary, i.e., if they reported to having had contact with a person with diarrhea in the seven days prior to symptom onset. Cases were selected from the national reportable database by simple random sampling; in Lower Saxony an attempt to interview all case-patients was launched. Case selection was done irrespective of whether patients also had been interviewed with an exploratory questionnaire. At least one age and telephone exchange-matched control subject was selected for each case-patient by sequentially adding 2 to each case-patients' telephone number. Control subjects were eligible if they were in the same age-group as their matched case-patient (0–5 years, 6–17 years, 18–59 years, 60 years or older), had no gastrointestinal symptoms after 1 October 2001, and had not traveled abroad in the seven days prior to the onset of symptoms of the matched case. Telephone interviews were conducted by state health departments, local health departments, and the RKI. Data were analyzed with Epi Info V6.04c (Centers for Disease Control and Prevention, Atlanta, GA). Investigation by the food safety authorities The local food safety authority inspected company A's production facility and took samples from already packaged ("in-house") chocolates, and from ingredients from its suppliers. Beginning December 11, a nationwide chocolate sampling of German chocolates in grocery stores was initiated by the Federal Ministry of Consumer Protection, Food and Agriculture, and was assisted by the Federal Institute for Risk Assessment (Bundesinstitut für Risikobewertung, "BfR"). On December 18, when a chocolate leftover from brand A tested positive, the investigations were tailored to German chocolates from company A. The BfR examined quantitatively four Salmonella positive chocolate leftovers and five chocolates from grocery stores using the most probable number technique [ 7 ]. Molecular subtyping For comparison by the use of PFGE, S . Oranienburg isolated from stool specimens were sent to the NRC from laboratories in Germany and, on Enter-net request, from other countries. Furthermore, isolates from chocolates were submitted from state or private food laboratories in Germany as well as from Canada and the Czech Republic. PFGE-analysis was carried out according to Prager et al [ 8 ]. Results Epidemiologic investigation Descriptive epidemiology In 2001, the RKI received 50 reports of S . Oranienburg up to reporting week 42, but 462 reports in the following 23 weeks (15 October 2001–24 March 2002, "outbreak period", figure 1 ). Thus, an excess of 439 S . Oranienburg reports were registered assuming a background rate of one report per week. The median age was 15 years (range: 0–92 yrs), 240 (52%) patients were female. There was no difference in the gender distribution within the single 10-year age-bands ( P = .51). All 16 states of Germany reported S . Oranienburg cases during the outbreak period, with the highest incidence in the state of Schleswig-Holstein (1.78/100,000) bordering on Denmark. In total, 206 of the 440 German counties were affected with a median of one report and a maximum of 16 from the city of Hamburg during the outbreak period. Figure 1 Disease onset (n = 362) of reported (n = 462) S . Oranienburg cases from reporting week 42/2001 to reporting week 12/2002 (outbreak period). The asterisk indicates the week when the (first) public warning was issued, and the incriminated chocolate products were recalled Sixty exploratory questionnaires were received from eight states by the end of 2001. Forty-three (88%) of 49 patients with information on chocolate consumption had a symptom onset after 1 October 2001. Of the 34 who gave information as to where they bought the chocolate, 21 (62%) explicitly reported chain X. Some reported exclusively having eaten chocolate bars from brand A, among them a two-year-old child. On 18 December 2001, two months after the initial outbreak alert, a leftover consumed reportedly by this child in the seven days before symptom onset tested positive for S . Oranienburg. International case-finding On December 11, one day after the Enter-net request was distributed, Denmark was the first country to respond. Twelve cases of S . Oranienburg had been reported in Denmark from October 18 through December 10, compared with only two cases in 2001 before October 18. None of the clustered cases were travel-related [ 9 ]. Exploratory patient interviews had already been conducted at the time of the Enter-net request. At this point in time the investigators in Denmark, without knowledge of the German S . Oranienburg problem, independently suspected German chocolate bought in chain X as the source of the Danish outbreak. Chocolate was the only food item that all patients reported eating. The majority stated purchasing chocolate in chain X, which, although German, operates internationally [ 9 ]. In the next few days, an increase in the number of S . Oranienburg infections was reported from other countries such as Austria, Belgium, Finland, Sweden, The Netherlands (figure 2 ), and Canada [ 10 ]. As it became apparent that German chocolate was contaminated with S . Oranienburg, patient interviews were conducted that showed that several patients remembered having consumed German chocolate [ 10 ], except in Canada where all of the patients denied this consumption. Figure 2 Number of S . Oranienburg infections reported to the Enter-net database from participating countries, except Germany Case-control study Sixty cases and 62 controls from five states were enrolled in the matched case-control study. Interviews were conducted with a median delay of 37 days (range: 12–64 days) after disease onset in case-patients. Twelve case-control pairs were excluded from chocolate-specific analysis, nine because the case-patients could have been secondary, one due to illness in September, and two where the control subjects could not remember whether they had eaten chocolate. Of the 48 cases and 50 controls that were analyzed, 24 (50%) case-patients and 32 (67%) control subjects were female, 22 (46%) case-patients were younger than 10 years. Ten (21%) of the case-patients reported to have suffered from bloody diarrhea and 14 (29%) were hospitalized (table 1 ). Results of the preliminary analysis were available on December 14. All 48 case-patients ate chocolate in the seven days before symptom onset, but this also applied to 43 (86%) of the control subjects. Three variables relating to the seven-day period prior to symptom onset of the case-patient were significantly associated with disease (table 2 ). The first variable was having shopped at chain X (matched odds ratio [MOR]: 4.2; 95% confidence interval [CI]: 1.2–23.0). The second variable was having consumed chain-X-chocolate (MOR: 5.0; 95% CI: 1.1–47.0). Eleven (25%) of 44 case-patients gave such an exposure history, six of whom reported having consumed either brand A chocolate exclusively (which could be inferred from the flavor of chocolate eaten), or were uncertain whether they had also eaten another kind of chocolate (n = 3). The third variable was having eaten (any kind of) chocolate on a daily basis (MOR: 4.8; 95% CI: 1.3–26.5). None of the other variables including all those relating to the seven days before the interview were significantly associated with illness. Table 1 Clinical characteristics of S . Oranienburg cases (n = 48) analyzed in a case-control-study, December 2001 Symptoms Frequency, n (%) Diarrhea 41 (85) Fever > 38,5°C 27 (56) Vomiting 17 (35) Hospitalization 14 (29) Antimicrobial medication 12 (25) Visible blood in stool 10 (21) Table 2 Significant risk factors for S . Oranienburg-associated illness in Germany, October-December 2001 Exposure Cases exposed (n/N, %) * Controls exposed (n/N, %)* MOR Exact 95% CI P -value Ate chocolate bought at chain X 11/44 (25) 2/45 (4) 5.0 1.1, 47.0 0.04 Daily consumption of chocolate 22/48 (46) 12/50 (24) 4.8 1.3, 26.5 0.01 Shopped at chain X 31/44 (71) 19/45 (42) 4.2 1.2, 23.0 0.03 MOR = Matched odds ratio, CI = Confidence interval * Proportion and percentages of cases and controls exposed ignoring matching Public health action On December 18, the finding of S . Oranienburg in a chocolate leftover of a patient led to an immediate public warning and recall of all chocolates of this brand with specific production numbers by company A. The recall was extended to other products from company A a few days later. Chocolates included in the German recall were promptly withdrawn from the market in other European countries as well as in Canada. In Canada, Finland, and Sweden, samples from withdrawn chocolates tested positive for S . Oranienburg [ 10 ]. Investigation by the food safety authorities The local food safety authority in Germany did not identify hygienic deficiencies at the production facility. Samples obtained in the beginning of December 2001 from in-house chocolates (n = 12), as well as from cocoa (n = 3) and cocoa powder (n = 7) from a supplier of company A tested negative. This applied also to German chocolates sampled in grocery stores until 18 December (on that day a leftover tested positive). Overall, S . Oranienburg was found in 18 (5%) of 381 chocolates that were tested and reported to BfR during the outbreak period. S . Oranienburg was isolated from two different brands of company A; all positive chocolates were produced during the same week in August 2001. Estimates of the number of Salmonella in the tested chocolates ranged between 1.1 and 2.8 per gram. Molecular subtyping From October 2001 through January 2002, the NRC received 98 S . Oranienburg isolates from human cases of gastroenteritis originating in Germany (n = 52), Austria (n = 19), Belgium (n = 8), Canada (n = 6), Denmark (n = 4), The Netherlands (n = 4), Sweden (n = 4), and the Czech Republic (n = 1). Furthermore, 15 chocolate isolates were sent to the NRC for PFGE-analysis from Germany (n = 12), Canada (n = 2), and the Czech Republic (n = 1). They came from an in-house sample, from leftovers of chocolates consumed by patients in their incubation period, and from chocolates sampled in grocery stores in Germany. The PFGE profiles of S . Oranienburg isolates from patients with symptom onset after 1 October 2001 (outbreak period) in Germany and in the other countries mentioned above, except Canada, were indistinguishable (figure 3 ), but differed from S . Oranienburg isolates from German patients with symptom onset before October. All 15 chocolate isolates showed PFGE profiles indistinguishable from human isolates of the outbreak period. Figure 3 Comparison of human S . Oranienburg isolates from the outbreak- period with strains of this serovar received sporadically before the outbreak by the use of PFGE (digested with Xba I, Bln I, and Spe I) lanes: 1–5: isolates from the outbreak period 6–13: isolates before the outbreak period S: molecular reference Discussion We describe an international outbreak of S . Oranienburg and present several lines of evidence that German chocolate from company A was the vehicle of infections. S . Oranienburg, a rare serotype in food as well as in humans in Germany, was isolated from retail-sampled chocolates of two brands produced by company A, from chocolate leftovers that had been consumed by patients before symptom onset, and from an in-house sample of company A obtained prior to the outbreak. In a case-control study, S . Oranienburg infection was significantly associated with the consumption of chain-X-chocolate (proxy for chocolate from company A) in the week prior to symptom onset, but not in the seven days before the interview. Case-patients were more likely than control subjects to report eating chocolate daily, likely indicating an increased probability of having been exposed to contaminated chocolate. Furthermore, patient isolates from the outbreak period shared a PFGE profile with isolates from chocolates but differed from isolates of patients who became sporadically diseased with S . Oranienburg before the outbreak. In addition, the food histories and microbiological results from S . Oranienburg patients in several other countries pointed to the same source [ 9 , 10 ]. Salmonella infections after consumption of contaminated chocolate, although rare, have been known since the 1960's [ 11 ]. Common to all reported chocolate-outbreaks, including ours, was that the epidemics were propagated in time, widely disseminated geographically, and affected large number of persons, predominantly children [ 12 - 17 ] (table 3 ). In addition, only very small numbers of Salmonella have been recovered from chocolates in these outbreaks, suggesting a very low infectious dose. Estimates of the number of S . Oranienburg cells per gram in this outbreak ranged from 1.1–2.8. However, we cannot exclude that the bacteria were unevenly distributed in the chocolate(products) and that those parts carrying many viable cells were not tested quantitatively. In chocolate, the low moisture (water activity a w : 0.4–0.5) and high sugar content does not favor bacterial growth, but significantly increases thermal resistance [ 11 , 18 , 19 ]. In addition, it has been speculated that the food matrix protects Salmonella against the acidic conditions of the stomach [ 11 ], which could imply that only few salmonellae are necessary to cause illness. Table 3 Overview of published chocolate outbreaks due to Salmonella contamination Year Country Serovar Vehicle* Source of contamination cfu/g No. of affected persons Peak of outbreak Age of cases 1970 Sweden S . Durham Chocolate products (n>1), Cocoa powder / 110 Dec-May 53% ≤15 years 1973 – 1974 USA, Canada S . Eastbourne Chocolate balls from Canada Cocoa beans 2.5 200 Dec-Feb. 3 years (median) 1982 England, Wales S . Napoli Chocolate bars from Italy Unknown 2–23 272 May-Aug 58% ≤ 15 years 1985 – 1986 Canada S . Nima Chocolate coins from Belgium Unknown / / Dec-Jan ? 1987 Norway, Finland S . Typhimurium Chocolate products, (n = 3) from Norway Avian contamination speculated ≤1 349 Mar-May 6 years (median) 2001– 2002 Germany, other European countries S . Oranienburg Two chocolate brands from Germany Unknown 1.1–2.8 439 Oct-Dec 15 years (median) * In each outbreak, the identified vehicles were traced to a single manufacturer Company A produced several dozen tons of chocolate per day. All positive samples were produced in the same week. However, S . Oranienburg reports above an expected baseline of 1–2 reports per week in Germany were received for five months. The protracted nature of chocolate-associated outbreaks probably reflects both the long shelf-life of chocolate [ 20 ] and the long survival of Salmonella in these products [ 15 , 21 ]. S . Oranienburg was isolated from chocolates five months after manufacture. In an S . Napoli outbreak in England and Wales, this interval was 12 months. The number of affected persons reported in chocolate-associated Salmonella outbreaks has grown steadily over the years (table 3 ). Among other factors, this may parallel advances in food-processing technologies and improvements of national surveillance systems. Taken together, the chocolate industry faces a difficult situation because: raw ingredients (e.g., cacao beans, milk powder) can carry Salmonella spp., the low water activity and high fat content in chocolate increases thermal resistance so that temperatures reached during chocolate production (even after considerable overheating [ 19 ]) do not necessarily destroy Salmonella , a small number of Salmonella may be sufficient to cause disease, even with low-level contamination, chocolate can affect large number of persons (often children) scattered over a wide area, and thus, has the potential to cause serious public health consequences. It is noteworthy that the case-control study did not identify further products as risk factors. This applied also to the second contaminated brand of company A, which was included in the recall. The small proportion of study cases (25%) mentioning having eaten chain-X-chocolate lends support to the hypothesis of more contaminated brands (even from other manufacturers), which could be one explanation for the continuing case-occurrence. However, inaccuracies due to lack of brand awareness may have played a particular role in this outbreak, and the time-delay between disease onset and interview (median: 37 days) may have contributed to an inaccurate recall of cases and their guardians. Furthermore, cases with a disease onset in 2002 may have occurred as a result of a diminished impact of the public warning due to the Christmas season. For example, chocolate gifts received or given for Christmas may not have been thoroughly enough checked for best-before dates stated in the public warning. Identification of vehicles in foodborne outbreaks can become difficult if the exposure is common. Consumption of a wide variety of German chocolates was reported by all case-patients (and 88% of explored patients), but also from 86% of the control subjects in the week prior to onset of illness. When groups are (nearly) universally exposed or a more specific hypothesis cannot be tested, often the best one can do is to establish a "dose-response-relationship" [ 22 ], i.e., unravel differences in the frequency of consumption of the incriminated food between cases and controls. Consequently, the observation that a higher proportion of cases reported eating chocolate on a daily basis added to the evidence that chocolate was the vehicle in this outbreak. Furthermore, the Danish data provided powerful supplementary evidence because consumption of German chocolate was particularly common in Germany but unusual in Denmark. Therefore, in multinational outbreaks, international collaboration provides an important means for disclosing the common source of infections, particularly when the contaminated food is very popular in one (likely the source) country (e.g., [ 23 , 24 ]). Multinational collaboration facilitated by Enter-net helped in preventing contaminated chocolate from entering the market in Canada, Finland and Sweden, thereby averting human illness. Furthermore, by rapid electronic exchange and comparison of PFGE profiles, the Canadian cluster of human cases could be classified as unrelated to this outbreak. No source or point of contamination was identified. Hygienic deficiencies had not been observed at the production facility of company A, which used a modern production method. This included an extra heating of the milled cocoa beans by a special heat-steam treatment with 125–130°C as an additional safeguard. Samples from in-house chocolates and from ingredients tested negative. However, no environmental samples and very few samples of raw ingredients (n = 10) were obtained. In a S . Eastbourne outbreak in Canada/USA in 1973/74, 286 environmental samples and 98 chocolate samples from the production-line were examined. No in-line chocolate sample tested positive and overall only 6 (1.6%) samples were positive (bean processing rooms [n = 4], and samples from a molding plant [n = 2]] [ 12 ]. Therefore, source investigations in chocolate-outbreaks should include extensive sampling in the production environment to increase the likelihood of determining possible points of contamination. In this outbreak, it remains unclear whether the salmonellae survived the heating or (re)contaminated the chocolate afterwards. Consequently, long-term preventive measures to render chocolate-production safer could not be implemented. An Enter-net urgent inquiry was sent after the first results of molecular subtyping suggested a link between human cases and chocolate from company A. Until then, investigators in Germany and Denmark had worked independently unaware that the outbreak extended outside of their respective countries. An earlier inquiry, ideally as early as an outbreak was suspected by the investigating countries (or as an increase was noted in the Enter-net database), may have speeded up hypothesis generating, and thus, may have helped in earlier identification of the vehicle, thereby preventing illnesses. Finally, a public warning or recall of company A products did not occur before a brand A leftover tested positive although the confluence of information – the results of the case-control study, the Danish investigations, and the subtyping comparison between human isolates and the in-house sample – had already pointed to company A products as the source of the outbreak. Yet, no specific product or lot had been identified at the time. For this reason, a recall or a public warning were considered excessive responses by the German food safety authority. However, relying on microbiological confirmation in leftovers, if available for testing, is disputable (directionality of contamination unclear) and is dangerous in unopened food packages because critical time can elapse before a positive culture in food is obtained [ 25 ]. Therefore, it has been argued that public health action should be based on well-performed epidemiological investigations encompassing clear statistical associations with a specific exposure [ 25 - 27 ]. Such data are easiest to obtain when only one (ideally distinct) vehicle is involved that is infrequently consumed. Nonetheless, when food-production leads to more than one contaminated foodstuff, or when popular foods are vehicles of infection, hypothesis generating or testing can become intricate. Unfortunately, these instances appear conducive to affect large areas/populations. Therefore, we believe that clear associations even with surrogates of exposure suffice to justify public health actions (e.g., extensive source investigations) provided they plausibly fit other lines of evidence. Conclusions To our knowledge, this is the largest reported chocolate-associated outbreak, the seriousness being emphasized by the hospitalizations (29%) and self-reported bloody diarrhea (21%) of the study cases. Despite the use of improved production technologies, the chocolate industry continues to carry a small risk of manufacturing Salmonella -containing products. For the future, awareness among German food safety authorities must be heightened for the need to base public health action not exclusively on laboratory confirmation in food, and to conduct timely and comprehensive source investigations to enhance food safety in the long-run. The international scale of this outbreak shows how easy it is to distribute a contaminated product across many countries. This underlines the necessity of mechanisms for international surveillance and information dissemination such as Enter-net to ensure that international outbreaks can be dealt with rapidly and in an appropriate manner. Similar networks should be set up or, if existing, should be connected (possibly overseen by WHO), to allow rapid communications to other parts of the world when it is clear that a contaminated product is distributed internationally. Competing interests The author(s) declare that they have no competing interests. Authors' contributions DW was the principal investigator of the German part of this outbreak; he carried out the statistical analysis of the case-control study, and drafted the manuscript. JD and FF were responsible for the outbreak investigation including the case-control study for the state ("Bundesland") of Lower Saxony. UvT was responsible for the outbreak investigation including the case-control study for the state ("Bundesland") of Northrhine-Westfalia. GF was responsible for the outbreak investigation including the case-control study for the state ("Bundesland") of Hamburg. SE conducted the Danish part of this outbreak investigation. AMH was responsible for the outbreak investigation including the case-control study for the state ("Bundesland") of Hesse. PR detected the outbreak and conducted microbiological investigations. RP and HT conducted the PFGE-analysis. ISTF coordinated the Enter-net inquiries and investigations. SB conducted interviews in the case-control study, designed the database and helped in the analysis. EB conducted quantitative analysis of Salmonella in chocolate. EW coordinated food safety investigations in this outbreak. AE conducted the Canadian part of this outbreak investigation. AS conducted the Finnish part of this outbreak investigation. YA conducted the Swedish part of this outbreak investigation. MHK was instrumental in the design of the case-control study. AA coordinated the German part of this outbreak investigation, broadened the scope of this outbreak by prompting an urgent Enter-net inquiry, and helped in designing the case-control study and drafting the manuscript. All authors participated in revising the manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC552305.xml |
544567 | Covariates of corticotropin-releasing hormone (CRH) concentrations in cerebrospinal fluid (CSF) from healthy humans | Background Define covariates of cerebrospinal corticotropin-releasing hormone (CRH) levels in normal humans. CRH CSF was measured in 9 normal subjects as part of an intensive study of physiological responses stressors in chronic pain and fatigue states. CRH CSF was first correlated with demographic, vital sign, HPA axis, validated questionnaire domains, baseline and maximal responses to pain, exercise and other stressors. Significant factors were used for linear regression modeling. Results Highly significant correlations were found despite the small number of subjects. Three models were defined: (a) CRH CSF with blood glucose and sodium (explained variance = 0.979, adjusted R 2 = 0.958, p = 0.02 by 2-tailed testing); (b) CRH CSF with resting respiratory and heart rates (R 2 = 0.963, adjusted R 2 = 0.939, p = 0.007); and (c) CRH CSF with SF-36 Vitality and Multidimensional Fatigue Inventory Physical Fatigue domains (R 2 = 0.859, adjusted R 2 = 0.789, p = 0.02). Conclusions Low CRH CSF was predicted by lower glucose, respiratory and heart rates, and higher sodium and psychometric constructs of well being. Responses at peak exercise and to other acute stressors were not correlated. CRH CSF may have reflected an overall, or chronic, set-point for physiological responses, but did not predict the reserves available to respond to immediate stressors. | Background Corticotropin – releasing hormone (CRH) plays a major role in regulating the hypothalamic – pituitary – adrenal (HPA) axis, acute responses to stressors, and other neurological functions [ 1 ]. The cerebrospinal concentrations of CRH (CRH CSF ) and neuropeptide Y (NPY), another important neuropeptide involved in pain, autonomic and stress responses [ 2 , 3 ], were measured in 9 normal humans to better understand these functions. Subjects were studied as part of a large scale investigation of subjective and objective (functional magnetic resonance imaging) responses to pain testing, exercise, and other stressors [ 4 ]. Subjects completed questionnaires that assessed pain, anxiety, depression, coping skills, and other psychometric variables. Normal control, fibromyalgia, [ 5 ], chronic fatigue syndrome [ 6 ], and veterans of the first Persian Gulf War [ 7 ] were included. This pilot investigation focuses on the statistical strategy used to analyze the control population (n = 9), and to develop methods for effective evaluation of patient populations. Three multiple linear regression models [ 8 ] were defined that predicted CRH CSF based on (a) metabolic, (b) autonomic, and (c) psychometric measures. Results Demographics The mean age for the group was 35.2 yr (95% C.I.: 30.5 to 39.9) years. There were 2 females, 5 African-Americans, 3 Caucasians, and 1 Caucasian Hispanic (TABLE 1 ). These subjects were a very normal and healthy group based upon their histories and the strict exclusion criteria. NPY CSF Mean NPY CSF was 121.9 (87.8 to 156.0) pg/ml. NPY did not correlate with any variable. Variables correlated with CRH CSF Significant covariates of CRH CSF could be grouped into: metabolic, autonomic function, and perceptional and cognitive functions (TABLE 3 ). Serum glucose was positively, and sodium negatively, correlated with CRH CSF (FIGURE 1 ). Explained variances (R 2 ) were 0.97 and 0.66, respectively. Glucose and sodium were also negatively correlated (R 2 = 0.85). Resting norepinephrine levels at 2 time points and heart rate at 4 time points were colinear and positively correlated to CRH CSF . CRH CSF was negatively correlated with the Holter monitor-derived measure of log heart rate summed for the daytime, and the threshold temperature causing an initial, mild sensation of burning pain (Stressor I). The latter indicated that subjects with higher CRH CSF perceived the burning pain of the cutaneous forearm hyperthermic stimulation at a lower threshold temperature than their peers. This may indicate an increased sensitivity to nociceptive stimuli. However, there was no correlation with deep pressure – induced pain. Negative correlations were also found with the SF-36 Vitality and SES Manage Symptoms domains. High scores were normal for these questionnaires, with lower scores indicating dysfunction. Effects in males All parts of the study were completed by at least 5 males. Analysis of the male subgroup gave some information about the role of gender. The pattern of significant covariables was different from the total group (TABLE 4 ). Respiratory rate was the only vital or physiological sign related to CRH CSF . Perceptions of vulnerability, physical functioning and self-efficacy were more highly related. Statistical models Sets of the significant metabolic, autonomic, and perceptual variables were grouped and analyzed by 3 linear regression models in order to detect significant relationships between independent variables and CRH CSF . The 3 models had high significance levels (TABLE 5 ). A metabolic model related CRH CSF to glucose and sodium. An autonomic model linked CRH CSF to resting respiratory and heart rates. The optimum perceptual model predicted CRH CSF based on SF-36 Vitality and MFI Physical Fatigue. The latter 2 domains were not the same as those in TABLES 3 and 4 because many of these variables were co-linear or surrogates of one another. This was reinforced by the similarity of R 2 and p values where these domains were substituted for Vitality and Physical Fatigue. These results were remarkable because it is very unusual for models with such small numbers of observations (n ≤ 9) to be so significant (p < 0.02). The explained variances were very high (R 2 > 0.85) suggesting that the factors may be causally connected. CRH CSF did not correlate with variables associated with maximum exercise, heat – and pressure – (dolorimetry) induced pain, or other rapid onset stressors (TABLE 1 ). Discussion Despite the small numbers of normal subjects in this pilot investigation, the data, the explained variances for relationships, and final 3 models predictive of CRH CSF were highly significant. The models were of importance, since they reflected the specific neurological functions of this neurohormone. CRH and NPY were co-expressed in the hypothalamus [ 2 ], but NPY CSF did not correlate with CRH CSF or any other variable. CRH and the hypothalamic-pituitary-adrenal axis maintain numerous systemic functions. Our metabolic model showed a tightly correlated relationship between 2 pm serum glucose, sodium and CRH CSF . Relatively higher CRH CSF levels were associated with elevated serum glucose levels. When glucose is elevated, it is pumped into cells along with sodium ions [ 9 ]. In our model, this may have been reflected by the reduced serum sodium concentrations (FIGURE 1 ). These measurements were taken at different times, suggesting that the CRH CSF set a long-term operating range for this system. The variables were interrelated. CRH CSF in the low normal range was inferred from a low glucose and relatively high sodium. Other reports also suggest a role of CRH and energy balance [ 10 ]. Neuroendocrine responses such as these rely solely on CRH type 1 (CRH1) receptors and the HPA axis [ 11 ]. The other CRH receptor gene, CRH2, has 3 splice variants (α, β and γ) but only CRH2α is expressed in the brain. CRH1 and CRH2α receptors have nonoverlapping distributions, but mediate many similar defensive behaviors suggesting that they act in parallel neural circuits. Different stressors may act by separate circuits and have distinct feedback and control systems [ 11 - 13 ]. CRH1 receptors in the central nucleus of the amygdala may participate in conditioned fear responses [ 12 ]. These CRH neurons may project to the hippocampus and CRH1 receptors in the locus ceruleus to induce defensive behaviors and autonomic reflexes [ 13 , 14 ]. Dorsal raphe nucleus neurons may release CRH that acts on inhibitory CRH2α receptors in the lateral septum [ 15 ]. Neurons from the lateral septum tonically inhibit periaqueductal grey regions that induce similar defensive behaviors. These nuclei are probably additional sources of CRH in the CSF. Resting respiratory and pre-exercise heart rates and CRH CSF were positively correlated. The statistical model indicated that a low CRH CSF was predicted by low respiratory and heart rates. Respiratory and cardiac functions are rigorously controlled by brainstem and other nuclei that integrate incoming signals of plasma O 2 , CO 2 and H + concentrations, activity needs, anxiety and other stressors. Efferent cardiovascular and other autonomic reflexes are modulated by CRH in man [ 16 ]. These central nervous system effects may be due to, or highly correlated with, CRH CSF . This is supported by studies in mice that genetically overexpress CRH. They develop chronic stress – like autonomic and physiological alterations [ 17 ]. Stressors acting via conditioned fear responses may involve CRH1 receptors in the central nucleus of the amygdala. Some of these CRH neurons project to locus ceruleus neurons [ 17 ] that activate autonomic reflexes and defensive behaviors such as "freezing" (immobility) in rodents [ 14 ]. An example of this valuable defense would be the freezing of prey in the presence of a predator. Immobility would allow the prey's camouflage to blend into the surroundings without generating motion – induced visual cues for the predator. In humans, excessive, aberrant or dysregulated manifestions of defense behaviors such as freezing may contribute to the immobility, inertia, or even catatonia that contribute to the clinical picture of depression [ 1 ]. Extrapolation of these concepts suggests that elevated CRH may be related to anxiety, depression, or other disorders associated with chronic stress responses. If so, then lower, but normal, CRH CSF should be present in persons lacking these stressor states. This was supported by the negative correlations of CRH CSF with scores for the SF-36 Vitality and Change in Health, Self Efficacy Scale Manage Symptoms, MIQ Vulnerability, and MFI Physical Functioning domains. Each scale has an idealized "normal" end of the range of scores. Deviation towards either higher (MFI) or lower (SF-36) ends of the scales provides an estimate of dysfunction. For each of these domains, the lower CRH CSF were associated with more normal scores, while higher CRH CSF was associated with scores that were beginning to shift away from the normal pole of each scale. These trends were further supported by the statistical model where the optimum covariates of CRH CSF were SF-36 Vitality and MFI Physical Fatigue domains. The model predicted that CRH CSF would be in the low normal range when Vitality and Physical Fatigue domain scores were high (normal). Taken together, these results confirm the consistent physical status, mental coping skills, and general health of these subjects. Studies of intraventricular CRH injection in primates support our findings [ 33 ]. The CRH diffused to brain regions that led to 3 types of behavioral changes. Externally oriented behaviors such as locomotion and environmental exploration were significantly decreased. Anxiety – related self – clasping was increased. Depression – like behaviors of avoidance of social contact, huddling, slouching, and wall facing were seen only in social settings. There were high interindividual differences in responses, but a key element was the social context in this study. This social context is lacking in rodent studies where animals are typically studied in isolation. Conclusion This small but intensively studied group of normal humans demonstrated surprisingly robust relationships between CRH CSF and metabolic, autonomic, and psychometric measures. These are novel findings in humans, but are consistent with data on CRH in acute and chronic stress models, and proposed CRH neural circuits. It will now be of great interest to contrast these statistical models identified for normal subjects with the other chronic pain and fatigue patient subsets to determine if CRH CSF correlates with different sets of variables. Methods Subjects Nine normal, healthy subjects (2 females) gave informed consent for this paid, Institutional Review Board – approved protocol. They had a comprehensive screening evaluation to exclude: severe physical impairment, morbid obesity, autoimmune/inflammatory diseases, cardiopulmonary disorders, uncontrolled endocrine or allergic disorders, malignancy, severe psychiatric illnesses (e.g., schizophrenia, substance abuse), factors known to affect the HPA axis or autonomic function (cigarette smoking, daily intake of caffeine exceeding the equivalent of 2 cups of coffee), or medication use. Subjects had a history and physical examination, were administered the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) (SCID II) [ 19 , 20 ], Composite International Diagnostic Interview (CIDI) [ 21 , 22 ] and Center for Epidemiologic Studies Depression scale (CESD) [ 23 ] to detect psychiatric co-morbidities, and completed the self – report Short Form 36 (SF-36) [ 24 , 25 ], Self Efficacy Scale (SES) [ 26 ], Meaning of Illness Questionnaire (MIQ) [ 27 ], and Multidimensional Fatigue Inventory (MFI) [ 28 ]. Study protocol Subjects were admitted to the G-CRC on the evening of Day 1. An 18-gauge catheter was inserted antecubitally and infused with normal saline at 50 ml/hr. A Holter monitor was attached to monitor autonomic regulation of cardiac rhythm [ 29 - 31 ]. Upon awakening, or at 6:30 am, subjects stayed in bed until they had "pre-awakening" blood tests drawn. Breakfast was served between 7:00 and 8:00 a.m. and baseline Day 2 blood samples drawn at 8:30 a.m. Vital signs, questionnaire responses, and other blood samples were drawn from the catheter to assess HPA axis and stress system responses before, during and after Day 2 stressors. Pain responses were tested by pressure applied to each subject's thumbnail and heat to the forearm applied in random staircase testing paradigms (Stressor I) [ 32 , 33 ]. Stressor II was a series of cognitive challenges including the Benton Visual Retention Test of visual-spatial memory; Digit Span recitation; and Pig Latin, a test of verbal working memory [ 34 ]. Lunch was served, followed by a 2 to 2.5 hours rest break to allow post-prandial catecholamine levels to reequilibrate. At 2 pm, serum glucose, electrolytes, plasma catecholamines and other analytes were measured. Subjects then squeezed an isometric hand grip dynamometer to test autonomic function and muscular fatigue [ 35 ] (Jamar, Sammons Preston, Bollingbrook, IL) (Stressor III). After 30 min of rest, they had a sub-maximal exercise test on an electronically-braked cycle ergometer (Sensormedics, Yorba Linda, CA). The test was graded in 3 min stages and ended when the subjects' heart rates reached 85% of their age-predicted maximum (Stressor IV). Lumbar punctures were performed 30 min later (approximately 4 pm) (Stressor V). Subjects sat while sterile technique was used to prepare the skin over the L4–L5 lumbar region, infiltrate the subcutaneous and deep tissues with 2% lidocaine, and insert a 22G spinal catheter. CSF was collected as 3 to 4 aliquots of about 2 ml each. Catheters were withdrawn and subjects allowed to rest in their preferred position for 30 min. Lumbar punctures and CRH and NPY radioimmunoassays (RIA) Tubes of CSF were immediately placed on ice and then centrifuged at 4°C. The supernatants were rapidly frozen at -80°C. Tube 2 or 3 was removed from the freezer and thawed at 4°C. Peptides were extracted by precipitating high molecular weight proteins by adding an equal volume of 100% ethanol, 0.1 M acetic acid, 0.2% sodium bisulfite [ 36 , 37 ]. The supernatant was dried (SpeedVac, Thermo Savant, Holbrook, NY), resuspended in phosphate buffered saline with 1% bovine serum albumin (RIA buffer; Peninsula Laboratories, Inc., San Carlos, CA.). Samples and standard amounts of each peptide were aliquoted and peptide specific rabbit antibodies added. After overnight incubation at 4°C, I 125 -CRH or -NPY was added, tubes gently vortexed, and again incubated overnight. Goat anti-rabbit antibodies were added, tubes incubated, and immune complexes precipitated by centrifugation. Radioactivity was counted for the standards, and the concentrations for each sample interpolated from the standard curves. Spiking CSF with fixed amounts of I 125 -peptides and performing the RIA shifted the curves to the left by the anticipated concentrations. Standard curves were reproducible to within 10%. Plasma catecholamine levels were measured by HPLC (Mayo Clinic Laboratories, Rochester, MN). Statistical methods All the data from the provocation studies, questionnaire domains, blood work, and neuropeptide analysis were entered into a SAS spreadsheet (SAS, Carey, NC) using sequential hand or scanner entry followed by data checking routines. Means with 95% confidence intervals were reported. Simple correlations between our outcome measure (CRH CSF ) and other objective and subjective patient variables were conducted as an exploratory tool. Given the small sample size (n = 9), both parametric and non-parametric correlations were used to evaluate the robustness of the correlation coefficients. Partial and intra-class correlations examined the structure of the relationships between CRH CSF and the independent variables. Original data were always checked to make sure that correlations were not spurious, due to outliers, colinearity, or sets of virtually identical scores. Our data included a very wide range of covariates which spanned the physical and psychological attributes of the patients. Scatter plots were used to determine the response slopes with CRH CSF . Independent variables with flat slopes or high scatter that did not generate correlations with CRH CSF , that were highly correlated with each other (collinear), or that had < 5 recorded values were excluded from further analysis. The remaining suitable independent variables fell into 3 categories: metabolic, autonomic and psychometric function. Three 3 separate linear regression models [ 8 ] were constructed to predict CRH CSF , the dependent variable. Separate models were required because of the low degrees of freedom available for the analyses, and to ensure that the explained variance (R 2 ) was not inflated due to overloading the model with potentially collinear variables. Variance inflation factors and tolerance values were incorporated to optimize the selection of the most independent combinations of variables that also maximized the R 2 and p values. The linear regression procedure took into account the multiple comparisons for each model. Abbreviations CESD, Center for Epidemiologic Studies Depression scale; CFS, Chronic Fatigue Syndrome; CIDI, Composite International Diagnostic Interview; CRH, corticotropin-releasing hormone; CSF, cerebrospinal fluid; MFI, Multidimensional Fatigue Inventory; MIQ, Meaning of Illness Questionnaire; NPY, neuropeptide Y; R 2 , explained variance; SCID II, Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV); SF-36, Short Form 36; SES, Self Efficacy Scale Authors' contributions JNB directed the RIA studies, correlation of clinical variables, and wrote the manuscript. HM confirmed the accuracy of the study results by double data entry methods, vetted the SAS database, performed the statistical analysis, and co-wrote the manuscript. GW performed the CRH and NPY assays and maintained the repository of these samples. DJC was the Principal Investigator for the overall study and helped review the draft and final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544567.xml |
549043 | Mental health first aid responses of the public: results from an Australian national survey | Background The prevalence of mental disorders is so high that members of the public will commonly have contact with someone affected. How they respond to that person (the mental health first aid response) may affect outcomes. However, there is no information on what members of the public might do in such circumstances. Methods In a national survey of 3998 Australian adults, respondents were presented with one of four case vignettes and asked what they would do if that person was someone they had known for a long time and cared about. There were four types of vignette: depression, depression with suicidal thoughts, early schizophrenia, and chronic schizophrenia. Verbatim responses to the open-ended question were coded into categories. Results The most common responses to all vignettes were to encourage professional help-seeking and to listen to and support the person. However, a significant minority did not give these responses. Much less common responses were to assess the problem or risk of harm, to give or seek information, to encourage self-help, or to support the family. Few respondents mentioned contacting a professional on the person's behalf or accompanying them to a professional. First aid responses were generally more appropriate in women, those with less stigmatizing attitudes, and those who correctly identified the disorder in the vignette. Conclusions There is room for improving the range of mental health first aid responses in the community. Lack of knowledge of mental disorders and stigmatizing attitudes are important barriers to effective first aid. | Background Surveys in many countries have found that mental disorders have a high prevalence and are a major cause of disability in the population [ 1 - 3 ]. For example, the Australian National Survey of Mental Health and Wellbeing found that close to one in five adults met the criteria for a mental disorder at some time during the 12 months before the survey [ 3 ]. The most common mental disorders were anxiety (10%), depressive (6%) and substance use disorders (8%). These disorders are so prevalent that everyone in the community can expect to have close contact with someone experiencing a mental disorder. How people initially respond to others with a mental disorder may influence their recovery. For example, it is known that many people with mental disorders get no professional help [ 4 ] and that help-seeking is more likely if relatives or friends suggest it [ 5 ]. It is also known that recovery can be assisted if family members are supportive and not critical [ 6 ]. This type of initial help from a person's social network can be defined as mental health first aid. There has been no previous research on mental health first aid knowledge in the population. Previous mental health literacy surveys have assessed knowledge and beliefs about mental disorders and their treatment [ 5 ], but these surveys have not assessed the public's intentions for mental health first aid responses to hypothetical cases of persons with mental disorders. Thus it is not known which mental health first aid responses are currently adequate and which need improving. Hence a relevant question was asked in a recent national survey of mental health literacy in Australia. Methods The Australian survey In 2003–2004 a household survey was carried out of Australian adults aged 18 or over by the company AC Nielsen. Households were sampled from 250 census districts covering all states and territories and metropolitan and rural areas. Up to 5 call backs were made to metropolitan selections and 3 to non-metropolitan selections. To achieve a target sample of 4,000 interviews with adults aged 18 years or over, visits were made to 28,947 households. The outcome of these visits was: no contact after repeated visits 14,630; vacant house or lot 306; refused 7,815; person sampled within household temporarily unavailable 1,132; no suitable respondent in household 287; did not speak English 383; incapable of responding 213; and unavailable for the duration of the survey 181. The achieved sample was 3998 persons, with 1001 receiving the depression vignette, 999 the depression with suicidal thoughts vignette, 997 the early schizophrenia vignette, and 1001 the chronic schizophrenia vignette. Interview content The interview was based on a vignette of a person with a mental disorder. On a random basis, respondents were shown one of four vignettes: a person with major depression, one with major depression together with suicidal thoughts, a person with early schizophrenia, and one with chronic schizophrenia. All vignettes were written to satisfy the diagnostic criteria for either major depression or schizophrenia according to DSM-IV and ICD-10. The vignette with depression and the one with early schizophrenia were written to satisfy these diagnostic criteria at a minimal level, so that we could ascertain the public's reaction to cases of a developing disorder which had reached the point where intervention was needed. The vignette of the person with depression together with suicidal thoughts was identical to the depression vignette in all respects except the suicidal thoughts and was designed to assess how this symptom affected the public's response. The chronic schizophrenia vignette was designed to assess the response to someone with a severe long-standing disorder, where acceptance seemed less likely. Respondents were also randomly assigned to receive either male ("John") or female ("Mary") versions of the vignette. These vignettes (John version) are shown in Table 1 . Table 1 Case vignettes used in the survey Disorder Vignette Depression John is 30 years old. He has been feeling unusually sad and miserable for the last few weeks. Even though he is tired all the time, he has trouble sleeping nearly every night. John doesn't feel like eating and has lost weight. He can't keep his mind on his work and puts off making decisions. Even day-to-day tasks seem too much for him. This has come to the attention of his boss, who is concerned about John's lowered productivity. Depression with suicidal thoughts John is 30 years old. He has been feeling unusually sad and miserable for the last few weeks. Even though he is tired all the time, he has trouble sleeping nearly every night. John doesn't feel like eating and has lost weight. He can't keep his mind on his work and puts off making any decisions. Even day-to-day tasks seem too much for him. This has come to the attention of John's boss who is concerned about his lowered productivity. John feels he will never be happy again and believes his family would be better off without him. John has been so desperate, he has been thinking of ways to end his life. Early schizophrenia John is 24 and lives at home with his parents. He has had a few temporary jobs since finishing school but is now unemployed. Over the last six months he has stopped seeing his friends and has begun locking himself in his bedroom and refusing to eat with the family or to have a bath. His parents also hear him walking about his bedroom at night while they are in bed. Even though they know he is alone, they have heard him shouting and arguing as if someone else is there. When they try to encourage him to do more things, he whispers that he won't leave home because he is being spied upon by the neighbour. They realize he is not taking drugs because he never sees anyone or goes anywhere. Chronic schizophrenia John is 44 years old. He is living in a boarding house in an industrial area. He has not worked for years. He wears the same clothes in all weathers and has left his hair to grow long and untidy. He is always on his own and is often seen sitting in the park talking to himself. At times he stands and moves his hands as if to communicate to someone in nearby trees. He rarely drinks alcohol. He speaks carefully using uncommon and sometimes made-up words. He is polite but avoids talking with other people. At times he accuses shopkeepers of giving information about him to other people. He has asked his landlord to put extra locks on his door and to remove the television set from his room. He says spies are trying to keep him under observation because he has secret information about international computer systems which control people through television transmitters. His landlord complains that he will not let him clean the room which is increasingly dirty and filled with glass objects. John says he is using these "to receive messages from space". After being presented with the vignette, respondents were asked a series of questions to assess their recognition of the disorder in the vignette, their beliefs about treatment and long-term outcomes, beliefs about causes and risk factors, stigmatizing attitudes, awareness of mental disorders in the media, contact with people like those in the vignette, and the health and sociodemographic characteristics of the respondent. The questions relevant to the present paper are described below. To assess recognition of the problem in the vignette, respondents were asked: "What would you say, if anything, is wrong with John?". Responses of "depression" were counted as correct for the first two vignettes above, and responses of "schizophrenia" or "psychosis" for the second two. To assess mental health first aid responses, respondents were asked the open-ended question: "Imagine John is someone you have known for a long time and care about. You want to help him. What would you do?". Answers were recorded verbatim by the interviewer. To assess contact with people like those in the vignette, respondents were asked: "Has anyone in your family or close circle of friends ever had problems similar to John's?"; "Have they received any professional help or treatment for these problems?'; "Have you ever had problems similar to John's?"; "Have you received any professional help or treatment for these problems?"; and "Have you ever had a job that involved providing treatment or services to a person with a problem like John's?". Those that said "yes" to these questions were respectively labelled in the analyses reported below as "carers", "consumers" or "professionals". To assess stigma, respondents were asked a series of nine questions designed to elicit their attitudes towards the person in the vignette (personal stigma) and nine items concerning what they thought others in the community would believe about the person in the vignette (perceived stigma) [ 7 ]. Personal stigma items were of the form: "Please indicate how strongly you agree or disagree with each statement. People with a problem like John's could snap out of it if they wanted. Strongly agree, Agree, Neither agree nor disagree, Disagree, Strongly disagree". Perceived stigma items were of the form: "Now we would like you to tell us what you think most other people believe. Please indicate how strongly you agree or disagree with the following statements. Most people believe that people with a problem like John's could snap out of it if they wanted. Strongly agree, Agree, Neither agree nor disagree, Disagree, Strongly disagree". Sociodemographic characteristics recorded included age group (coded as 18–39, 40–59 and 60+ years), gender, and education (dichotomized as bachelor's degree or higher versus lower-level qualifications). Content analysis of responses to open-ended question Responses were coded according to the categories identified from an earlier study where the same question was administered as part of a randomized controlled trial of Mental Health First Aid [ 8 ]. Responses were coded with a "yes" or "no" in each category, such that multiple categories were possible. The categories were: A. Encourage professional help-seeking B. Listen to / talk to / support person C. Listen to / talk to / support family D. Assess problem / assess risk of harm E. Give or seek information F. Encourage self-help Responses coded into category A. Encourage professional help-seeking , were subcoded into multiple categories to identify the type of professional help recommended. These categories were: A1. GP / doctor unspecified A2. Counsellor A3. Psychiatrist A4. Psychologist A5. Mental health team / services A6. Other mental health professionals A7. Unspecified professionals and other professionals A8. Accompany person (eg. Offer to go with him/her) A9. Contact help on their behalf Examples of responses coded into category B included "support, understanding and caring, someone to talk to him", "talk to her about it", "listen", "be there for him". Responses coded into category C were the same for category B, but referred to giving support, listening to, or talking to the sufferer's family. For example, "talk to her family", "contact relatives", "ask advice of parents", "support his parents". Responses coded into category D included "keep an eye on her, make sure she is safe", "make a contract with her so if she wants to harm herself she rings me first", "find out what is the real problem behind the behaviour, focus on the problem". Responses coded into category E included "ring the local health authority to get advice", "talk to other people who have been in the situation", "speak to health professionals and get best advice", "get some brochures from community health and give them to him". Responses coded into category F included "I suggest that he have a holiday/exercise/change jobs", "try and help him get into something he is interested in", "support groups", "do something for herself to get out of the situation". Inter-rater reliability of the coding was assessed by a second rater who independently coded 100 responses which were randomly selected using the SPSS Select Cases procedure [ 9 ]. Statistical analysis Inter-rater reliability of the content coding was assessed using kappa. Kappa values were interpreted according to Altman [ 10 ] as follows: 0.8–1.0 very good; 0.6–0.8 good; 0.4–0.6 moderate; 0.2–0.4 fair; and <0.2 poor. The frequency of codings was analysed by pooling across male and female versions of each vignette and percent frequencies calculated. Percentages were calculated applying survey weights to give better population estimates. Standard errors of these percentages were estimated using the Complex Samples procedure in SPSS 12.0 [ 9 ]. This procedure takes account of sampling weights and geographic clustering in the sample. For simplicity of exposition, the standard errors are not reported for each estimate. However, they were always <2%. Such a standard error implies that a difference of 4% between vignettes was always statistically significant at the P < 0.05 level. Multiple logistic regressions were then conducted to examine the levels of association between participants suggesting particular treatment options and their sociodemographic and mental health experience attributes. The following predictor variables were included: age group; consumer, carer and professional status, including whether or not professional help was obtained; and levels of perceived stigma and personal stigma. Two vignette measures – type of vignette provided and whether respondents correctly identified the problem portrayed in that vignette – were also included in the analyses. Each logistic regression was also adjusted to take into account of sampling weights and clustering method applied in this survey. These analyses were undertaken using STATA 8 [ 11 ]. Results Reliability of coding open-ended responses Inter-rater reliability was assessed for a randomly chosen 100 responses. Kappa was very good or good for encourage professional help-seeking (0.89), listen/talk/support person (0.70), listen/talk/support family (1.00), encourage seeing doctor (0.98), encourage seeing counsellor (0.93), encourage seeing psychiatrist (0.94), encourage seeing psychologist (0.88), and accompanying the person to a professional (0.95). It was moderate for give or seek information (0.48), encourage seeing unspecified and other professionals (0.56), and contact professional on their behalf (0.56). Kappa was fair for encourage self-help (0.34) and poor for assess problem/risk of harm (0.15). Kappa could not be calculated because of zero frequencies from the first rater for the categories of mental health team/services and other mental health professional. To better understand the reasons for the fair and poor agreement with two of the codes, positive and negative agreement were examined separately [ 12 ]. In both cases, negative agreement was high (0.95 for both), but positive agreement was low (0.38 and 0.15 respectively). The low positive agreement resulted because the second rater used these codes much less frequently than the first rater. However, despite this difference, neither rater used these two codes frequently, suggesting a low frequency of these responses in the population. Frequencies of mental health first aid responses Table 2 shows the percentage frequency of each category of response. The most common responses for all vignettes were to encourage professional help-seeking and to listen/talk/support the person. The differences between the vignettes were comparatively small. For the chronic schizophrenia vignette, there was a greater frequency of encouraging professional help-seeking and giving or seeking information, and a lesser frequency of listen/talk/support the person. Assessing the problem/risk of harm was more common for the depression vignettes than for the schizophrenia vignettes. Listen/talk/support family was more common for the early schizophrenia vignette, which was the only one to specifically mention family members. Table 2 Percentage of respondents who mentioned various first aid responses Response Depression Vignette Depression/Suicidal Vignette Early Schizophrenia Vignette Chronic Schizophrenia Vignette Encourage professional help-seeking 58.6 62.4 57.5 66.0 Listen/ talk/ support person 69.1 73.4 70.3 65.6 Listen/ talk/ support family 2.2 2.8 8.5 2.4 Assess problem/ risk of harm 17.3 14.8 9.7 6.8 Give or seek information 5.5 7.1 9.1 12.9 Encourage self-help 12.0 10.8 12.3 10.2 Table 3 shows types of professionals mentioned by respondents who encouraged professional help-seeking. The most commonly mentioned was GP/doctor unspecified. This recommendation was more common for the depression vignettes than for the schizophrenia vignettes. Conversely, psychiatrists were mentioned more for the schizophrenia vignettes. Table 3 Percentage of respondents who mentioned encouraging help-seeking with various types of professionals Type of Professional Depression Vignette Depression/ Suicidal Vignette Early Schizophrenia Vignette Chronic Schizophrenia Vignette GP/ doctor unspecified 40.1 35.9 26.9 27.4 Counsellor 8.1 8.3 7.6 7.3 Psychiatrist 2.9 3.3 6.4 7.3 Psychologist 2.2 2.5 2.5 3.1 Mental health team/services 0.1 0.4 1.5 2.3 Other mental health professional 0.2 0.2 1.0 1.1 Unspecified professionals or other professionals 14.1 21.1 20.9 29.9 Table 4 gives the percentage frequency of ways of encouraging professional help-seeking. Accompanying the person was more common for the chronic schizophrenia vignette, while contacting the professional on the person's behalf was more common for both the schizophrenia vignettes. Table 4 Percentage of respondents who mentioned ways of encouraging professional help-seeking Type of Professional Depression Vignette Depression/ Suicidal Vignette Early Schizophrenia Vignette Chronic Schizophrenia Vignette Accompany person 8.7 11.0 9.7 16.7 Contact professional on their behalf 3.0 4.1 12.8 15.9 Predictors of responses Table 5 shows the predictors of first aid responses from the multiple logistic regressions. Taking predictors with P <0.01, encouraging professional help-seeking was more likely in response to the chronic schizophrenia vignette, from women, and from those who correctly recognized the problem in the vignette. It was less likely from consumers who had not sought help and respondents high on personal stigma. Listening/talking/supporting the person was more likely from consumers who had sought help. Listening/talking/supporting the family was more likely in response to the chronic schizophrenia vignette, and less likely from those high on personal stigma. Assessing the problem/ risk of harm was less likely in response to either schizophrenia vignette and from people aged 60+. Giving or seeking information was more likely in response to either schizophrenia vignette and from those who perceived stigma in others, but less likely from those high in personal stigma. Encouraging self-help was more likely from respondents high in personal stigma and less likely from those who correctly recognized the problem in the vignette. Table 5 Odds ratios (and P-values) from multiple logistic regression analyses predicting first aid responses Predictor Encourage professional help-seeking Listen/ talk/ support person Listen /talk/ support family Assess problem/ risk of harm Give or seek information Encourage self-help Type of vignette Depression 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 Depression/ suicidal 1.13 (0.284) 1.19 (0.149) 1.28 (0.433) 0.82 (0.154) 1.21 (0.360) 0.87 (0.385) Early schizophrenia 0.95 (0.636) 1.13 (0.298) 4.67 (0.000) 0.46 (0.000) 1.83 (0.002) 1.07 (0.671) Chronic schizophrenia 1.66 (0.000) 0.88 (0.266) 1.00 (0.997) 0.32 (0.000) 2.83 (0.000) 0.75 (0.100) Sociodemographic characteristics Age 18–39 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 Age 40–59 1.04 (0.646) 0.89 (0.138) 0.91 (0.645) 0.94 (0.628) 0.87 (0.341) 1.04 (0.759) Age 60+ 0.97 (0.753) 0.79 (0.010) 0.85 (0.515) 0.64 (0.002) 0.66 (0.031) 1.04 (0.809) Female gender 1.32 (0.000) 1.20 (0.024) 0.92 (0.663) 0.74 (0.015) 0.99 (0.937) 0.87 (0.247) University degree 1.20 (0.065) 1.07 (0.428) 0.92 (0.660) 1.19 (0.218) 1.01 (0.934) 1.06 (0.676) Experience with mental disorders Consumer – not sought help 0.50 (0.000) 1.47 (0.061) 1.56 (0.291) 1.54 (0.047) 0.61 (0.212) 1.77 (0.010) Consumer – sought help 0.88 (0.314) 1.49 (0.000) 0.47 (0.024) 1.02 (0.920) 0.78 (0.247) 1.12 (0.514) Carer – not sought help 0.78 (0.149) 1.08 (0.666) 0.48 (0.195) 1.11 (0.658) 1.01 (0.965) 1.17 (0.503) Carer – sought help 1.22 (0.032) 1.02 (0.841) 1.09 (0.668) 0.83 (0.187) 1.05 (0.741) 0.98 (0.900) Professional 0.92 (0.403) 1.20 (0.068) 1.19 (0.400) 0.97 (0.853) 0.96 (0.801) 1.06 (0.679) Stigma Personal stigma 0.95 (0.000) 0.99 (0.494) 0.93 (0.001) 1.01 (0.612) 0.95 (0.000) 1.06 (0.000) Perceived stigma 1.01 (0.112) 0.99 (0.300) 1.00 (0.883) 1.00 (0.876) 1.04 (0.003) 0.99 (0.580) Correct recognition of disorder in vignette 1.60 (0.000) 1.05 (0.550) 1.22 (0.320) 1.27 (0.095) 1.32 (0.060) 0.70 (0.007) 1 Reference category Table 6 shows predictors of encouraging help-seeking with various types of health professionals. Encouraging the person to see a GP or doctor was more likely in response to either schizophrenia vignette, from women and from those who correctly recognized the problem in the vignette, while it was less likely in those high on personal stigma. Encouraging help-seeking from a counsellor was less likely from those aged 60+. Encouraging help-seeking from a psychiatrist was more likely in response to either schizophrenia vignette, while encouraging help-seeking from a psychologist was more likely from the university educated. Encouraging help-seeking from a mental health team/services was more likely in response to the chronic schizophrenia vignette and from those with professional experience in the area of mental health. Encouraging help-seeking from other mental health professionals was less likely in respondents high on personal stigma, while encouraging it from unspecified professionals was more likely in response to the depression/suicidal vignette, to either schizophrenia vignette, and from respondents who perceive stigma in others, while it was less likely in those high on personal stigma. Table 6 Odds ratios (and P-values) from multiple logistic regression analyses predicting encouragement of help-seeking from various types of professionals Predictor GP/ doctor unspecified Counsellor Psychiatrist Psychologist Mental health team/ services Other mental health professionals Unspecified professionals or other professionals Type of vignette Depression 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 Depression/ suicidal 0.77 (0.012) 1.05 (0.797) 1.04 (0.895) 1.13 (0.688) 1.92 (0.597) 1.28 (0.784) 1.76 (0.000) Early schizophrenia 0.55 (0.000) 0.92 (0.697) 2.23 (0.002) 1.42 (0.281) 13.79 (0.013) 3.68 (0.102) 1.72 (0.000) Chronic schizophrenia 0.60 (0.000) 1.16 (0.461) 2.66 (0.000) 1.76 (0.062) 20.34 (0.004) 7.12 (0.024) 2.96 (0.000) Sociodemographic characteristics Age 18–39 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 1.00 1 Age 40–59 1.19 (0.055) 0.88 (0.366) 1.35 (0.096) 0.96 (0.887) 0.50 (0.112) 1.03 (0.949) 0.93 (0.476) Age 60+ 1.35 (0.010) 0.49 (0.001) 1.55 (0.044) 1.24 (0.505) 1.00 (1.00) 0.54 (0.447) 0.74 (0.017) Female gender 1.34 (0.000) 1.29 (0.08) 0.74 (0.061) 1.06 (0.823) 1.00 (1.00) 1.22 (0.676) 1.10 (0.294) University degree 0.96 (0.638) 1.34 (0.054) 0.87 (0.473) 2.29 (0.001) 1.80 (0.132) 1.89 (0.148) 1.12 (0.264) Experience with mental disorders Consumer – not sought help 0.69 (0.074) 0.63 (0.237) 0.38 (0.126) 0.86 (0.813) - 2 - 2 0.61 (0.056) Consumer – sought help 0.91 (0.365) 1.23 (0.227) 0.82 (0.457) 1.64 (0.060) 1.02 (0.967) 1.22 (0.735) 0.88 (0.334) Carer – not sought help 0.87 (0.477) 1.10 (0.759) 1.53 (0.255) 0.96 (0.938) 0.88 (0.907) 0.78 (0.817) 0.97 (0.904) Carer – sought help 1.08 (0.389) 1.32 (0.074) 1.46 (0.025) 1.51 (0.110) 1.05 (0.908) 1.08 (0.842) 1.01 (0.916) Professional 1.00 (0.985) 0.95 (0.751) 0.891 (0.570) 1.17 (0.570) 2.42 (0.008) 1.01 (0.990) 0.98 (0.828) Stigma Personal stigma 0.97 (0.005) 1.00 (0.969) 1.00 (0.835) 1.00 (0.917) 0.92 (0.120) 0.89 (0.007) 0.95 (0.000) Perceived stigma 0.99 (0.079) 1.01 (0.468) 1.00 (0.969) 0.98 (0.439) 1.04 (0.188) 1.05 (0.083) 1.03 (0.000) Correct recognition of disorder in vignette 1.43 (0.000) 1.49 (0.019) 1.17 (0.379) 0.74 (0.171) 1.60 (0.255) 3.00 (0.072) 1.24 (0.031) 1 Reference category 2 Could not be calculated Table 7 shows predictors of ways of encouraging professional help-seeking. Accompanying the person to professional help was more likely in response to the chronic schizophrenia vignette and from women. Contacting the professional on the person's behalf was more likely in response to either schizophrenia vignette and from respondents who perceived stigma in others, while it was less likely from those high on personal stigma. Table 7 Odds ratios (and P-values) from multiple logistic regression analyses predicting ways of encouraging professional help-seeking Predictor Accompany person Contact professional on their behalf Type of vignette Depression 1.00 1 1.00 1 Depression/ suicidal 1.21 (0.285) 1.68 (0.071) Early schizophrenia 1.08 (0.646) 5.27 (0.000) Chronic schizophrenia 2.26 (0.000) 8.01 (0.000) Sociodemographic characteristics Age 18–39 1.00 1 1.00 1 Age 40–59 1.07 (0.549) 0.97 (0.828) Age 60+ 0.90 (0.485) 0.63 (0.017) Female gender 1.43 (0.003) 1.06 (0.683) University degree 0.86 (0.220) 1.31 (0.046) Experience with mental disorders Consumer – not sought help 0.51 (0.060) 0.33 (0.068) Consumer – sought help 1.19 (0.228) 0.71 (0.136) Carer – not sought help 0.83 (0.491) 1.16 (0.620) Carer – sought help 0.98 (0.893) 1.12 (0.424) Professional 1.21 (0.109) 1.12 (0.427) Stigma Personal stigma 0.99 (0.659) 0.95 (0.000) Perceived stigma 0.99 (0.283) 1.03 (0.008) Correct recognition of disorder in vignette 1.16 (0.280) 1.37 (0.028) 1 Reference category Discussion The most common first aid responses were found to be encouraging professional help-seeking and listening/talking/supporting the person. Nevertheless, these responses were far from universal, with 32–44% not mentioning professional help and 27–34% not mentioning listening/talking/supporting. Given the likely helpfulness of these first aid responses, they need greater promotion in the community. Other first aid responses were mentioned only by a minority. Of particular concern is the low percentage assessing risk of harm for the person in the depression/suicidal vignette. Asking about suicidal intentions is often recommended as a response [ 13 , 14 ], although there is no evidence on whether this actually helps prevent suicide. Previous research has investigated first aid responses of young people to suicidal intent in their peers, finding that many would not tell a responsible adult about it [ 15 ]. However, we are unaware on any previous research on adults' responses to suicidal intent in someone they know. Encouraging self-help was another minority response, but was associated with stigma and lack of recognition of the mental disorder in the vignette. Respondents appear to have suggested self-help as an alternative to professional help, rather than as a complement to it. We have previously reviewed the evidence on self-help interventions for depression and anxiety disorders and found that some have support [ 16 , 17 ]. Such interventions need to more widely promoted, but not as a substitute for professional help. When correlates of first aid responses were examined, most variables had at least one significant association. However, the variables that most often predicted first aid responses were female gender, low personal stigma and correct recognition of the disorder in the vignette. The latter two predictors indicate potential barriers to providing first aid. Respondents who saw the person in the vignette as having negative attributes were less likely to respond by encouraging professional help-seeking or providing personal support. Efforts to reduce stigma in the community may therefore facilitate greater first aid. People who did not recognize the disorder showed a similar pattern of responses. These people lack knowledge of mental disorders, at least to the extent of being able to apply a psychiatric label. Therefore community education about how to recognize these disorders may also facilitate helpful first aid responses. One approach to improving public responses to people with a mental disorder is an individual training course in mental health first aid [ 18 ]. Such a course has been developed and teaches an action plan with five steps of mental health first aid: (1) Assess the risk of suicide or harm; (2) Listen non- judgementally; (3) Give information and encouragement; (4) Encourage person to get appropriate professional help; and (5) Encourage self-help strategies. [ 19 ]. Two randomised controlled trials of this course, one in a work place environment [ 20 ] and the other with members of the public in a rural area [ 8 ], have shown benefits of the training: better recognition of disorders, changes in beliefs about treatment to be more like those of professionals, decreased social distance, increased confidence in providing help, and increase in actual help provided. The workplace trial also found improved mental health in course participants [ 20 ]. The study has two limitations which must be acknowledged. The major one is that the study has assessed intended first aid to a hypothetical person in a case vignette. Whether these intentions would be implemented in practice is unknown. Intentions might be seen as placing an upper limit on responses, such that if a respondent fails to state an intention, this is unlikely to be seen in practice. An alternative approach would have been to ask the respondent how they had treated actual people they knew with mental health problems. However, the disadvantage of this alternative would have been the lack of standard situations. A second limitation is that the inter-rater reliability of coding some of the first aid responses was low. Conclusions about these responses must be viewed with caution. On the other hand, the strengths of the study are the large representative sample, the open-ended responses which did not constrain the respondents to particular alternatives, and the ability to compare responses to a series of standard scenarios. Conclusions There is room for improving the range of mental health first aid responses in the community. Lack of knowledge of mental health and stigmatizing attitudes are important barriers to effective first aid. Competing interests The author(s) declare that they have no competing interests. Authors' contributions AFJ was involved in securing funding for the survey, had a major role in the design of the survey and the interview questionnaire, carried out the descriptive statistical analysis and had a major role in writing the manuscript. KAB provided research assistance with the survey, coded all the responses, and wrote the method relevant to this coding. KMG was involved in the design of the study, developed the stigma scales and wrote some of the manuscript. BAK developed the coding scheme, was the second rater for inter-rater reliability, and wrote some of the manuscript. RAP did the regression analyses and wrote the section of the Method describing this. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549043.xml |
549719 | The functional landscape of mouse gene expression | Background Large-scale quantitative analysis of transcriptional co-expression has been used to dissect regulatory networks and to predict the functions of new genes discovered by genome sequencing in model organisms such as yeast. Although the idea that tissue-specific expression is indicative of gene function in mammals is widely accepted, it has not been objectively tested nor compared with the related but distinct strategy of correlating gene co-expression as a means to predict gene function. Results We generated microarray expression data for nearly 40,000 known and predicted mRNAs in 55 mouse tissues, using custom-built oligonucleotide arrays. We show that quantitative transcriptional co-expression is a powerful predictor of gene function. Hundreds of functional categories, as defined by Gene Ontology 'Biological Processes', are associated with characteristic expression patterns across all tissues, including categories that bear no overt relationship to the tissue of origin. In contrast, simple tissue-specific restriction of expression is a poor predictor of which genes are in which functional categories. As an example, the highly conserved mouse gene PWP1 is widely expressed across different tissues but is co-expressed with many RNA-processing genes; we show that the uncharacterized yeast homolog of PWP1 is required for rRNA biogenesis. Conclusions We conclude that 'functional genomics' strategies based on quantitative transcriptional co-expression will be as fruitful in mammals as they have been in simpler organisms, and that transcriptional control of mammalian physiology is more modular than is generally appreciated. Our data and analyses provide a public resource for mammalian functional genomics. | Background Tissue-specific gene expression has traditionally been viewed as a predictor of tissue-specific function: for example, genes specifically expressed in the eye are likely to be involved in vision. But microarray analysis in model organisms such as yeast and Caenorhabditis elegans has established that coordinate transcriptional regulation of functionally related genes occurs on a broader scale than was previously recognized, encompassing at least half of all cellular processes in yeast [ 1 - 5 ]. Consequently, gene expression patterns can be used to predict gene functions, thereby providing a starting point for the directed and systematic experimental characterization of novel genes [ 1 - 10 ]. As an example, it was observed in yeast that a group of more than 200 genes involved primarily in RNA processing and ribosome biogenesis is transcriptionally co-regulated, in addition to being constitutively expressed at some level [ 11 ]. Application of statistical inference methods led to the prediction that the uncharacterized genes in this co-regulated group were likely to be involved in RNA processing and/or ribosome biogenesis [ 5 , 9 ]. Subsequent experimental analysis using yeast mutants validated that many of these predictions were in fact accurate [ 9 ]. To date, this approach has only been extensively applied to relatively simple model organisms such as yeast and C. elegans . Its general utility in mammals has not yet been established with respect to the proportion of either genes or functional categories to which it can be effectively applied. Nor has it been formally examined how the use of quantitative transcriptional co-expression for inference of gene function compares to the more traditional approach of inferring functions on the basis of tissue-specific transcription. The extent and precision of hypotheses regarding gene functions that can be drawn from expression analysis in mammals is an important and timely question, given the current absence of knowledge of the physiological functions of at least half of all mammalian genes. Given that distinct and coordinate expression of a group of functionally related genes implies an underlying pathway-specific transcriptional regulatory mechanism, identification of such instances would also represent a step towards delineating mammalian transcriptional networks. Here, in order to demarcate the general utility of using gene expression patterns to infer mammalian gene functions, and to use this information to begin characterizing genes discovered by sequencing the mouse genome [ 12 ], we used custom-built DNA oligonucleotide microarrays to generate an expression data set for nearly 40,000 known and predicted mouse mRNAs across 55 diverse tissues. Several criteria show that these data are reliable and consistent with other information about gene expression and tissue function. Cross-validation results from machine-learning algorithms show that patterns of gene co-expression within many functional categories are 'learnable' and distinct from patterns of other categories, thus proving that many functional categories are transcriptionally co-expressed and likely to be co-regulated. In contrast, tissue-specificity alone is a comparatively poor predictor of gene function, illustrating the importance of quantitative gene expression measurements. To exemplify this, we functionally characterized the highly conserved gene PWP1 , which is widely expressed. PWP1 is co-expressed with many RNA-processing genes in mouse, and we show that its yeast homolog is required for rRNA biogenesis. The data and the associated analyses in this paper will be invaluable for directing experimental characterization of gene functions in mammals, as well as for dissecting the mammalian transcriptional regulatory hierarchy. Results Expression analysis of mouse XM gene sequences In order to generate an extensive survey of mammalian gene expression, we analyzed mRNA abundance in 55 mouse tissues using custom-designed microarrays of 60-mer oligonucleotides [ 13 ] corresponding to 41,699 known and predicted mRNAs identified in the draft mouse genome sequence using gene-finding programs [ 12 , 14 ] (NCBI 'XM' sequences; approximately 39,309 are unique; for further details, see the Materials and methods section). Tissue collection was a collaborative effort among several labs in the Toronto area, each with expertise in distinct areas of physiology; consequently, the mouse tissues we analyzed were obtained from several different strains of mice which are typically used to study specific organs and cell types of interest (additional cell lines and fractionated cells from animals were also analyzed, but the results are not included here because the data appear to bear little relationship to the tissue of origin of the cells examined). Since it has previously been established that there is a high correlation in expression of orthologous genes between mice and humans [ 15 ], large variations in tissue-specific expression should not occur between individuals within the same species, although we cannot rule out subtle strain-specific differences. To maximize the fidelity of measurements, unamplified cDNA from at least 1 μg of polyA-purified mRNA was hybridized to each array, with fluor-reversed duplicates performed in each case. For most organs this required pooling RNA from multiple animals; for example, more than 50 mice were required to obtain sufficient prostate mRNA. Consequently, potential variations due to parameters such as circadian rhythms or individual dissections should have been minimized by averaging over multiple animals. All hybridizations were performed in duplicate. Data processing and normalization are described in detail in the Materials and methods. The data were processed so that each measurement reflects the abundance of each transcript in each tissue relative to the median expression across all 55 tissues; although the microarray spot intensities were used to determine which genes were detected as expressed (see below), the figures herein show the normalized, arcsinh-transformed and median-subtracted data, which for convenience we refer to as ratios. All of the data, together with tables detailing correspondence to genes in other cDNA and EST databases, annotations and other features of the encoded proteins, probe sequences, and other files used in our analyses below, are available as Additional data files and without restriction on our website [ 16 ]. Validation of expression data Four lines of evidence support the quality of our data and its consistency with existing knowledge of mammalian physiology and gene expression. First, we detected the expected patterns of expression for genes previously shown to be expressed specifically in each of the 55 tissues surveyed (Figure 1 ). This validates the accuracy of our dissections, and indicates that there was little cross-contamination between tissue samples. Second, there is a clear correspondence, albeit not absolute, between our data and two other mouse microarray data sets [ 15 , 17 ], which surveyed a subset of the genes and tissues that we have examined. Thirteen tissues and 1,109 genes were unambiguously shared among the three studies (Figure 2a ). Our data are more highly correlated with those of Su et al. [ 15 ], who also employed oligonucleotide array technology, whereas Bono et al. [ 17 ] used spotted cDNAs (Figure 2a ). Furthermore, our data are more highly correlated with either of the two other studies than the two other studies are to one another. (It should be noted that these previous studies did not examine the use of transcriptional co-expression to predict gene function, which is the focus of the present study.) Third, our array data are consistent with RT-PCR analysis. We tested for expected tissue-specific expression of 107 genes (a mixture of characterized and uncharacterized) in 18 selected tissues. In this analysis a single primer pair was tested for each gene. (It is possible that the predicted exon structures for many of the poorly characterized XM genes are incorrect: there was a clear correspondence between whether a product was obtained and whether there was an EST or cDNA in the public databases, which would indicate correct gene structures – see Materials and methods.) Among the 55 primer pairs that could result in amplification, 53 (96%) gave a correct-product size in the tissue(s) expected on the basis of our array data, and 47 (85%) produced amplification most strongly or exclusively in the expected tissue(s) (Figure 2b and data not shown). Although RT-PCR is semi-quantitative, there is an obvious correspondence between the left and right panels in Figure 2b , confirming that our microarray measurements are largely consistent with a more conventional expression analysis method. Fourth, in the analyses detailed in the following sections, we show that the annotations of genes expressed preferentially in each tissue correspond in many cases to known physiological functions of the tissue, further confirming the accuracy of the dissections and the microarray measurements. Moreover, sets of functionally related genes were often observed to display uniform expression profiles, a result that is highly unlikely to occur by chance. Definition of 21,622 confidently detected transcripts In order to establish rigorously which genes are expressed in each tissue sample, we used the 66 negative-control spots on our arrays (corresponding to 30 randomly generated sequences, 31 mouse intergenic or intronic regions, and five yeast genes). We considered the XM genes to be 'expressed' only if their intensity exceeded the 99th percentile (that is, all but 1%) of intensities from the negative controls (Figure 3a ). 21,622 transcripts satisfied this criterion in at least one sample. There were 1,790 transcripts that were detected in every sample, and manual inspection verified that many of these have traditional 'housekeeping' functions (for example, ribosomal proteins, actin and tubulin). There were 4,475 transcripts detected in only one of the 55 samples (Figure 3b ). Most of the 21,622 genes, however, were expressed in multiple tissues (Figure 3b ). Each of the tissues expressed fewer than half of the 21,622 genes (Figure 3c ). The number of genes detected in each sample was slightly lower than the conventional estimate of 10,000 genes expressed per cell (for example, we detected 6,094 different transcripts in embryonic stem (ES) cells, the only pure cell population examined, whereas a recent study using sequence tags indicated approximately 8,400 different transcripts in human ES cells [ 18 ]). This level of detection is not unexpected, for several reasons. First, tissues are mixtures of cell types, such that low-abundance, cell-type-specific transcripts may be diluted below the array detection limits of 1 in 1,000,000 [ 13 ]; second, the arrays did not include every single mouse gene; and third, our threshold for expression was conservative. The full 21,622 × 55 data matrix is found in the Additional data files. Figure 4a shows a clustering analysis of the 21,622 expressed genes in the 55 surveyed tissues, which illustrates that distinct tissues with related physiological roles also tend to have similar overall gene expression profiles. For example, all components of the nervous system featured higher expression of a common subset of transcripts, as did all components of the lower digestive tract. Correspondence between gene and tissue function To examine the relationships among tissues and gene functions, we asked whether genes carrying specific Gene Ontology 'Biological Process' (GO-BP) categories, which reflect the physiological function of a gene, were preferentially expressed in each of the tissue samples, using a statistical test (Wilcoxon-Mann-Whitney; WMW). A selection of the WMW scores are shown in Figure 4b , and expression patterns of all genes in all GO-BP categories can be seen in the Additional data files and at the Toronto gene expressions website [ 19 ]. This analysis revealed that the preferentially expressed GO-BP categories typically reflected known functions of the tissue, sometimes with surprising resolution. For example, while the category 'synaptic transmission' scored highly in all neuronal tissues, 'learning and memory' was highest in cortex and striatum; 'locomotor behavior' was highest in cortex, midbrain, and spinal cord; 'response to temperature', in the trigeminal nucleus of the brainstem; and 'neurogenesis', in both adult central nervous system and embryonic heads (Figure 4d ). While the WMW test may not have captured all of the categories relevant to each brain tissue, this finding does illustrate that our data contain differential expression of genes involved in distinct high-level neural functions. Further investigation of several tissue-associated GO-BP categories that were initially unanticipated revealed that they are easily rationalized; for instance, lung, bladder, skin, and intestines all express immune-related categories, presumably because they are exposed to the environment and infiltrated by immune cells (see for example, [ 20 ]). Correspondence between gene function and transcriptional co-expression An alternative way to ask whether gene regulation corresponds to gene function is to examine the correlations among the transcript levels of genes, independent of the tissue-source information. An initial confirmation that patterns of transcript abundance correspond to gene functions comes from simply examining the behavior of all genes within distinct functional categories. For example, Figure 5 shows the expression of individual genes in 17 categories that exemplify ways in which gene expression relates to gene function (similar diagrams for all GO-BP categories can be seen in the Additional data files and at the Toronto gene expressions website [ 19 ]). There are prominent patterns that are distinctive of a subset of genes in each category. The fact that not all of the genes within each annotation category conformed to a single pattern could result from imperfections in the annotations or the measurements, or could be due to the correspondence between gene function and gene expression being less than absolute. While highly tissue-specific expression of genes in a category was observed in some cases (such as 'pregnancy' genes in placenta or 'fertilization' genes in testis), it was much more common that genes within a category were expressed across multiple functionally related tissues (for example, 'bone remodeling' in all bone tissues), consistent with the results shown in Figure 4b . In other instances, genes within a single annotation category were subdivided into multiple expression patterns: for example, 'cell-cell adhesion' contains three distinct groups of genes with elevated expression in skin-containing samples, neural tissues, and digestive tract, respectively. Consistent with a previous study [ 21 ], we observed coordinate regulation of genes within distinct biochemical pathways; Figure 5 includes the examples 'polyamine biosynthesis' and 'serine biosynthesis'. Moreover, a number of functional categories corresponding to basic cellular or biochemical functions which are traditionally thought of as 'housekeeping' (since they are required for cell viability) were in fact coordinately regulated across tissues: Figure 5 shows genes in the category 'RNA splicing', which are expressed most highly in neural and embryonic tissues, perhaps reflecting the higher levels of gene expression and alternative mRNA splicing known to occur in these tissues. Interestingly, subsets of genes in the categories 'cytokinesis', 'microtubule-based movement', 'oxidative phosphorylation', and 'M phase', all of which might be considered as central to cellular physiology, were also expressed in distinctive patterns among mouse tissues. We also asked more generally whether groups of co-expressed transcripts were associated with specific GO-BP categories. Figure 4c shows that this is indeed the case: any given 'cluster' of genes with correlated expression levels is more likely than not to be associated with a local enrichment of one or a few annotation categories, and manual analysis suggests that tissue-specific expression often reflects the known physiological role(s) of the tissues in which the genes are expressed (examples are shown in Figure 4d ). False-discovery rate analysis (see the Materials and methods section) confirmed that over 58% of the 21,622 genes were co-regulated with a set of genes significantly enriched for at least one GO-BP category. For the 7,387 GO-BP annotated genes, over 66% were co-expressed with a set of genes significantly enriched for at least one GO-BP category; in over 25% of these instances, the most significant category was one of its existing annotations. Random permutation analysis (that is, repeating the analysis with randomized gene identities) established a false discovery rate [ 22 ] of less than 1% for these analyses (see Materials and methods for details). Hence, quantitative co-expression of functionally related genes appears to be a general phenomenon in mammals. Using transcriptional co-expression to predict mouse gene functions It stands to reason that a gene expressed in a specific tissue is likely to be functioning in that tissue. Therefore, we next asked how accurately mammalian gene functions can be predicted on the basis of gene expression profiles. There are many anecdotal examples in which the tissue-specific or cell-type-specific expression of a gene has been used to aid in discovering its function, and this approach has been advocated in previous analyses of mouse tissue expression data (see for example, [ 15 ]). Our data indicate that the expression of most mouse genes shows some degree of tissue restriction, but most of the genes are not expressed in a highly tissue-specific manner (Figure 3b ). Furthermore, most tissues express genes from multiple functional categories (Figure 4b ), and genes from many functional categories are expressed across many tissues (Figure 5 ), which could make it difficult to distinguish genes in these categories on the basis of expression in one or a few tissues. In addition, defining tissue specificity involves drawing thresholds to form lists, rather than using the quantitative expression information directly to draw functional inferences. An alternative strategy is to generate functional predictions on the basis of transcriptional co-expression [ 23 , 24 ], which we show (above) often reflects gene function (Figure 5 ). This approach utilizes quantitative measurements and places no restriction on tissue-specificity, allowing all expressed genes to be treated equally in the analysis. Furthermore, the use of quantitative co-expression allows the application of sophisticated computational tools that have been optimized for the general problem of classification on the basis of features within a data matrix [ 25 ]. We examined the extent to which this approach is effective for our data, and we show (below) that it yields almost universally superior predictions of gene function in comparison to using information regarding simple tissue specificity or tissue restriction. In this analysis, we used support vector machines (SVMs) [ 26 ]. An SVM is a machine-learning algorithm (a computer program) that has previously been shown to work well for the prediction of gene functions in yeast on the basis of microarray expression data [ 25 ] but which has not, to our knowledge, been used extensively to predict gene functions from mammalian expression-profiling data. The theory and implementation of SVMs have been described elsewhere in detail [ 25 , 26 ]. Briefly, an SVM outputs a 'discriminant value' for each gene in each category, and this value reflects relative confidence that the gene is in the category in question. The SVM considers each functional category separately, and the discriminant value is assigned on the basis of where the gene lies relative to other genes within the 'gene expression space' (for example, analysis of 55 samples results in 55 different coordinates). If the gene lies in a region where there is a high proportion of genes that are known to be in the category in question, this will lead to a high discriminant value. SVMs are conceptually related to clustering analysis in the sense that the discriminant values are derived from similarity among expression profiles. But in clustering analysis, genes are grouped solely on the basis of their expression levels; in contrast, SVMs use the known classifications (that is, knowledge regarding which genes are in the category and which are not) in order to map the initial gene expression space into a one-dimensional space (the discriminant values) in which the two classes are optimally distinguished. Importantly, the discriminant values output by an SVM can be processed to obtain an estimate of the probability that the prediction for each gene in each category is correct (that is, an estimate of precision), on the basis of how well previously annotated genes in the given category can be distinguished from previously annotated genes that are not in the category. This is accomplished by a three-fold cross-validation strategy, in which the analysis is run three times, each time with a different one-third of the annotations masked so that the SVM algorithm does not know whether or not they are in the category when it is assigning discriminant values. Any given discriminant value is then converted to a precision value by simply asking what proportion of the masked genes with discriminant values above the given discriminant value really are in the category in question. The proportion of known genes in the category that are identified by the SVM as being in the category is also obtained at each discriminant value, and is referred to as recall. For all subsequent analyses we used precision and recall as our primary measures of success. We trained separate SVMs for each of the 992 GO-BP categories. This revealed that genes in hundreds of categories could be recognized with precision greater than 50% (Figure 6a ). Typically, not all of the genes in a category could be recognized (the curves in Figure 6a correspond to recall of 10% through 40%); this is due to the fact that not all genes within any given category display the characteristic expression pattern (Figure 5 ). As a control, when the gene labels were randomized, only zero to fifteen categories (depending on the randomization run) achieved 10% precision and 10% recall simultaneously (black dotted line at the bottom of Figure 6a ). Therefore, this analysis demonstrates that, in a blind test, the known genes in many functional categories can be distinguished on the basis of the expression profiles of other genes that are members of the same functional category. This implies that there are distinct regulatory mechanisms that control these pathways, and indicates that correlation-based methods can be used to predict the functions of uncharacterized genes in mammals. Predicted functions for unannotated genes are supported by sequence features We next used these trained SVMs (Figure 6a ) to predict functions for the 12,123 unannotated genes for which we detected expression in our data. The number of genes with at least one predicted function (that is, one GO-BP category) is shown in Figure 6b at varying precision thresholds (blue line). All of the predictions with precision above 15% are listed in the Additional data files. To make the outputs easier to peruse manually, we grouped 587 GO categories into 231 'superGO' categories, by combining categories that resulted in the same set of predicted genes and that were manually verified to be physiologically related. Figure 6b (red line) confirms that the number of unannotated genes that are predicted to have some function by an SVM with 'superGO' categories are similar to those with the original GO categories, although the number of categories has been compressed. In order to provide a set of 'highest priority' predictions, we singled out those with the highest estimated precision. Among the unannotated genes (that is, those carrying no annotation in GO-BP), 1,092 (representing 117 superGO categories) were associated with precision values of 50% or greater; thus, on the basis of the analysis above, each of these genes is more than 50% likely to be involved in the given biological process. Figure 7 shows the original microarray data for these 1,092 genes, sorted by the predicted categories. Predictions were made for genes expressed in all of the tissues analyzed, and represent a wide spectrum of biological processes. While some predictions correspond to expression in a single tissue (for example, the 56 genes predicted in 'vision' were predominantly expressed in the eye), such cases were unusual. Rather, most of the predictions were based on expression in multiple functionally related tissues (for example, the five genes predicted in 'regulation of cell migration' were characterized primarily by high expression in colon, large intestine, and small intestine) or more complex patterns (for example, genes predicted in 'CNS/brain development' were preferentially expressed in all adult neural tissues as well as in embryonic heads). Many predictions were found to be in categories related to the cell cycle and RNA processing. These genes tended to be expressed constitutively, but were most highly expressed in embryonic tissues, presumably because of rapid cell growth during development. However, many other predictions relate to neural functions, the immune response, muscle contraction, small-molecule metabolism, and other aspects of adult physiology. All of the individual predictions are provided in a table in the Additional data files, together with the expected precision and other information regarding the gene and the encoded protein, and these can be sorted by gene or by functional category. Among the 1,092 unannotated genes, 488 (45%) have no overt sequence features suggesting physiological or biochemical function (that is, they have no similarity to previously characterized proteins or known functional domains; they are listed in Additional data files; and also see Materials and methods). Examination of the remaining 55% provided evidence that many of the predictions are likely to be correct. First, a handful of genes that were not annotated in our version of GO have in fact been characterized in the literature. For example, SVMs correctly predicted that phospholamban, the regulator of the Ca 2+ -ATPase in cardiac sarcoplasmic reticulum [ 27 ] is involved in 'muscle contraction or development'. Other genes are similar to characterized genes in other species: for example, the mouse homolog of the yeast 'Extra Spindle Poles' ( ESP1 ) gene was predicted by SVM to function in 'mitotic cell cycle', 'cytokinesis', and 'microtubule based process', consistent with the function of its yeast counterpart [ 28 ]. A more comprehensive and objective analysis was enabled by the fact that, in an independent sequence-based analysis we conducted (see Materials and methods), known protein domain structures were encoded by 461 (42%) of these 1,092 unannotated genes (listed in the Additional data; see also the Materials and methods section) [ 29 ]. These provided further independent support for many of the predictions, since neither the primary sequences nor the domain features of the unannotated genes played a part in the predictions. In many cases, the domains also augment the predicted physiological function with a potential biochemical mechanism. For example, 3 of the 11 genes predicted in the category 'acyl-CoA/fatty acid/peroxisome' encode a short-chain dehydrogenase motif, suggesting that they are metabolic enzymes. Among the 86 unannotated genes predicted to function in 'microtubule-based process' are 4 with chromosome-segregation ATPase domains, one with an intermediate filament protein domain, one with a kinesin-motor domain, one with a myosin heavy-chain domain, and one with a tropomodulin domain, all of which are consistent with microtubule- and/or cytoskeleton-related functions. Of the four proteins predicted in 'skeletal development', one encodes a fibrillar collagen carboxy-terminal domain, and one encodes a collagen triple-helix repeat. Some of the relationships between predictions and domains are striking on the simple basis of their numbers: 7 of the 95 genes predicted in 'humoral immune response' encode an immunoglobulin domain; 13 of the 87 genes predicted in 'chromosome organization/DNA packaging' have high mobility group (HMG) domains, and 23 of the 149 genes predicted in 'RNA processing/ribosome biogenesis' encode helicase domains, RNA-binding domains, or RNA-modifying motifs. Table 1 lists a selection of statistically significant associations between the different prediction classes shown in Figure 7 and protein domains. Comparisons among data sets for predicting gene functions Although there was a significant correlation among the three different mouse tissue-specific data sets compared in Figure 2a , there were also many cases in which the three data sets disagreed in their assessment of relative abundance of individual genes in different tissues (Figure 2a and data not shown). We reasoned that the SVM cross-validation analysis could provide an objective measure of the quality of the different data sets: since random measurements lead to very poor predictions (Figure 6a ), any errors in the data would tend to degrade the precision and recall values. While our manuscript was in preparation, an additional data set was released by Su et al. , the authors of reference [ 15 ]. Their newer data [ 30 ] include measurements of 36,182 known and predicted genes over 61 tissues, measured in duplicate using custom-built Affymetrix arrays, and are thus similar in scope to our data set. Figure 6c shows a comparison between cross-validation results from running SVMs on the three data sets: ours, that of Su et al. [ 30 ], and that of Bono et al. [ 17 ], with each restricted to the 13 tissues and 1,800 genes common to all three, and the same GO-BP annotations used for all three data sets. Figure 6c shows that, although our data fare slightly better, the power of our data set and that of Su et al. [ 30 ] for predicting GO-BP categories are comparable. This confirms that distinct and coordinate regulation of many mammalian functional pathways is authentic because it is observed in two independent data sets. Comparison of tissue-specificity with co-expression for predicting gene functions We used two different approaches to ask how well tissue specificity can predict the functional classes of genes, in comparison to co-expression. First, from our data we compiled three sets of lists: genes that are expressed in each of the 55 individual samples; genes that are expressed highest in each of the individual 55 samples and in groups of functionally related samples (for example, treating all neural tissues as a single group); and also genes that are expressed uniquely in individual samples. All of these lists (175 in total) are compiled in the Additional data files. For each of the 992 GO-BP categories, we assessed the precision and recall for each of these lists (that is, whether these lists can distinguish genes in the category from those not in the category), and then identified the best precision value and its associated recall value for that category. Figure 6d shows a histogram of the difference between SVM precision and tissue-specificity precision, at the same recall value, for each GO-BP category. The vast majority of data points are greater than zero ( P < 10 -76 ; two-sided pairwise t test), indicating that co-expression patterns can be used (by SVMs) to predict gene functions significantly better than tissue-specificity alone. It is possible that improved results might be obtained by other ad hoc procedures for sorting the genes in different ways, or by more automated procedures for generating large numbers of lists. However, an alternative analysis suggests that this is unlikely: when we re-ran SVMs with the matrix of 1s and 0s indicating which gene is expressed (or not) in each tissue, rather than the matrix of quantitative expression values, the resulting predictions were inferior (dotted magenta line in Figure 6a ). In theory, if any combination of on/off information about gene expression in different tissues was informative for identifying genes in any category, it would have been identified by the SVMs. The result we obtained indicates that the quantitative measurements contain critical information reflecting functions of genes that is not, for the most part, contained in the binary (expressed/not expressed) information. Validation of predictions by de novo functional analysis Finally, we asked whether new functional predictions could be confirmed by directed experimentation. Among the genes we predicted to function in RNA processing and ribosome biogenesis was PWP1 , which encodes a protein that includes WD40 repeats and which has previously been found to be up-regulated in pancreatic cancer tissue [ 31 ]. In our data, PWP1 was most highly expressed in embryonic tissues, as is characteristic of most genes annotated as 'RNA processing' by GO-BP (Figure 8a ). The encoded protein Pwp1p is highly conserved across eukarya (Figure 8b ) but to our knowledge it has not been functionally characterized in any species, although it has been found in the human nucleolus [ 32 ], and in yeast it is essential for cell growth [ 33 ]. We created a titratable-promoter allele of yeast PWP1 , and found that cells depleted for Pwp1p displayed a striking reduction in 25S rRNA (Figure 8c ), confirming the involvement of this gene in RNA processing and ribosome biogenesis. Given that WD40 repeats are thought to be protein interaction domains, we also asked whether Pwp1p physically associates with other proteins. We found that epitope-tagged yeast Pwp1 protein co-purified with known trans -acting ribosome biogenesis factors, as well as with several ribosomal protein subunits (Figure 8d ), consistent with a direct role in ribosome biosynthesis. Discussion Simultaneous gene discovery, network mapping, and functional inference The data presented here and the resulting inferences for the physiological roles of mammalian genes significantly extend previous microarray-based analyses of mammalian gene expression [ 7 , 15 , 17 , 21 , 23 , 24 , 30 ]. First, the data support the notion that there are thousands of mouse genes that are not represented in current cDNA databases [ 12 , 34 - 36 ]. Amongst all 21,622 confidently detected transcripts (Figure 4 ), 5,600 were not associated with a cDNA; 3,551 of these had EST but not cDNA support, indicating that many of them correspond to bona fide genes (Figure 2b ). Moreover, inferences for the physiological roles of these transcripts can be obtained by analysis of quantitative expression levels across tissues; the 1,092 unannotated genes for which we made high-confidence predictions (Figure 7 ) contain 54 with no EST or cDNA support, and an additional 114 that have only EST support. Second, our estimate that more than 58% of all transcripts are regulated together with genes in specific functional categories is much higher than previous estimates. One analysis, based on cursory analysis of early yeast and Xenopus expression data, suggested that only 5–10% of all genes fall within 'synexpression' groups [ 2 ]. Our results represent a minimum estimate of the correspondence between gene function and gene expression, because shortcomings in either the annotations or the data would tend to reduce these figures. Our results indicate that there are regulatory pathways that control many distinct biological processes in mammals, and that it is already possible to interpret the expression patterns of the majority of mammalian genes in a functionally meaningful way by comparing them to the patterns of the subset of genes that are already annotated. Moreover, it may be more straightforward than was previously anticipated to apply the same computational techniques to mammalian microarray data that are now being applied to identify 'network modules' and regulatory mechanisms in far simpler organisms, such as yeast (see for example, [ 37 ]). These potential applications represent an obvious future extension of the work presented here. Third, while previous analyses using microarray expression data to predict gene functions [ 1 - 10 ] have focused on the fact that genes in large or general categories can be recognized (for example, ribosome biogenesis, translation, or proteolysis), we show quantitatively that this methodology is applicable to a much wider variety of functional categories, many of which are specific to higher organisms (for example, the category 'Pregnancy/embryo implantation' is specific to mammals; Figures 5 , 7 ). Fourth, our analysis shows that the use of quantitative gene expression measurements to infer mammalian gene functions is more powerful than the traditional approach of using information on simple tissue specificity. Genes in many GO-BP categories were precisely identified by SVM using quantitative co-expression, but not on the basis of tissue specificity or tissue restriction (Figure 6a,6d ). It appears from our results that genes in functional categories with more widespread expression (such as 'epidermal differentiation', 'regulation of cell migration', and 'apoptotic program'), categories corresponding to basic cellular functions (such as 'oxidative phosphorylation', 'RNA processing', and 'DNA replication'), and even categories that describe interactions among different cell types ('taxis', 'glycoprotein metabolism', 'cell-cell adhesion') can all be recognized and distinguished from genes in other categories on the sole basis of their coordinate expression across many tissues (Figures 5 , 7 ), even though they are expressed in many tissues (and in some cases all tissues, as in the case of mRNA splicing and other 'housekeeping' functions). PWP1 is an example of a gene that is widely expressed but has a pattern of expression that was predictive of its function (Figure 8 ). A strategy and resource for mouse functional genomics Analysis of mutant phenotypes is one of the most powerful and definitive ways to study gene functions. Many of the predicted gene functions (see Figure 7 and the Additional data files) in turn predict specific mutant phenotypes; for example, mutation of genes predicted to function in 'vision' would be expected to display defects in sight or eye morphology, while mutation of those predicted to function in 'RNA processing/ribosome biogenesis' might be lethal embryonically, but with alterations in RNA profiles or ribosome content. We have already initiated efforts to validate predicted gene functions in animals: of the XM genes on our array, 2,917 are already represented in collections of publicly available gene-trap ES cell lines [ 38 ] (indicated on the right in Figure 7 ). It will become increasingly straightforward to test these predictions as RNA interference methods are refined and the collection of mapped mouse mutants expands [ 38 - 40 ]. All of our predictions are listed in the Additional data files and will provide guidance for future efforts in mammalian functional genomics and/or support for other functional studies. In contrast to other available mouse tissue-specific data sets (such as those described in [ 30 ]), all of our data, as well as the SVM predictions, can be downloaded anonymously without restriction from the Additional data files or from our website [ 16 ] and can be freely copied, modified, and propagated. The oligonucleotide sequences are provided, so that copies of our array design can be obtained and modified by other labs, and our expression data can be mapped to any clone collection or updated sequence annotation by batch BLAST. Many other supporting files are provided, including GO annotations, maps to gene-trap collections, and genomic locations of the probes. To facilitate perusal of the data, we have also created a web tool [ 19 ] that displays subsets of the gene expression data together with functional information and SVM predictions. This tool currently supports queries originating with a gene of interest, a functional category of interest, or a region of the chromosome of interest, which may facilitate the use of gene expression patterns and predicted gene functions in identifying genes that confer mapped traits [ 41 ]. We anticipate expanding and refining this resource to mirror both additional data and updated annotations. Conclusion We have created an extensive mouse expression data set and asked whether quantitative gene expression patterns correspond to functional gene categories. Our major finding is that most tissues express many functional categories, consistent with the fact that they contain many different cell types performing many different functions, but that many different functional pathways are coordinately expressed in a quantitative manner across tissues such that many categories display one or more distinctive patterns. For example, embryonic heads contain many cell types, and consequently express genes in a variety of categories including 'CNS/brain development', 'M phase', 'skeletal development', and 'microtubule-based process', yet an SVM can distinguish genes in these categories because they are differentially regulated across all 55 tissues in a way that is characteristic for each functional category (Figures 5 , 7 ). The simplest explanation for this observation is that there are discrete factors or sets of factors that control each coordinately regulated pathway. We conclude that functional genomics strategies that rely on quantitative transcriptional co-expression will be as fruitful in mammals as they have been in simpler organisms, and that transcriptional control of mammalian physiology may be more modular than is generally appreciated. Materials and methods Mouse mRNA isolation Mouse tissues were isolated from the following strains: ICR (whole brain, testis, skeletal muscle, heart, lung, liver, embryo at 15 days, embryo at 12.5 days, embryo at 9.5 days, mammary gland, placenta at 9.5 days and placenta at 12.5 days); CD1 (Charles River Laboratory, Wilmington, USA; cortex, cerebellum, striatum, hindbrain, midbrain, bone marrow, knee, teeth, mandible, calvaria, femur (bone marrow flushed diaphyses), tongue surface, snout, large intestine, thyroid, aorta, brown fat, lymph node, olfactory bulb, adrenal gland, prostate, digits, trachea, trigeminal nucleus); C3H (The Jackson Laboratory, Bar Harbor, USA; salivary gland, thymus, ovary, uterus, tongue, stomach, small intestine, spleen, colon, uterus, pancreas, epididymis, eye, bladder, skin); C57BL/6 (The Jackson Laboratory, Bar Harbor, USA; spinal cord); Black Swiss (NTac:NIHBS; Embryonic heads); and R1 (ES cells). With the exception of embryonic tissues and ES cells, tissues were harvested from 3–6 month-old mice. Following recommended University of Toronto protocols, mice were euthanized by barbiturate injection and tissues were dissected as quickly as possible (within 10 minutes), snap-frozen in liquid nitrogen, and preserved at -80°C until use. RNA was extracted using homogenization and Trizol reagent (Invitrogen, Carlsbad, USA) following the instructions from the manufacturer, and mRNA was purified as described previously [ 3 ]. Microarray probe design A FASTA file of 42,192 known and predicted mRNAs (XM sequences) was obtained from Deanna Church at NCBI on July 9, 2002 and is posted as Additional data file 1. Interspersed repeats and low complexity DNA sequences were masked with Repeat-Masker [ 42 ]. The 500 nucleotides from the 3' end of each mRNA were extracted and 10 non-overlapping T m -balanced probes were generated using PrimerX [ 43 ] with default settings. The most unique among the 10 was identified on the basis of having the highest ΔG difference between the first (identical) and second most significant BLAST hits among the 42,192 initial XM mRNAs. Then, 41,699 probe sequences (those for which probes could be designed using this procedure) were submitted for oligonucleotide microarray production (Agilent Technologies, Palo Alto, USA). These arrays are manufactured using an ink-jet process, in which oligonucleotides are synthesized on the array by direct deposition of phosphoramidites [ 13 ]. The specificity, sensitivity, and reproducibility of these 60-mer arrays has been described elsewhere in detail [ 13 ]. Among the probes on the array, 40,822 were unique; those that were not unique can be attributed primarily to gene duplications, predominantly pseudogenes of GAPDH, ribosomal proteins, and retrovirus-like sequences. To minimize the impact of redundancy on statistical analyses, we collapsed the data from 1,928 duplicated probes and XM sequences that were in these sequence families (including 100 probes duplicated between the two array designs) into 525 groups that shared identical probe sequence and/or were both annotated and regulated in the same way. We also mapped all of the XM sequences to the current version of the mouse genome (Build 32) and to three cDNA databases (UniGene, RefSeq, and Fantom II; see below) and identified 1,991 XM sequences in which XM sequences adjacent on the chromosome also mapped to the same cDNA; these were collapsed into 904 groups. The Additional data files include a table mapping the 41,699 probes against the 39,309 presumed distinct transcripts. Labeling and hybridization The mRNA (1–2 μg) was reverse-transcribed with random nonamer primers (1 μg per reaction) and T 18 VN (0.25 μg per reaction) to synthesize cDNA. The reaction contained a 1:1 mixture of 5-(3-aminoallyl) thymidine 5'-triphosphate (Sigma, St. Louis, USA) and thymidine triphosphate (TTP) in place of TTP alone. The cDNA products were bound to QIAquick PCR Purification columns (Qiagen, Hilden, Germany) following the manufacturer's instructions, washed three times with 80% ethanol, and eluted with water. Purified cDNA was reacted with N-hydroxy succinimide esters of Cy3 or Cy5 (Amersham Pharmacia Biotech, Piscataway, USA) following the manufacturer's instructions. Hydroxylamine-quenched Cy-labeled cDNAs were separated from free dye molecules using QIAquick columns. Mixed labeled cDNAs were added to hybridization buffer containing 1 M NaCl, 0.5% sodium sarcosine, 50 mM methyl ethane sulfonate (MES), pH 6.5, 33% formamide and 40 μg herring sperm DNA (Invitrogen, Carlsbad, USA). Hybridizations were carried out in a final volume of 0.5 ml injecting into an Agilent hybridization chamber at 42°C on a rotating platform in a hybridization oven (Robbins Scientific Corporation, Sunnyvale, USA) for 16–24 h. Slides were then washed (rocking for approximately 30 seconds in 6 × SSPE, 0.005% sarcosine, then rocking for approximately 30 seconds in 0.06 × SSPE) and scanned with a 4000A microarray scanner (Axon Instruments, Union City, USA). Hybridizations were performed in duplicate with fluor reversal: that is, each mRNA sample was examined in duplicate, once in the Cy3 channel and once in the Cy5 channel, on separate arrays. Each array was hybridized with two samples simultaneously, each from an individual tissue. Essentially identical results were obtained from single-channel data from the same mRNA sample analyzed on different arrays, which were distinct from individual channels on the same arrays analyzed with a different mRNA. The organization of the hybridizations, and the data for individual channels, are given in the Additional data files. Image processing and normalization TIFF images were quantitated with GenePix (Axon Instruments). Individual channels were spatially detrended (that is, overall correlations between spot intensity and position on the slide were removed) by high-pass filtering (see [ 44 ]) using 10% outliers. We applied variance stabilizing normalization (VSN) [ 45 ] using 25% of the genes to normalize all single channels to each other. We manually identified and removed measurements that were inconsistent between dye-swaps, by either removing data from residual artifacts apparent on microarray images or removing the higher of the two disparate intensity measurements (in order to minimize false-positive detections). Measurements were transformed to arcsinh values (which are similar to natural log values, but are defined for negative numbers which emerge from the VSN) and for each measurement the median across all arrays was subtracted to obtain relative expression ratios for each gene in each tissue compared to all tissues. Remaining inconsistencies between dye-swaps were addressed by removing the higher of any two measurements that differed by more than two arcsinh units (in order to further minimize false-positive detections). The dye-swap arcsinh values were then averaged between replicates and among multiple probes detecting the same sequence. Clustering and manual analysis indicated that ratios below zero were generally not biologically meaningful (and probably stem largely from measurement error among low-intensity spots); hence ratios below zero were set to zero for all analyses using median-subtracted arcsinh values (Figures 1 , 2 , 4 , 5 , 6 , 7 and SVM analyses). Missing values (fewer than 0.01% of all data points) were set to zero. Median-subtracted arcsinh values correspond approximately to the following ratios (arcsinh = linear): 0 = 1/1; 1 = 2.7/1; 2 = 7.5/1; 3 = 20/1, 4 = 55/1; 5 = 155/1, 6 = 405/1. Annotations Mouse GO-BP annotations were downloaded from the Gene Ontology website [ 46 ] and the European Bioinformatics Institute (EBI) [ 47 ] and both were mapped to XM gene sequences by sequence identity to the annotated source sequences. The full annotation database is on our website [ 16 ]. Fewer than 0.01% of these annotations were derived from gene expression (IEP code); we confirmed that removal of these genes had no appreciable impact on statistical analysis or the SVM analysis, and hence the use of these annotations to analyze gene expression is not circular. The Mouse Genome Informatics (MGI) annotations are reported to be manually compiled, whereas the EBI annotations include automated sequence-based annotations (for example, potassium channels are annotated as being in 'ion transport' and the mouse homolog of the yeast Tim8 protein, which is a translocase of the inner mitochondrial membrane, is annotated as being in 'mitochondrial translocation'). All GO-BP annotations were propagated up all possible edges of the GO graph. Redundant GO-BP categories were excluded. Categories with fewer than three genes among the 21,622 expressed genes were excluded from our analysis since they are not appropriate for the statistical tests we used, and those with more than 500 genes were excluded because they are not specific to distinct physiological processes. False-discovery rate analysis Each gene was associated with a co-regulated group consisting of the 50 annotated genes with the highest Pearson correlation coefficient relative to it. Annotation enrichment of this group in each GO-BP category was scored using the hypergeometric P value [ 48 ]. The minimum value of this score across all GO-BP categories was used as the measure for significant enrichment in any GO-BP category. P values were assigned to these measures using a permutation scheme on the gene labels. The statistical significance of the P values was evaluated using the Benjamini-Hochberg (BH) linear step-up procedure [ 22 ] to ensure a false discovery rate (FDR) of less than 1%. For annotated genes, a second measure was computed: the minimum among its annotated categories of the hypergeometric P values of its co-regulated group. A gene-specific permutation scheme associated P values with these scores and the FDR was also controlled at 1%. Cluster diagonalization Starting with an initial hierarchical clustering (agglomerative, average linkage, based on Pearson correlation coefficient), rows were divided into groups by removing a small number of links at the highest levels of the tree and grouping together all rows contained within the same disconnected subtree. Each row group was then associated with the column that contained the maximum expression value averaged over all the profiles in the group. The row groups were then sorted in increasing order of their associated column numbers. Support vector machines We used the SVM software package Gist [ 49 ] version 2.0.8 in Linux with parameter settings '-radial -zeromeanrow -diagfactor 0.5'. Precision was established by three-fold cross validation. Identification of corresponding clones in cDNA and EST databases We identified the closest corresponding mouse mRNAs in FANTOM II [ 50 ] (60,770 sequences); RefSeq [ 51 ] (16,601 sequences); UniGene [ 52 ] (87,495 sequences); and Ensembl [ 53 ] (32,911 sequences) using BLASTN with a threshold of E-60. We identified corresponding mouse mRNAs in dbEST [ 54 ] (3,939,961 sequences) using BLASTN with a threshold of E-20. Identification of genes common to other microarray data and Spearman rank correlations For Figure 2a , mRNA sequences were downloaded from [ 55 ] (for Su et al. data [ 15 ]) and [ 56 ] (for Bono et al. data [ 17 ]). The Su et al. [ 15 ] gene expression data were downloaded from [ 57 ] (9,977 sequences represented on the array) and the Bono et al. [ 17 ] data, from [ 58 ] (54,005 sequences represented on the array). The selected 41,699 NCBI mRNAs were used in a BLAST search against these two mRNA databases; a BLAST comparison between the two databases was also performed to retain only genes for which the closest sequence to each XM gene is also the closest sequence between the two other databases. All BLAST searches were performed with threshold E-60, and the best hit was selected for the multiple blast results. The 1,109 genes that have common hits in all the BLAST results and with gene expression data available were selected for the gene expression analysis. The 1,109 genes from all three datasets were normalized to make them comparable. To facilitate comparison, in the Bono et al. [ 17 ] dataset, each gene was median-centered in each tissue by subtracting its median expression value across all 13 common tissues. The Su et al. [ 15 ] data were arcsinh-transformed before median-centering. The data from the study described here that was used in the comparison was not zeroed, as it was in other analyses, and was median-centered using the median calculated only on the 13 common tissues, rather than all 55. The Spearman rank correlation coefficients of each pair of tissues among all three studies were transformed to Z-scores by multiplication by sqrt(1108) and then converted to P values using the cumulative probability density of a standard normal distribution. For Figure 6c , an alternative mapping strategy was employed: our probe sequences, the Bono et al. [ 17 ] clone sequences, and the Su et al. [ 30 ] probe sequences were associated with 30,832 MGI sequences by mapping directly to corresponding MGI/GenBank sequences; 1,800 genes were identified in which a reciprocal best match between the probe sequences and the MGI sequence was identified in all three studies. RT-PCR Primer pairs were designed to have a matching T m (59°C) and sequences are listed in the Additional data files. RT-PCR assays were performed using the OneStep RT-PCR Kit (Qiagen). Reactions were performed in 25 μl volumes containing 0.5 ng polyA + mRNA, 7.5 units porcine RNAguard (Amersham) and 300 pM each of the forward and reverse primers. After 30 rounds of amplification, the reaction products were separated on 2% agarose gels stained with ethidium bromide. Inverted black-and-white images of the gels were recorded using a Syngene gel documentation system and GeneSnap software (Synopics, Frederick, USA). In total, 107 primer pairs were tested. Of the 57 XM genes tested that corresponded to a known cDNA, 42 were among those that were amplified (74%). Of the 25 tested that corresponded to an EST but not to a known cDNA, 12 were amplified (48%). However, of the 25 tested that did not correspond to a cDNA or EST, only one was amplified (4%). Identification of genes associated with gene traps Six different gene-trap resources were searched to identify genes associated with gene trap ES cell lines. For BayGenomics [ 59 ], Centre for Modeling Human Disease (CMHD) [ 60 ], University of California Resource of Gene Trap Insertions [ 61 ], and Fred Hutchinson Cancer Research Center (FHCRC) [ 62 ], the gene-trap sequence tags were downloaded from the website and searched against the selected 41,699 mRNA sequences using BLASTN. For the German Genetrap Consortium (GGTC) [ 63 ] and Mammalian Functional Genomics Centre (MFGC) [ 64 ], the web-based BLAST servers were used to search the 41,699 mRNA sequences against their gene-trap sequence databases. The hits with lengths equal to or larger than 50 nucleotides, and identity equal to or larger than 98%, were considered to be associated with the gene-trap ES-cell lines. RNA extraction, northern blotting, affinity purification, and mass spectrometry The TetO 7 - PWP1 and isogenic wild-type control strains were created and analyzed as previously described for other essential yeast genes [ 9 ]. Briefly, strains were exposed to 10 μg/ml doxycycline (Sigma) for a total of 24 h before harvesting for RNA extraction. RNA extraction and northern blotting were performed using standard protocols and oligonucleotide probes as described previously [ 9 ]. TAP purification of Pwp1p was performed as previously described [ 9 ] using 1.3l culture volumes; gel-purified proteins were identified by MALDI-TOF mass spectrometry. Additional data files There are 40 Additional data files comprising all the raw data; they are also available on our website [ 16 ]. A web tool for querying and browsing the data online is also available [ 19 ]. Supplementary Material Additional data file 1 XM (predicted mRNA sequences) from NCBI Click here for additional data file Additional data file 2 XP (encoded protein sequences) from NCBI Click here for additional data file Additional data file 3 Array Design 1 (spot map) Click here for additional data file Additional data file 4 Array Design 2 (spot map) Click here for additional data file Additional data file 5 Array Master file - 41,699 probes; the Master file contains probe sequences, columns listing the closest sequences in RIKEN, ENSEMBL, Refseq, and Unigene; GenBank description, EST overlap, GO-BP annotations, Domain (most significant) Click here for additional data file Additional data file 6 Array information after removal of redundant probes: probe combinations Click here for additional data file Additional data file 7 Array information after removal of redundant probes: master file - 39,309 presumed distinct transcripts Click here for additional data file Additional data file 8 Hybridization records Click here for additional data file Additional data file 9 Figure 1 data Click here for additional data file Additional data file 10 Figure 2b data Click here for additional data file Additional data file 11 Figure 4a data Click here for additional data file Additional data file 12 Figure 4b data Click here for additional data file Additional data file 13 Figure 4c data Click here for additional data file Additional data file 14 Figure 4d data Click here for additional data file Additional data file 15 Figure 5 data Click here for additional data file Additional data file 16 Figure 7 data Click here for additional data file Additional data file 17 Supplementary figures S1-3 Click here for additional data file Additional data file 18 Data: 41,699 probes - single channel, arcsinh intensities Click here for additional data file Additional data file 19 Data: 41,699 probes - replicates combined, arcsinh intensities Click here for additional data file Additional data file 20 Data: 41,699 probes - median-subtracted, zeroed Click here for additional data file Additional data file 21 Data: 39,309 presumed distinct transcripts - median-subtracted, zeroed Click here for additional data file Additional data file 22 Data: 21,622 presumed distinct transcripts - median-subtracted and zeroed, expressed above 99% of negative-control spots Click here for additional data file Additional data file 23 Data: binary matrix of expression above 99% of negative-control spots Click here for additional data file Additional data file 24 Annotations: GO annotations among 39,309 presumed distinct transcripts (12,543 annotated genes, 47,900 annotations) Click here for additional data file Additional data file 25 Annotations: GO annotations among 21,622 presumed distinct transcripts (9,499 annotated genes, 37,876 annotations) Click here for additional data file Additional data file 26 Annotations: superGO annotations Click here for additional data file Additional data file 27 Annotations: map between GO and superGO annotations Click here for additional data file Additional data file 28 SVM Predictions: GO-BP SVM Predictions, 15% precision Click here for additional data file Additional data file 29 SVM Predictions: GO-BP SVM Predictions, 50% precision Click here for additional data file Additional data file 30 SVM Predictions: superGO SVM Predictions, 15% precision Click here for additional data file Additional data file 31 SVM Predictions: superGO SVM Predictions, 50% precision Click here for additional data file Additional data file 32 RT-PCR primer sequences Click here for additional data file Additional data file 33 XM genes with gene trap lines Click here for additional data file Additional data file 34 XM motifs Click here for additional data file Additional data file 35 779 known and putative DNA-binding transcription factors among XM genes Click here for additional data file Additional data file 36 Table of accession numbers of genes common to Zhang, Su, and Bono data Click here for additional data file Additional data file 37 Tech report on spatial detrending Click here for additional data file Additional data file 38 Map locations of XM genes (BLASTed against Build 32) - retained only if top hit of both XM sequence and array probe overlap (30,387 presumed distinct transcripts) Click here for additional data file Additional data file 39 SVM functional predictions for 7,147 unannotated mapped transcripts (may be useful for positional cloning), also see [19] Click here for additional data file Additional data file 40 175 lists of genes that are expressed in individual tissues, highest in individual tissues, or specific to individual tissues Click here for additional data file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549719.xml |
506783 | Accounting for regional background and population size in the detection of spatial clusters and outliers using geostatistical filtering and spatial neutral models: the case of lung cancer in Long Island, New York | Background Complete Spatial Randomness (CSR) is the null hypothesis employed by many statistical tests for spatial pattern, such as local cluster or boundary analysis. CSR is however not a relevant null hypothesis for highly complex and organized systems such as those encountered in the environmental and health sciences in which underlying spatial pattern is present. This paper presents a geostatistical approach to filter the noise caused by spatially varying population size and to generate spatially correlated neutral models that account for regional background obtained by geostatistical smoothing of observed mortality rates. These neutral models were used in conjunction with the local Moran statistics to identify spatial clusters and outliers in the geographical distribution of male and female lung cancer in Nassau, Queens, and Suffolk counties, New York, USA. Results We developed a typology of neutral models that progressively relaxes the assumptions of null hypotheses, allowing for the presence of spatial autocorrelation, non-uniform risk, and incorporation of spatially heterogeneous population sizes. Incorporation of spatial autocorrelation led to fewer significant ZIP codes than found in previous studies, confirming earlier claims that CSR can lead to over-identification of the number of significant spatial clusters or outliers. Accounting for population size through geostatistical filtering increased the size of clusters while removing most of the spatial outliers. Integration of regional background into the neutral models yielded substantially different spatial clusters and outliers, leading to the identification of ZIP codes where SMR values significantly depart from their regional background. Conclusion The approach presented in this paper enables researchers to assess geographic relationships using appropriate null hypotheses that account for the background variation extant in real-world systems. In particular, this new methodology allows one to identify geographic pattern above and beyond background variation. The implementation of this approach in spatial statistical software will facilitate the detection of spatial disparities in mortality rates, establishing the rationale for targeted cancer control interventions, including consideration of health services needs, and resource allocation for screening and diagnostic testing. It will allow researchers to systematically evaluate how sensitive their results are to assumptions implicit under alternative null hypotheses. | Background Cancer mortality maps are important tools in health research, allowing the identification of spatial patterns, clusters and disease 'hot spots' that often stimulate research to elucidate causative relationships [ 1 , 2 ]. In most spatial analysis software a statistical pattern recognition approach has been implemented whereby a statistic ( e.g. spatial cluster statistic, autocorrelation statistic, etc .) quantifying a relevant aspect of spatial pattern is calculated. The value of this statistic is then compared to the distribution of that statistic's value under a null spatial model. This provides a probabilistic assessment of how unlikely an observed spatial pattern is under the null hypothesis [ 3 ]. Waller and Jacquez [ 4 ] formalized this approach by identifying five components of a test for spatial pattern. 1. The test statistic quantifies a relevant aspect of spatial pattern ( e.g . Moran's I , Geary's c , LISA, a spatial clustering metric, etc .) 2. The alternative hypothesis describes the spatial pattern that the test is designed to detect. This may be a specific alternative, such as clustering near a focus, or it may be the omnibus "not the null hypothesis". 3. The null hypothesis describes the spatial pattern expected when the alternative hypothesis is false ( e.g . Complete Spatial Randomness, often called CSR). 4. The null spatial model is a mechanism for generating the reference distribution. This may be based on distribution theory, or it may use randomization ( e.g . Monte Carlo) techniques. 5. The reference distribution is the distribution of the test statistic when the null hypothesis is true. CSR is the null hypothesis employed by most, if not all, statistical tests for spatial pattern, and is the workhorse of almost all spatial statistical software. Examples of statistics used in these tests include spatial autocorrelation ( e.g . Moran's I and Geary's c); its local counterpart ( e.g . LISA); geographic boundary statistics ( e.g . boundary count and mean length), and a host of techniques for identifying hot spots, cold spots and foci. While CSR is useful in some situations, it often is not a relevant null hypothesis for highly complex and organized systems such as those encountered in the environmental and health sciences [ 5 , 6 ]. For such fields CSR may be not relevant because spatial randomness rarely, if ever, occurs – some spatial pattern is almost always present. Hence in many situations rejecting CSR has little scientific value because CSR does not describe any plausible state of the system. As emphasized by Ord and Getis [ 7 ], Type I errors may abound when statistical tests are applied without regard to the global autocorrelation structure. For example, locations would be identified as hot spots simply because they lie in areas of generally high (or low) values, which would lead one to blend together local peaks and clusters of high (low) values. Even when health professionals are interested in identifying areas with generally high (or low) disease rates, it is still important to account for spatial autocorrelation to avoid an over-identification of the number of significant spatial clusters or outliers. In summary what are needed are realistic models that incorporate background pattern – the spatial and multivariate structure found when the null hypothesis is true. The term " Neutral Model " captures the notion of a plausible system state that can be used as a reasonable null hypothesis ( e.g . "background variation"). The problem then is to identify spatial patterns above and beyond that incorporated into the neutral model, enabling, for example, the detection of cancer clusters beyond background or regional variation in the risk of developing cancer. Neutral models can be generated using simulation techniques developed in the field of geostatistics [ 8 ] which provides a set of statistical tools for analyzing and mapping data distributed in space and time. In particular, sequential Gaussian simulation (SGS) allows one to generate realizations of the spatial distribution of rates that reproduce the sample histogram and spatial patterns displayed by the data, and also account for any auxiliary data or information on the local trend [ 9 ]. The objective of this paper is to present geostatistical approaches to generating neutral models that account for the spatial dependence of cancer rates, their regional background and spatially heterogeneous population sizes. These models are then used for the detection of local clusters and anomalies in cancer rates. The new methodology is applied to the analysis of the geographical distribution of lung cancer in three counties of Long Island, New York, which have been investigated under the CSR hypothesis in a previous issue of this journal [ 10 , 11 ] Methods Data The use of neutral models in cluster analysis will be illustrated using the lung cancer data analysed in [ 10 , 11 ]. This section briefly summarizes the salient features of this dataset, and readers are referred to the above papers for a detailed description. The New York State Department of Health (NYSDOH) published the cancer incidence data online as part of their Cancer Surveillance Improvement Initiative, . Data have been released on the following four cancers: breast (female only), colorectal (female and male), lung (female and male), and prostate. These data represent newly diagnosed cancer cases in the period 1993–7 assigned to the patient's residence at diagnosis, and they are calculated as the number of cancers for each 100,000 people in the population. To protect patient privacy, the NYSDOH data provided case counts referenced to ZIP codes rather than individual residences. While ZIP codes are somewhat arbitrary spatial units of analysis with respect to potential health and environmental factors, they provide a convenient way to group the population and preserve confidentiality. The methods presented here do not depend on the spatial unit of aggregation and the reader may use census geography if that is their preference. As in the earlier analysis [ 10 , 11 ], the focus of this study is on the 214 ZIP codes within Nassau, Queens and Suffolk County on Long Island. Because cancer incidence is related to age, NYSDOH calculated the expected cancer incidence for each ZIP code using the ZIP code's age structure and the average incidence by age class for New York State (direct adjustment). We thus are using an external standard (the state average) rather than an internal standard (the average for Long Island), to calculate the expected incidence. A standardized morbidity ratio (SMR) has been calculated by dividing the observed value by the age-adjusted expected incidence. An SMR value of 1.0 indicates that the observed incidence is the same as expected, lower than 1.0 indicates that fewer than expected cases of cancer occurred, and greater than 1.0 indicates that more than expected occurred. Local cluster analysis under CSR Jacquez and Greiling [ 10 ] identified significant clustering and spatial outliers in SMR using Anselin's local Moran test [ 12 ] in the ClusterSeer™ Software . The local Moran test evaluates local clustering or spatial autocorrelation by computing the contribution of each location to the Moran's I statistics for the whole study area. Its null hypothesis is that there is no association between SMR values in neighboring ZIP codes. The working (alternative) hypothesis is that spatial clustering exists. For each ZIP code, referenced geographically by its centroid with the vector of spatial coordinates u = (x, y), the LISA (Local Indicator of Spatial Autocorrelation) statistic is computed as: where z( u ) is the SMR for the ZIP code being tested, which is referred to as the "kernel" hereafter. z( u j ) are the values for the J( u ) neighboring ZIP codes that are here defined as units sharing a common border or vertex with the kernel u (1-st order queen adjacencies). All values are standardized using the mean m and standard deviation s of the SMR data (here 214 values). Since the standardized values have zero mean, a negative value for the LISA statistics indicates a spatial outlier where the kernel value is much lower or much higher than the surrounding values (e.g. SMR is below the global zero mean while the neighborhood average is above the global zero mean, or conversely). Cluster of low or high values will lead to positive values of the LISA statistics (e.g. both kernel and neighborhood averages are jointly above zero or below zero). In addition to the sign of the LISA statistics, its magnitude informs on the extent to which kernel and neighborhood values differ. To test whether this difference is significant or not, a Monte Carlo simulation is conducted, which traditionally consists of sampling randomly and without replacement the global distribution of observed rates (i.e. sample histogram), then computing the corresponding simulated neighborhood averages. This operation is repeated many times (e.g. L = 999 draws) and these simulated values are multiplied by the kernel value to produce a set of L simulated values of the LISA statistics at location u : with z (l) ( u j ) = F -1 [p (l) ( u j )], F[.] is the sample cumulative distribution function (cdf), and p (l) ( u j ) is a random number uniformly distributed within 0 and 1. This set represents a numerical approximation of the probability distribution of the LISA statistics at u , under the assumption of spatial independence. The observed LISA statistics, LISA( u ), can then be compared to the probability distribution, allowing the computation of the probability of not rejecting the null hypothesis (so-called p -value). Following Jacquez and Greiling [ 10 ], we used an adjusted significance level α = 0.01101 to account for the fact that the multiple tests (i.e. 214 in this study) are not independent since near ZIP codes share similar neighbors. This significance level was obtained using the Bonferroni adjustment which amounts at dividing the significance level α = 0.05 by the average number of neighbors in each test. Thus, every ZIP code where the p-value is lower than 0.01101 will be classified as a significant spatial outlier (HL: high value surrounded by low values, and LH: low value surrounded by high values) or cluster (HH: high value surrounded by high values, and LL: low value surrounded by low values). A typology of neutral models The use of CSR as the null hypothesis means that the distribution of cancer rates is assumed to be spatially random (no autocorrelation) with uniform risk over the study area. In most cases, however, mortality rates are spatially correlated while the risk of developing cancer varies regionally as a result of changes in environmental exposure or other demographic, social, and economic factors. Another weakness of the above test is that it does not consider whether ratio data are based on many or a few cases, thereby ignoring the instability of rates computed from small population sizes. The basic idea of the proposed approach is to generate neutral models that are more realistic in the sense that they incorporate presence of spatial autocorrelation, non-uniform risk, and account for spatially heterogeneous population sizes. Table 1 provides a typology of neutral models that could be used for inference regarding numerator and denominator, including incidence and prevalence, as well as mortality rates. Model I corresponds to the CSR case, while model II reproduces the spatial correlation of the cancer rates. Model III reflects the situation where environmental exposures or other factors make the risk non-uniform. Models IV through VI allow one to account for the impact of population size on the stability of observed rates. Unlike Model I these more complex neutral models can not be generated simply by shuffling randomly the SMR data across the 214 ZIP codes, and geostatistical simulation techniques to generate each type of model are described below. Table 1 Typology of neutral models. Models differ according to the reproduction of spatial correlation, the incorporation of non-uniform risk, and the filtering of noise caused by spatially varying population sizes. Risk Accounting for Population size No Yes Uniform, spatially random I IV Uniform, spatially correlated II V Heterogeneous, spatially correlated III VI Normal score transform of SMR data The simulation techniques used in this paper assume a multiGaussian distribution for the variable under study, which requires a prior normal score transform of SMR data to ensure that at least their univariate distribution (histogram) is normal. Normal score transform is a graphical transform that allows one to normalize any distribution, regardless of its shape. It can be seen as a correspondence table between equal p -quantiles z p and y p of the z-cdf F(z) (cumulative histogram) and the standard Gaussian cdf G(y). In practice, the normal score transform proceeds in three steps: 1. The N original data z( u α ) (i.e. SMR data) are first ranked in ascending order. Since the normal score transform must be monotonic, ties in z -values must be broken, which has been done randomly as implemented in GSLIB software [ 13 ]. 2. The sample cumulative frequency of the datum z( u α ) with rank k is then computed as = k/N - 0.5/N. 3. The normal score transform of the z -datum with rank k is matched to the -quantile of the standard normal cdf: y ( u α ) = φ ( z ( u α )) = G -1 [ F ( z ( u α ))] = G -1 [ ] Local cluster analysis under spatial neutral model (Model II) Model II aims to reproduce the pattern of spatial correlation displayed by the data that is here quantified using the normal score semivariogram [ 8 , 14 ] which plots the average squared difference between normal score transformed SMR data as a function of the separation distance and direction between ZIP codes: Here | h | corresponds to the Euclidian distance between two ZIP codes. Note the following discussion can be readily generalized to other distance measures that could be more appropriate to capture contiguity of entities of complex shape: our methodology is general and does not depend on a particular formulation of the distance measures. Following previous simulation studies [ 9 ] and in order to account for the noise induced by small population sizes, each pair has been assigned a weight proportional to the square root of the population size, , where n( u α ) is the size of the population at risk in the ZIP code with centroid u α . Following an earlier analysis of the data [ 10 ], the population in ZIP codes was estimated using 2000 US census numbers. Spatial neutral models are generated using Sequential Gaussian Simulation (SGS) which proceeds as follows (see [ 8 ] p. 380 for more details): 1. Fit a permissible function [ 9 ] to the experimental semivariogram (Equation 3). The modeling was here performed using least-square regression [ 15 ]. All semivariogram models were bounded, that is a sill is reached for a given distance referred to as the range of influence. The covariance models were then derived by subtracting the semivariogram model from the sill value. 2. Define a random path (i.e. using a random number generator) visiting each ZIP code location u α only once. 3. At each location u α determine the mean and variance of the Gaussian probability distribution of y-values as: where y (l) ( u i ) are normal scores simulated at locations previously visited along the random path and located within a search radius from u α , m Y is the stationary mean of the variable Y (which is zero following the normal score transform), and C( u i - u α ) is the covariance function of the normal score variable Y for the separation vector h i α = u i - u α . λ i are kriging weights obtained by solving the following system of linear equations (simple kriging, SK): 4. Draw a simulated value from the conditional cumulative distribution function (ccdf) of probability and add it to the data set. In other words, the simulated value at u α is , where p (l) is a random number between 0 and 1. 5. Proceed to the next location along the random path, and repeat the two previous steps. 6. Loop until all N locations (i.e. N = 214 here) are simulated. 7. Transform the simulated normal scores {y (l) ( u α ); α = 1,..., N} so that the target histogram (in this case the global distribution of observed rates, F[.]) is reproduced: z (l) ( u α ) = F -1 [p (l) ( u α )] with p (l) ( u α ) = G[y (l) ( u α )] The procedure is repeated using a different random path and set of random numbers to generate another realization. Note that these realizations account for only the histogram and semivariogram model of the SMR data (global conditioning), but they are non-conditional to the SMR data themselves (e.g. location of zones of high or low SMR values). Once the L sets of N simulated SMR values, {z (l) ( u α ); α = 1,..., N} have been generated, Equation (2) is applied to each member of this set to compute the simulated values of the LISA statistics at each location u . The simulated LISA values form the empirical distribution of the LISA statistics, allowing the calculation of the p-value for the test of hypothesis. Local cluster analysis under a locally constrained spatial neutral model (Model III) The simulation of neutral model II is conducted using a stationary mean for SMR values, which is unrealistic for situations where environmental exposure or other factors make the risk non-uniform. In this instance the researcher wishes to detect spatial pattern above and beyond this non-uniform risk. For example, one might want to detect clusters of melanoma beyond those that are explained by the north-south gradient in solar radiation. Non-uniform risk can easily be accounted for in the simulation procedure by replacing the stationary mean m Y in Equation 4 by locally varying means m Y ( u α ), that is using the following estimate for the mean and variance of the Gaussian ccdf: where C R ( u i - u α ) is the covariance function of the residual normal score variable [Y( u α ) - m Y ( u α )] for the separation vector h i α = u i - u α , and the kriging weights are obtained by solving the following system of linear equations (simple kriging with local means, SKlm): The first step in the generation of model III is the computation of the local means m Y ( u α ), which define the reference background risk the user wants to consider for the null hypothesis. In this paper a smooth model of background risk values was obtained by using the following kriging estimator of the local means of observed SMR data: The kriging weights are calculated in two-steps. First, the following "kriging of the local mean" system [ 8 ] is solved: Then, to incorporate data reliability due to spatially varying population size directly into the geostatistical filter the kriging weights are rescaled, following [ 9 ], as: This rescaling is applied separately to the negative and positive kriging weights, keeping constant the overall contribution of these two sets of weights; that is the sum of positive (negative) kriging weights is the same before and after rescaling, which ensures that the unbiasedness constraint in system (11) is still satisfied. Note that although the population size is incorporated in the estimation of the local means, it is not accounted for directly into the test of hypothesis, which will be achieved using Models IV through VI introduced below. Once the local means of the normal score transformed SMR data have been estimated, they are subtracted from the SMR values and the semivariogram of residuals is estimated and modelled. Then, the simulation is performed using SGS and SKlm. Last, the L realizations are used to derive the empirical probability distribution of the LISA statistics and the p-value of the test is computed. Accounting for population size in local cluster analysis (Models IV to VI) The neutral models introduced so far ignore the fact that cancer rates estimated over small areas, such as United States ZIP code areas or census tracts, tend to be less reliable [ 16 , 17 ], hence larger fluctuations among simulated rates are expected at these locations. If ignored, large differences in population size decrease the ability of Moran's I to detect true clustering. There are essentially three approaches to incorporate population sizes in cluster detection: 1) randomly shuffle the cases rather than the rates (i.e. under a heterogeneous Poisson model the cases are allocated to each area using hypergeometric sampling [ 18 ]), 2) use a modified version of the test statistics (i.e. Oden's I pop [ 19 ] or Waldhör's I [ 20 ]), and 3) transform or standardize the rates first, then compute the LISA statistics on the results (i.e. Empirical Bayes Index [ 21 ], Cressie's transform [14 p.385–402], or any other smoothing algorithm [ 17 , 22 ]). In this paper, the third approach has been adopted and the noise caused by small population sizes was filtered using a variant of the estimator introduced in equation 10: The kriging weights are calculated in two-steps. First, the following system is solved: with g 0 = b 0 × (1- δ ( u i - u α )) where b 0 is the nugget variance in the weighted semivariogram model of SMR data, and δ ( u i - u α ) = 0 if u i = u α and 1 otherwise. Then, to incorporate data reliability (i.e. population size) directly into the geostatistical filter the kriging weights are rescaled according to Equation 12. The ability of the proposed approach to reconstruct the underlying disease risk from observed mortality rates has been tested in extensive simulation studies [ 9 ]. Results and discussion Generating spatial neutral models Figures 1 and 2 (top graphs) show the geographic distribution of lung cancer in males and females (aggregated to the ZIP code level), in Long Island, New-York. Middle graphs show the experimental weighted semivariograms computed in four directions from the normal score transforms of SMR data. For both males and females SMR normal scores exhibit a range of autocorrelation of about 15 km, with smaller variability (i.e. lower semivariogram values) observed along the NW-SE direction. The spatial anisotropy is less pronounced for female lung cancer and an isotropic model was fitted (solid black line). Regional background is further revealed once the noise and short-range variability of SMR data has been removed using factorial kriging (Equation 10) and the semivariogram model fitted to sampled values (solid line in middle graphs), see Figures 1 and 2 (bottom graph). High SMR values are recorded mainly along the Southern shore of the Island for both genders, while differences between males and females are more striking for low value: the lowest SMR values are observed in the westernmost part of Long Island for females and slightly more to the east for the males. These maps of regional background were subtracted from the original SMR maps, and the spatial autocorrelation of the corresponding residuals was quantified using the experimental semivariograms displayed in Figure 3 . Since some of the spatially correlated variability is captured by the regional background, the residual semivariograms show lower sills and shorter ranges relatively to the SMR semivariograms of Figures 1 and 2 . Figure 1 Geographic distribution and spatial variability of male lung cancer. The fill color in each ZIP code represents the SMR, with green indicating relatively low SMR and purple representing relatively high SMR (categories correspond to deciles of the histogram of rates). From these rates, a population-weighted semivariogram is computed in four directions. The semivariogram model (solid line) is used to filter the noise and short-range variability of observed SMR, yielding a smooth map of SMR local means (regional background). Figure 2 Geographic distribution and spatial variability of female lung cancer. The fill color in each ZIP code represents the SMR, with green indicating relatively low SMR and purple representing relatively high SMR (categories correspond to deciles of the histogram of rates). From these rates, a population-weighted semivariogram is computed in four directions. The semivariogram model (solid line) is used to filter the noise and short-range variability of observed SMR, yielding a smooth map of SMR local means (regional background) Figure 3 Residual semivariograms for male and female lung cancer. The regional background displayed at the bottom of Figures 1 and 2 is subtracted from the maps of SMR data, and the spatial variability of these residuals is characterized by population-weighted semivariograms computed in four directions. One hundred realizations of neutral Models I through III were generated using Sequential Gaussian simulation and the information displayed in Figures 1 to 3 . The first two realizations of each model for male lung cancer are displayed in Figure 4 . The two top maps (model I), which were obtained by shuffling randomly the 214 ZIP code SMR data in Figure 1 (top map), illustrate the simplistic nature of CSR as null hypothesis in cluster detection. Spatial patterns are reproduced by the middle maps (Model II) where one notices groups of low and high simulated SMR values the position of which changes from one realization to the next since the simulation is not conditioned locally to the observed rates. The regional background displayed in Figure 1 (bottom graph) is incorporated in Model III, which reduces fluctuations among realizations and led, for example, to high SMR values being consistently simulated in the central part of Long Island. Figure 4 Different neutral models for male lung cancer. The fill color in each ZIP code represents the simulated SMR, with green indicating relatively low SMR and purple representing relatively high SMR (categories correspond to deciles of the histogram of simulated rates). Simulated maps (realizations) of the spatial distribution of lung cancer SMR data are generated under the assumption of complete spatial randomness (Model I), or created using geostatistical simulation in order to reproduce the spatial autocorrelation displayed by observed rates (Model II) as well as the regional background, i.e. SMR local means (Model III). Accounting for population size in spatial neutral models Population in Long Island ZIP codes can vary substantially, ranging from 445 to 105,723, with a mean of 23,298, see Figure 5 (top graph). Population sizes also display a strong spatial pattern, with a gradient from highly populated ZIP codes in the western part of Long Island to the sparsely populated eastern part. The scattergrams in Figure 5 illustrate how the population size impacts the magnitude of fluctuations among SMR data. As the ZIP codes become less populated the variability among SMR values increases, which reflects the smaller reliability of the rates inferred from small populations at risk and makes problematic the later detection of clusters or spatial outliers. Figure 5 Geographic distribution of population size (male + female) and its impact on stability of SMR values. The fill color in each ZIP code represents the 2000 population size, with green indicating sparsely populated ZIP codes and purple representing larger population sizes (categories correspond to deciles of the histogram of sizes). The scatterplots illustrate the larger spread of measured SMR for ZIP codes with low population and how the extreme rates recorded in these ZIP codes are smoothed by geostatistical filtering. Using factorial kriging and the SMR semivariogram models displayed in Figures 1 and 2 , the noise caused by small population sizes was geostatistically filtered from SMR maps: compare filtered maps in Figure 6 with original maps shown at the top of Figures 1 and 2 . While the noise filtering does not change the mean of the SMR data, their standard deviation decreases: 0.290 to 0.238 for males and 0.355 to 0.329 for females. The larger decrease observed for male SMR values is caused by the higher amount of noise (i.e. relative nugget effect) reflected as the discontinuity at the origin of the semivariogram. The scattergrams at the bottom of Figure 5 indicate that the geostatistical filtering changes mainly the extreme rates recorded for sparsely populated ZIP codes. Then, one hundred realizations of neutral Model IV through VI were generated using Sequential Gaussian simulation and the filtered SMR maps statistics. Figure 6 Geostatistical filtering of male and female lung cancer data. The fill color in each ZIP code represents the noise-filtered SMR, with green indicating relatively low SMR and purple representing relatively high SMR (categories correspond to deciles of the histogram of filtered rates). Local cluster analysis under various neutral models Female Figures 7 and 8 show the results of the cluster analysis for female SMR values under the neutral models I through VI, while Table 2 lists the exact number of ZIP codes classified as significant clusters of high values (HH) or low values (LL), and outliers (LH and HL). Table 2 also indicates how the p-value varies among neutral models, highlighting the fact that depending on the assumption being made, the size and locations of clusters/outliers can change. Figure 7 Results of the local cluster analysis conducted for female lung cancer using neutral models I to III. The fill color in each ZIP code represents the classification into significant low-low or high-high clusters, as well as high-low or low-high outliers. Yellow indicates ZIP codes that have not been found significant using an adjusted significance level α = 0.01101. Figure 8 Results of the local cluster analysis conducted for female lung cancer using neutral models IV to VI. The fill color in each ZIP code represents the classification into significant low-low or high-high clusters, as well as high-low or low-high outliers. Yellow indicates ZIP codes that have not been found significant using an adjusted significance level α = 0.01101. Table 2 Number of significant zip codes for the different types of cluster/outliers and neutral models. Results are reported for female lung cancer. Numbers between parentheses indicate zip codes that have similar classification under the reference Model I (CSR). Summary statistics for the p-values are also provided. Neutral Model Type Model I Model II Model III Model IV Model V Model VI High-High 7 0(0) 4(2) 10(5) 0(0) 3(2) High-Low 1 0(0) 2(0) 0(0) 0(0) 1(0) Low-High 2 0(0) 4(0) 2(2) 0(0) 1(0) Low-Low 18 1(0) 2(2) 31(18) 4(4) 6(4) P-value Mean 0.178 0.230 0.405 0.166 0.237 0.394 CV 85.9% 62.5% 71.3% 95.3% 62.2% 76.9% Under the CSR model (Model I), results similar to the ones reported in [ 11 ] were found. First, the local Moran test identified a single, large cluster of low SMR extending through portions of Flushing in the north and Jamaica in the south. Next to this cluster is the only high-low outlier: Oakland Gardens (11364) with a SMR of 1.116. Sayville (11782) is a significant spatial outlier with low SMR (72% of the New York average), though its SMR has a wide confidence interval resulting from the small number of observed cases there (15,896 habitants). Thus, while statistically distinct from its neighbors, it does not have significantly reduced risk. This is also the case for the second low-high outlier, Manorville (11949, 11,384 habitants), which has a SMR close to one but is located in the western part of Long Island where background rates are higher. Several local clusters of high SMR values occurred in the more central portions of Long Island. There is a cluster in north-mid Long Island, made up of two significant local clusters centered on Bayville (11709) and Mill Neck (11765). This cluster has about 60–70% higher SMR than the New York state average. A large cluster in south central Long Island is composed of four local clusters centered on Ronkonkama (11779), Central Islip (11722), Islip Terrace (11752), and East Islip (11730). This cluster has an SMR about 40% higher than the New York state average. Further east is a third cluster of high female lung cancer incidence centered on Mastic (11950) and including several adjacent ZIP codes. Its SMR is about 60% higher than the New York state average. Accounting for spatial autocorrelation (i.e. Model II) leads to a substantial reduction in the size of significant clusters compared to the CSR assumption. In fact only one ZIP code is a significant low-low cluster under Model II: Saint Albans (11412) which was the center of the Southern low-low cluster detected under CSR. The scattergram in Figure 9 (left graph) shows that the use of spatially correlated neutral models leads to larger p-values on average (0.23 vs. 0.18), and those are highly correlated with the ones obtained under CSR (Model I). These larger p-values cause a substantial reduction in the size of significant ZIP codes, since fewer units exceed the adjusted significance level α of 0.01101. The reason for that increase in p-values is illustrated for the ZIP code #11364 (Oakland Gardens) which was the only unit classified as high-low outlier under neutral model I. Figure 10 (left top graph) shows the distribution of simulated values of the LISA statistics for that ZIP code. Clearly, the variance of the distribution is much larger than the results obtained under CSR, while both means are equal to zero. The spatial autocorrelation of simulated rates increases the likelihood that the J neighboring values are jointly small or high, causing the neighborhood average, hence the LISA value, to exhibit much larger fluctuations among realizations. Consequently, the probability that the observed LISA statistics falls in the tails of the simulated distribution decreases, leading to a larger p-value (0.061 versus 0.003) and a ZIP code that is no longer a significant outlier. Figure 9 Scatterplots of the p-values obtained when conducting the local cluster analysis under CSR assumption (Model I) or more complex neutral models. Model III reproduces the pattern of spatial correlation as well as the regional background of SMR values, while Model II accounts only for the spatial correlation. Figure 10 Histograms of the values of the LISA statistics simulated for ZIP code #11364 (Oakland Gardens) under different neutral models. The black dot denotes the observed LISA statistics which lies inside the 0.95 probability interval for all models except Models I and IV developed under the CSR assumption. The map of significant ZIP codes at the bottom of Figure 7 bears little resemblance with the maps obtained under the neutral models I and II. This is expected since Model III addresses a different question, namely the detection of local departures from the regional background. Thus, in general, one would expect HL and LH outliers to be more frequent than spatial clusters HH or LL. Also the local constraining of the neutral models to the regional background causes less variation among realizations, leading to the J neighboring values being consistently either small or large across the realizations. Thus the distribution of 999 simulated LISA values is expected to be narrower than for the two previous models with a shift in the mean. This is illustrated for the ZIP code #11364 in Figure 10 (left bottom graph). Because this unit is located in a low-valued area, the use of neutral models reproducing the regional background yields smaller simulated LISA values (average = -0.14 instead of 0.0). In high-valued areas, the shift is expected to be in the opposite way, leading to a larger range of p-values observed across the area, see the scattergram in Figure 9 (right graph). Table 2 and Figure 9 indicate that the p-values are of larger magnitude (average: 0.405 versus 0.23 for Model II) and weakly correlated with the ones obtained under CSR. For female lung cancer, the same numbers of ZIP codes (6) were classified as significant outliers or clusters under neutral model III. The two low-low clusters are Springfield Gardens (11413) and Saint Albans, which was the only significant unit under neutral model II. These ZIP codes are both located in the western part of Long Island with low background SMR values, and in the same area the following three low-high outliers are found: Bellerose (11426), Little Neck (11362), and New Hyde Park (11040) surrounding the high-high cluster Glen Oaks (11004). The last low-high outlier is found in Shelter Island Heights (11965) in the eastern part of Long Island, though its SMR (72% of the New York average) has a wide confidence interval resulting from the small number of observed cases there (1,080 habitants). The two high-low outliers are found in central Long Island characterized by a low SMR background level: Ridge (11961) and Bayport (11705) with SMR values 20 to 40% higher than the New York state average. Three more clusters of high SMR (1.15 to 1.20) are found in the North western part of Long Island, next to the large group of low SMR recorded in Flushing and Jamaica: Bayville (11709), Mill Neck (11765), and Glen Cove (11542). For all three types of model, accounting for population size through geostatistical filtering leads to a larger number of ZIP codes classified as clusters and fewer outliers, see Figure 8 and Table 2 . This result can be explained by the smoothing of local fluctuations, in particular the ones recorded in sparsely populated ZIP codes, yielding larger and more compact clusters, such as for Model IV. Figure 10 (right column) also shows that this smoothing halves the standard deviation of the distributions of simulated LISA statistics. Major differences between Models I and IV include bigger and more compacts clusters of low and high SMR values, the disappearance of two sparsely populated high-high clusters (Bayville and Mill Neck, with 7,134 and 732 habitants, respectively), and the classification of a former high-low cluster into a low-low cluster (Oakland Gardens) since the filtered rate becomes slightly lower than the global mean. A similar trend is observed for spatially correlated neutral models where the filtering increases the number of significant low-low clusters from one to four, all located in the eastern part of Long Island. The comparison of Models III and IV indicates the disappearance of a few sparsely populated ZIP codes which were classified as spatial outliers prior to filtering: HL (Bayport, 8,006 habitants), LH (Shelter Island Heights, Bellerose and Little Neck, with 1,080, 18,726 and 17,502 habitants). The only remaining LH cluster is New Hyde Park which has 39,156 habitants. The HH cluster (Glen Oaks, 14,682 habitants) also disappeared. Conversely, three other ZIP codes with populations ranging from 776 to 21,282 became significant LL clusters under Model IV: East Marion (11939), Woodbury (11797), and Cambria Heights (11411). Across all six types of neutral models, only one out of 214 ZIP codes is consistently classified into the same category: the low-low cluster at Saint Albans (11412) which has a SMR = 0.82 and a population of 37,452. The stability of this cluster under alternative specifications of the statistical null hypothesis can be used by cancer surveillance and control efforts to quantify the degree of confidence associated with this cancer cluster. Male Results of the cluster analysis for male lung cancer are displayed in Figures 11 and 12 and reported in Table 3 . Model I (CSR assumption) yields the same results as the one reported in [ 11 ]. Three local clusters of low SMR values were identified, centred on Great Neck (ZIP 11024), Roslyn (11576), and Huntington (11743), all in the northwest portion of Long Island. These clusters are typified by lung cancer SMR values that are 50–75% of the New York State average. A large cluster of lung cancer SMR 20–60% higher than the New York average was identified in central Long Island. Cutchogue (11935) was found a significant high-low outlier although its small population (3,444) impacts the reliability of the morbidity ratio. The two low-high outliers are Moriches (11955) and Rockaway Park (11694). Figure 11 Results of the local cluster analysis conducted for male lung cancer using neutral models I to III. The fill color in each ZIP code represents the classification into significant low-low or high-high clusters, as well as high-low or low-high outliers. Yellow indicates ZIP codes that have not been found significant using an adjusted significance level α = 0.01101. Figure 12 Results of the local cluster analysis conducted for male lung cancer using neutral models IV to VI. The fill color in each ZIP code represents the classification into significant low-low or high-high clusters, as well as high-low or low-high outliers. Yellow indicates ZIP codes that have not been found significant using an adjusted significance level α = 0.01101. As for female lung cancer, accounting for spatial autocorrelation (i.e. Model II) leads to fewer significant ZIP codes compared to the common CSR assumption. Only two units are now significant: high-high cluster at Shirley (11967) and a low-high outlier at Rockaway Park (11694). Changes are also substantial when looking at results obtained under neutral model III. We found two high-high clusters: Shirley (11967) and Queens Village (11429), one low-low cluster: Corona (11368), and one low-high outlier: Elmont (11003). Accounting for population size in the cluster analysis (Model IV) enhances the size and compactness of the two major clusters of low and high SMR values, while the classification of three sparsely sampled ZIP codes (Cutchogue, Moriches and Rockaway Park with 3,444, 2,652 and 19,278 habitants respectively) changed from spatial outliers to clusters. A new cluster of high SMR values (SMR = 1.29) is also found in Lindenhurst (11757). Using spatially correlated neutral models the geostatistical filtering (Model V) reveals a new cluster of low SMR values in Port Washington (11050) and Great Neck (11024) with lung cancer SMR values that are 70% of the New York State average. Comparison of Models III and VI indicates that besides increasing the size of clusters identified under Model II geostatistical filtering leads to the identification of a new low-low cluster: Cold Spring Harbor (11724, with a SMR 63% below the New York state average) and one low-high outlier: Springfield Gardens (11413, SMR = 0.80). Across all six types of neutral models, only one out of 214 ZIP codes is consistently classified into the same category: the high-high cluster at Shirley (11967) which has a SMR = 1.157 and a population of 24,942. How many realizations are needed? The use of randomization in test of hypothesis relies on the assumption that the space of solution is sampled fairly exhaustively and uniformly (equally-probable realizations [ 23 ]). It is thus necessary to investigate how conclusions change as a function of the number of neutral models generated. For example, Figure 13 shows the influence of increasing the sample size from 99 to 999 on the average difference in terms of p-value and classification of ZIP codes into significant outliers and clusters (the reference is the results obtained using 99 realizations). All curves exhibit a plateau within this range of sampling intensity, although the asymptotic behavior depends on the type of neutral models. This result indicates that for this case study enough realizations of neutral models were generated to yield stable classifications of ZIP codes. Figure 13 Impact of the number of realizations and type of neutral models on the stability of local cluster analysis results. The left graph displays the absolute value of the average change in p-value as the number of realizations increases from 99 to 999. The right graph shows the number of ZIP codes that are classified differently as the number of realizations increases from 99 to 999. Conclusions Cancer mortality maps are used by public health officials to identify areas of excess and to guide surveillance and control activities. Maps of incidence as well as mortality are used as input to disease clustering procedures whose purpose is to identify local areas of excess and deficit. While some controversy revolves around the utility of these techniques, it is indisputable that the finding of a confirmed cancer cluster is often of considerable concern. The accurate quantification of local excesses and deficits, as well as regional trends and differences in cancer incidence and mortality, is therefore a problem of considerable practical importance. Arguably one of the biggest problems facing spatial epidemiology and exposure assessment is that of identifying geographic pattern ( e.g. hotspots, coldspots, clusters, etc) above and beyond background variation. Most, if not all, environmental contaminants and diseases with potential environmental causes occur at a background level in the absence of a pollution- or disease-generating process. Nonetheless, this background pattern is typically ignored in spatial analyses that employ null hypotheses of complete spatial randomness. Because some spatial dependency is expected at background levels, CSR often is an inappropriate null hypothesis. When should the different types of neutral models be employed? The 6 types of neutral models presented here represent permutations of whether or not population size is accounted for, and 3 types of underlying risk models. As a rule of thumb one should employ that neutral model or those neutral models that most closely correspond to the spatial pattern expected in the absence of the alternative spatial process. So, for a cluster study one would select those neutral models that specify the risk function deemed most likely in the absence of spatial clustering. When working with rates spatial heterogeneity in the size of the at-risk population should always be accounted for, and selections from neutral models of types IV through VI are appropriate. When in doubt about which neutral model to employ, it makes sense to use several in order to determine how sensitive the results are to specification (and misspecification) of the null hypothesis. To the authors' collective knowledge, CSR is rarely if ever encountered in real-world biological systems. It is an apt descriptor of the "snow" that used to appear on late-night television when the programming day was over. It thus seems that neutral model types I and IV will seldom be appropriate. They perhaps will prove most useful for evaluating the extent of bias in past studies that employed CSR. The approach presented in this paper enables researchers to assess geographic relationships using appropriate null hypotheses that account for the spatial correlation and background variation modeled from the observed rates and any ancillary information (i.e. exposure model). An immediate consequence of using more realistic (i.e. spatially correlated) neutral models are larger p-values, leading to a substantial reduction in the number of ZIP codes declared significant outliers or clusters across Long Island. This result confirms earlier findings that CSR often leads to an over-identification of the number of significant spatial clusters or outliers. These false positives have potentially serious consequences in that it can lead to public alarm and demands for investigation by already stretched state health departments. The drop in the number of significant units is however accentuated by the use of an adjusted significance level (Bonferroni adjustment) to account for the correlation between the tests conducted at neighboring ZIP codes. Further research should investigate the redundancy between the use of spatially correlated neutral models and adjusted significance level, which might lead to an "under-identification" of the number of significant spatial clusters or outliers. When the constraint of local conditioning of neutral models is superimposed to the reproduction of spatial autocorrelation (i.e. model III), the approach allows one to detect local departures from the conditioning background specified by the user. In this paper, this background was identified to the regional variability of SMR data which was estimated geostatistically. Future research will investigate the use of exposure models for local conditioning of neutral models, leading to the detection of clustered or isolated geographical units that depart significantly from the cancer rates expected from exposure data. A similar approach has recently been implemented whereby the regional background observed in the past has been incorporated into the geostatistical simulation of neutral models [ 24 ]. This new methodology allowed one to identify geographic pattern above and beyond background variation displayed in prior time intervals for cervix cancer mortality rates. Another issue, which often impacts the results of cluster analysis, is the lack of reliability of rates inferred from small populations. If ignored, large differences in population size decrease the ability of Moran's I to detect true clustering/departures from spatial randomness. A geostatistical smoother, which accounts for the spatial pattern of SMR data (i.e. anisotropic variability, range of autocorrelation), has been applied to eliminate the random variability that appeared as a nugget effect on the experimental semivariograms. The smoothing of local fluctuations, in particular the ones recorded in sparsely populated ZIP codes, resulted in the detection of larger and more compact clusters of low or high SMR values as well as the disappearance of some unreliable spatial outliers. Geostatistical filters are very flexible and could be used to filter short-range variability in addition to the noise created by small population sizes. In this case, the focus of the analysis would be on the regional background of the data, allowing the detection of regional clusters. The neutral models and methods in this paper make possible, for the first time ever, evaluation of the sensitivity of the results of cluster or boundary analyses to specification of the null hypothesis. Within a study, this will provide detailed quantification of the reliability of the results, and will identify those areas that are stable (i.e. always classified as a member of a cluster or not) or whose classification is highly sensitive to specification of the null hypothesis. This end result will be a spatially explicit analysis of potential false positives and false negatives. Authors' contributions Authors PG and GMJ collaborated intensely on all aspects of the manuscript, from research design to data preparation. PG carried out most of the geostatistical and local cluster analysis and drafted the manuscript. Both authors read and approved the final manuscript. Table 3 Number of significant zip codes for the different types of cluster/outliers and neutral models. Results are reported for male lung cancer. Numbers between parentheses indicate zip codes that have similar classification under the reference Model I (CSR). Summary statistics for the p-values are also provided. Neutral Model Type Model I Model II Model III Model IV Model V Model VI High-High 8 1(1) 2(1) 15(8) 2(1) 4(1) High-Low 1 0(0) 0(0) 0(0) 0(0) 0(0) Low-High 2 1(1) 1(0) 1(1) 0(0) 2(0) Low-Low 14 0(0) 1(1) 24(14) 2(1) 4(2) P-value Mean 0.185 0.259 0.368 0.166 0.251 0.371 CV 83.4% 54.8% 71.8% 92.9% 58.9% 80.5% | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC506783.xml |
509301 | The Zebrafish moonshine Gene Encodes Transcriptional Intermediary Factor 1γ, an Essential Regulator of Hematopoiesis | Hematopoiesis is precisely orchestrated by lineage-specific DNA-binding proteins that regulate transcription in concert with coactivators and corepressors. Mutations in the zebrafish moonshine (mon) gene specifically disrupt both embryonic and adult hematopoiesis, resulting in severe red blood cell aplasia. We report that mon encodes the zebrafish ortholog of mammalian transcriptional intermediary factor 1γ (TIF1γ) (or TRIM33), a member of the TIF1 family of coactivators and corepressors. During development, hematopoietic progenitor cells in mon mutants fail to express normal levels of hematopoietic transcription factors, including gata1, and undergo apoptosis. Three different mon mutant alleles each encode premature stop codons, and enforced expression of wild-type tif1γ mRNA rescues embryonic hematopoiesis in homozygous mon mutants. Surprisingly, a high level of zygotic tif1γ mRNA expression delineates ventral mesoderm during hematopoietic stem cell and progenitor formation prior to gata1 expression. Transplantation studies reveal that tif1γ functions in a cell-autonomous manner during the differentiation of erythroid precursors. Studies in murine erythroid cell lines demonstrate that Tif1γ protein is localized within novel nuclear foci, and expression decreases during erythroid cell maturation. Our results establish a major role for this transcriptional intermediary factor in the differentiation of hematopoietic cells in vertebrates. | Introduction Hematopoiesis involves the coordinated processes of cell proliferation and differentiation of a relatively small number of progenitor cells into billions of circulating red and white blood cells ( Thisse and Zon 2002 ). Hematopoiesis in vertebrates, from zebrafish to humans, is an evolutionarily conserved program that produces two waves of stem or progenitor cells that differ both in their embryonic origins and in the lineages of differentiated blood cells produced ( Palis and Yoder 2001 ; Orkin and Zon 2002 ; Galloway and Zon 2003 ). The first, or primitive, wave of hematopoiesis originates from ventral mesoderm and gives rise to progenitor cells that differentiate in embryonic blood islands. The primitive wave of hematopoiesis produces a burst of embryonic erythrocytes and macrophages. The second, or definitive, wave of hematopoiesis arises from self-renewing stem cells that develop primarily in the intraembryonic aorta–gonad–mesonephros region. These definitive hematopoietic stem cells seed the later developing marrow spaces, to produce all lineages of adult blood cells, including definitive erythrocytes, myeloid cells, and lymphocytes. We have undertaken a genetic approach to characterize genes that control hematopoiesis using the zebrafish as a model system ( Thisse and Zon 2002 ). As part of a large-scale forward genetic screen, we previously identified bloodless zebrafish mutants that failed to express the erythroid transcription factor gata1 normally in embryonic hematopoietic precursors ( Ransom et al. 1996 ). We named one of these zebrafish genes moonshine (mon), and another group named a noncomplementing allele vampire ( Weinstein et al. 1996 ). Here, we have determined that mutations in the mon gene cause a disruption in both primitive embryonic and definitive adult hematopoiesis, resulting in a severe loss of erythroid cells. Erythroid progenitor cells in mon mutants are initially present, but fail to express normal levels of hematopoietic transcription factors and undergo apoptosis. Positional cloning identifies the mon gene as the zebrafish ortholog of mammalian transcriptional intermediary factor 1γ (TIF1γ), a member of the TIF1 family of transcriptional coactivators and corepressors ( Le Douarin et al. 1995 ; Friedman et al. 1996 ; Kim et al. 1996 ; Venturini et al. 1999 ; Peng et al. 2002 ). The three members of the vertebrate TIF1 family (α, β, and γ) are large nuclear proteins that each contain an N-terminal RBCC or TRIM domain ( Reymond et al. 2001 ) composed of a RING finger, two B-boxes, and a coiled-coil domain. TIF1 family members also contain a C-terminal plant homeodomain finger and bromodomain that are characteristic of chromatin remodeling factors. TIF1α has been shown to associate with a variety of ligand-bound nuclear hormone receptors ( Le Douarin et al. 1995 ) and function as a coactivator for retinoic acid receptors ( Zhong et al.1999 ). TIF1β has been shown to act as a corepressor for the large family of Krüppel-associated box (KRAB) domain zinc-finger transcription factors ( Friedman et al. 1996 ; Abrink et al. 2001 ). In contrast, TIF1γ does not associate directly with either nuclear receptors or KRAB domains that bind to the other TIF1 family members ( Venturini et al. 1999 ; Abrink et al. 2001 ). Biochemical studies also demonstrate that TIF1γ forms both homo-oligomers and hetero-oligomers with TIF1α but not with TIF1β ( Peng et al. 2002 ). The murine Tif1α and Tif1γ genes have not yet been subjected to gene targeting experiments, whereas analysis of mouse mutants demonstrates that Tif1β is required for postimplantation embryogenesis and mesoderm induction in particular ( Cammas et al. 2000 ). Taken together, these studies suggest that a major function of TIF1 family members is to link DNA-binding proteins with other coactivators or corepressors during development. Our studies establish that tif1γ functions as an essential regulator of embryonic and adult hematopoiesis in vertebrates. Cell transplantation studies demonstrate that tif1γ acts in a cell-autonomous manner during embryonic hematopoiesis. The tif1γ gene is expressed specifically in ventral mesoderm and hematopoietic progenitors, then downregulated as erythroid maturation occurs. Tif1γ protein localizes to a novel class of nuclear bodies in both primary mouse embryo fibroblasts and erythroleukemia cell lines. Taken together, our studies demonstrate that Tif1γ is required for normal erythroid cell development and survival. Results The Zebrafish mon Gene Is Essential for Both Primitive and Definitive Erythropoiesis In order to determine when the mon gene is required in development, we first examined hematopoietic gene expression and apoptosis in zebrafish homozygous mon mutant embryos. During embryogenesis, homozygous zebrafish mon mutants have no red blood cells (RBCs) visible in circulation ( Ransom et al. 1996 ; Weinstein et al. 1996 ). The mon mutants initiate expression of gata1 in hematopoietic cells around the five-somite stage, similar to wild-type embryos (data not shown); however, based on TUNEL staining, the differentiating erythroid cells undergo programmed cell death from the 12-somite stage to 22 h postfertilization (hpf) ( Figure 1 A and 1 B, arrows). At 12 somites, gata1 expression is only slightly reduced. By 18–22 hpf, hematopoietic-specific markers such as gata1, scl, gata2, and ikaros are not detected in the embryonic blood island ( Figure 1 A and 1 B; unpublished data). The hematopoietic cells are thus correctly specified early during the development of mon mutant embryos, but these precursors undergo cell death. Based on expression of c- myb and rag1 ( Figure 1 B, arrows), mon mutants have normal myeloid and lymphoid development, respectively. In addition to the deficit of RBCs in mon mutants, there is a prominent loss of fin-fold and tail mesenchyme ( Ransom et al. 1996 ). TUNEL staining of mon mutants demonstrates extensive apoptosis of mesenchymal cells in the trunk and tail bud regions ( Figure 1 A and 1 B, arrows). The mon gene is thus required for normal development and survival of both committed erythroid progenitor cells and posterior mesenchymal cells. Figure 1 Zebrafish mon Mutants Have Severe Defects in Primitive Hematopoiesis (A) Whole-mount TUNEL assays reveal that ventral-posterior mesodermal cells undergo apoptosis in homozygous mon tg234 mutant embryos. Whole-mount in situ hybridization of gata1 detected at the 12- and 18-somite stage in genotyped embryos. Posterior views, anterior to the left. (B) Extensive apoptosis is visible in the trunk and tail (arrowhead) and also in hematopoietic cells of the embryonic blood island at 22 h of development (arrow). Whole-mount in situ hybridization at 22 hpf including scl, gata2, gata1, ikaros, and myb in mon tg234 mutants. Expression of myb is greatly reduced in the blood islands because of a loss of erythroid cells, but embryonic macrophages are still present (arrows). The expression of rag1 in thymic T-cells appears normal in mon mutants at 5 d postfertilization (arrow heads). Lateral views of 22 hpf and 5-d-old embryos. We next examined definitive hematopoiesis in rare surviving homozygous adult zebrafish mon mutants. Mutations in mon are generally lethal by 10 to 14 d of development ( Ransom et al. 1996 ), although rare mon homozygous mutants (approximately 1 in 500 bloodless embryos) of all tested alleles survive to adulthood. Adult mon mutants show cardiac hypertrophy, presumably due to the severe anemia leading to a high output state ( Figure 2 ). In wild-type zebrafish, the adult site of hematopoiesis is the kidney ( Al Adhami and Kunz 1977 ), which contains erythroid, lymphoid, and myeloid populations at various stages of differentiation ( Bennett et al. 2001 ). In mon homozygous mutants, there is a severe block in maturation at the proerythroblast stage ( Figure 2 ), whereas the differentiation of myeloid cells is normal (unpublished data). This demonstrates that the mon gene product acts during both primitive and definitive erythropoiesis. Figure 2 Zebrafish mon Mutants Also Have Severe Defects in Definitive Hematopoiesis Adult phenotype of wild-type and mon mutants. A rare surviving mon tb222 homozygous adult shows significant cardiomegaly in comparison to a wild-type age-matched control. Wright–Giemsa stained marrow of wild-type adult in comparison to a homozygous mutant. Note the dramatic reduction of terminally differentiated erythroid cells and the presence of abnormally large megaloblastic proerythroblasts in the mon tb222 mutant marrow. Positional Cloning Identifies mon as the Zebrafish Ortholog of Mammalian TIF1γ We identified the mon gene by positional cloning using a panel of 2,200 diploid mutants collected from Tübingen background (TU)/WIK strain hybrid parents carrying the mon tg234 allele. The mon mutant gene was positioned on Chromosome 8 between microsatellite markers z987 and z11001 ( Figure 3 A) ( Knapik et al. 1998 ). For positional cloning purposes, over 12,000 polymorphic markers were screened using amplified fragment length polymorphism (AFLP) ( Ransom and Zon 1999 ), and 36 markers within the interval were isolated. One of these, MA3, was found to be 0.3 cM from the gene ( Figure 3 A) and was utilized as the starting point of a chromosomal walk. A critical P1 bacterial artificial chromosome clone (PAC), 107N19, was obtained that spanned the genetic interval. Two simple sequence conformation polymorphism (SSCP) markers found on this PAC clone flank the critical genetic interval. The marker 80M12-T7 maps two recombinants out of 4,400 meioses telomeric of the mutation, and the marker 157J23-T7 maps one recombinant centromic of the mutation ( Figure 3 A). The end sequences and SSCP markers of PAC 107N19 are found in the zebrafish genomic sequence contig ctg23107 ( http://www.ensembl.org/Danio_rerio/ ) containing a predicted zebrafish TIF1 family gene. This PAC was hybridized to a kidney cDNA library, resulting in the isolation of four clones that represented the same gene. Figure 3 Positional Cloning Identifies the mon Gene as Zebrafish tif1γ (A) Physical map of the mon locus on zebrafish Chromosome 8. Microsatellite markers z987 and z11001 were used to initially identify recombinants in a panel of 2,200 diploid mon tg234 homozygous mutants. The AFLP marker MA3 was used to initiate a chromosomal walk in PAC libraries. The critical PACS that were isolated to encompass the mon locus are indicated by numbers above bar. The PAC 107N19 defines the critical interval for the mon gene. This PAC was used as a probe to screen cDNA libraries and to identify zebrafish tif1γ cDNAs. Numbers below the bar indicate the number of recombinants identified by SSCP analysis. (B) Clustal-W–generated phylogentic tree of zebrafish ( Danio rerio [Dr]) Tif1γ and Tif1α peptide sequences in comparison to TIF1 family members: human (Hs) TIF1α, TIF1β, and TIF1γ; mouse (Mm) Tif1α, Tif1β, and Tif1γ;; and fly (Dm) bonus. (C) Diagrams illustrating the structure of the Tif1γ-predicted peptide and the three identified point mutants. RING finger (RING), B-boxes (B1 and B2), plant homeodomain finger (PHD) and bromodomain (BROMO). Numbers below the first diagram indicate the percent identity shared between each of these domains in zebrafish and human TIF1γ. The predicted truncated proteins are indicated. (D) DNA sequence chromatograms showing the three ENU-induced point mutants in comparison to wild-type control sequences The mon gene encodes a member of the TIF1 family of transcriptional cofactors ( Figure 3 B and 3 C). The coding sequence of mon is most similar to human TIFγ ( Le Douarin et al. 1995 ; Friedman et al. 1996 ; Venturini et al. 1999 ), and the locations of exon boundaries are conserved between the zebrafish and human genes (unpublished data). The mon locus on zebrafish Chromosome 8 is also predicted to be syntenic to the region of human Chromosome 1p that contains the TIF1γ gene based on the conserved locations of 12 other orthologous gene pairs, including NRAS, mapped to these regions in human and zebrafish ( Barbazuk et al. 2000 ). Therefore, based on sequence similarity and chromosomal location, the zebrafish mon gene is the likely ortholog of the human TIF1γ gene. We have identified ethyl-nitrosourea (ENU)-induced point mutations in three alleles of mon ( Figure 3 C and 3 D), each of which generates a premature stop codon. The mon tb222b and mon tg234 alleles have a severe phenotype with no circulating blood cells. In contrast, the mon m262 allele has 10–100 circulating blood cells by 48 hpf, in comparison to the approximately 3,000 RBCs in the circulation of wild-type or heterozygous embryos at the same time point. The mon m262 allele was found to encode a premature stop codon at position E40, which would encode a putative protein of only 40 amino acids. Although this mutation would be expected to lead to a complete loss of mon gene product, another methionine is found downstream at amino acid position 104. In vitro translation experiments in reticulocyte lysates demonstrate reinitiation of translation from this methionine (unpublished data). Therefore, the hypomorphic larval phenotype of the mon m262 allele is likely due to partial loss of mon function or expression. The presence of mutations in each of the mon alleles indicates that defective Tif1γ function is the cause of the mon phenotype. In order to determine whether tif1 γ is expressed in hematopoietic mesoderm, we next examined zebrafish embryos by whole-mount in situ hybridization ( Figure 4 A). tif1γ mRNA is expressed maternally and is found throughout the embryo during blastula stages. During gastrulation and epiboly stages, zygotic expression of mon is highest in the mesendoderm of the germ ring. At tail bud and early somite stages a high level of tif1γ expression delineates a horseshoe-shaped population of ventral/lateral mesoderm that will give rise to blood and also expresses stem cell leukemia hematopoietic transcription factor (scl) ( Liao et al. 1997 ). This group of cells continues to express tif1γ and scl while it converges and forms the embryonic blood island ( Detrich et al. 1995 ). The tif1γ gene is also highly expressed in the central nervous system as well as the mesenchyme of the trunk and tail. Homozygous mon tg234 mutants have a greatly reduced amount of tif1γ mRNA in all tissues consistent with nonsense-mediated message decay. Thus, zebrafish tif1γ is specifically expressed in ventral mesoderm and putative hemangioblasts prior to and during the embryonic stages when hematopoietic progenitors are undergoing apoptosis in mon mutants. We also compared the expression of zebrafish mon to mouse Tif1γ ( Figure 4 A and 4 B). Mouse Tif1 γ is highly expressed in erythroid blood islands of the yolk sac, and it is subsequently expressed in the fetal liver at a high level, and in other tissues, including the central nervous system. Taken together these results strongly suggest that zebrafish mon and mouse Tif1γ are orthologs that function during hematopoiesis. Figure 4 The mon/tif1γ Gene Is Highly Expressed in Hematopoietic Mesoderm (A) In situ hybridization of zebrafish embryos demonstrating the embryonic expression of tif1γ. tif1γ is initially expressed as a maternal mRNA. Increased expression of tif1γ in ventral-lateral mesoderm begins between the one- to three-somite stages and increases through early development. By five somites, tif1γ is strongly expressed in lateral stripes of mesoderm that also express scl. At 22 hpf tif1γ is expressed broadly in the brain, spinal cord, trunk, and tail mesenchyme, but is at much higher levels in hematopoietic cells of the blood island. Zebrafish tif1α is also broadly expressed but relatively more uniform in most tissues, in comparison to tif1γ. Tif1α is weakly expressed at early somite stages in hematopoietic mesoderm and uniformly expressed at 22 hpf, including expression in the blood islands. Expression of scl at five somites and 22 hpf highlights the embryonic blood island in comparison to tif1γ expression. (B) In situ hybridization of mouse embryos detects broad expression of Tif1γ at embryonic day 8.5 including the yolk sac blood islands (arrow). AT embryonic day 12.5, there is high level expression in the fetal liver (arrow) and broad expression in the embryonic brain, spinal chord, gut, and muscle. Given that mammalian TIF1γ has been shown to form hetero-oligomers with Tif1α ( Peng et al. 2002 ), we searched for additional TIF1 family members in zebrafish to compare with tif1γ. Using zebrafish expressed sequence tag (EST) sequences, we designed primers to RT-PCR amplify a TIF1-related cDNA from embryonic 10-hpf and 24-hpf RNA. This cDNA encodes a predicted zebrafish ortholog of human TIF1α based on predicted amino acid sequences (see Figure 3 B). In addition, zebrafish tif1α ESTs map to LG4 in a region predicted to be syntenic to the region of human Chromosome 7 that contains the TIF1α gene based on the conserved locations of eight other orthologous gene pairs, including SEMA3A, mapped to these regions in human and zebrafish ( Barbazuk et al. 2000 ). We next compared the embryonic expression pattern of tif1α mRNA to tif1γ by in situ hybridization. Like mammalian TIF1α ( Le Douarin et al. 1995 ; Niederreither et al. 1999 ), the predicted zebrafish tif1γ gene is broadly expressed (see Figure 4 A). At five somites, zebrafish tif1α does not display the relatively high expression in the horseshoe-shaped region of hematopoietic mesoderm seen with tif1γ. At later stages, tif1α is evenly expressed throughout most of the embryo, including the developing blood islands. Therefore, tif1α is coexpressed in the same cells with tif1γ and may therefore be available to form hetero-oligomers in vivo. Forced Expression of tif1γ Rescues Hematopoiesis in mon Mutants To further confirm that a mutation in the zebrafish tif1γ gene is responsible for the mon mutant phenotype we performed embryo rescue experiments ( Figure 5 A; Table 1 ). Microinjection of synthetic wild-type mon mRNA at the one-cell stage rescues the formation of embryonic erythrocytes in genotyped mutant embryos without causing obvious defects in embryonic patterning or organogenesis. At 4 d of development, 70% ( n = 10) of mon tg234 mutants show significant (greater than 200 cells in comparison to a wild-type estimate of 3,000 cells) rescue of circulating hemoglobinized RBCs in comparison to control sibling mutants ( n = 75). Based on the correction of the jagged fin-fold phenotype ( Ransom et al. 1996 ), the mesenchymal cells are rescued to a similar extent as the anemia (unpublished data). Overexpression of mon did not result in expanded blood cell numbers in wild-type embryos and was not toxic at doses that rescue the phenotype of mon mutants (unpublished data). Since there were no expanded or ectopic blood populations in the embryos, these rescue experiments suggest that mon functions as a permissive factor required for hematopoiesis. Figure 5 Overexpression of Wild-Type tif1γ mRNA or Marrow Transplantation Rescues Embryonic Hematopoiesis in mon Mutants (A) mon tg234 mutants are rescued by injection of mRNA-encoding wild-type Tif1γ protein. At 4 d of development, large numbers of RBCs are visible in the circulation of wild-type zebrafish, shown here by o-dianisidine staining of hemoglobin. Uninjected mon ttg234 homozygous mutants are completely bloodless. Injection of 100 pg of wild-type tif1γ mRNA rescues erythropoiesis in mutant embryos. o-dianisidine-stained larvae are shown in ventral views to highlight blood in vessels. (B) Transplantation of wild-type zebrafish marrow cells carrying a gata1:GFP transgene into 2-d-old embryos reconstitutes erythropoiesis, but not viability, in mon tg234 homozygous mutants. Still frames from movies of live embryos at day 3 posttransplant highlight less than 100 GFP + RBCs in circulation (top). Transplanted cells were observed to proliferate resulting in thousands of donor-derived erythrocytes 7 d later (bottom). Arrows indicate the hearts of control and transplanted zebrafish. See Videos S1–S4 . Table 1 Overexpression of tif1γ mRNA Rescues mon Mutants: Hematopoietic Phenotypes Synthetic tif1γ mRNA (100 pg) was injected at the one-cell stage into embryos of the indicated genotypes. For the mon embryos, circulating cells where counted each day through 4 d, when the embryos were fixed and stained with o-dianisidine to detect hemoglobin in mature RBCs. Normal embryos contain approximately 3,000 circulating cells at these time points. Results are given as number of embryos with the indicated phenotype. Numbers in parentheses represent percentage of total embryos analyzed Marrow Transplantation Rescues Erythropoiesis in mon Mutants The high levels of tif1 γ expression in erythroid cells suggest that it functions as a cell-autonomous regulator of gene expression in hematopoietic cells. In order to test this hypothesis, we transplanted wild-type adult zebrafish kidney marrow cells carrying a gata1:green fluorescent protein (GFP) transgene into 48-hpf mon mutant embryos ( Figure 5 B; Table 2 ). The gata1:GFP transgene makes use of the gata1 promoter to drive GFP expression and can thus be used to mark donor-derived erythroid cells ( Long et al. 1997 ). Untransplanted mutant embryos have no embryonic blood cells in circulation. Following transplantation, mutant host embryos were observed daily for 2 wk. Of 191 mutant embryos injected, 129 (68%) showed GFP + cells in circulation 2 d later. Many recipients showed robust increases in donor-derived cells over the observation period. Of 81 recipients initially scored as having less than ten GFP + cells at day 2 posttransplant, 13 (16%) of these demonstrated a marked increase in erythroid cells with 100–1,000 GFP + cells in circulation 6 d later. By day 10, these transplanted embryos showed approximately 3,000 cells in circulation, similar to the number of blood cells in normal embryos. Despite robust reconstitution of blood cells, mutant recipients did not inflate their swim bladders and thus failed to survive longer than nontransplanted sibling controls, all dying by 3 wk of age. In contrast, 13/35 (37%) heterozygous mon tg234 transplants survived to early adulthood. Similar transplants of wild-type cells can fully rescue vlad tepes (gata1) mutants ( Traver et al. 2003 ). Therefore, the results of cell transplantations suggests that tif1γ plays a cell-autonomous role in erythroid cells, and its role in nonhematopoietic tissues, such as trunk mesenchyme or the nervous system, is also required for embryo survival. Table 2 Marrow Transplantation Rescues Hematopoiesis But Not Survival in mon Mutants: Embryos with Transplanted Erythroid Cells Between 100 and 1,000 kidney marrow cells from adult gata1:EGFP transgenic donors were injected per zebrafish embryo at 48 hpf. Individual transplanted embryos were anesthetized and visualized for GFP + erythroid cells. By 10 d posttransplantation the indicated number of embryos had an estimated 100 to 3,000 GFP + cells in circulation. At 3 mo the indicated number of fish were alive. The relative percentage of embryos is shown in parentheses Tif1γ in Punctate Nuclear Foci Is Developmentally Regulated In order to examine the subcellular distribution of Tif1γ protein, we generated an affinity-purified rabbit polyclonal antiserum directed against the C-terminal 15 amino acids conserved in human TIF1γ and mouse Tif1γ. Immunofluorescence of mouse embryo fibroblast nuclei with the anti-Tif1γ antiserum demonstrates that Tif1γ is localized in small nuclear foci ( Figure 6 A). The localization of Tif1γ protein appears different from the more diffuse nuclear patterns typically seen in studies of Tif1α ( Remboutsika et al. 2002 ) or TIF1β ( Cammas et al. 2002 ). A recent report demonstrates that TIF1β associates with heterochromatin-containing foci after retinoic acid treatment or serum starvation ( Cammas et al. 2002 ). Thus, localization or expression of the TIF1 proteins may be regulated during distinct developmental processes or by environmental cues. The nuclear foci that contain Tif1γ do not colocalize with two markers of heterochromatin, HP1α protein and DAPI staining of DNA ( Figure 6 A). Furthermore, Tif1γ does not colocalize with promyelocytic leukemia gene product (PML) nuclear bodies, DNA repair complexes that contain Mre11, or transcriptional complexes containing TFII-B (unpublished data). We next examined the expression of Tif1γ protein during the differentiation of G1E cells, a murine erythroleukemia cell line that can terminally differentiate into erythrocytes when a Gata1:estrogen receptor fusion protein is stabilized in response to estrogen exposure ( Weiss et al. 1997 ). Western blot analysis demonstrated that Tif1γ protein expression decreases with terminal erythroid differentiation ( Figure 6 B). Consistent with this finding, after 24 hpf, zebrafish mon mRNA expression falls during the terminal maturation of the primitive erythroid cells (unpublished data). In two different murine erythroleukemia cell lines (MEL and G1E), Tif1γ is also expressed in nuclear foci, and even though the overall Tif1γ protein level is reduced, this nuclear foci localization does not change with differentiation (unpublished data). This provides further support for the hypothesis that Tif1γ acts within novel nuclear foci, during erythroid differentiation. Figure 6 Mammalian Tif1γ Protein Localizes to Nuclear Bodies Distinct from Heterochromatin (A) Deconvolved immunofluorescence images of a mouse embryonic fibroblast cell nucleus stained with an anti-Tif1γ antibody and stained with a monoclonal antibody directed against HP1α. This is also compared to DAPI staining. The merged images of the nucleus show that Tif1γ does not colocalize with the HP1α or DAPI staining of heterochromatin while HP1α and DAPI staining overlap. (B) G1ER mouse erythroleukemia cells express high levels of Tif1γ protein as detected by Western blot analysis. Expression of Tif1γ decreases during Gata1-dependent erythroid maturation induced by β-estradiol treatment to induce a Gata1–ER fusion protein. Discussion The zebrafish is an excellent model system to elucidate the molecular machinery controlling gene expression during hematopoiesis ( Thisse and Zon 2002 ; Galloway and Zon 2003 ). As part of a large-scale forward genetic screen, we originally identified a complementation group of independent mutant alleles in the zebrafish gene that we named moonshine ( Ransom et al. 1996 ). Positional cloning was used to identify the mon gene, establishing a critical role for a transcriptional intermediary factor, Tif1γ, during hematopoietic development. The mon Gene Encodes the Zebrafish Ortholog of Mammalian TIF1γ Our results strongly support the conclusion that we have positionally cloned the zebrafish mon gene correctly, and it is the ortholog of mammalian Tif1γ. Tif1γ is present in the critical genetic interval encompassing a single approximately 50-kb PAC clone defined by linkage analysis (see Figure 3 ). Sequence analysis indicates that zebrafish tif1γ is most similar in predicted amino acid sequence and intron/exon structure compared to the predicted orthologous human and mouse genes. Zebrafish tif1γ is located in a region of zebrafish Chromosome 8 syntenic to the region of human Chromosome 1 containing TIF1γ. We identified point mutations in tif1γ from three different alleles of mon that each result in premature stop codons and mRNA decay. In addition, tif1γ/Tif1γ is highly expressed in hematopoietic cells throughout embryogenesis in both zebrafish and mouse (see Figure 4 ). And as predicted, forced expression of wild-type tif1γ mRNA efficiently rescues hematopoiesis in mon mutants and does not perturb hematopoiesis in wild-type embryos (see Figure 5 ). We have also cloned the predicted zebrafish ortholog of tif1α, which is more uniformly expressed in zebrafish embryos like mammalian TIF1α ( Le Douarin et al. 1995 ; Niederreither et al. 1999 ) (see Figures 3 A and 4 A) and may therefore be available to form hetero-oligomers with Tif1γ protein in developing hematopoietic cells. Comparing available zebrafish and mammalian TIF1-predicted amino acid sequences, it appears that the Tif1γ orthologs are the most highly conserved family members while the Tif1α sequences are relatively more divergent. We have not found a Tif1β ortholog, thus far, in the zebrafish or fugu genome or EST sequences. It is possible that Tif1β, like the KRAB domain transcription factors it binds to, may be present only in tetrapods ( Urrutia 2003 ). However, more complete genome sequences will be needed to confirm this hypothesis. Based on our analysis of zebrafish mon mutants, it is reasonable to predict that Tif1γ, the most evolutionarily conserved TIF1 family member, plays a similarly essential role in human and mouse hematopoiesis. Mutations in tif1γ Cause Apoptosis of Erythroid Progenitors Our examination of hematopoietic gene expression, apoptosis, and marrow histology in mon mutants demonstrates that early erythroid progenitors are formed in homozygous mutants, but they fail to properly differentiate and instead undergo programmed cell death (see Figure 1 ). The expression of gata1 appears to initiate normally in the committed erythroid cells of mon mutants. However, the cells are abnormal prior to the complete loss of gata1 expression. TUNEL-positive apoptotic cells are abundant by the 12-somite stage of development, and by 22 hpf all hematopoietic gene expression is extinguished. The expression of marker genes, including scl and gata2, characteristic of hematopoietic stem cells and primitive hematopoietic progenitors, are also not detected in the embryonic blood islands of mutants at 22 hpf. This indicates that the mutant hematopoietic cells are not blocked prior to commitment to the erythroid lineage, but instead develop as abnormal erythroid cells and undergo apoptosis, similar to gata1- deficient erythroid cells ( Fujiwara et al. 1996 ; Lyons et al. 2002 ). Defective erythropoiesis and severe anemia were also observed in rare surviving homozygous mutant mon adults, demonstrating that tif1γ is also required in definitive hematopoiesis (see Figure 2 ). The zygotic phenotypes of mon mutants may not reveal the function of maternally inherited Tif1γ. Maternally expressed zebrafish Tif1γ may play roles in hematopoiesis or other aspects of organogenesis that are not detectable due to the presence of wild-type mRNA in eggs laid by heterozygous mothers. Analysis of the offspring of homozygous mon mutant female zebrafish will aid in defining the function of this maternal mRNA. The present analysis of zygotic mon mutants provides data that are consistent with the conclusion that tif1γ is essential for erythropoiesis but do not rule out essential functions in other hematopoietic lineages. The hematopoietic phenotype of mon mutants resembles the loss of erythroid cells seen in both mouse Gata1 knockout embryos and zebrafish vlad tepes (gata1) mutant embryos ( Fujiwara et al. 1996 ; Lyons et al. 2002 ). In an effort to determine if there is a genetic relationship between mon and gata1, we tested their ability to rescue erythropoiesis. Both injection of gata1 mRNA into mon homozygous mutant embryos and injection of tif1γ mRNA into gata1 knock-down embryos failed to rescue hematopoiesis (unpublished data). We also tested for a direct interaction between Tif1γ and Gata1 proteins by coimmunoprecipitation and yeast two-hybrid assays and found no association (unpublished data). Although the mutations in each of these genes arrest cells at a similar stage of development, our results suggest that gata1 and tif1γ act independently. This does not rule out the possibility that parallel genetic pathways involving gata1 and tif1γ, operating together, regulate gene transcription within blood cells. The Role of Tif1γ in Primitive and Definitive Erythropoiesis Taken together, our data suggest that tif1γ is required as a permissive cofactor for the erythroid lineage-specific control of hematopoietic gene expression. We reasonably predict that Tif1γ protein functions as a transcriptional intermediary factor in hematopoietic progenitor cells given that both TIF1α ( Zhong et al. 1999 ) and TIF1β ( Friedman et al. 1996 ; Abrink et al. 2001 ) have been shown to act as intermediary factors that positively or negatively regulate gene transcription. These studies indicate that TIF1α and TIF1β act as scaffolds that link different classes of DNA-binding proteins and chromatin-associated proteins into larger regulatory complexes. Tif1γ is detected within nuclear foci (see Figure 6 ), which, based on our analysis, do not appear to correspond to several types of previously described nuclear bodies, including PML bodies. Localization of Tif1γ to these nuclear bodies may be regulated by posttranslational modification such as SUMO modification that is required for PML to form PML nuclear domains ( Zhong et al. 2000a , 2000b ; Best et al. 2002 ). These foci may serve as assembly points where Tif1γ forms multisubunit complexes with DNA-binding transcription factors and their other essential coactivators or corepressors, during the early stages of erythroid differentiation. It will be important to determine the identity of Tif1γ-interacting proteins in nuclear foci and establish how they function with Tif1γ to regulate blood cell development. Materials and Methods Zebrafish and mouse strains and studies Zebrafish were maintained and staged as described ( Westerfield 1998 ). The alleles mon tb222b and mon tg234 were generated in a large-scale screen for ENU-induced mutations ( Ransom et al. 1996 ) on the TU, whereas the mon m262 allele was derived on the AB strain and was originally called vampire ( Weinstein et al. 1996 ). Mapping strains were constructed by mating to WIK or SJD polymorphic strains. Linkage analysis was performed on haploid or diploid embryos obtained from TU/SJD or TU/WIK hybrids. In situ hybridization and stainings of embryos were done as described ( Thompson et al. 1998 ; Liao et al. 2002 ). In situ hybridization of mouse embryos was performed as described ( Kingsley et al. 2001 ). Genomic DNA isolation, genotyping, AFLP analysis, and chromosomal walking were each performed as previously described ( Brownlie et al. 1998 ; Ransom and Zon 1999 ). A complete list of primers for genetic mapping, RT-PCR, and sequencing of mon are available on request. mRNA expression constructs, morpholinos, and microinjection The full-length mon cDNA was subcloned into EcoRI and XhoI sites in the pCS2 + vector. Synthetic mRNA was transcribed in vitro, and microinjection was performed essentially as described ( Liao et al. 2002 ). Cell transplantation Whole kidney marrow cells were isolated from adult gata1:EGFP transgenic donors, resuspended in 0.9X phosphate-buffered saline + 5% fetal bovine serum, and injected into the sinus venosus of 2-d-old mon tg234 −/− and control embryos. Between 10 2 and 10 3 kidney marrow cells were injected per embryo. Individual transplanted embryos were anesthetized and visualized daily under an inverted fluorescent microscope (DM-IRE2; Leica, Wetzlar, Germany) for GFP + cells over a span of 12 d. On day 13 posttransplant, all surviving larvae (12/129; 9%) were placed in tanks and monitored for survival. Antibodies, immunostaining, and immunoblots Antisera against the human C-terminal TIF1γ sequence RRKRLKSDERPVHIK was generated in rabbits (Genemed Synthesis, South San Francisco, California, United States) and affinity purified. Mouse embryonic fibroblasts grown on coverslips were immunostained with HP1α (Chemicon, Temecula, California, United States) and Tif1γ antisera simultaneously. In brief, cells were fixed in 4% paraformaldehyde for 5 min, washed with phosphate-buffered saline, and blocked with 5% serum (PBST) for 30 min. After incubation with the primary antibodies (PBST, 60 min) cells were washed three times with PBST and incubated with secondary antibodies (Jackson Laboratory, Bar Harbor, Maine, United States) followed by three washes in PBST. Cells were embedded with Vectashield/DAPI and analyzed using an Axioplan 2 microscope (Zeiss, Jena, Germany). Digital images were processed using the Volocity 1.0 software (Improvision, Lexington, Massachusetts, United States). G1E cell differentiation experiments were performed essentially as described ( Weiss et al. 1997 ). Supporting Information Transplantation of wild-type zebrafish marrow cells carrying a gata1:GFP transgene into 2-d-old embryos reconstitutes erythropoiesis, but not viability, in mon tg234 homozygous mutants. Movies of live embryos at day 3 posttransplant highlight less than 100 GFP + RBCs in circulation. Transplanted cells were observed to proliferate, resulting in thousands of donor-derived erythrocytes 7 d later. Movies present GFP-fluorescent images of live zebrafish larvae. Video S1 Untransplanted Control mon tg234 Homozygous Mutants Had No Fluorescent Cells in Circulation at 3 Days of Development (13.7 MB MOV). Click here for additional data file. Video S2 One Day after Transplantation, Less Than 100 GFP + Erythrocytes Were Visible in the Circulation of Three mon tg234 Homozygous Mutants (11.3 MB MOV). Click here for additional data file. Video S3 Untransplanted Control mon tg234 Homozygous Mutants Had No Fluorescent Cells in Circulation at 9 Days of Development (7.9 MB MOV). Click here for additional data file. Video S4 Seven Days after Transplantation, Thousands of Donor-Derived Erythrocytes Were Visible in the Circulation of a Representative mon tg234 Homozygous Mutant (11.2 MB MOV) Click here for additional data file. Accession Numbers The GenBank ( http://www.ncbi.nlm.nih.gov/Genbank ) accession numbers for the genes and gene products discussed in this paper are fly bonus (AAF19646), human TIF1α (015164), human TIF1β (Q13263), human TIFγ (Q9UPN9), human TIF1γ (Q9UPN9), mon (AY59853), mouse Tif1α (Q64127), mouse Tif1β (AAH58391), and mouse Tif1γ (NP444400). The cDNA sequences of zebrafish mon/tif1γ and tif1α have been deposited in GenBank under the accession numbers AY598453 and AY598454, respectively. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC509301.xml |
553972 | Individual patient data meta-analysis : Cervical stitch (cerclage) for preventing pregnancy loss in women | Background Cervical cerclage is a surgical procedure involving suturing the cervix with a purse type stitch to keep it closed during pregnancy. This procedure has been used widely in the management of pregnancies considered at high risk of preterm delivery. Several observational studies into the efficacy of cervical cerclage have claimed high rates of successful pregnancy outcome in women with a poor obstetric history attributed to cervical incompetence. However, a recent aggregate data Cochrane review found no such conclusive evidence from seven included randomised studies. Current data suggests that cervical cerclage is likely to benefit women considered to be 'at very high risk' of a second trimester miscarriage due to a cervical factor, however identifying such women remains elusive and many women may be treated unnecessarily. Undertaking an individual patient data (IPD) meta-analysis of the studies will allow us to investigate whether treatment is more effective in particular subgroups. Such an analysis will also provide a more powerful analysis of the predictors of preterm delivery and pregnancy loss, including ultrasound measurement of cervical length, and will allow a more complete analysis of 'time to event' outcomes. Methods/Design The analysis will include data from randomised trials comparing the intervention of elective cerclage versus no cerclage or bedrest to prevent miscarriage or pre-term labour. A specific list of data will be requested for each trial, including demographic and obstetric history data. The primary outcomes of interest will be neonatal mortality/morbidity. Attention will also be given to secondary outcomes such as time from randomisation to delivery, preterm delivery before 32 weeks and maternal morbidity. An intention to treat analysis will be performed, with attention paid to assessing clinical and statistical heterogeneity. Multilevel models with patients and trials as the two levels will be explored to investigate treatment effect on various outcomes. Patient-level covariates will be incorporated into the models in an attempt to account for statistical heterogeneity as well as to investigate interactions with treatment effect. Discussion Predictive models generated from our analysis should lead to more effective counselling of women at risk and a more cost effective use of cerclage. | Background Cervical cerclage is a surgical procedure carried out during pregnancy. The operation involves suturing the neck of the womb (cervix) with a purse type stitch to keep the cervix closed. This surgical procedure has been used widely in the management of pregnancies considered to be at high risk of preterm delivery. Several observational studies in the last 50 years have claimed high rates of successful pregnancy outcome in women that had a poor obstetric history attributed to cervical incompetence. However, a recent Cochrane review found no conclusive evidence from seven included randomised studies that inserting a cervical stitch in women perceived to be at risk of preterm birth or second trimester pregnancy loss attributed to cervical factors, reduces the risk of pregnancy loss, preterm delivery or morbidity associated with preterm delivery (Drakeley 2003)[ 1 ]. In the Cochrane review, the data for important clinical outcomes including preterm delivery and maternal infection showed significant heterogeneity due to inconsistency in clinical definitions used, including the cut off gestational age defining preterm delivery, and different patient populations studied. Practically, methods of undertaking a meta-analysis of several studies may involve collecting either aggregate data, or data on each patient individually. The advantages of the latter approach, described as the 'yardstick' (Chalmers 1993)[ 2 ] include (i) a more complete analysis of 'time of event' outcomes and (ii) a more powerful analysis of whether treatment is more or less effective in particular subgroups (Stewart 1993)[ 3 ]. One of the main concerns regarding current evidence related to cervical cerclage and other interventions for preventions of preterm delivery is a possibility that the 'primary outcomes' may have been selected to give results in greatest accord with the a priori beliefs of the authors. The evidence to support this phenomenon of within-study selective reporting comes from empirical research, which demonstrates discrepancies between research protocols and subsequent publications (Hahn 2002 [ 4 ], Williamson 2005 [ 5 ], Chan 2004 [ 6 ]). Individual patient data (IPD) meta-analysis has the capacity to overcome these problems. Currently available data suggest that cervical cerclage is likely to be of benefit for women considered 'at very high risk' of second trimester miscarriage due to a cervical factor e.g. greater than two second trimester losses or progressive shortening of the cervix on ultrasound. However, predicting those women who will miscarry due to a cervical factor remains elusive and many women may be treated unnecessarily. The use of IPD will allow us to investigate predictors of preterm delivery including ultrasound measurement of cervical length and other woman-cerclage interactions. IPD meta-analysis will allow an investigation of the hypothesis that the effect of cerclage is greater on extreme preterm delivery. In addition, an IPD meta-analysis has greater power than a single trial for examining subgroups. The efficacy of a treatment may depend on several factors. For aggregate data, a meta-analysis stratifying by the absolute risk in the control group may be the only method possible for accounting for these multiple factors simultaneously. This analysis is 'flawed and produces seriously misleading results' (Sharp 1996)[ 7 ]. A regression analysis of IPD allows the relation between treatment effect and risk score, derived from these multiple risk factors, to be investigated thereby avoiding these problems. Methods/Design Objectives The aim of this project is to undertake an IPD meta-analysis of randomised trials of cervical cerclage. Specific objectives are as follows. 1. To estimate the effect of cervical cerclage on gestational age at delivery. 2. To investigate whether cervical cerclage is more likely to prevent extreme prematurity (<28 weeks) or delivery at later gestations. 3. To investigate risk factors for preterm delivery. 4. To investigate interactions between risk factors and cervical cerclage. 5. To model the effect of cervical cerclage and other risk factors on neonatal and maternal morbidity. Criteria for considering studies for this IPD meta-analysis The types of studies considered for inclusion in the analysis will be all randomised trials comparing cervical cerclage with expectant management or no cerclage during pregnancy. The previous Cochrane review (Drakeley 2003) [ 1 ] identified eight eligible trials with 2,513 randomised women (Rust 2000 [ 8 ], Althuisius 2000 [ 9 ], Althuisius 2001 [ 10 ], Rush 1983 [ 11 ], Lazar 1984 [ 12 ], Dor 1982 [ 13 ], MRC/RCOG 1988 [ 14 ], Meekai To et al. [ 15 ]). We have agreement in principle to provide IPD from six of these trials (Meekai To et al. [ 15 ], MRC/RCOG 1988 [ 14 ], Rust 2000 [ 8 ], Althuisius 2000 [ 9 ], Althuisius 2001 [ 10 ], Rush 1983 [ 11 ]), accounting for 1919(78%) of all women randomised. The remaining trialists and trialists of any further trials identified as eligible will be approached at the start of the project and we anticipate their willingness to collaborate. Already, two further trials have been identified (Berghella 2004 [ 16 ], Ezechi 2003 [ 17 ]) and the authors have agreed to provide IPD from these trials. The data collected in the studies will relate to women with confirmed, or suspected of having, cervical incompetence who desire future pregnancies and women who present as an emergency and are thought to have a diagnosis of cervical incompetence. The intervention investigated in the studies will be elective cerclage by whichever method (Shirodkar technique, McDonald technique, transabdominal and transvaginal methods), versus no cerclage or bed rest as interventions to prevent miscarriage or pre-term labour as defined in the original Cochrane review (Drakeley 2003)[ 1 ]. Search strategy for identification of studies The methods of trial identification described in the original Cochrane review (Drakeley 2003)[ 1 ] (see below) will be adopted and updated to December 2004. The original review has drawn on the search strategy developed for the Pregnancy and Childbirth Group. The full list of journals and conference proceedings as well as the search strategies for the electronic databases, which are searched by the Group on behalf of its reviewers, are described in detail in the 'Search strategies for the identification of studies section' within the editorial information about the Cochrane Pregnancy and Childbirth Group. Briefly, the Trials Search Coordinator searches on a regular basis MEDLINE, the Cochrane Controlled Trials Register and reviews the Contents tables of a further 38 relevant journals received via ZETOC, an electronic current awareness service. In addition, handsearches will be performed on congress proceedings of the International and European society meetings of feto-maternal medicine, recurrent miscarriage and reproductive medicine. Whenever possible, investigators will be contacted to ask about any additional studies potentially eligible for inclusion. Trial eligibility and methodological quality assessment Two reviewers will independently assess eligibility of identified randomised controlled trials for inclusion in the review. Any difference of opinion will be resolved by discussion. The methodological quality of each trial will be assessed by summarising the method of generation of randomisation list, method of allocation concealment, and potential impact of losses to follow-up. Quasi-randomised studies in which allocation was transparent (e.g. use of alternative allocation or medical record numbers) were excluded in the original review. Data collection The following data for each woman/infant pair will be requested from all trials: date of randomisation and gestational age, maternal demographics and obstetric characteristics at randomisation including cervical length on ultrasound, fibronectin and bacterial vaginosis data if available, treatment allocated, complications during pregnancy including ruptured membranes, maternal pyrexia or chorioamnionitis, date of delivery, gestational age at delivery and all neonatal data including birthweight, length of stay at NICU and morbidity related to prematurity. The following methodological data will also be requested for all trials: method of generation of randomisation list, method of concealment of randomisation, stratification factors and blinding methods. Data will be accepted either in electronic (floppy disk/CD/internet) or paper form. A desired format and coding will be specified but trialists may supply data in the most convenient way open to them, providing details of coding are sent with the data. Data validation strategy A copy of the original data sent (before checking) will be held in a separate file. The following procedures will then be performed and documented for all trial data supplied. Trial details will be crosschecked against any published report of the trial. Range and consistency checks will be applied – missing data, errors and inconsistencies will be followed up with a nominated individual. The chronological randomisation sequence will be reviewed. The balance of prognostic factors will be checked, taking into account of factors stratified for in the randomisation procedure. Outcome measures The primary outcome of interest will be neonatal mortality/morbidity. Choice of primary outcome is about what should determine clinical decision-making. However it is recognised that trials to date may have insufficient power and there is a need to consider secondary outcomes of time from randomisation to delivery, preterm delivery before 32 completed weeks (<32+0 weeks) and maternal morbidity as defined in the Cochrane Review (Drakeley 2003)[ 1 ]. We will aim to obtain all neonatal and maternal morbidity outcome data collected in each trial and not just those reported in publications. Reporting of these outcomes in the original trial report is not an eligibility requirement for this review. Data analysis Data on all randomised patients will be requested to perform an intention-to-treat analysis as far as possible. Clinical heterogeneity will be assessed by reviewing the differences across trials in characteristics of randomised patients. Initially, an aggregate data analysis will be undertaken although treatment effect estimates will be obtained from the individual patient data. Binary outcomes will be summarised in terms of odds ratios or relative risks, depending on the degree of heterogeneity observed. Time-to-event outcomes will be summarised in terms of the log (hazard ratio). The I square statistic and chi-square test for statistical heterogeneity will be applied to these summary data. Regression models, stratified by trial, will be used to explore the effects of treatment, risk factors and treatment-covariate interactions on the various outcomes of interest. These will include Cox and accelerated life models for time-to-event outcomes (Tudur-Smith 2004 [ 18 ], Williamson 2002 [ 19 ]) and logistic regression models with trial indicator variables for binary outcomes (Whitehead 2002) [ 20 ]. Factors other than treatment to be investigated are gestational age at randomisation, maternal demographics, obstetric characteristics including obstetric history, cervical length on ultrasound, fibronectin, bacterial vaginosis, multiple pregnancy. Two-level multilevel regression models will be fitted with patients corresponding to level one units and trials as level two units for the various outcomes of interest, adopting the relevant approach for continuous, binary, categorical and time to event outcomes as applicable. Trial effects will be represented by fixed effects whilst treatment effects will be represented by random effects in an attempt to reflect the assumed similar (but not identical) treatment effect across trials. Patient-level covariates (as listed above) will then be incorporated into the model in an attempt to account for some of the remaining statistical heterogeneity. An attempt will be made to incorporate these covariates first of all by assuming their effect to be constant across all trials and subsequently by assuming some heterogeneity in the covariate effect across trials by modelling them either as fixed or random effects. Finally, treatment-covariate interactions will be investigated by including additional variables and adopting a similar approach. If IPD are not available for some trials, the potential for bias will be investigated as follows. The reasons for not being able to obtain the data will be assessed for the potential for bias. Results using aggregate data from these trials will be compared with results using aggregate data from trials where IPD have been supplied, and any difference investigated. The analysis plan will be reviewed in light of the availability of IPD but prior to any comparative analyses. Discussion Predictive models generated by our analysis should allow more effective counselling of women at risk of preterm delivery and thus more cost effective use of cerclage. Competing interests ZA and PRW were authors of a paper that will be included in the IPD meta-analysis (To, 2004). ZA was an author of the non-IPD systematic review on this topic (Drakeley, 2003)[ 1 ]. The authors declare that they do not have any other competing interests. Authors' contributions PRW conceived the idea, CTS drafted the initial protocol, and all authors commented on and approved this final version. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC553972.xml |
535349 | Are both sympatric species Ilex perado and Ilex canariensis secretly hybridizing? Indication from nuclear markers collected in Tenerife | Background Intra-specific and intra-individual polymorphism is frequently observed in nuclear markers of Ilex (Aquifoliaceae) and discrepancy between plastid and nuclear phylogenies is the rule in this genus. These observations suggest that inter-specific plastid or/and nuclear introgression played an important role in the process of evolution of Ilex . With the aim of a precise understanding of the evolution of this genus, two distantly related sympatric species collected in Tenerife (Canary Islands), I. perado and I. canariensis , were studied in detail. Introgression between these two species was previously never reported. One plastid marker (the atpB-rbcL spacer) and two nuclear markers, the ribosomal internal transcribed spacer (ITS) and the nuclear encoded plastid glutamine synthetase ( nepGS ) were analyzed for 13 and 27 individuals of I. perado and I. canariensis , respectively. Results The plastid marker is intra-specifically constant and correlated with species identity. On the other hand, whereas the nuclear markers are conserved in I. perado , they are highly polymorphic in I. canariensis . The presence of pseudogenes and recombination in ITS sequences of I. canariensis explain this polymorphism. Ancestral sequence polymorphism with incomplete lineage sorting, or past or recent hybridization with an unknown species could explain this polymorphism, not resolved by concerted evolution. However, as already reported for many other plants, past or recent introgression of an alien genotype seem the most probable explanation for such a tremendous polymorphism. Conclusions Data do not allow the determination with certitude of the putative species introgressing I. canariensis , but I. perado is suspected. The introgression would be unilateral, with I. perado as the male donor, and the paternal sequences would be rapidly converted in highly divergent and consequently unidentifiable pseudogenes. At least, this study allows the establishment of precautionary measures when nuclear markers are used in phylogenetic studies of genera having experienced introgression such as the genus Ilex . | Background Aquifoliaceae comprise one genus, Ilex [ 1 ] and approximately 400 species. The fossil record indicates that the genus was cosmopolitan during the Eocene. It is now largely extinct in Australia, Europe and Africa where only few species persist. Most diversity is currently found in South-America and in Southeast-Asia. They are evergreen or deciduous trees or bushes living in warm-moist-temperate, sub-tropical, tropical or montane-tropical areas. The molecular phylogeny of the genus Ilex [ 2 , 3 ] shows that systematic relationships are still not well understood. The plastid phylogeny (inferred from the atpB-rbcL spacer, rbcL and trnL-F ) is highly correlated with the geographic distribution of extant species. Four chloroplast clades are found: one exclusively Eurasian clade, one exclusively American clade and two different North-American/Asian clades (one of them comprising most of the deciduous species among other evergreen species). On the other hand, the nuclear phylogeny (inferred from ribosomal ITS and the 5S RNA spacer) is incongruent with the plastid phylogeny, suggesting frequent interlineage hybridizations. The nuclear phylogeny is not correlated with the geographic distribution of extant species. Any of the plastid or the nuclear phylogeny corroborates previous morphological or biosystematic studies [ 3 ]. Using chloroplast RFLPs, trnL-trnF sequencing and nuclear ITS sequencing, a study of Asian Ilex of the Bonin Island and of the Ryukyu Island [ 4 ] confirmed that hybridization played a role in this region, leading to interspecific introgressions independently observed on both Islands. RAPD data indicate that the Japanese species Ilex leucoclada M. is highly polymorphic [ 5 ]. During its history, the genus Ilex probably experienced frequent incomplete lineage sorting and nuclear and/or cytoplasmic introgression, making the study of its history very complex. Few data are reported on the chromosome number of Ilex [ 6 ]. The basic haploid number is 20, with deviation to 17, 18 and 19. From the 27 chromosome numbers available for the genus Ilex , three species are tetraploid ( I. anomala, I. verticillata and I. argentina ) and one species is hexaploid ( I. pedunculosa ), indicating probable hybridizations between species having divergent genomic background (alloploidy). A previous study [ 3 ] showed that individuals of many species of Ilex contain polymorphic nuclear sequences (ITS and 5S rDNA spacer). Except for I. purpurea and I. guianensis , only one individual was studied per species. The sampling being too low for a correct evaluation of this intraspecific polymorphism, an exhaustive study of one plastid marker (the atpB-rbcL spacer) and two nuclear markers (the ribosomal internal transcribed spacers, ITS, and the nuclear encoded plastid glutamine synthetase, nepGS ) was undertaken on several individuals of I. perado and I. canariensis collected in Tenerife (Canary Islands). These species were chosen because, based on DNA data, they are not closely related [ 2 , 3 ] and are growing sympatrically in Canary Islands. Both species are morphologically variable but few characters allow species identification. I. canariensis is endemic of Canary Islands, whereas I. perado has a wider distribution in Spain, Portugal, North-Africa and Canary Islands. Natural or artificial hybridization between both species was never reported. The data show that, contrarily to I. perado , I. canariensis has highly polymorphic ITS and nepGS sequences. The aim of this study was to (1) explain the polymorphism observed in ITS of I. canariensis by an investigation of its pattern of substitution and its functionality, (2) determine the evolutionary mechanisms responsible of this polymorphism and (3) focus on ITS evolution and consequences for phylogenetic reconstruction of the genus Ilex . Results ITS polymorphism Figure 1 (inset) shows the unique plastid atpB-rbcL spacer phylogram obtained from the alignment of the sequences of I. perado and I. canariensis collected in Tenerife. All individuals of I. perado have the same atpB-rbcL spacer sequence. For I. canariensis , 26 individuals have the same atpB-rbcL spacer sequence and three substitutions are observed (in a T-rich variable region) for specimen 39. This plastid marker perfectly agrees with species determination. Figure 1 Most parsimonious atpB-rbcL phylogram (in the inset) and one of the most parsimonious ITS phylogram. Branches conserved in the strict consensus ITS tree are thicker. Bootstrap values are indicated bellow the branches. ITS sequences of 43 Ilex species [3] are indicated by their species name followed by their GC content (in %) and by their DNA accession code in GenBank. ITS sequences of I. perado and I. canariensis collected in Tenerife are indicated by a species code ("per" and "can") and 3 numbers: "can 41_1_45" means I. canariensis specimen 41, clone 1, with a GC content of 45%. Clone 0 means that the sequence was read directly from the PCR product. Specimen "can 28B" represents the shorter PCR product found in specimen 28 of I. canariensis , cut out from agarose gel (see Figure 3). The average GC content is indicated for each clades of the ITS tree. Circled numbers refer to clades discussed in the text (see Figure 4). Bold characters indicate ITS sequences with no substitution at conserved 5.8S sites (see results). Black dots indicate ITS sequences studied in more details (see results). The asterisk indicates an alternative position of the GC 45% clade. As the ITS sequences found in I. canariensis are extremely polymorphic, it was interesting to observe their relationships with available ITS sequences previously investigated species by Manen et al. [ 3 ]. Figure 1 shows one of the most parsimonious ITS tree of I. perado and I. canariensis sequences found in Tenerife altogether with 43 ITS sequences of other Ilex species. Thick bars indicate internal branches conserved in the consensus tree. The closest possible outgroups for the genus Ilex , are Helwingia and Phyllonoma . However these genera are so isolated systematically that their use to root Ilex should be taken with care. On 13 individuals of I. perado , only one substitution (a transition) is observed in ITS. On the other hand, ITS sequences of most individuals of I. canariensis are polymorphic and few sequences are identical. The divergence between all ITS sequences observed in I. canariensis is much higher than between available ITS sequences of all other species investigated. The GC content is 57% for I. perado and from 45 to 62% for I. canariensis . Regarding their GC content, three groups of ITS are found in I. canariensis : a clade with 45% GC, a clade with 61% GC in average and several clades with 53–54% GC in average. The GC content of other investigated species range from 55 to 61% (Figure 1 ). The ITS sequences of I. canariensis are distributed in two groups in the phylogram represented in Figure 1 . One group forms a large clade conserved in the consensus tree but not sustained by bootstrap statistics. Another group forms a small clade (GC 45%) which branches variably: as indicated in Figure 1 , or at the position indicated by the asterisk. The sequences of the GC 45% clade have a 110 bp deletion in ITS 1 and are suggested to represent pseudogenes. In many ITS PCR products of individuals of I. canariensis , a shorter PCR band is visible on ethidium bromide agarose gel electrophoresis altogether with the main ITS band (Figure 2 ). In specimen 28, this electrophoretic band has been cut out and directly sequenced (sample "can 28B"). It has a sequence very close to both the cloned sequences of the GC 45% clade found in specimens 35 and 41. Thus these putative pseudogenes seem rather common in I. canariensis . During selection of the clones to be sequenced, the longest PCR fragments were favored with the aim to select functional sequences. Thus, as the shorter pseudogene band seems frequent in I. canariensis , Figure 1 underscores this class of ITS sequences represented by the clade GC 45%. Figure 2 Example of agarose gel electrophoresis of ITS PCR products of individuals of I. canariensis . The line numbers represent individuals of I. canariensis . The star indicates the position of the expected functional ITS band and the dot indicates the position of the GC 45% ITS pseudogene band. Secondary structure of ITS 2 The secondary structure of both ITS regions is involved in the processing of the rRNA precursor and is thus constrained for this function. In angiosperms, an ITS 2 secondary structure has been proposed and comprises 6 conserved regions (C1 to C6) which are involved in common pairing relationships on the structure [ 7 ]. In order to determine which ITS sequences found in I. canariensis are functional, the secondary structure of ITS 2 was investigated from selected sequences representing a good sampling of ITS (sequences marked with a dot in Figure 1 : per 1_0_57, can 28_B_45, can 36_1_53, can 44_4_54, can 90_2_62 and can 39 3 53). Figure 3 shows functional secondary structures found using Mfold [ 8 ]. Only the sequence of per 1_0_57, can 90_2_62 and can 39_3_53 provide an apparently functional ITS 2 secondary structure showing the common pairing relationships of conserved C1 to C6 regions according to Hershkovitz and Zimmer [ 7 ]. No such ITS 2 secondary structure was found for sequences can 28B_45, can 36_1_53 and can 44_4_54, suggesting pseudogenes. Figure 3 Functional secondary structures of some ITS 2 sequences of I. perado and I. canariensis according to Hershkovitz and Zimmer [7]. The flanking coding regions (3'end of 5.8S and 5'end of 25S) are indicated in bold characters. Conserved regions (C1 to C6) are indicated. Pattern of substitution The GC content, indicated for all ITS sequences of Figure 1 , suggests that the pattern of substitutions is biased towards A or T for sequences of the GC 45% clade and for sequences of the GC 53–54% clades, as expected for pseudogenes. A reconstructed ancestral sequence of Ilex was calculated by maximum likelihood from the ITS data of Manen et al. [ 3 ] and used to investigate the pattern of substitution on selected ITS sequences. Table 1 shows that a higher rate of substitution is observed for ITS sequences found in the GC 45% and GC 53–54% clades than for the I. perado sequences and for the I. canariensis sequences found in the GC 62% clade. This increased rate is statistically significant according to the Kruskal-Wallis rank test [ 9 ]. Table 1 Substitution patterns of I. perado and I. canariensis ITS sequences from a reconstructed maximum likelihood ancestral ITS sequence of Ilex . Rate nmC mC Chi2 per 1_0_57 0.045 4/226 (1.77%) 11/174 (6.32%) * can 28_B_45 0.188 24/226 (10.62%) 37/174 (21.26%) ** can 39_3_53 0.080 11/226 (4.87%) 19/174 (10.92%) * can 44_4_54 0.061 12/226 (5.31%) 16/174 (9.20%) ns can 36_1_53 0.090 10/226 (4.42%) 26/174 (14.94% *** can 90_2_62 0.055 0/226 (0.00%) 2/174 (1.15%) ns Rate: Kimura 2-parameter distance from the reconstructed ancestral sequence. nmC: Number of C>T substitutions / number of non-methylated cytosines on both DNA strands of the reconstructed ancestral ITS sequence of Ilex . mC: Number of C>T substitutions / number of methylated cytosines on both DNA strands of the reconstructed ancestral ITS sequence of Ilex . The corresponding ratios of C>T substitutions are indicated between brackets. Chi2: Chi-square homogeneity test between expected and observed C>T substitutions at methylated cytosines (ns: non significant; *, ** and ***: significant at 0.05, 0.01 and 0.001 levels, respectively). As expected for pseudogenes [ 10 , 11 ], the observed rate of deamination-like substitutions at methylated cytosine sites (CpG and CpNpG sites) is higher than the expected rate of C -> T and G -> A substitutions at non-methylated sites for can 28B_45, can 39_3_53, can 44_4_54 and can 36_1_62 (Table 1 ). A chi-square homogeneity test [ 9 ] indicates that this is highly significant for can 28B_45 and can 36_1_53, which certainly represent pseudogenes. Substitutions at conserved sites of the 5.8S rDNA The alignment of fifty 5.8S sequences (modified from Muir et al. [ 11 ]) shows that 59 sites are conserved in vertebrates, invertebrates, fungi and plants and are expected to be functionally constrained. Substitutions observed at these sites would suggest non-functional pseudogenes. Contrarily to I. perado , many ITS sequences found in I. canariensis have substitutions at some of these conserved sites. Only the GC 61% clade of I. canariensis comprises non-substituted conserved 5.8S sites (sequences indicated in bold in Figure 1 ). Sequences of the GC 45% clade have 10–11 substitutions. Sequences of the GC 53–54% clades have 2 to 7 substitutions. Three sequences of the GC 61% clade have only one mutation, which may be PCR artifacts [ 11 ] and two sequences (can_20_1_58 and can_27_2_59) with a lower GC content (58 and 59%, respectively) have 2 mutations. Thus ITS sequences of I. canariensis having a GC content higher than 60% are expected to be functional genes, all other sequences with lower GC content are suspected to be pseudogenes. Recombinations Most ITS sequences of I. canariensis experienced frequent recombinations: in the entire ITS matrix of I perado and I. canariensis , the DnaSP program [ 12 ] detects 19 minimum possible recombination events (RM). From 0 to 8 minimum possible recombination events are calculated in the different ITS clusters (Figure 4A ). No recombination was detected in I. perado and in clade 2 of I. canariensis . Clades 5 and 6 are highly recombined. An example of recombined ITS sequences of I. canariensis from clade 5 is shown in Figure 4B , where only informative sites are shown. In order to exclude the possibility that the observed pattern of substitution is the result of homoplasy and to confirm that these sequences are actually recombined, maximum likelihood tests were carried out. Using PIST [ 13 ], the maximum-likelihood score of the sequences represented in Figure 4 is compared with the scores of 1000 simulated clonal sequences along the calculated maximum-likelihood tree and under the specified model of evolution (see methods). The observed score q (0.554) was greater than for all 1000 clonal replicates (mean value 0.381, higher value 0.505), indicating a history of recombination (significance 0.001). Figure 4 Recombination evidence in ITS sequences. A: Minimum number of recombination events in ITS clades (numbered as in Figure 1) calculated using the DnaSP program [12]. "perado": I. perado clade. "canariensis": GC 61% clade of I. canariensis representing functional ITS sequences. B: An example of obvious recombined ITS sequences found in I. canariensis clade 5. Only informative nucleotides are represented. Homologous sequence fragments have the same color. Stars indicate the recombination points found by maximum likelihood (program LARD) for sequences can 90_6_54, can 25_2_54 and can 90_4_53 (see results). The LARD maximum-likelihood method [ 14 ] was applied to find the breakpoints in the alignment, which gave the highest likelihood under an evolutionary model incorporating recombination. Only 3 sequences can be analyzed with this program. Three ITS sequences shown in Figure 4 were submitted to LARD: can 90_6_54, can 25_2_54 and can 90_4_53. Two recombination points were located by the program (at the left of positions 242 and at the left of position 582) in accordance with the delimitation indicated in Figure 4 . There is no recombination point between positions 392 and 455 for these particular 3 sequences. Nuclear encoded plastid glutamine synthetase (nepGS) data There is no polymorphism in nepGS of I. perado . On the other hand, I. canariensis shows polymorphism for this gene. Thirty sites differentiate I. perado from I. canariensis , of which eight are polymorphic in I. canariensis , either heterozygous or homozygous (Figure 5 ). For all of these eight polymorphic sites, always one of the alleles is shared with I. perado . Figure 5 Alignment of the nuclear encoded plastid glutamine synthetase ( nepGS ) of I. perado and I. canariensis . Only variable nucleotides are represented. Polymorphic sites of I. canariensis are boxed. R = A or G; Y = C or T; M = A or C; M = A or C; W = A or T; K = G or T. Discussion The high polymorphism of ITS sequences observed in I. canariensis is frequently reported for other plant groups [ 15 , 16 ]. It might have several origins: an incomplete lineage sorting from ancestral polymorphism or an horizontal transfer (introgression) through inter-specific hybridization (alloploidy), both of them not resolved by concerted evolution. Before the discussion on the origin of this polymorphism, the characterization and the fate of these different ITS sequences will be first examined. The genome of Ilex canariensis contains ITS pseudogenes High polymorphism of ITS has been explained by the presence of divergent pseudogenes in Gossypium, Nicotiana, Tripsacum, Exospermum, Zygogonum , Zea [ 10 ], Quercus [ 11 ], Leucaena [ 17 ], Adinauclea, Haldina, Mitragyna [ 18 ] and others. Thus, this could also be the case for I. canariensis . Individual criteria are not sufficient to identify pseudogenes unambiguously [ 17 ] and different criteria were chosen: GC content, secondary structure of ITS 2, rate of substitution, pattern of substitution at methylated cytosine sites and substitutions at highly conserved sites of the 5.8S rDNA. ITS sequences with a GC content of 45% are unambiguously pseudogenes and satisfy to all other criteria. Moreover they have a large deletion in the ITS 1 region, which make these sequences certainly non-functional. The deletion allows an easy detection of this pseudogene on agarose gels and it is observed in many individuals of I. canariensis (Figure 2 ). Other classes of ITS sequences with a GC content of 53–54% are also suspected to be pseudogenes by one or the other criteria but not by all of them, as expected regarding their relatively high GC content. For instance, some ITS 2 sequences of the GC 53–54% class still have a typical angiosperm secondary structure (for instance can_39_3_53, Figure 3 ), but have (1) an increased rate of nucleotide substitution, (2) deamination-like substitutions or (3) mutations at normally highly conserved 5.8S rDNA sites. Only the GC 61 % clade contains functional ITS sequences. Thus, it can be considered that the functional ITS GC content is 57 % for I. perado and above 60 %. for I. canariensis . It is interesting to note that most I. canariensis individuals of the GC 61 % clade never have ITS pseudogenes in the GC 45 % or GC 53–54 % clades. This is probably because these individuals do not contain pseudogenes. For other individuals, a PCR selection for pseudogenes occurred, as reported for Nicotiana [ 10 ], in which ITS sequences with a weak secondary structure (pseudogenes) are preferentially used as templates. The inclusion of dimethylsulfoxide (DMSO) in PCR reactions [ 10 , 19 ], but see [ 18 ], would allow amplification of functional ITS sequences in these individuals of I. canariensis . In conclusion, the high divergence found in ITS sequences of I. canariensis with a GC content lower than 60% (clades 1, 2, 3, 4, 5 and 6) could be explained by a release of evolutionary constraint and a subsequent high rate of substitution. Indeed, ITS sequences have functional constraints in relation with the processing of the rRNA precursor producing the functional 18S, 26S and 5.8S subunits. ITS sequences of Ilex canariensis are recombined Evidence for recombination in divergent sequences is not obvious. It is difficult to recognize homoplasy generated by recombination from actual homoplasy (parallel history). Statistical methods (based on linkage desequilibrium, neutrality tests and substitution distribution along the locus) are still too rudimentary to precisely describe the recombination events in the set of ITS sequences found in I. canariensis . Moreover, recombinants could result from "jumping" PCR reaction [ 20 - 23 ], where prematurely terminated extension products can act as primer on paralogous templates. This has been shown on nepGS for Oxalis [ 24 ] and on four low-copy genes for Gossypium [ 25 ]. The minimum number of recombination events (RM) calculated with DnaSP [ 12 ] underestimates the total number of recombination events [ 26 ]. Thus, there is no doubt that I. canariensis ITS sequences experienced intra-molecular recombinations (Figure 4 ). The factor RM has been also calculated for PCR products of each individual in order to detect possible jumping PCR artifacts. In few of them (specimens 20, 22, 24, 27, 28 and 38) recombinants have been detected in ITS sequences resulting from a unique PCR reaction (data not shown). This could be the result of jumping PCR. However most of them are multiple recombinants and not simple recombinants as it is expected in jumping PCR [ 10 ]. As an example, the alignment represented in Figure 4 shows that specimen 90 comprises two different recombined ITS sequences resulting from the same PCR reaction, that could be the result of jumping PCR. DnaSP did not detect recombination between the four cloned ITS sequences of individual 90 because recombined fragments are paralogous sequences fragments found in other individuals. Moreover, the recombinants result from at least three crossover events and are suggested not PCR artifact. Thus, they represent true organismal intra-molecular recombinations. The distribution of informative characters shown in Figure 4 , as well as the use of programs PIST and LARD based on maximum-likelihood analyses, demonstrate unambiguously that sequences of clade 5 (Figure 4 ) experienced recombination events. This can not be generalized for other clades. Although DnaSP suggests recombination, an alignment demonstrating recombination, as for clade 5, was not possible for other clades, even with the help of PIST and LARD. This could be explained by the recent origin of the recombination events observed in clade 5 and by the fact that mutations did not yet obscured the recombined orthologous fragments. In this respect it is to be noticed that clade 5 shows much longer branches than other clades. This may indicate that, in clades with relatively shorter branches, mutations (or concerted evolution) did homogenize the recombined fragments, mimicking clonal divergence. Thus it can be considered that most clades also comprise recombined ITS sequences, as DnaSP suggests, but of more ancient origin than those of clade 5, and homogenized by mutation or concerted evolution. Recombination in highly polymorphic ITS sequences seems a rule in plants. This is not surprising because the mechanisms of concerted evolution in rDNA arrays are based on crossing-over and gene conversion. It has been reported in Begonia [ 27 ], Microseris [ 28 ], Quercus [ 11 ], Amelanchier [ 29 ], Paeonia [ 30 ], Buddia , Gossypium , Nicotiana , Tripsacum [ 10 ], Armeria [ 31 ] and others. In addition to the high rate of substitution of pseudogenes, at least some ITS sequences experienced recombination. This explains why the divergence between ITS sequences of I. canariensis is much higher than between ITS sequences of all other species investigated, knowing that, according to their GC content (see Figure 1 ), they all are potentially functional. This also explains the absence of a bootstrap support for a monophyletic clade of I. canariensis ITS sequences because of long branch problems due to accelerated rate of substitution and more certainly to recombination. The origin of the ITS polymorphism in I. canariensis Two evolutionary mechanisms could produce the observed ITS polymorphism: an ancestral polymorphism escaping lineage sorting or a past or recent introgression of an alien genotype escaping concerted evolution. Because of the influence of concerted evolution, ancestral polymorphism is not the most likely explanation of ITS polymorphism [ 31 ]. On the other hand, a growing number of reports shows that ITS polymorphism is attributable to interspecific hybridization, although the parents are not always identifiable [ 15 , 16 ]. Assuming that multiple ITS sequences found in I. canariensis are the result of experienced hybridization with another species, or an ancient polymorphism with incomplete sorting, the determination of the identity of the putative hybridizing species or the finding of genetic relationships of the putative polymorphism is not obvious. This is because ITS sequences enclosed in non-functional clusters have dramatically diverged from the putative functional sequences and are recombined. All available ITS sequences of 43 other species of Ilex , representing a good sampling of the genus [ 2 , 3 ] were incorporated in the phylogenetic analysis, altogether with all ITS sequences found in I. perado and I. canariensis of Tenerife (Figure 1 ). Most functional (above 60% GC) and non-functional (53–54% GC) ITS clades of I. canariensis group together but with no bootsrap support. They group with an American lineage ( I. brevicuspis, I. anomala, I. microdunta, I. integerrima, I. theezans, I. guianensis, I. brasiliensis and I. cassine ). Only the GC 45% clade does not group with the bulk of I. canariensis ITS sequences. Its position is not defined and varies in the vicinity of a Eurasian lineage ( I. latifolia , I. leucoclada , I. maximocziana , I. rugosa and I. perado ). Thus data do not support a particular relationship of most I. canariensis ITS pseudogenes with another Ilex species, except for the pseudogenes with a GC content of 45%, that are frequently observed in I. canariensis . In the case of hybridization involving the island species I. canariensis , the most probable candidate would be the sympatric species I. perado . It can not be ruled out however that the distribution of I. canariensis was much wider in the past [ 32 , 33 ] and that this hybridization may have occurred with another unknown or extinct species of the Eurasian lineage represented here by I. latifolia, I. leucoclada, I. maximocziana, I. rugosa and I. perado . Pseudogene sequences (particularly the ITS sequences of clade GC 45%) being too divergent and of different nucleotide composition, the observed relationship of clade GC 45% with the group of species comprising I. perado is questionable because of possible spurious long branch attraction. However, the data of the nuclear encoded plastid glutamine synthetase (a nuclear single copy locus) are not conflicting with an introgression of I. perado in I. canariensis . All the eight polymorphic sites observed in I. canariensis always comprise one allele shared with I. perado . Another possibility is that these ITS pseudogenes represent a relictual ancestral polymorphism in the course of elimination by lineage sorting or concerted evolution. In fact ancestral polymorphism could also be the result of ancient introgressions. The data accumulated here do not allow a definitive conclusion. If a putative cryptic hybridization between I. perado and I. canariensis is confirmed, the introgression would be unidirectional because ITS sequences of I. perado do not show any polymorphism. This situation is reminiscent of the unilateral hybridization observed between Begonia formosana and B. aptera , where on 60 ITS sequences analysed in natural or artificial hybrids, 58 sequences are clustering with the ovule donor B. formosana , and only 2 are found clustering with the pollen donor B. aptera [ 27 ]. Unidirectional interspecific hybridization linked to unilateral incompatibility is frequently described in plants. However, this is not the only mechanism that can explain unidirectional hybridization. The flowering time of I. perado precedes the one of I. canariensis , thus the loading of still living I. perado pollen grains on young effective I. canariensis stigmates is more favored than the contrary. Moreover, there are much more male than female I. perado plants in Tenerife [ 34 , 35 ]. These evidences could explain the proposed unidirectional introgression. Conclusions This study was undertaken with the aim to study and overcome the problem of ITS polymorphism found in many species of Ilex [ 3 ]. Introgression [ 3 , 4 ] and high polymorphism [ 5 ] have already been shown in several species of Ilex . Thus, precautionary measures should be taken when studying nuclear ITS sequences in the genus, particularly the search for recombinant and pseudogenes. Particular PCR conditions should be used [ 10 , 19 , 23 ]. Razafimandimbison and al. [ 18 ] were however unable to find PCR conditions to amplify a functional ITS sequence in 2 species of Rubiaceae. Amplified ITS sequences should be checked for function. A measure of the GC content (above 55% for a functional ITS sequence in Ilex is recommended. This study will probably make phylogenetic interpretations easier and will certainly help to the understanding of the complex evolutionary history of Ilex [ 3 ]. Methods Material Thirteen individuals of I. perado ssp. platyphylla Webb. & Berth. and 27 individuals of I. canariensis Poir. were collected in Tenerife (Las Mercedez, Aqua Garcia and Aqua Mansa) and genomic DNA was extracted from dry leaves as previously reported [ 2 ]. Sequencing The plastid atpB-rbcL spacer was sequenced for the 40 specimens of Tenerife according to Cuénoud et al. [ 2 ]. In a first experiment all ITS sequences (ITS 1, 5.8S and ITS 2) were directly sequenced from the PCR fragment according to Manen et al. [ 3 ]. All individuals of I. perado and five individuals of I. canariensis produced a perfectly readable ITS sequence with no polymorphism. On the other hand, 22 individuals of I. canariensis produced an unreadable highly polymorphic ITS sequence. Cloning in E. coli was necessary and four clones per individual were sequenced. For all specimens, the nuclear encoded plastid glutamine synthetase ( nepGS ) was amplified and sequenced according to Emshwiller and Doyle [ 36 ]. Polymorphisms were also observed in most individuals of I. canariensis , but as indels are not involved in the polymorphism, sequences were readable and polymorphic sites were coded according to the international code for nucleotide polymorphism (see Figure 5 ). Sequences are deposited at GenBank ( atpB-rbcL spacer: AJ786512-AJ786551; ITS: AJ786413-AJ786504; nepGS : AJ809595-AJ809628). Phylogenetic analysis The ITS sequences of I. canariensis and I. perado were aligned with ITS sequences of 43 species of Ilex previously studied in Manen et al. [ 3 ] and Phyllonoma and Helwingia , the closest relatives of the genus Ilex (for the ITS matrix see additional file 1 ). Plastid atpB-rbcL spacer and nuclear nepGS matrices only comprise the sequences found in Tenerife for I. perado and I. canariensis . Maximum parsimony trees were calculated from the atpB-rbcL spacer matrix, the ITS matrix and the nepGS matrix, using PAUP 4.0b10 [ 37 ] (heuristic search, TBR branch swapping with 10 random additions of sequences, only keeping the first 100 most parsimonious trees). Bootstrap statistics of the ITS tree were calculated from 1000 replications with the same method, except that the first 10 most parsimonious trees were kept. Secondary structure of ITS 2 The secondary structure of ITS 2 was investigated using the minimum free-energy program Mfold [ 8 ], which has the advantage to provide sub-optimal folding. Sequences were constrained to force the pairing of the 5'-end of the 26S and the 3'-end of the 5.8S regions according to the results of Hershkovitz and Zimmer [ 7 ]. Pattern of substitution of ITS In order to examine the pattern of substitution of ITS sequences found in I. perado and I. canariensis , these sequences were compared with a reconstructed Ilex ancestral sequence. This ancestral sequence was determined by maximum likelihood [PAUP 4.0b10 with base frequencies, ti/tv ratio, proportion of invariable sites and gamma shape parameter estimated under the HKY model (Hasegawa et al. 1985) allowing for different rate of transitions and transversions as well as unequal base frequencies] using the unique maximum parsimony tree obtained from the ITS matrix of Manen et al. [ 3 ] based on 45 species of Ilex and Helwingia and Phyllonoma as outgroup. For comparisons, the number of substitutions (Kimura 2-parameter) was calculated from this ancestral sequence for I. perado and I. canariensis ITS sequences using PAUP. The high frequency of deamination-like substitutions (C -> T and G -> A at CpG and CpNpG sites) is typical for pseudogenes and was also calculated from the reconstructed Ilex ancestral sequence for I. perado and I. canariensis ITS sequences, according to Buckler et al. [ 10 ] and Muir et al. [ 11 ]. This frequency was compared with the frequency of C -> T and G -> A substitutions at non-methylated sites. Substitutions at conserved sites of the ribosomal 5.8 S subunit Based on the alignment of fifty 5.8S sequences including vertebrates, invertebrates, fungi and plants (modified from Muir et al. [ 11 ]), 59 totally conserved sites were determined. The number of substitutions observed at these invariant sites in all I. perado and I. canariensis 5.8S sequences was calculated. Detection of recombinations The minimum number of recombination events RM [ 26 ] was calculated using the DnaSP program [ 12 ] in the entire ITS matrix of I. perado and I. canariensis and in different ITS clusters observed in I. canariensis . This program is based on linkage desequilibrium, neutrality tests and substitution distribution along the locus. ITS sequences showing strong evidence of recombination, as detected by DnaSP, were submitted to PIST [ 13 ] to calculate the probability of recombination by maximum likelihood: the tree score of these sequences is compared with the tree scores of 1000 simulated clonal sequences along the specified tree under the specified model of evolution. The score of recombined sequences will tend to have larger score than the simulated clonal sequences because richer in conflicting phylogenetic information. A maximum likelihood tree of selected I. canariensis ITS sequences (see results) was constructed with base frequencies, ti/tv ratio, proportion of invariable sites and gamma shape parameter estimated under the HKY model [39] and these parameters were used to calculate the tree scores of simulated clonal sequences. The recombination points of three selected sequences (see results) showing evidence of recombination were calculated by maximum likelihood using LARD [ 14 ] with the HKY model [ 38 ]. The program calculates a maximum likelihood unrooted tree of 3 sequences and searches for a tree with a better score assuming a recombination point in the input sequences. After the calculation of one recombination point, the sequence alignment was truncated at this point to search for other potential recombination points. Supplementary Material Additional File 1 ITS matrix in NEXUS Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC535349.xml |
520744 | Iterative class discovery and feature selection using Minimal Spanning Trees | Background Clustering is one of the most commonly used methods for discovering hidden structure in microarray gene expression data. Most current methods for clustering samples are based on distance metrics utilizing all genes. This has the effect of obscuring clustering in samples that may be evident only when looking at a subset of genes, because noise from irrelevant genes dominates the signal from the relevant genes in the distance calculation. Results We describe an algorithm for automatically detecting clusters of samples that are discernable only in a subset of genes. We use iteration between Minimal Spanning Tree based clustering and feature selection to remove noise genes in a step-wise manner while simultaneously sharpening the clustering. Evaluation of this algorithm on synthetic data shows that it resolves planted clusters with high accuracy in spite of noise and the presence of other clusters. It also shows a low probability of detecting spurious clusters. Testing the algorithm on some well known micro-array data-sets reveals known biological classes as well as novel clusters. Conclusions The iterative clustering method offers considerable improvement over clustering in all genes. This method can be used to discover partitions and their biological significance can be determined by comparing with clinical correlates and gene annotations. The MATLAB © programs for the iterative clustering algorithm are available from | Background Clustering is one of the most common methods for discovering hidden structure in micro-array gene expression data. Clustering of samples has been used to discover new disease taxonomies [ 1 - 3 ]. Cluster analysis is often performed with hierarchical [ 4 ], K-means [ 5 ] or Self-Organizing Map [ 6 ] algorithms, using the entire set of genes as the basis for calculating pair-wise distances between samples. This gives equal weights to the expression of all genes and may be effective in cases where there is a large difference between subsets of samples ( e.g. comparing samples of normal and cancerous tissues). Many diseases, though, are characterized by small numbers of genes that differentiate between different disease states. Giving equal weight to relevant and irrelevant genes will obscure this difference. Figure 1 shows an example, where clustering in all genes masks the biological differences between samples with BRCA1 and BRCA2 mutation (data from Hedenfalk et al [ 7 ]) In this article we propose an iterative algorithm, where we initially do a clustering using all the genes. This clustering (which gives a binary partition of the samples) is used to select genes that differentiate between the two clusters. The clustering is done again, but this time, only in the set of genes that was selected in the previous iteration. This alternation between clustering and feature selection continues until there is no change in the set of genes (and partition) between two iterations. The final gene set is removed, and the process repeated on the remaining genes to find other partitions. The algorithm generates a set of binary partitions, along with corresponding sets of genes which differentiate the clusters present in these partitions. Similar approaches have been used in other algorithms. Ben-Dor et al [ 8 ] use simulated annealing to efficiently search the space of all binary sample partitions. Xing and Karp [ 9 ] use a Normalized Cut algorithm to restrict the search to only the promising partitions and use a similar method of iteration between clustering and feature selection. Von Heydebreck et al [ 10 ] and Tang et al [ 11 ] present algorithms that select sample partitions and corresponding gene sets by defining a measure of partition quality and then using greedy search (in the former) and simulated annealing (in the latter) to maximize this measure. Iteration between cluster analysis and gene selection is also used in the "gene shaving" algorithm of Hastie et al [ 12 ]; though their goal was clustering of genes rather than samples. Algorithm We use a Minimal Spanning Tree (MST) based algorithm [ 13 , 14 ] for clustering along with the Fukuyama-Sugeno clustering measure. Gene selection is done on the basis of the two-sample t-statistic with pooled variance. In the next three subsections we will look in detail at the clustering and feature selection aspects before presenting the formal algorithm. Minimal spanning trees Let V = { x 1 , x 2 ..., x N } be a set of points with distances d ij = d ( x i , x j ) defined between all x i and x j . A tree on V is a graph with no loops whose vertices are elements of V and edge lengths are d ij . A minimal spanning tree (MST) is a tree that connects all points such that the sum of the length of the edges is a minimum. An MST can be efficiently computed in O(N 2 ) time (including distance calculations) using either Prim's [ 13 ] or Kruskal's [ 14 ] algorithm. Deletion of any edge from an MST results in two disconnected trees. Assuming the length of the deleted edge to be δ and denoting the sets of nodes in the two trees as V 1 and V 2 , we have the property that there are no pairs of points ( x 1 , x 2 ), x 1 ∈ V 1 , x 2 ∈ V 2 such that d ( x i , x j ) < δ . Define the smallest distance between any two points, one in V 1 and the other in V 2 , as the separation between V 1 and V 2 . Then we have the result that the separation is at-least δ . The significance of this result is that by deleting an edge of length δ we are assured of a partition where the two clusters have a separation of at-least δ . This means that if we are interested in looking at all binary partitions with large separations between the clusters, it is sufficient to look at partitions obtained by deleting edges of the MST. Instead of looking at all possible binary partitions (which number 2 N -1 -1) our algorithm looks only at partitions obtained by deleting single edges from the MST (which number N -1). Minimal Spanning Trees were initially proposed for clustering by Zahn [ 15 ]. More recently, Xu et al have used MST for clustering gene expression data [ 16 ]. Clustering measure To compare the partitions obtained by deleting different edges of the MST, we use the Fukuyama-Sugeno clustering measure [ 17 ]. Given a partition S 1 , S 2 of the sample index set S , with each S k containing N k samples, denote by μ k the mean of the samples in S k and μ the global mean of all samples. Also denote by the j -th sample in cluster S k . Then the Fukuyama-Sugeno (F-S) clustering measure is defined as Small values of FS(S) are indicative of tight clusters with a large separation between clusters. We have considered various other clustering measures. The ideal clustering measure should show local minima at each viable partition and have good performance even with a large number of noisy features. We have found the Fukuyama-Sugeno (F-S) measure to give the best performance in these two respects (Supplementary data – Additional file 1 ). Feature selection For a given partition with two clusters, we can ask if a particular gene shows sufficient differential expression between samples belonging to the different clusters. A gene which is very differently expressed in samples belonging to different clusters can be said to be relevant to the partition or to support the partition. There can be many ways of measuring a gene's support for a partition. Here we use the two sample t-statistic with pooled variance. The t-statistic is computed for each gene to compare the mean expression level in the two clusters. Genes with absolute t-statistic greater than a threshold T thresh are selected. The percentile threshold parameter P thresh ∈ (0,100) is used to compute T thresh . T thresh is the P thresh /2-th percentile of a random variable distributed according to Student's t-distribution with mean zero and N -2 degrees of freedom ( N is the number of samples). Here we use the t-statistic as a heuristic measure of the contribution of each gene to the selected partition; no statistical significance is implied. The condition for selection of a gene becomes stricter with each iteration. In the first iteration we choose genes with absolute t-statistic greater than T thresh /2. This cutoff increases linearly with the number of iterations until it reaches T thresh . This is done so that we do not lose any useful genes by putting a too-stringent selection criterion before the partition has evolved close to its final form. The algorithm Initially, an MST is created using all the genes; then each binary partition obtained by deleting an edge from the tree is considered as a putative partition. The partition with the minimum value of the F-S clustering measure is selected. The t-statistic is used to select a subset of genes that discriminate between the clusters in this partition. In the next iteration, clustering is done in this set of selected genes. This process continues until the selected gene subset converges (remains the same between two iterations), resulting in a set of genes and the final partition. Having identified a partition and the associated set of genes, these selected genes are removed from the pool of genes. This prevents the algorithm from detecting the same partition the next time. The whole process repeats in the pool of remaining genes to find other partitions. The inputs to the algorithm are the gene expression matrix { x s , g }, the maximum number of partitions to be found MaxN p and percentile threshold P thresh . P thresh is used to compute T thresh . The outer loop of the algorithm runs as long as the number of discovered partitions is less than MaxN p . The set of selected genes F is initialized to be the set of all genes Fset and the cutoff t is initialized as T thresh /2. In the inner loop, an MST is created using the genes in F , and for all partitions obtained by deleting single edges from this MST, the F-S measure is calculated. For the partition P * with the lowest F-S measure, genes are selected from F based on the t-statistic. These selected genes form the new gene set F new . If F new ≠ F , the cutoff t is increased and another iteration of the inner loop is performed. If F new = F , this means that the gene set has remained unchanged between two iterations and the current partition P * along with the current gene set F is output. The number of discovered partitions is increased and another iteration of the outer loop is performed. Since this is an unsupervised method, the partitions picked might be indicative of biological differences that are relevant, irrelevant (like age or sex of patients) or unknown. We control the detection of chance partitions ( i.e. generated due to noise and not due to any biological difference) by requiring a minimum of 2 M (1 - P thresh /100) genes in support of a partition ( M is the total number of genes); the algorithm is terminated if there are fewer. P thresh plays an important part in the kind of partitions that are extracted. A value of P thresh close to 100 will preferentially extract partitions that are supported by genes with large differential expression between the two clusters. A smaller value of P thresh will pick up partitions that are supported by larger number of genes with lower differential expression between the clusters. P thresh cannot be interpreted as a measure of the statistical significance of the partitioning since we are doing both the partitioning and the feature selection on the same set of samples. Here we only use P thresh as a parameter for selecting genes. Algorithm 1: Algorithm for iterative clustering Input MaxN p , P thresh , x s , g ; Fset ← {1, 2..., n}; N p ← 0; /*Number of currently discovered partitions*/ Compute T thresh ; While N p < MaxN p do F ← Fset ; T ← T thresh /2; While 1 do If length of F < 2 M (1 - P thresh /100) then /*Not enough genes support partitions*/ exit; end Create MST in feature set F with metric d ; Delete edges one at a time and calculate F-S measure for each ensuring binary partition; Find partition P * with the lowest F-S measure; Compute t-statistic t g for all genes g ∈ F for this partition; Set F new to the set of genes {g : | t g | > t }; If F new = F AND t = T thresh then /*Feature set has converged */ output P * and F ; /*Remove genes in F from Fset* / Fset ← Fset \ F ; N p = N p + 1; break; else F ← F new ; Increase t ; end end Results Synthetic data We first tested the algorithm on synthetic data to compare its performance against a hierarchical clustering method at detecting planted partitions. We also estimated the probability of detection of spurious partitions created by noise ( i.e. the false detection rate). For both iterative clustering and hierarchical clustering, we found that the probability of detecting the true partition depended only on the Euclidean distance between the clusters in the partition, and for a fixed distance, is relatively insensitive to the number of signal genes (Supplementary data – Additional file 2 ). Figure 2 shows the results of a logistic regression analysis of the dependence of probability of detection of the true partition on the distance between the clusters for both clustering methods. Independent of the total number of genes N , iterative clustering detects the planted partition when the two clusters are separated by about half the distance compared to hierarchical clustering. For genes with similar levels of differential expression, this means that the iterative clustering method will detect clusters supported by a quarter of the number of genes required for detection by hierarchical clustering. The false detection rate was found to be very low: 0.012 for the correlation and 0.011 for the Euclidean distance. Microarray data To test whether classes with strong biological significance can be discovered without knowledge of the class labels, we tested the algorithm on three publicly available sets of micro-array data. 1. BRCA mutation data reported by Hedenfalk et al [ 7 ] with 6512 cDNA clones of 5361 genes for 7 samples with BRCA1 mutation, 8 samples with BRCA2 mutation and 7 with sporadic breast cancer. 2. Leukemia data-set reported by Golub et al. [ 6 ]. Expressions for 7070 genes are provided for 47 acute lymphoblastic leukemia (ALL) samples and 25 acute myeloid leukemia (AML) samples. 3. Lymphoma data-set reported by Alizadeh et al. [ 1 ] containing 46 samples of tissues with diffuse large B-cell lymphoma (DLBCL). Expressions for 4026 genes were measured for each of these samples. It must be noted that if class labels are already available and the goal is to discover genes that differentiate between samples of different classes, then class comparison and class prediction methods exist that are more suitable [ 21 ]. Such methods make use of the prior information (in the form of class labels) to detect genes that are significantly differentially expressed between the various classes. The expression of these genes can be used to develop classifiers that predict the class of new samples. Our iterative method is for cases where no a-priori class labels are assigned. Nevertheless, we have used data for which class labels are known so that there is a ground truth to which the results of the iterative method can be compared. This is similar to what has been done by other authors for validating the results of unsupervised clustering algorithms [ 8 - 11 ]. A-priori gene filtering and normalization performed were similar to that done for the dataset by the original authors. The iterative algorithm was then run with maximum number of partitions N p = 10 and P thresh = 0.999. Table 1 shows the distribution of BRCA1 and BRCA2 samples present in the two clusters for the first four partitions discovered in the BRCA dataset. The fourth partition obtained from the BRCA data separates samples with BRCA1 and BRCA2 mutations with one misclassification. Figure 1 shows the result of hierarchical clustering on the BRCA data-set. The tree structured clustering using all the genes fails to differentiate between samples with BRCA1 and BRCA2 mutations. Figure 3 shows hierarchical clustering using only the genes selected by the iterative clustering method (61 genes). BRCA1 and BRCA2 samples are separated into different branches of the tree with only one misclassification. With the Leukemia data-set, the first partition obtained matches well with ALL-AML classification, with one cluster containing 46 ALL samples (out of 47 total) and 1 AML sample while the second cluster contains 24 AML samples (out of 25 total) and 1 ALL sample (Table 2 ). To see whether the gene set obtained in support of the partition correlating with the AML/ALL classification truly separates ALL and AML samples, we used a split-sample method. The iterative algorithm was used on part of the dataset (38 samples, corresponding to the "training set" used in [ 6 ]) and several partitions were obtained. We did not obtain exactly the same partitions as when the whole dataset was used, but the second partition corresponded well to the ALL/AML classification. It contained one cluster with 25 ALL and no AML samples and another cluster with 11 AML and 2 ALL samples. There were 252 genes that were selected in support for this partition. If the 252 selected genes were truly discriminatory between the ALL and AML samples, then we should be able to separate the two classes in unknown data by unsupervised clustering using these genes. This was verified by clustering the rest of the samples in the dataset (containing 34 samples corresponding to the "testing set" used in [ 6 ]) using these genes. Since the iterative algorithm uses a combination of MST and F-S measure to do clustering, we performed the validation using a similar clustering method. An MST was created using the 252 genes and then the edge to be deleted selected according to minimum F-S measure. This identical to the clustering method used in the inner loop of the iterative clustering algorithm (Algorithm 1). We obtained two clusters with the first cluster containing 20 ALL and 1 AML samples while the second cluster contained 13 AML and no ALL samples. This almost-complete separation of ALL and AML in the testing data shows that the genes selected by the iterative clustering are truly supportive of the partition discovered in the training data. The biological differences present in the Lymphoma data-set were originally detected using hierarchical clustering [ 1 ] after manual selection of genes. We have included our results using the iterative method to show how successful the iterative clustering algorithm is in picking out these disease subclasses (Table 3 ). The third partition best corresponds to the subclasses discovered by Alizadeh et al. One cluster has 24 GC B-like DLBCL samples and 7 Activated B-like DLBCL samples while the other cluster has 16 Activated B-like DLBCL samples. The results from the iterative clustering algorithm is compared to that obtained by Overabundance Analysis (OA) [ 8 ] (Table 4 ) and CLIFF [ 9 ] (Table 5 ). Ben-Dor et al use the Jaccard index [ 20 ] to measure the similarity of the partitions discovered by OA to the true biological classes. For comparison, we calculated the same index for partitions discovered by iterative clustering. The Jaccard index ranges from 0 for complete mismatch to 1 for complete match. Both OA and iterative clustering pick out partitions corresponding to the ALL/AML classification, though OA detects it as the fourth partition while iterative clustering detects it as the first partition. There is a small but definite improvement in the Jaccard index for the results obtained for the Lymphoma data by iterative clustering as compared to OA. Compared to CLIFF, iterative clustering picks a partition in the Leukemia data that is marginally better (2 misclassified as compared to 3 for CLIFF). Discussion We have presented a clustering method that uses a minimal spanning tree to lead the search for partitions of samples that form good clusters. Iteration between minimal spanning tree cluster analysis and feature selection is used to converge onto partitions that form well separated clusters and gene subsets that support these partitions. At the convergence of each set of iterations, the result is a partition of the samples and a set of genes that support them. These genes are removed from the pool of genes before searching for other partitions. This removes genes that obscure other partitions supported by smaller numbers of less differentially expressed genes. Genes that support more than one partition will be selected in favor of the partition for which their support is stronger. Testing on synthetic data shows that the algorithm picks out planted clusters with high accuracy and low false positive rate. Application of the algorithm to breast cancer, leukemia and lymphoma data returns partitions with very well separated clusters, some of which have a strong biological significance. The results are comparable to those obtained by other similar algorithms, and superior to those obtained by standard hierarchical clustering. The kind of partitions discovered depends very much on the value of P thresh . Values of P thresh close to 100 will give preference to partitions that are supported by a small number of very highly differentially expressed genes. On the other hand, smaller values of P thresh will preferentially detect partitions that are supported by a large number of genes differentially expressed to a lesser degree. If the first application of the algorithm returns several partitions that are correlated with each other, then we could suspect that there is one partition that is supported by a large number of genes and run the algorithm again with a smaller value of P thresh to detect all these genes. We have not been able to specify a single value of P thresh that works in all cases, although the range of values we used ( P thresh = 99.9-99.95) works well in most situations. The partitions and supporting gene sets detected by the use of this algorithm must be further analyzed using gene annotation and clinical data to determine whether they are biologically relevant and worth further investigation. The significance of the detected partitions must be further investigated by evaluating the clinical correlates of patients in different clusters. Clinical observations made on patients, like survival duration, response to therapy and grade of tumor can be compared among the clusters obtained to see if there are any detected partitions whose clusters are correlated with clinical features. Another, complementary, approach is to analyze the genes that are differentially expressed between two clusters for regulatory relationships with each other or prior known influence on the disease in question. Software tools for searching gene annotations [ 18 ], and exploring PubMed and GeneCards for prior published relationships among given genes [ 19 ] are available. The results of these two approaches can help the biologist to formulate hypotheses about the significance of the partitions as well as the role of the selected genes in influencing the course of the disease. Examples of this process can be seen in [ 1 ] and [ 2 ]. Methods Synthetic data were created by generating normally distributed expression profiles for each gene. Each planted partition is supported by a fraction of the genes (called signal genes) which were differentially expressed between the two clusters. Each signal gene is differentially expressed to the same extent. The expressions were normally distributed; x s , g ~ N (0,0.5 2 ) for samples s belonging to cluster 1 and x s , g ~ N ( c ,0.5 2 ) for samples belonging to cluster 2. The rest of the genes are not differentially expressed and are called noise genes and are distributed according to a normal distribution N (0,0.5 2 ). If we have k signal genes, each differentially expressed by c between the two clusters, the Euclidean distance between the cluster-means will be . Sets of synthetic data were generated for number of genes M = 1000 and M = 10000 with varying fraction of signal genes ε and distance between cluster means D c . Each set of data was analyzed both by the iterative and the hierarchical clustering method (using average linkage). The iterative clustering method was used to obtain the first partition discovered using the Euclidean distance ( P thresh = 99.95). Hierarchical clustering was used to obtain a tree, and the branch at the highest level was split to produce a partition. The match of these two partitions with the true partition was calculated and the detection accuracy was assigned 1 if the match was greater than 75% and 0 otherwise. A logistic regression analysis was used to model the dependence of the probability of detection on the distance between the clusters D c . To estimate the false detection rate, the algorithm was run on synthetic data containing 10 – 100 samples with P thresh = 99.9. Each sample is a 1000-dimensional vector drawn from a multivariate normal distribution. Thus any clusters detected can be expected to be spurious clusters formed by chance. For the micro-array data, the iterative method was used to detect the first 10 partitions, ( P thresh = 99.9) using (1-correlation coefficient) as the distance measure for the MST. For the BRCA data, we also clustered the data using standard hierarchical clustering using centered correlation as the distance metric [ 4 ] to compare the results of our algorithm with that obtained by clustering with respect to all genes. Supplementary Material Additional File 1 Comparison of clustering measures Synthetic data was created with 100 samples and 1000 genes containing clusters embedded in the first 50 genes. The other 950 genes were normally distributed noise. There are three clusters in the first 50 genes: Samples 1 through 20, samples 21 through 70 and samples 71 through 100. For each binary partition of the points S 1 = {1, 2..., i}, S 2 = {i+1, i+2..., 100}, we calculated the clustering measure. The figure shows the value of the measure for each split point. It can be seen that the Average Linkage and Xie-Beni [22] measures have weak minima and they suffer from extreme values for unbalanced splits. The Log-Likelihood measure has performance similar to the F-S measure but has extreme values for unbalanced splits. Click here for file Additional File 2 Detection of true partition for different data parameters Sets of synthetic data were generated for 1000 and 10000 total number of genes with varying fraction of signal genes ε and distance between cluster means D c . The figure shows detection of planted partition for various values of ε and D c . Blue points are data for which the percentage match between the first discovered partition and the planted partition is less than 75%. The red points are data for which the match is greater than 75%. Detection (match > 75%) depends only on the distance between the clusters for both hierarchical and iterative clustering. Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC520744.xml |
547907 | Quetiapine augmentation of SRIs in treatment refractory obsessive-compulsive disorder: a double-blind, randomised, placebo-controlled study [ISRCTN83050762] | Background Although serotonin reuptake inhibitors are effective in the treatment of OCD, many patients fail to respond to these agents. Growing evidence from open-label and placebo-controlled trials suggests a role for augmentation of SRIs with atypical antipsychotics in OCD. Quetiapine is generally well tolerated and previous open-label data has produced mixed results in OCD and additional controlled data is needed. Methods We undertook a double-blind, randomised, parallel-group, flexible-dose, placebo-controlled study of quetiapine augmentation in subjects who had responded inadequately to open-label treatment with an SRI for 12 weeks. Following informed consent and screening, forty-two subjects were randomised to either placebo or quetiapine for six weeks. Results There was significant improvement from baseline to endpoint on the Yale-Brown Obsessive-Compulsive Scale in both the quetiapine and placebo groups (quetiapine, n = 20, p < 0.0001; placebo, n = 21, p = 0.001) with 40% (n = 8) of quetiapine and 47.6% (n = 10) of placebo treated subjects being classified as responders. Quetiapine did not demonstrate a significant benefit over placebo at the end of the six-week treatment period (p = .636). Similarly quetiapine failed to separate from placebo in the subgroup of subjects (n = 10) with co-morbid tics. Quetiapine was generally well tolerated. Conclusions In this study, quetiapine augmentation was no more effective than placebo augmentation of SRIs. A number of limitations in study design make comparisons with previous studies in this area difficult and probably contributed to our negative findings. Future work in this important clinical area should address these limitations. | Background Obsessive-compulsive disorder (OCD) is a prevalent, chronic and disabling disorder [ 1 ]. Controlled pharmacotherapy studies have established superiority of serotonin re-uptake inhibitors (SRI's) over noradrenaline reuptake inhibitors and over placebo in OCD and these currently form the cornerstone of pharmacotherapy management [ 2 ]. Despite the considerable advances made with the introduction of the SRI's into clinical practice, 40–60% of subjects still fail to respond adequately to initial therapy [ 3 , 4 ]. From this it is clear that a need exists to pursue more effective treatments for those with OCD who fail to respond or respond inadequately to SRI's. To this end, preliminary evidence supports a role for the addition of atypical antipsychotics to SRIs in OCD. These agents combine serotonin-dopamine antagonism with the advantage of being well tolerated including a low potential for inducing motor side-effects. To date a number of open-label studies have suggested that augmenting SRI's with atypical antipsychotics is an effective strategy for treatment-refractory OCD. These include support for risperidone [ 5 - 7 ], olanzapine [ 8 - 13 ], and more recently amisulpride [ 14 ] and quetiapine [ 15 - 18 ]. A single open-label study using quetiapine as augmentation showed lack of effect in a small sample using low doses [ 19 ]. The outcome of the first controlled study in this area with the antipsychotic haloperidol demonstrated preferential benefit for refractory OCD subjects with co-morbid tic disorder [ 20 ]. In two subsequent studies the efficacy of risperidone in SRI refractory OCD has also been reported [ 21 , 22 ]. Interestingly the former study [ 21 ], did not replicate the particular advantage for subjects with co-morbid tic disorder. Efficacy has also been shown for quetiapine [ 23 ] and olanzapine [ 24 ] using similar designs, but the effects on co-morbid tic disorders were not reported. In contrast a recent controlled study using olanzapine failed to demonstrate efficacy over placebo in a six week study [ 25 ]. Despite some mixed evidence in this area, in general the available literature appears to support the use of relatively short trials with low doses of antipsychotic agents as augmentation to SRIs. Quetiapine has a particularly interesting profile in that it is the only available antipsychotic with significant 5-HT1D effects and this serotonin receptor subtype has been implicated in OCD [ 26 , 27 ]. Our objective was to examine the effects of quetiapine augmentation in subjects with OCD who had failed to respond adequately to a 12 week trial of an SRI, employing a double-blind, placebo-controlled, six week study design. Methods Patients Forty two subjects aged 18–65 years inclusive were recruited in our multi-centre study comprising five sites in South Africa and one in Canada. Recruitment took place between May 2002 and November 2003. Prior to commencement, all sites in the study received approval from their relevant Research Ethics committees/Institutional review boards and regulatory authorities. All subjects provided written informed consent prior to the commencement of any study-related procedures. Diagnosis was confirmed using the MINI Neuropsychiatric Interview (Version 5.00, 1998) [ 28 ] to ensure compatibility with the Diagnostic and Statistical Manual for Mental Disorders, Fourth Edition (DSM-IVTR)[ 29 ] criteria for OCD. Subjects with any co-existing Axis I disorder were excluded unless the co-morbid condition was deemed to be secondary to the OCD. Female subjects of childbearing potential were required to use adequate contraception and were not permitted to breastfeed while on the study. Subjects were excluded if they suffered from unstable medical conditions including renal or hepatic insufficiency, epilepsy or had suffered previous brain injury or undergone brain surgery. Taking medication that was deemed likely to interact with quetiapine or any other psychoactive substance was grounds for exclusion. Study Design All subjects were treated and monitored by investigators for the minimum twelve week duration of SRI-alone treatment phase before inclusion into this study. This was to ensure that patients met criteria for duration of SRI treatment which included at least 6 weeks on the maximum tolerated dose of the relevant SRI. Sample size calculations were based conservatively on similar work in this area [ 20 , 21 ]. Accordingly, for the primary outcome variable (YBOCS), clinically meaningful differences between treatment groups of 6.67 with a standard deviation (SD) of 6 would be detected with a power of 80% at a 5% significance level with a sample size of 14 in each of the treatment groups. The larger sample recruited reflects the anticipation of a 33% drop-out rate in the double-blind treatment phase. A double-blind, randomised, parallel-group six-week augmentation with quetiapine or matching placebo of the SRI to which participants had not responded adequately, was undertaken. Specific SRI's, mean doses and dose range are provided in Table 1 . Non-responsiveness to an SRI was defined as either an improvement score on the clinical global impression scale of minimally improved (3) or worse (4,5,6), or less than 25% reduction in Yale Brown Obsessive-compulsive score following twelve weeks of treatment. Inadequate response, as defined above, to at least one SRI administered for a minimum of 12 weeks of which 6 weeks was either at the maximum tolerated dose or alternatively the manufacturer's recommended maximum daily dose. SRI doses were maintained at the same level throughout the double-blind treatment phase. For assignment to either quetiapine or placebo groups, we used a computer generated randomization schedule supplied by the sponsoring pharmaceutical company which also packaged the medication. This procedure ensured blinded, balanced allocation to each treatment group across all the study sites. All investigators remained blind to this schedule until closure of the study. No incidents requiring investigators to break the blind occurred through the course of the study. Table 1 SRI's used by subjects for failed treatment trial prior to inclusion in the study. Drug N Mean maximum tolerated dose (mg/day) Median daily dose (mg/day) Range Fluoxetine 13 60 60 60 Citalopram 10 61 60 10 Paroxetine 4 65 60 20 Fluvoxamine 10 290 300 100 Sertraline 1 200 200 0 Clomipramine 2 250 250 0 Treatment At baseline participants were randomly allocated to receive treatment with either quetiapine or placebo using a computer generated schedule and numbered dispensing wallets. A flexible dosing schedule was initiated at 25 mg per day for one week and then doubled weekly to the start of week 4. Based on Clinical Global Impression of Improvement (CGI-I) scores of minimally improved or worse, clinicians were permitted to increase the dose to a maximum of 300 mg per day for the final two weeks of the study. In addition to clinical measures of improvement, clinicians also considered patient tolerability in their decision to adjust doses. Following completion of the treatment phase, subjects were withdrawn from study medication while continuing their SRIs. All subjects were then followed up for any adverse effects. Ratings Patients were assessed by clinicians at baseline and on completion of weeks 2, 4 and 6. Telephonic assessments were performed on completion of weeks 1 and 3. Symptoms of obsessive-compulsive disorder were measured by the same clinician where possible at all study visits using the Yale Brown Obsessive Compulsive Scale YBOCS) [ 30 , 31 ]. A global assessment of severity and improvement was made by clinicians at all assessment points using the Clinical Global Impressions scale of Severity (CGI-S) and Improvement (CGI-I) [ 32 ]. Depression was rated using the 10-item Montgomery-Asberg Depression rating scale (MADRS) [ 33 ]. For a measure of patient-rated disability we used the Sheehan Disability scale (SDS) [ 34 ]. For subjects with tics, frequency and severity were rated using the Yale Global Tic Severity Scale (YGTSS) [ 35 ]. Our primary outcome measures for OCD symptoms were (1) the change in YBOCS score from baseline to endpoint and (2) the clinical global impression of improvement (CGI-I) at endpoint. In the final analysis, treatment response was defined as a 25% or greater reduction in YBOCS score and a CGI-I of 1 (very much improved) or 2 (much improved) from baseline to endpoint. Secondary outcome measures included the MADRS, SDS and YGTSS (in subjects with co-morbid tics). Statistical analysis Thirty-nine of the forty-two randomised subjects successfully completed the six-week treatment phase. Two subjects withdrew from the study prematurely (Week 1 and Week 4) due to severe levels of sedation. In both of these cases at least one week of study medication had been taken and at least one post-baseline clinical assessment was completed. Both of these subjects were included in the final analysis using data from the last observation carried forward (LOCF). The single subject not included in the efficacy analysis completed the study, but was found not to have correctly fulfilled the study definition of treatment refractoriness and was excluded. Twenty subjects were allocated to the quetiapine arm and twenty-one to the placebo arm. Student's t-tests were used to determine any baseline differences in the groups for age, gender, number of previous trials of SSRI's, severity of symptoms in relation to OCD, depressive symptoms, and CGI-S. Analysis of variance was undertaken with group and tics as factors. All tests were two-tailed with p-values of less than 0.05 considered significant. Results Study sample characteristics For the final analysis, our sample comprised 19 men and 22 women. Baseline characteristics of two treatment groups did not differ with respect to age (years) (quetiapine group 33.8(SD 9.66), placebo group 31.81 (SD 12.14); p = 0.57), gender (p = 0.29), number of previous adequate SRI trials (quetiapine 1.55 (SD 1), placebo 1.62 (SD 1.02) (p = 0.83), baseline severity of OCD (CGI-severity, p = 0.47; YBOCS, p = 0.33), depressive symptoms (p = 0.91), patient-rated disability (p = 0.28) or the presence (n = 11, p = 0.66 and severity (p = 0.87) of co-morbid tics. Treatment outcomes For the primary outcome measure of severity (YBOCS), quetiapine (p < 0.0001) and placebo (p = 0.001) augmentation of an SRI significantly improved symptoms of OCD. However quetiapine did not demonstrate significant benefit over placebo at the end of the six-week treatment period (F = .19; p = .636) (Figure 1 ). The mean reduction in YBOCS scores for the combined group was 7.15 points (quetiapine = 7.10; placebo 7.19). Forty percent (n = 8/20) of subjects on quetiapine were classified as responders (YBOCS reduction of >25% from baseline and CGI-improvement score of 1 or 2) while 47.6% (n = 10/21) of subjects on placebo were classified as responders. A higher number of previous SRI trials for the each treatment group did not correlate with the degree of change on the YBOCS or the response status. Table 2 provides details of individual subject SRI doses, baseline clinical severity ratings and response status. Table 3 provides a summary of baseline and change scores for each of the primary and secondary outcome variables. Figure 1 YBOCS change for treatment groups Quetiapine and placebo groups improved significantly, without significant between group differences (F = 0.19; p = 0.636) Table 2 Baseline characteristics of treatment groups Treatment Group SRI baseline SRI Dose baseline (mg/day) Previous SRI trials Total YBOCS – Baseline Total YBOCS – Week 6 Endpoint dose (mg/day) % CHANGE CGI – Improvement Response status Quetiapine 1 Paroxetine 60 1 33 29 50 -12.00 3 N/R 2 Citalopram 60 2 25 23 200 -8.00 4 N/R 3 Fluvoxamine 300 1 32 27 300 -16.00 3 N/R 4 Citalopram 70 1 27 25 25 (E/W)* -7.00 4 N/R 5 Clomipramine 250 2 22 27 100 23.00 5 N/R 6 Paroxetine 60 5 35 32 300 -9.00 4 N/R 7 Fluoxetine 80 3 25 16 300 -36.00 2 R 8 Sertraline 200 1 18 3 50 -83.00 1 R 9 Fluoxetine 20 1 21 20 300 -5.00 4 N/R 10 Fluoxetine 60 1 22 17 300 -23.00 2 N/R 11 Citalopram 60 1 32 17 300 -47.00 1 R 12 Fluoxetine 80 1 32 14 300 -56.00 1 R 13 Fluvoxamine 300 1 30 12 50 -60.00 1 R 14 Fluoxetine 60 1 25 7 50 -72.00 1 R 15 Citalopram 60 1 27 24 200 -11.00 4 N/R 16 Fluvoxamine 300 1 24 12 150 -50.00 2 R 17 Fluvoxamine 200 2 22 22 50 .00 4 N/R 18 Fluvoxamine 300 2 25 22 25 -12.00 4 N/R 19 Fluoxetine 60 2 24 25 25 (E/W)* 4.00 4 N/R 20 Clomipramine 250 2 27 12 300 -56.00 2 R Placebo 1 Fluvoxamine 300 2 32 25 300 -22.00 3 N/R 2 Fluvoxamine 300 1 30 28 300 -7.00 3 N/R 3 Fluoxetine 80 2 34 38 300 12.00 4 N/R 4 Paroxetine 60 1 22 18 300 -18.00 1 N/R 5 Citalopram 60 1 28 26 300 -7.00 4 N/R 6 Fluoxetine 60 5 26 24 300 -8.00 3 N/R 7 Fluoxetine 60 1 23 23 300 .00 4 N/R 8 Fluoxetine 60 1 27 10 300 -63.00 1 R 9 Fluoxetine 40 1 32 18 300 -44.00 2 R 10 Citalopram 60 2 24 23 300 -4.00 4 N/R 11 Citalopram 60 1 35 23 300 -34.00 2 R 12 Fluvoxamine 300 3 22 14 300 -36.00 2 R 13 Citalopram 60 1 28 10 50 -65.00 2 R 14 Fluoxetine 60 1 28 4 50 -86.00 2 R 15 Citalopram 60 1 26 19 100 -27.00 2 R 16 Fluoxetine 60 1 27 12 100 -56.00 1 R 17 Fluoxetine 20 1 26 18 200 -31.00 2 R 18 Fluvoxamine 300 2 23 35 200 52.00 6 N/R 19 Citalopram 60 2 26 9 100 -65.00 2 R 20 Paroxetine 80 1 32 29 300 -9.00 3 N/R 21 Fluvoxamine 300 3 31 25 100 -19.00 2 N/R *E/W = Early withdrawal Table 3 Summary scores (baseline) and change scores for primary and secondary outcome variables. Quetiapine Placebo YBOCS (baseline) 26.4 (SD4.6) 27.7(SD3.9) YBOCS (change at week 6) -7.1(SD7.2) -7.2(SD8.4) YBOCS % change -26.9% -26% CGI-Severity (baseline) 5.2 (SD0.8) 5.3 (SD0.8) CGI-Severity (week 6) 4.1 (SD1.4) 4.1(SD1.5) MADRS (baseline) 10.6 (SD 4.8) 10.71 (SD 9.8) MADRS (change at week 6) -2.6 (SD 6.5) -3 (SD 8.3) SDS (baseline) 17.9 (SD 5.3) 19.6(SD4.7) SDS (change at week 6) -5.3(SD5.6) -6.1 (SD4.8) YGTSS (baseline) 24.7 (SD 19.3) 22.6(SD 22.3) YGTSS (change at week 6) -4.5(SD 5.1) -9.4 (SD 14.6) YGTSS % change -18.2% -41.6% Of the 11 subjects with co-morbid tics, six were randomised to quetiapine. Endpoint data was missing for one subject on quetiapine. Of the remaining 10 subjects, 3 (quetiapine n = 2 (33%); placebo n = 1(20%)) were classified as YBOCS responders. The reduction in the YGTSS did not differ significantly between treatment groups with tics (quetiapine -4.5, placebo -9.4; F = 2.8, p = .46). Severity ratings for depressive symptoms (MADRS) were low at baseline (mean 10.6, SD 4.8), showed little change over the study period, and at week 6 remained similar for both groups (quetiapine = 8.2, SD 4.8; placebo = 7.7, SD 6.1). The mean daily dose at week 6 for the quetiapine group was 168.75 mg (SD 120.82) compared to 228.57 mg (SD 99.46) per day for those on placebo. Quetiapine responders (187.5 mg, SD 124.6) did not differ significantly from quetiapine non-responders (156.25 mg, SD 122.1) in their mean daily dose at Week 6 (p = .585). Furthermore, within the quetiapine group, participants receiving ≥ 200 mg/day (10/20 at week 6 demonstrated non-significant differences (F = 6.837, p = .988) and a marginally lower percentage reduction in YBOCS at endpoint (26.7%, SD 20.34) compared to those receiving a dose ≤ 200 mg/day (26.9%, SD 36.24) at endpoint. Tolerability Quetiapine was generally well tolerated and no serious adverse events (SAE's) were reported through the course of the study period. Two patients on quetiapine withdrew from the study due to severe sedation (Week 1 and Week 4) that was judged to be drug related. Otherwise adverse events were in the mild to moderate range and were mostly self-limiting. No subjects on placebo withdrew from the study. Table 4 provides a list of the adverse events and their frequencies in the respective study groups. Table 4 Percentage of subjects for each treatment group reporting adverse events Adverse event Quetiapine (%, n) Placebo (%, n) Sedation 75% (15) 33.3%(7) Dry mouth 15% (3) 0 Headache 15% (3) 38% (8) Fatigue 15% (3) 19% (4) Irritability 10% (2) 4.7% (1) Impaired concentration 10% (2) 0 Dizziness 5% (1) 14.3% (3) Nausea 5% (1) 9.5% (2) Increased appetite 5% (1) 9.5% (2) Delayed ejaculation 5% (1) 0 Weight gain 5% (1) 0 Worsening mood 5%(1) 4.7%(1) Memory difficulties 5%(1) 0 Muscle aches 5%(1) 0 Abdominal tenderness 5%(1) 0 Slurred speech 5%(1) 0 Discussion Our findings indicate that both quetiapine and placebo significantly reduced symptoms in subjects with OCD who had failed to respond adequately to 12 weeks of an SSRI and, that the difference between groups was not significant. Similarly in the subgroup with co-morbid tics, no preferential benefit was noted for quetiapine. Interestingly, the high placebo response was similar to that seen in a recent failed controlled trial of olanzapine [ 25 ], but stands in contrast to the positive studies in this area in which low placebo response rates were seen when demonstrating efficacy of quetiapine [ 23 ], risperidone [ 21 ], and olanzapine [ 24 ]. It is likely that features of study design or specific study population characteristics may have contributed to this finding and these are discussed below. First, the duration of a therapeutic trial of an SRI prior to augmentation with an antipsychotic should be of adequate dose and duration. In our study the majority of participants had failed only the single trial of an SRI on which they continued during the study (63.4% mean 1.59). Notably only six weeks of this treatment was required at the maximum tolerated dose. Despite the notion that an optimum trial of pharmacotherapy in OCD is 12 weeks, it may be argued that higher and ultimately effective doses of an SRI had not been maintained for an optimum duration prior to randomization. Given that therapeutic doses of SRIs in OCD are usually on the upper end of the dose range, it seems feasible that the high placebo response rate may reflect a response to SRI's once they had been administered at these higher doses for the additional six weeks of the study. It seems possible that the recent study by Shapira et al [ 25 ] may have been impacted by similar factors. In a second and related point; the number of previous SRI trials in the subgroup receiving quetiapine did not predict a poorer response to treatment. This effect is probably related to the lack of statistical power to detect these differences in a group in which the low number of previous SRI trials was a distinguishing characteristic. Certainly, previous positive studies in this area have used relatively more refractory groups based on the number of previously failed SRI trials. Taken together with the first point above, we suggest that future work in this area should consider longer periods at maximum tolerated doses of SRI's prior to categorisation of subjects as treatment refractory. Third, the use of a slow up-titration resulted in a relatively low mean daily dose being administered for the majority of the study. For instance, a mean daily (week 6) dose of 168 mg/day in the quetiapine group (median 175 mg) had only been achieved for the final two weeks of the study. These doses are comparably low to those used in the negative single-blind study using low dose quetiapine by Sevincok et al [ 19 ]. In contrast the positive study using quetiapine by Denys et al [ 23 ] employed a more rapid up-titration and a fixed-dose design. This meant that subjects were exposed to 200 mg daily doses that were generally well tolerated, from the start of week 3. The authors of this study were able to show significant YBOCS differences between groups from the end of week 4. Similarly Mc Dougle et al [ 21 ], using risperidone, began treatment on 1 mg per day for one week and permitted weekly 1 mg incremental increases for 6 weeks. They found that by the beginning of week 2, most subjects were on or around the mean daily dose for treatment responders (2.2 mg). Despite the significant improvement in the quetiapine group demonstrated in our study, the apparent lack of benefit of doses higher than 200 mg per day may seem surprising, however, we cannot rule out the possibility that administering these higher doses for an adequate duration would have changed the outcome. In addition it must be noted that in our study the quetiapine group reported high rates of sedation (n = 15, 75%) and a 10% (n = 2) rate of premature withdrawal was experienced. As such it seems likely that a more aggressive up-titration schedule might have resulted in even higher rates of withdrawal. By comparison, rates of sedation were equally high, but did not appear to restrict use of the more rapid up-titration in the study by Denys et al [ 23 ]. Certainly evidence of efficacy using lower doses has been demonstrated in studies of 6 and 8 weeks duration [ 21 , 23 , 24 ], and it seems that therapeutically adequate doses should probably be reached earlier than week 4 in a 6 week study. Fourth, the impact of repeated clinical assessments and rating of relatively small changes in clinical severity combined with regular dose increases, may conceivably have increased the placebo response rates resulting from increased optimism, a tendency to over-report improvements and belief that higher doses are more likely to be more effective than lower doses. This may be particularly true for the placebo-treated group that were considerably less likely to report sedation as an adverse event and as such were more likely to have their treatment dose increased at each visit. Our results differ, with respect to placebo response, from a considerable literature that suggests a consistently lower placebo response rate in treatment trials in OCD than in other mood and anxiety disorders. While we believe that the reasons (1–3) discussed above probably provide the main reasons for our finding, the impact of repeated assessments and the potential effect thereof cannot be entirely discounted. Conclusions Despite significant improvement in each of the study groups, response to quetiapine augmentation in SRI non-responders, failed to separate from placebo treated subjects at the end of the six week treatment phase. A number of limitations in study design make comparisons with previous studies in this area difficult and probably contributed to our negative findings. Future work in this important clinical area should address these limitations. Competing interests Dr. Stein has received research grants and/or consultancy honoraria from Astrazeneca, Eli-Lilly, GlaxoSmithKline, Lundbeck, Orion, Pfizer, Pharmacia, Roche, Servier, Solvay, Sumitomo, and Wyeth. Authors' contributions PDC drafted manuscript and analysed data, BV protocol design, SS protocol design, JEM drafting of manuscript, statistical design, data acquisition, MVM protocol design, DJS principal investigator, protocol design, drafting of manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC547907.xml |
544598 | Catechol-O-Methyltransferase (COMT) Val108/158 Met polymorphism does not modulate executive function in children with ADHD | Background An association has been observed between the catechol-O-methyltransferase ( COMT ) gene, the predominant means of catecholamine catabolism within the prefrontal cortex (PFC), and neuropsychological task performance in healthy and schizophrenic adults. Since several of the cognitive functions typically deficient in children with Attention Deficit Hyperactivity Disorder (ADHD) are mediated by prefrontal dopamine (DA) mechanisms, we investigated the relationship between a functional polymorphism of the COMT gene and neuropsychological task performance in these children. Methods The Val 108/158 Met polymorphism of the COMT gene was genotyped in 118 children with ADHD (DSM-IV). The Wisconsin Card Sorting Test (WCST), Tower of London (TOL), and Self-Ordered Pointing Task (SOPT) were employed to evaluate executive functions. Neuropsychological task performance was compared across genotype groups using analysis of variance. Results ADHD children with the Val/Val , Val/Met and Met/Met genotypes were similar with regard to demographic and clinical characteristics. No genotype effects were observed for WCST standardized perseverative error scores [F 2,97 = 0.67; p > 0.05], TOL standardized scores [F 2,99 = 0.97; p > 0.05], and SOPT error scores [F 2,108 = 0.62; p > 0.05]. Conclusions Contrary to the observed association between WCST performance and the Val 108/158 Met polymorphism of the COMT gene in both healthy and schizophrenic adults, this polymorphism does not appear to modulate executive functions in children with ADHD. | Background Attention Deficit Hyperactivity Disorder (ADHD) is a childhood psychiatric disorder characterized by symptoms of inattention, impulsivity and motor hyperactivity afflicting 6–8% of school-aged children in North America [ 1 , 2 ]. Although ADHD is a disorder with complex and heterogeneous etiology, genetic factors appear to play a significant role in predisposing and perpetuating the development of the disorder as evidenced by twin [ 3 , 4 ], family [ 5 - 7 ], and adoption studies [ 8 ]. Association studies have implicated several susceptibility loci including a 40-base pair (bp) allele of the Variable Number of Tandem Repeats (VNTR) polymorphism of the SLC6A3 gene [ 9 ] and a 48-bp repeat polymorphism of the DRD4 gene [ 10 ]. Attempts to replicate these findings have met with modest success possibly owing to the clinical heterogeneity characteristic of the disorder [ 11 ]. One method that may act to augment the strength of these associations would be to identify endophenotypic intermediates conferring risk for the development of ADHD rather than attempting to identify direct linkages between genetic variations and the behavioural manifestation of the disorder. Theories of dysregulated dopamine (DA) pathways in ADHD have been supported by the efficacy of dopamine agonists in reducing the core symptoms of the disorder [ 12 ]. The mesocortical DA pathway appears to be integral to prefrontal cortex (PFC)-mediated cognitive functioning, specifically working memory [ 13 ], through the enhancement of task-related neural activity via D1 receptor activation [ 14 ]. Both PET [ 15 ] and SPECT [ 16 ] imaging studies support a neuromodulatory role for DA in the PFC during tasks of executive function. In addition, administration of DA agonists to the rat PFC acts to enhance working memory in these animals [ 17 ]. Consistent with this line of thinking, children with ADHD show deficits in performance of tasks of executive function [summarized in a meta-analysis by Sergeant et al. (2002)] [ 18 ] and significant improvement of performance under methylphenidate [ 19 , 20 ]. These findings have prompted the hypothesis that the overt symptoms of ADHD are the manifestation of an underlying deficiency in a range of PFC-mediated cognitive domains, including working memory, planning, and set shifting, collectively regarded as executive function [ 21 - 23 ]. The hypothesized role of a dysfunctional mesocortical dopaminergic pathway in the development of symptoms of ADHD has encouraged the investigation of candidate genes involved in this pathway including SLC6A3 [ 9 ], DRD4 [ 10 ] and, more recently, the catechol-O-methyltransferase ( COMT ) gene [ 24 ]. The COMT , encoded by a gene located on chromosome 22q11, catalyzes the degradation of catecholamines, most importantly DA [ 25 ]. A functional polymorphism of this gene, involving a substitution of Valine ( Val ) for Methionine ( Met ) at codon 108/158 ( Val 108/158 Met ), results in a 4-fold variation in enzyme activity, with individuals homozygous for either the Val or Met allele exhibiting either reduced or preserved levels of DA respectively [ 26 ]. Although the dopamine transporter (DAT) is the predominant means of DA termination in most dopaminergic neurons [ 27 ], considerable evidence exists to suggest that the DAT may play a reduced role within the PFC [ 28 - 32 ], where other clearance mechanisms may be implicated. Comparison of DA metabolite levels within discrete brain loci in both rats [ 33 ] and monkeys [ 34 ], as well as the measurement of DA levels in COMT knock-out mice [ 35 ], suggest an important functional role for COMT in the PFC. If COMT is indeed inextricably linked to DA metabolism within the PFC, it is reasonable to assume that variations in enzyme activity, as dictated by the Val 108/158 Met polymorphism, may modulate the performance of tasks of executive functioning in healthy individuals, as well as individuals with reduced PFC basal dopamine levels. In support of this assumption, associations have been reported between the Val 108/158 Met polymorphism and performance on the Wisconsin Card Sorting Test (WCST) in healthy adults [ 36 , 37 ]. In adults with Schizophrenia, a disorder characterized by dopaminergic hypofrontality [ 38 ], associations have also been observed between the COMT polymorphism and WCST performance [ 39 - 41 ]. Although one study reported an association between the COMT polymorphism and ADHD using a haplotype relative risk design [ 24 ], this study failed to investigate any indices of executive function and several other studies failed to replicate this finding [ 3 , 42 - 44 ]. Given the putative role of COMT in DA metabolism within the PFC [ 33 - 35 ], we hypothesized that the Val 108/158 Met polymorphism of the COMT gene will be associated with alterations in performance on tasks of executive function, a behavioural index of PFC integrity and function [ 45 ]. Since dysfunctional DA neurotransmission [ 46 ] and deficient neuropsychological task performance [ 18 ] are both characteristic of children with ADHD, we further hypothesized that this association would be evident within this particular clinical population. Specifically, ADHD children expressing the high enzymatic activity Val allele (H), resulting in reduced PFC DA neurotransmission [ 26 ], will show more pronounced deficits in neuropsychological task performance than their low enzymatic activity Met allele (L) counterparts. In order to test this hypothesis, we used three measures of executive function: the WCST [ 47 ], a measure of set-shifting ability capable of differentiating between ADHD children and controls [ 18 ] and associated with the COMT polymorphism in normal [ 36 , 37 ] and schizophrenic adults [ 39 - 41 ]; the Tower of London (TOL) [ 48 ], a measure of planning ability, which consistently differentiates ADHD children from controls [ 18 ], and the Self-Ordered Pointing Task (SOPT) [ 49 ], a measure of working memory also capable of differentiating between ADHD children and controls [ 18 ]. Methods Subjects 118 children were recruited from the Disruptive Behaviour Disorders Program and the children outpatient clinic at the Douglas Hospital. They were referred to these specialized care facilities by school principals, community social workers, and paediatricians. Inclusion criteria required children to be between the ages of 6 and 12 years of age, meeting DSM-IV diagnosis criteria for ADHD [ 50 ]. Diagnosis of ADHD was based on a structured clinical interview of parents using the DISC-IV (parental report) [ 51 ], school reports, teacher interviews, and clinical observation of the child. In the majority of cases, mothers were the primary informants for the collection of clinical information. Written reports from the child's school were also available in the majority of cases. Parents completed the Child Behavioural Checklist (CBCL) [ 52 ], a scale that assesses a variety of behavioural domains, and the Conners' Global Index for parents (CGI-P) [ 53 ]. Teachers also completed the Conners' Global Index (CGI-T) [ 54 ]. Assessments were made while children were free of medication. Exclusion criteria included a history of mental retardation, with an IQ less than or equal to 70 as measured by the WISC-III [ 55 ], and history of Tourette Syndrome, pervasive developmental disorder, psychosis or any medical condition or impairment that may interfere with the child's ability to complete the study. Neurocognitive assessment A comprehensive neuropsychological test battery assessing different aspects of the central executive functions was administered to all children by trained research personnel. All children were assessed subsequent to a one-week medication "wash-out" period. Children were permitted to take breaks upon request and, in some cases, testing was carried out over two sessions. On average, the testing procedure lasted 1.5 hours. The research protocol was approved by the Research Ethics Board of the Douglas Hospital. Parents were explained the study and provided written consent. Children were also explained the study and gave their assent to participate as well. Tests were selected according to their ability to tap into various performance domains of executive function. We restricted the number of tests in each domain in order to balance comprehensiveness with the co-operation of patients. Abstraction and concept formation were evaluated by means of the WCST (perseverative errors) [ 47 ]. In this task, children are required to sort cards according to three different criteria (colour, number, or shape of signs presented on cards). Feedback on whether the child achieved a correct or incorrect match is given after each trial. The matching criterion changes after ten consecutive correct matches and the child has to identify the new matching criterion using the feedback (correct/incorrect) provided to them. Evidence of the reliability and validity of the WCST with various normal and clinical populations has been reported in several studies [ 18 ]. Planning capacity was evaluated using the TOL [ 48 ]. This test is used to assess planning and problem solving aspects of executive functioning. The validity and reliability of the TOL has been reported in numerous studies [ 18 ]. Standardized administration and scoring procedures as well as normative data have been developed for paediatric populations [ 56 ]. Visual Working Memory was evaluated using the abstract version of the SOPT [ 49 ]. In this task, series of matrices of 6, 8, 10, and 12 images are presented to the child. The child is asked to select, by pointing, one different image on each page. Errors occur when the child points to images previously selected on the preceding pages. Each set is presented to the child three times. Successful performance on this task involves working memory as well as planning and monitoring skills. Shue & Douglas (1992) have reported significant differences in performance between ADHD children and normal controls on the SOPT [ 57 ]. Molecular genetics The Val 108/158 Met polymorphism of the COMT gene was genotyped using a PCR based method as previously described [ 26 ]. The PCR was performed in a 25 μl total reaction volume containing 1X PCR buffer, 200 uM dNTPs, 200 ng of primers (5'-GCGATGGTGGCACTCCAAGC; 5'-TTGGAGAGGCTGAGGCTGAC), 1 unit of Taq DNA polymerase, and 100 ng of genomic DNA. PCR products were electrophoresed on agarose-TAE gel along with 1 kb ad 100 bp DNA ladders, visualized under UV-light and coded according to the length of the PCR product. Genotypes were called by two independent and experienced technicians who were blind to all clinical data. No disconcordance in any of the readings was noted. Children were stratified according to genotype only after all neuropsychological task data was collected. Statistical analyses The Val 108/158 Met polymorphism consists of both the low-activity Met (L) and high-activity Val (H) alleles. Subjects were stratified into three groups: two homozygous genotype groups (LL, HH) and one heterozygous genotype group (HL). A one-way analysis of variance (ANOVA) where genotype (LL, HL, HH) was the independent variable and neuropsychological task performance (standardized WCST perseverative error score, standardized TOL total item score) was the dependent variable was performed. For the SOPT, no normalized scores are available and testing procedures involve several levels of difficulty (4). We therefore used a two-way, repeated measure, mixed design analysis of covariance (ANCOVA), where genotype and level of task difficulty were the between and within subjects independent variables, respectively, neuropsychological task performance (SOPT raw error score) was the dependent variable, and age was the covariate. As the TOL also involves multiple levels of task difficulty (12), we repeated the analysis for this test using the same statistical approach as that applied to the SOPT. A one-way ANCOVA, where genotype was the independent variable and age was the covariate, was performed on all other non-standardized measures of neuropsychological task performance (WCST number of categories completed, WCST number of trials to first category, TOL number of problems solved). An investigation of linkage and within-family association between quantitative phenotypes (standardized WCST perseverative error score, standardized TOL error score, and SOPT error score) was conducted utilizing the Quantitative Trait Disequilibrium Test (QTDT) statistical software package [ 58 ]. Results Table 1 shows clinical and demographic information for the children stratified according to genotype [n = 23 for LL (19.5%), n = 66 for HL (56.0%) and n = 29 for HH (24.5%)]. The three groups were similar with regard to age, average household income, severity of behavioural problems as assessed by the CBCL, and mean number of inattention items, mean number of hyperactivity items and distribution of ADHD subtypes according to the DISC-IV. No significant differences existed between the groups in IQ as measured by the WISC-III. Our sample was characterized by a high prevalence of comorbid disorders, particularly oppositional defiant disorder and conduct disorder. The frequency of these disorders was equally distributed between the genotype groups. The proportion of subjects who had never received medication for ADHD within each genotype group was also remarkably similar. Although a significant effect of gender was observed between genotype groups (χ 2 = 7.39; df = 2, p = 0.02), this result was treated as a type I error (false positive) due to the absence of female subjects with the HH genotype and given the relative lack of female representation across all genotype groups. However, given the previously observed association between gender and several polymorphisms at the COMT loci [ 59 ], increasing the sample size to achieve a more comparable gender representation and distribution would be a valuable revision to the present design. Table 1 Demographic and clinical characteristics of children with ADHD separated according to COMT genotype LL (23) HL (66) HH (29) p-value Gender (M/F) 20/3 52/14 29/0 χ 2 = 7.39, df = 2 p = 0.02 Age 9.2 (2.0) 9.0 (1.8) 9.3 (1.7) F 2,115 = 0.21, p = 0.81 IQ 97.2 (13.7) 97.5 (13.5) 95.6 (13.8) F 2,98 = 0.17, p = 0.84 CBCL (total score) 68.0 (9.8) 70.9 (10.4) 68.9 (8.9) F 2,112 = 0.87, p = 0.42 Income (% less than 20 K) 32 % 42 % 48 % χ 2 = 1.39, df = 2 p = 0.50 DISC-IV Inattention Items 7.3 (1.5) 6.9 (2.2) 7.2 (2.3) F 2,113 = 0.46, p = 0.63 DISC-IV Hyperactivity Items 5.9 (2.4) 6.4 (2.3) 6.4 (2.7) F 2,113 = 0.33, p = 0.72 DISC-IV ADHD Subtype (I/H/C) 10/3/10 14/13/39 7/3/19 χ 2 = 5.68, df = 2 p = 0.22 Comorbid ODD 13/23 50/66 20/27 χ 2 = 3.21, df = 2 p = 0.20 Comorbid CD 5/23 27/64 8/27 χ 2 = 3.57, df = 2 p = 0.17 Never Medicated 11/22 38/62 18/28 χ 2 = 1.17, df = 2 p = 0.56 CBCL = Child Behavioral Checklist. DISC-IV = Diagnostic Interview Schedule for Children fourth edition. ODD = Opposition Defiant Disorder, CD = Conduct Disorder. ADHD Subtypes: I = Inattentive, H = Hyperactive, C = Combined. Values are mean (SD). The genotype distribution conformed to a Hardy-Weinberg equilibrium (χ 2 = 0.42; df = 2, p = 0.81). 156 parents participated in the study and gave blood samples. Among these parents, 76 were heterozygous (M = 43 and F = 33) and transmitted the Val allele to their affected children in 28 occurrences, whereas this same allele was not transmitted in 29 occurrences [χ 2 = 0.02; df = 1, p > 0.05 (transmission disequilibrium)]. Conversely, parents transmitted the Met allele to their affected children in 29 occurrences, whereas this same allele was not transmitted in 28 occurrences [χ 2 = 0.02; df = 1, p > 0.05 (transmission disequilibrium)]. In addition, results from the QTDT revealed no evidence of linkage or within-family association between the three quantitative phenotypes and the COMT gene. A one-way ANOVA performed on these data revealed no significant difference between the LL, HL, and HH genotypes according to WCST standardized perseverative error scores [F 2,97 = 0.66, p > 0.05](Table 2 ) and TOL standardized total item scores [F 2,99 = 0.97, p > 0.05](Table 2 ). A repeated-measure, mixed design ANCOVA performed on these data revealed no effect of genotype on SOPT raw error scores [F 2,108 = 0.62, p > 0.05] (Table 2 ), TOL raw item scores [F 2,107 = 0.35, p > 0.05], and TOL time to complete each trial [F 2,108 = 0.04, p > 0.05]. No genotype by task interaction was observed for SOPT raw error scores [F 6,327 = 0.39, p > 0.05], TOL raw item scores [F 11,1199 = 1.63, p > 0.05], and TOL time to complete each trial [F 11,1210 = 1.65, p > 0.05]. A one-way ANCOVA performed on these data revealed no effect of genotype on WCST number of categories completed [F 2,96 = 1.94, p > 0.05], WCST number of trials to first category [F 2,96 = 1.04, p > 0.05] and TOL number of problems solved [F 2,112 = 1.04, p > 0.05]. No genotype effects were observed when the HL and HH genotype groups were combined into one category and contrasted with the LL genotype (recessive model) on WCST standardized perseverative error scores [F 1,98 = 1.11, p > 0.05], WCST number of categories completed [F 1,97 = 0.01, p > 0.05], WCST number of trials to first category [F 1,97 = 0.36, p > 0.05], TOL standardized total item scores [F 1,100 = 0.42, p > 0.05], TOL raw item scores [F 1,108 = 0.22, p > 0.05], TOL time to complete each trial [F 1,109 = 0.07, p > 0.05], TOL number of problems solved [F 1,113 = 1.33, p > 0.05] and SOPT raw error scores [F 1,109 = 0.85, p > 0.05]. Table 2 Neuropsychological task performance in children with ADHD LL (23) HL (66) HH (29) ES p-value WCST 96.3 (15.1) 99.1 (11.8) 100.6 (12.2) 0.31 F 2,97 = 0.67, p = 0.52 TOL 103.3 (16.5) 99.5 (15.1) 103.8 (12.6) 0.03 F 2,99 = 0.97, p = 0.38 SOPT 13.5 (6.9) 15.1 (8.8) 15.8 (8.2) 0.31 F 2,108 = 0.62, p = 0.54 WCST = Wisconsin Card Sorting Test standardized perseverative error score (LL: n = 21, HL: n = 56, HH: n = 23). TOL = Tower of London standardized score (LL: n = 20, HL: n = 55, HH: n = 27). SOPT = Self Ordered Pointing Task error score (LL: n = 23, HL: n = 63, HH: n = 26). ES = Effect size for LL vs. HH. Values are mean (SD). Discussion Previous studies have identified an association between the COMT polymorphism and a variety of indices reflecting executive control both in healthy [ 36 , 37 ] and schizophrenic adults [ 39 - 41 ]. The COMT appears to be important to the regulation of dopamine metabolism within the PFC [ 33 - 35 ]. Since the PFC and dopamine pathways have been hypothesized to play an important role in the pathogenesis of ADHD [ 9 - 11 , 60 , 61 ]), we conducted this study in an attempt to test whether the COMT Val 108/158 Met polymorphism, which is known to be associated with a significant change in the catabolic capacity of this enzyme, modulates the risk for ADHD or various indices of executive control. Contrary to our expectations and findings in both healthy [ 36 , 37 ] and schizophrenic adults [ 39 - 41 ], an association between the Val 108/158 Met functional polymorphism of the COMT gene and neuropsychological task performance reflecting executive control was not observed in children with ADHD. This result is consistent with the findings of a recent case-control study conducted by Mills et al. (2004), which, to our knowledge, is the only other study to investigate the relationship between the COMT Val 108/158 Met polymorphism and neuropsychological task performance in children with ADHD [ 62 ]. However, this study did not include the WCST, the measure responsible for producing the most consistent results in the previous literature. In addition, we did not identify a biased transmission of either of the two alleles from parents to affected offspring. The absence of an association between the COMT Val 108/158 Met polymorphism and behavioral indices of executive function in children with ADHD may be explained by the young age of the population of patients included in the present study. Indeed it is possible that, due to age-related changes in the functional importance of the COMT within the prefrontal cortex, this association is observable only in adults. This possibility is supported by data in both rats [ 63 - 65 ] and humans [ 66 , 67 ] suggesting that monoamine content and metabolism decrease with age. This age-related decrease may render functions dependent on monoamine content more prone to be dysfunctional at an older age. In addition, evidence from rat studies has indicated a positive correlation between aging and COMT activity [ 68 - 70 ]. This observation may suggest that the implication of the COMT in the catabolism of dopamine is developmentally regulated, with children relying less on this catabolic pathway than adults. Conversely, it has been reported that DAT density is inversely correlated with age [ 71 ]. Taken together, the presence of an inverse and direct correlation between age and DAT density on the one hand and COMT activity on the other hand, may suggest that dopamine metabolism relies more on the DAT than on COMT activity in children compared to adults. This hypothesis is compatible with the fact that several studies have identified an association between the DAT [ 9 , 60 , 72 - 74 ], but not the COMT , gene and ADHD. It is also possible that the negative result observed in the present study is due to a type II error (false negative) secondary to the lack of power of our sample to detect an association. However, using results from the WCST, the variable for which relevant genetic data already exists, we conducted a power analysis and determined that our sample size has sufficient power (80% at α = .05) to detect a mean difference of 11.2 on this measure. Furthermore, it is possible that some of the tests used in our assessment are mediated by the PFC but insensitive to PFC DA levels [ 75 ]. An additional limitation of the present study is that some genotype groups included few subjects. Increasing the sample size to achieve larger genotype groups would be necessary to reach firmer conclusions. This is particularly true for female subjects who were significantly underrepresented in the study (as is common to most clinical studies of ADHD). In order to generalize these negative results to females, a more comparable gender representation is required, particularly in view of some previous research indicating that the allelic distribution of the COMT may be gender dependent [ 59 ]. Conclusions This study does not support the involvement of the Val 108/158 Met polymorphism of the COMT gene in increasing the risk for ADHD or in modulating several indices of executive functions in children with ADHD. This result is contrary to previous findings in both healthy and schizophrenic adults and may be related to developmental specificities. Competing interests The author(s) declare that they have no competing interests. Authors' contributions ET performed the data analysis and drafted the manuscript. NG was involved in the conception of the study and provided clinical support. LBA provided clinical support and aided in data collection. PL provided clinical support. VM aided in neuropsychological testing and data collection. RD and ATZ performed the genotyping for the study and aided in data management. MTS coordinated the clinical aspects of the study and was involved in data management. CB provided clinical support. RJ was responsible for the conception of the study, drafting of the manuscript, and supervision of the research project. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544598.xml |
547913 | MILANO – custom annotation of microarray results using automatic literature searches | Background High-throughput genomic research tools are becoming standard in the biologist's toolbox. After processing the genomic data with one of the many available statistical algorithms to identify statistically significant genes, these genes need to be further analyzed for biological significance in light of all the existing knowledge. Literature mining – the process of representing literature data in a fashion that is easy to relate to genomic data – is one solution to this problem. Results We present a web-based tool, MILANO (Microarray Literature-based Annotation), that allows annotation of lists of genes derived from microarray results by user defined terms. Our annotation strategy is based on counting the number of literature co-occurrences of each gene on the list with a user defined term. This strategy allows the customization of the annotation procedure and thus overcomes one of the major limitations of the functional annotations usually provided with microarray results. MILANO expands the gene names to include all their informative synonyms while filtering out gene symbols that are likely to be less informative as literature searching terms. MILANO supports searching two literature databases: GeneRIF and Medline (through PubMed), allowing retrieval of both quick and comprehensive results. We demonstrate MILANO's ability to improve microarray analysis by analyzing a list of 150 genes that were affected by p53 overproduction. This analysis reveals that MILANO enables immediate identification of known p53 target genes on this list and assists in sorting the list into genes known to be involved in p53 related pathways, apoptosis and cell cycle arrest. Conclusions MILANO provides a useful tool for the automatic custom annotation of microarray results which is based on all the available literature. MILANO has two major advances over similar tools: the ability to expand gene names to include all their informative synonyms while removing synonyms that are not informative and access to the GeneRIF database which provides short summaries of curated articles relevant to known genes. MILANO is available at . | Background In the post-genomic era, biologists encounter a flood of information derived mainly from microarray experiments. The blessing of this wealth of information is accompanied by a great difficulty in identifying the biologically significant findings, which are often embedded in irrelevant information. Currently, there are several approaches to deal with this problem. One approach is to identify a category of genes which is overrepresented in the microarray output. This approach can be carried out using the Gene Ontology project (GO) which describes gene products in terms of their associated biological processes, cellular components and molecular functions [ 1 ]. The advantage of this approach is that it can be easily automated and thus can be used for quick screening of large outputs. On the other hand, this approach limits the analysis to the structure of the GO project and thus does not support the desire of many researchers to customize their analysis. A second approach involves searching the literature for information about each of the genes on the list. Although this approach is comprehensive, it suffers from many downsides: it is time consuming; there is no systematic way to integrate the information learned about each gene; usually one gets distracted with seemingly interesting comparisons early on during the literature search and thus does not give the genes at the end of the list the same weight that was given to genes that appear at the top of the list; there are multiple names and symbols for each gene and thus it is hard to extract the literature information for any particular gene since each author may refer to it differently. A third approach entails curated databases that have gathered all the known information pertaining to each gene. This approach is limited by the quality of the curation process. For example for studying the yeast Saccharomyces cerevisiae , there are excellent curated databases, such as the Yeast Proteome Database [ 2 ] and the Saccharomyces Genome Database [ 3 ], which contain all the known information about each gene. On the other hand in other organisms the curation procedure is at a less advanced stage and thus the information contained in the curated databases is still partial. We have developed an analysis tool that combines the advantages of all the mentioned approaches and overcomes some of the disadvantages. Our tool (MILANO – Microarray Literature-based Annotation) uses an automatic search of literature databases for performing custom annotation of the list of genes obtained from a microarray output. This is done by generating dynamic annotations for genes, built according to terms provided by the researcher. The program receives as input a list of gene identifiers obtained from any microarray experiment and a set of custom search terms. The program expands each gene identifier to its informative synonyms and searches literature databases for co- occurrences of every gene on the list with each of the custom terms. The program's output is an annotation table with the numbers of publications for each gene-term combination (hit-counts). This novel annotation format can be easily used within a web browser or a spreadsheet program to quickly identify genes within the list that are related to the terms provided by the researcher, and may be easily extended, as every hit-count in the annotation is a hyperlink to the query's results. The great advantage achieved by this method over standard static annotations, such as Gene Ontology (GO) annotations, is that the annotations are generated based on terms provided by the researcher, and therefore help in addressing the specific scientific question the researcher is pursuing. The program is able to search two literature databases, GeneRIF [ 4 ] and Medline [ 5 ]. GeneRIF contains ~90,000 short summaries of curated articles relevant to known genes. An initial search of the microarray results against the GeneRIF database provides results within minutes and is easily evaluated, thereby providing immediate insights to the microarray results. This search is followed by a comprehensive Medline search via Pubmed, allowing the identification of more subtle biological insights. To demonstrate the power of this strategy, we have analyzed a list of 148 genes affected by over-expression of p53 [ 6 ]. Our analysis assisted in retrieving from the list 11 known p53 targets, which are all the known targets in the list, and in identifying within the p53-affected genes a subset of putative p53 target genes that are known to be involved in apoptosis (43 genes), in cell cycle arrest (21 genes), and in Cancer (48 genes) as shown in Figure 3 . This example demonstrates the usefulness of our tool in narrowing down microarray results to a small list of genes involved in a specific biological activity. Implementation Web Interface MILANO is accessed through a familiar web form (Fig 1A ). A CGI (Common Gateway Interface)-based Perl [ 7 ] program is executed on submission, which creates the combined Boolean searches for the requested databases. The user can decide whether to provide gene symbols directly, or provide LocusLink/Gene numbers, which are expanded to synonyms as described below. Results, formatted as an HTML table, are displayed immediately on-screen for GeneRIF searches and sent by e-mail for Pubmed searches. Synonym expansion Gene aliases are collected from the LocusLink database file, downloaded from the NCBI ftp server [ 8 ]. We use an awk [ 9 ] program to extract gene symbols, aliases and product names. The alias collection is then processed by a Perl program that removes symbols that are shorter then three characters or that appear in a 23,000-word English dictionary, enhanced for scientific terms. This database is stored in a fashion than enables us to extract processed aliases for a gene by its LocusLink number. Pubmed searches Pubmed searches are performed by a Perl program which uses the NCBI eutilities esearch web service for accessing the Pubmed database [ 10 ]. There are limitations on when and how often we can query the NCBI server, so we integrated into the program a mechanism that makes sure that is does not make more than one query every three seconds. The Generic NQS (Network Queuing System) [ 11 ] ensures that jobs that include more than 100 queries run only between 9 p.m. and 5 a.m. ET. GeneRIF searches The GeneRIF collection is automatically downloaded weekly from the NCBI ftp server [ 12 ], and processed by a Perl program to include gene symbols from the synonym expansion database into every GeneRIF. The database is then indexed by a database server (SRS 7.1.3, Lion Bioscience AG), which provides a query interface for counting and displaying GeneRIF entries. Results Expanding the search terms One of the major problems in the automation of literature searches is the ambiguity in gene names [ 13 ]. Multiple names are used in the literature for any specific gene and thus it is not straightforward to define the Medline query that will find most of the relevant information on a gene. In order to overcome this problem we used the LocusLink database [ 14 ] to expand any gene symbol to all its synonyms. We also included in the expanded form the gene product name since many genes are mentioned in the literature by their product name and not by one of their symbols (for example most of the citations for the beta actin gene can be found by searching the Medline with the term "beta actin" and not with its official symbol " ACTB "). Although expansion of the search terms is a useful tool to increase the number of articles retrieved for each gene it also adds many irrelevant articles due to the fact that some of the gene symbols are not informative as Medline query terms. For example one of the aliases of the gene aquaporin_1 is CO , a term that is mostly mentioned as an abbreviation for Carbon mono-oxide, and one of the aliases of the gene CD36_antigen is FAT , which is found in over 100,000 articles, unrelated to CD36. In order to diminish this problem we filtered out from the list of gene symbols any term that was shorter than three characters and any term that is an English word. In order to check our name expansion strategy we conducted a Medline search for 16862 well-known human genes (all the genes that have an NM number indicating the identification of their full length mRNA), using three search strategies: using only the official symbol for each gene (Symbol), using the official symbol together with all its aliases and the gene product (Expanded) and using only the informative terms (Filtered). Using the Expanded search allowed the identification of literature information on about ~1900 additional genes over a query using the official symbol only (Table 1 ). Using the Filtered search terms allowed this addition without adding significantly to the number of queries that returned non-reasonable results. In addition to expanding the number of genes that were found in the literature, the Filtered search terms also increased the number of articles found per gene (from an average of 198 articles per gene found by searching with the symbol alone to an average of 451 articles per gene when searching with the filtered terms). These results indicate that our gene name expansion strategy achieves a higher percentage of relevant literature for each gene while limiting the addition of irrelevant information. Conducting automatic literature searches After expanding the search terms, MILANO performs an automatic search of literature databases, and retrieves the number of hits each query returned. MILANO performs Boolean searches in which one can search for co-occurrence of each of the primary terms (the expanded gene name) with user defined secondary terms (Figure 1 ). The program's output is a table (Figure 2 ) containing the number of publications for each gene-term combination (hit-counts). This table could serve as an annotation table, because the number of publications reflects the relationship between the genes to the secondary term used. For example a gene that has a role in DNA damage will appear in more articles about "DNA damage" or "gamma irradiation" than unrelated genes. In order to assist in further evaluation of the results, we have built the annotation table such that each number in the table is a hyper-link to the literature database and thus clicking on it will perform this specific search again and will open a window containing the actual abstracts found by this combination of search terms. Literature databases supported by the program The MILANO program can search two databases (Figure 1 ) – the full Medline database, currently containing more than 12,000,000 references, and the GeneRIF database that contains more than 90,000 short summaries of curated articles relevant to known genes. There are several advantages in using the GeneRIF database over the full Medline: the searches are quick and the results are obtained within minutes; each article is summarized by a sentence or two, reducing the amount of information that needs to be read; the curation procedure extracts from the papers only the information relevant to the gene, minimizing the cases in which two terms appear in the same abstract but are not related to each other; the GeneRIF entries are based on the full text of the articles and not only on the abstracts. However, since the curation procedure is an on-going process, the coverage of this database is only partial and thus information is missing and can be found only by performing a Medline search. For that reason our tool allows a combined search strategy in which both databases can be searched simultaneously. The GeneRIF database provides results within minutes and is easily evaluated, thereby providing immediate insights to the microarray results. In parallel a comprehensive Medline search can be done. Although this search takes longer and its results obtained by email, it allows the identification of more subtle biological insights. P53 To demonstrate the power of our literature-based annotation strategy, we analyzed a list of 148 genes affected by over-expression of p53 [ 6 ]. This list of genes was obtained by microarray experiments and nicely demonstrates the difficulty of microarray analysis since it contains many putative p53 target genes and their relevance to p53 cellular activity is not clear. Our first aim was to identify the known p53 target genes that were affected by p53 overproduction in this experiment. By using specific secondary terms, we were able to trim down the list of 148 genes to a much shorter list of genes highly enriched for known p53 target genes (Figure 3A ). In order to evaluate the number of target genes that were missed by our annotation strategy, we manually compiled a list of all known p53 target genes, ~60 genes. Eleven of these 60 genes were represented in the list of genes affected by over-expression. Our automatic annotation strategy found all of them. Moreover, the use of MILANO reduced the amount of articles per gene from an average of 2088 articles per gene in the initial list to 56 articles per gene in the limited list (Figure 3B ). The p53 example also demonstrates the usefulness of searching the GeneRIF curated database in which the use of the secondary term p53 allows filtering out most of the irrelevant genes without losing any known target gene (Figure 3A ). P53 is involved in apoptosis, cell cycle arrest and cancer. It is interesting to find out which of the genes affected by p53 is involved in these processes. Using MILANO we easily identified genes known to be involved in these processes (Figure 3C ), which helped the process of analyzing the microarray data. Comparison with other tools Recently, few literature mining tools has been developed, using a similar approach to the one presented here [ 15 - 17 ], however all of them suffer from the problem of inappropriate use of primary search terms. In order to demonstrate the advantage of using MILANO over the other tools, we have performed a comparative analysis of all these tools by looking at their performance on the 11 known p53 target genes described above. The software were run with these 11 genes as the primary search terms and "P53" as the secondary term and reported the number of occurrences of those terms. The results that are summarized in table 2 demonstrate that MILANO-GeneRIF search was the only method that revealed connections between all 11 genes and p53 and that the MILANO-Medline search gave the most comprehensive search results. PubMatrix [ 15 ] does not expand the primary search terms and thus it misses many literature occurrences. This problem is best demonstrated by its poor performance on the CDKN1A gene which is one of the most studied targets of p53. The synonym expansion methods used by MicroGENIE [ 16 ] improved the results regarding the CDKN1A gene, but missed the SFN gene completely, and gave non-informative synonyms to XRCC5 and TRAF4 ("Ku" and "TNF" respectively). BEAR GeneInfo [ 17 ] did not perform synonym expansion correctly for CDKN1A, and gave non-informative synonyms for PCNA and TRAF4 ("cyclin" and "h. mln62 mrna" respectively). When we attempted to analyze the full data set of 148 genes, some of the compared tools failed to give results due to errors. Discussion MILANO is a simple and intuitive literature search tool. It allows automatic Medline and GeneRIF searches followed by a quick survey of the results. Using this tool dramatically reduces the time needed to query literature databases. Moreover, due to its systematic nature, it assists in treating the 1 st and the 100 th query in an unbiased manner. The MILANO program uses all the published information for the annotation of each gene according to its co-occurrence in the literature with a user defined secondary search term. These features of MILANO makes it especially suitable for analyzing microarray results, since it can be used to annotate the results with terms defined by the user and not limited by preset terms such as the GO terms based annotation. We have demonstrated the power of our program by the analysis of a list of 148 genes that were deregulated in cells that overproduced the p53 tumor suppressor gene [ 6 ]. Frequently one of the first tasks in microarray data analysis is to determine the overlap between new results and results expected based on the literature. For example in analyzing the list of genes induced by over expression of p53 one expects to find known p53 target genes. Thus, we applied our automatic literature search tool in order to answer this question. We found that use of this tool dramatically shortens the time needed for such an analysis by allowing the researcher to focus on a relatively small subset of potential target genes and by reducing the amount of literature relevant to each gene (Figure 3 ). Our tool was also found useful in automatically sorting the target genes into functional groups. Based on the knowledge of p53 cellular functions we defined secondary search terms that fit p53's main activities – apoptosis and cell cycle arrest [ 18 ]. Using these terms allowed the quick identification, from the primary list, of a subset of genes that were not known to be involved in those processes and thus may be interesting for further research (Figure 3C ). Several literature mining approaches have been developed to integrate multiplex biological datasets into the context of published medical literature. A good example of such an approach is the PubGene program [ 19 ], which searches for literature co-occurrences of gene names in order to build a network among the genes. PubGene is useful for quickly realizing and viewing known relationships between genes, but it does not assist in annotating gene lists. To this end one needs an automatic literature searching tool that allows the use of flexible secondary terms with which co-occurrences are counted. Recently such tools have been built. PubMatrix [ 15 ] allows automatic Boolean searches to be performed on Pubmed using any list of primary and secondary terms. This tool carries out the search on the exact terms entered by the user thus in order to apply it to the analysis of microarray data, one has to translate each of the enriched spots to a name suitable for a Medline search. Two other tools – microGENIE [ 16 ] and BEAR GeneInfo [ 17 ] uses a very similar approach but in order to make it more compatible to microarray analysis, they allow the use of gene identifiers as input and provides the needed translation to gene names. During the translation the gene name is expanded to include its synonyms. All of these tools have improved the ability of researchers to quickly use the published literature to annotate lists of genes. However, they suffer from the limitations of any literature data search tool; the ambiguity of gene names and the partial information that can be retrieved by limiting the literature searches to abstracts [ 13 ]. MILANO's aim is to further improve the literature based automatic annotation approach by adding two essential features that address these limitations: Smart synonyms Each gene symbol is expanded to all its aliases, while removing non-informative terms, and the gene product name is added to the query. This addresses the synonym problem, while omitting many of the irrelevant results, thus reducing the polysemy problem (words with multiple meanings). The advantage of our synonym expansion scheme over the existing tools is demonstrated by the comparison presented in table 2 . The GeneRIF database In contrast to the existing tools, MILANO is able to search not only the Medline database, but also the GeneRIF database, which contains short summaries of articles relevant to known genes. The curation of GeneRIF is done by the National Library of Medicine's MeSH indexing staff, who have advanced degrees in the life sciences and use the full text of articles for the indexing process [ 4 ]. Using this database reduces the limitations of relying only on abstracts and aids in finding only relevant information about each gene. Nevertheless, the GeneRIF database suffers from the problems of all manually curated databases; it is partial and contains mistakes and biases introduced by the curation team. However, our ability to identify all of the p53 target genes within a group of p53-affected genes by using the GeneRIF database alone (Figure 3 ) demonstrates that, at least for well annotated genes, using such a database may be the ideal solution for annotating microarrays results. The quality of GeneRIF-based annotation depends on the amount of information entered for each gene in the GeneRIF database, which for many genes is insufficient (data not shown). However, its performance will improve as more information is incorporated into this database and we believe that in the future it will become the preferred annotation tool. Meanwhile, we recommend using MILANO for performing combined searches; searching the GeneRIF database provides quick results and searching the full Medline database allows a broader view that is not limited by the curation procedure. Conclusions We present MILANO , a literature mining tool that can help in annotating microarray results in light of all available literature using experiment-specific terms. In designing MILANO we focused on the accuracy of the search results by providing two novel features: i) Expansion of gene names to include in the literature searches all their informative synonyms, while removing non-informative synonyms; ii) Searching two literature databases – Medline and GeneRIF. While Medline encompasses all the literature and provides the most comprehensive results, it also contains many irrelevant articles. GeneRIF provides a subset of Medline articles that are relevant to known genes and thus avoids most of the irrelevant results often found in Medline searches. The usefulness of MILANO is demonstrated by the automatic analysis of a list of 148 p53 target genes. The use of literature mining dramatically reduced the time and effort required for a task such as identifying the known p53 target genes within this list. A search in GeneRIF immediately discovered the full list of target genes, with no false hits. Availability All software and databases are freely available and may be executed online at our web site: . The author will provide data, scripts and programs used on demand. We encourage users to install the software on their own servers, as we provide no assurance to the privacy or accuracy of the results. Authors' contributions RR designed and programmed the software. IS managed the project and drafted the manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC547913.xml |
524522 | Translation research: from accurate diagnosis to appropriate treatment | This review article focuses on the various aspects of translational research, where research on human subjects can ultimately enhance the diagnosis and treatment of future patients. While we will use specific examples relating to the asbestos related cancer mesothelioma, it should be stressed that the general approach outlined throughout this review is readily applicable to other diseases with an underlying molecular basis. Through the integration of molecular-based technologies, systematic tissue procurement and medical informatics, we now have the ability to identify clinically applicable "genotype"-"phenotype" associations across cohorts of patients that can rapidly be translated into useful diagnostic and treatment strategies. This review will touch on the various steps in the translational pipeline, and highlight some of the most essential elements as well as possible roadblocks that can impact success of the program. Critical issues with regard to Institutional Review Board (IRB) and Health Insurance Portability and Accountability Act (HIPAA) compliance, data standardization, sample procurement, quality control (QC), quality assurance (QA), data analysis, preclinical models and clinical trials are addressed. The various facets of the translational pipeline have been incorporated into a fully integrated computational system, appropriately named Dx2Tx. This system readily allows for the identification of new diagnostic tests, the discovery of biomarkers and drugable targets, and prediction of optimal treatments based upon the underlying molecular basis of the disease. | The Ultimate Goal Our systematic approach to translational medicine has been designed to achieve a vision shared by numerous "ambitious" investigators who are focused on applying bench-side discoveries to the clinical setting (Figure 1 ). We share a vision of future medicine, in which it will be possible to predict the likelihood (risk) of a clinical event during the course of an individual's lifetime, accurately diagnose the event in its earliest manifestation, and treat accordingly based upon the diagnosis. We believe that this will become reality through the combination of medical informatics and the multiplex "omics" technologies (that we broadly characterize as genomics and proteomics throughout this review) now at our disposal. Clearly, there are certain bioethical, political and fiscal roadblocks which need to be considered as we progress towards this goal, not limited to patient privacy, regulatory issues, health care reimbursement and ownership of intellectual property. In this review however, we will focus on the science and logistics relating to implementation of a successful translational research program. Figure 1 The goal of our translational research effort – From Diagnosis to Treatment (Dx2Tx). We have developed a systems approach to track pertinent clinical events within the lifespan of an individual subject. In the future, it may be possible to predict the risk of a clinical event in advance, accurately diagnose the event in its earliest manifestations, and treat based upon the underlying molecular/clinical traits. We believe the integration of medical informatics with cutting edge molecular technologies such as genomics and proteomics will expedite this transition to molecular-based medicine. While a degree of project specificity must be incorporated into the design of any research effort, many of the components of translational research are shared across apparently disparate disease areas (Figure 2 ). Although many facets of a research proposal (i.e. attaining funding and IRB approval) require specification of the particular hypothesis under evaluation, we also make every effort to collect samples (such as blood, urine and tissue) together with as much clinical data (both historical and longitudinal) as possible while maintaining patient confidentiality, privacy and safety at the forefront. In addition to testing the pre-conceived clinical hypothesis (for example, does an event have an underlying molecular basis), this approach readily allows for the discovery of additional genotype-phenotype patterns (hypotheses) that can be subsequently cross-validated in additional subjects and samples (i.e. this event does have an underlying molecular basis). This somewhat blinded yet methodical approach clearly requires a database and supporting analytical software that are intricately linked, such that analytical results can become a new query against the database content (Figure 3 ). While these aspects of the Dx2Tx system will be discussed in more detail throughout this review, it is essential to stress the importance of collecting and archiving standardized clinical data throughout the course of a patient's lifespan. It is difficult to identify statistically significant correlations within non-standardized datasets. Well documented clinical data is equally as important as clinical samples, a fact often overlooked in many research studies reported to date. Figure 2 Systematic overview of translational research. Commencing with human research subjects, we can transition through specimen and data collection, data analysis, preclinical models and ultimately clinical trails. The various facets of this pipeline will be discussed throughout this review. Figure 3 The Dx2Tx integrated solution stores and analyzes clinical (and preclinical), experimental and molecular data from a variety of disparate sources. For more information see Effective translational research requires a multi-disciplinary and team approach. Our initial concept meetings include HIPAA/IRB advisors, research nurses, surgeons (and/or other health care providers such as oncologists), pathologists and members of diagnostic service labs, information technologists (both biological and medical disciplines) and statisticians in addition to the principal investigators. Each member of the translational research team is critical to the overall productivity and success of the pipeline. In acknowledgement of the effectiveness of this approach, we would advise other establishments to involve key personnel across the various disciplines early. We also believe that smaller dedicated teams can often be more efficient and less bureaucratic than oversized consortium. IRB, HIPAA, Scheduling and Consenting The development of research protocols on human subjects that involve specimen and/or data collection must be approved by the investigational review boards (IRB). The IRB governs patient safety and risk for the hospitals and/or universities/institutions. In accordance with federal regulations, one of the requirements states that each human subject must be thoroughly informed about the research to be undertaken on their sample and/or data during a consenting process. The specificity of the consent should be such that it outlines the research protocol, study procedures and risks and benefits and meets the federal, state and local requirements, so that the subject can make an informed decision regarding participation in the study. In retrospect, a well informed patient can help facilitate the procurement of clinical data and samples in an efficient and longitudinal manner. While the typical testing of a hypothesis requires the collection of only a subset of the available clinical data (for example tumor stage, survival time), the emerging field of medical informatics and electronic charts allows for the potential collection of vast quantities of standardized clinical data that potentially harbors invaluable information with regard to additional "traits" or "phenotypes". As such, we currently collect upward of 9000 potential data fields per patient within our specific IRB-approved protocols, all of which can be populated during the normal course of patient care and stored within either electronic or paper records for each subject. In addition, each patient receives a standardized self-reporting questionnaire addressing histories (such as habits, family history, medication history, etc). Any protected fields are restricted through a security portal, which can be considered a suitable clinical gatekeeper (such as research nurse or electronic portal) (Figure 2 ). This security portal allows the flow of clinical data to the laboratory researcher as defined in the IRB-approved protocol, and also links each subject/sample with an anonymous ID. Patient data that is de-identified in this fashion is not subject to HIPAA regulations. It should be noted that some protected health information (such as zip code) may represent valuable information to the research team (for example when addressing potential local or regional causes of disease incidence). Patients are informed and provide written authorization for the inclusion of such potential identifiers during the consenting process; these are then collected and archived in secure locations. For example, we routinely include current home and work 5-digit ZIP codes and date of birth in our IRB-approved protocols. The latter is used to normalize patients based upon age of each clinical event, where date of birth represents time zero in a patient's life span. As such, every recorded event in a patient's life is tagged to a date and time, such that events can be readily interrelated. For more information on HIPPA and sharing of PHI for research purposes, see . In addition to IRB and HIPAA issues, patient recruitment can also be a major hurdle. The teamwork approach is critical here, together with patient/physician outreach and screening efforts as necessary. Our experience is the vast majority of patients are enthusiastic about inclusion in research studies so long as they are informed of the opportunity, particularly since we are careful to follow standard-of-care. For the most part, we are finding a willingness to participate for the hope of betterment for the whole, whether it is potentially helping their own families or others with their disease. At this level, the physician is the central figure and is uniquely positioned to introduce the protocol and support patient participation in the study. For studies where low participation is expected (predominantly due to low incidence), we typically include an IRB-approved public educational effort (for both physicians and potential patients) that maximizes recruitment. Our clinical research nurses work closely with the physician(s) and their office scheduling staff for notification of potential candidates in advance wherever possible. This team approach with the physician advocating the effort allows the greatest accrual particularly when there are several geographic sites for meeting potential protocol volunteers. On site consenting of high risk or suspected individuals within the outpatient clinic and hospital settings where these patients are typically evaluated should to be considered to enhance accrual. For patients with known disease, the best environment for accrual occurs in a multidisciplinary setting in which the key physician works closely with the medical oncologist, radiation oncologist, nurses, physician assistants, residents and research staff. This is particularly important for longitudinal sampling of patients on treatment protocols to minimize the loss of follow-up data. The logistics of a coordinated approach may differ between a university medical school setting and a community based practice due to the competing obligations of a busy private practice. The multidisciplinary approach is optimal in a single setting in which patients can be seen by the various disciplines, thereby, reducing the extent of patient relocation between specialty offices and allowing for centralized specimen and data collection. Nonetheless, in community-based practice, we have introduced standardized mechanisms allowing us to consent patients and collect clinical samples and data from multiple sites. Clinical Data Acquisition There are now abundant examples where clinical researchers are using the revolutionary "omics" technologies to demonstrate a clear association between the molecular make-up ("genotype") of clinical samples and well-defined clinical characteristics ("phenotype"). Near-complete genotypes can now be obtained through multiplex technologies such as sequencing, mutational screening (such as single nucleotide polymorphism (SNP) analysis) and gene/protein expression profiling. The absence of well annotated clinical information (such as medications, response to treatments, histories, environmental exposures, toxicities, etc), however, is evident from most studies reported to date. Rather than taking the reductionist approach at the time of data collection, our group instead chooses to gather all information relating to an individual's medical history as well as follow-up data as it becomes available and as approved by the IRB. It is relatively trivial to reduce the clinical data pool retrospectively during analysis as deemed necessary. Our Dx2Tx system date/time stamps individual events for a given subject/sample, such that multiple events can be temporally associated. The first step is to gather the information in its rawest form from a standardized source. This could be as complicated as a centralized medical informatics database (such as the Oracle-based clinical informatics system used by Spectrum Health Hospitals (Grand Rapids, Michigan) housing rich longitudinal clinical data that can be expressed in Health Level 7 standards for data portability), or it could be as uncomplicated as a locally maintained excel spreadsheet or access database. Regardless of the source, we attempt to collect as much information regarding a subject's history, diagnosis, treatment, and response assessment as possible. As highlighted above, the critical elements are reliability of the source data and standardization. While we do perform statistical analysis on parsed text from open string comments, we attempt to force objective measurements (such as integers, floating numbers, text pull downs, binary data, etc) wherever feasible. In the absence of a centralized clinical database utilizing standardized clinical nomenclature and data, the responsibility falls on the clinical members of the team to interpret clinical data within isolated databases and/or paper charts. Clearly, the navigation from paper to electronic medical records, and the generation of middleware that can link disparate databases, will greatly alleviate this burden that rapidly becomes a major rate-limiting step in the translational pipeline. Coupled with voice recognition and electronic data recording with inbuilt QC check-points, standardized digital data entry should become the norm in the not so distant future. Clinical data permitted under protocol should be, where possible, prospectively accumulated. The accuracy of the data, particularly those parameters such as the pathologic characteristics, clinical staging, dates of intervention, dates of intermediate endpoints (such as disease progression or death), must be unquestionable in order to identify clinical and molecular correlations among diverse datasets. The accuracy of these data will depend to a large extent upon the frequency of follow-up for parameters such as disease progression, and the degree to which the patients' status is investigated longitudinally. Patients on clinical trials, where the procedures and follow-up are specified as part of the protocol, will have the most robust and standardized data since interval radiographic and physical examinations and/or other clinical procedures and methods of data collection must be adhered to in order to avoid protocol violation. Moreover, the approach of the physician(s) who follows the patient and the availability of an existing mechanism that readily allows the retrieval of prospective data (i.e. a database) will greatly influence the accuracy of the data. Within oncology, staging issues are some of the more difficult to resolve, especially if inadequate staging is performed in the operating room (OR) or if one must rely entirely on clinical or radiographic data. For the majority of organ systems, the oncological staging system is specified in the AJCC Staging Handbook, but it is the responsibility of the investigator to know whether the pathologic and intraoperative details necessary for accurate staging were adequately performed. These include but are not limited to accurate size measurement of the primary tumor, careful intraoperative description of abutment or invasion of nearby structures, verification of negative margins, and extent of lymph node involvement. The patients used in the examples below were all part of either Phase II or Phase III mesothelioma protocols, with computerized tomographic surveillance after surgery every 3–4 months until death. As such, the accuracy of the clinical data was optimized. One surgeon performed the operations, obtained aggressive intraoperative staging in all cases, and took the responsibility for collecting samples and data from all participants in the study. This approach led to the development of a consistent standard of care philosophy for the surgical management of mesothelioma. Our Dx2Tx system performs detailed statistical analysis on both clinical and molecular data, thus leaving little room for error when it comes to standardized data collection and entry. Physicians are uniquely positioned to organize the various aspects of data collection. This endeavor begins with the design of a clinical protocol for the procurement of specimens and the collection of data from a patient population. In addition to identifying the experimental group of patients (i.e. those with a condition or trait of interest), it is important to ensure the inclusion of the proper control populations wherever possible. With respect to our mesothelioma protocols, the correct controls for early diagnostic strategies include age-matched individuals exposed to similar doses of asbestos in order to compare with equivalent individuals diagnosed with the established disease. The recruitment of these control subjects ideally begins in parallel with the recruitment of the experimental group. In the absence of the correct control group, the investigator must identify cohorts of individuals within existing cooperative group mechanisms, SPORES, or the Early Detection Research Network (EDRN). In order to circumvent this problem, the Karmanos Cancer Institute has established a National Center for Vermiculite and Asbestos Related Cancers through a cooperative clinical trial with the Center for Environmental Medicine in Southeast Michigan. This center, under informed consent, evaluates individuals who have been exposed to asbestos in the workplace or at home, and depending on history and pulmonary function data, these individuals undergo computerized tomographic scanning to establish a baseline radiographic evaluation of asbestos exposure and risk. These individuals also give written permission for periodic sample collection including blood and urine for studies of marker assessment for asbestos related malignancies. A major focus of our research is to identify the molecular causes of differential therapeutic response across patient populations (pharmacogenetics). It is generally appreciated that patients and their disease show significant variations in response to a given treatment [ 1 ]. It is therefore critical to develop standardized approaches for routinely monitoring adverse responses (for example using common toxicity criteria) as well as disease response (for example using the Response Evaluation Criteria in Solid Tumors, or RECIST criteria). While various reporting schema are in place under specific clinical protocols, we are working on a more systematic approach; the collection of raw clinical data that can be compiled to assess clinical response and toxicity. Accordingly, if the raw data is collected, the response of all patients undergoing defined treatments can be assessed in a longitudinal fashion. The compiled response and toxicity data can then be analyzed retrospectively with other clinical and molecular attributes as outlined below. Specimen Collection and Archiving We have introduced standard procurement procedures for various physiological samples including tissue, urine and blood, all of which are now routinely collected under our IRB-approved protocols. These have been optimized for both ease and reliability of clinical collection as well as maintaining the integrity of the sample for subsequent histopathological and molecular analysis. DNA from peripheral blood mononuclear lymphocytes (PBML) obtained from whole blood can be screened by SNP analysis, to identify possible genetic markers of a clinical event [ 2 - 4 ]. Once validated, these SNP markers could serve as a genetic test to predict the risk of the event in prospective subjects. Plasma or serum (we prefer non-clotted plasma) and urine can be screened for proteomic markers of a clinical incident, and used in future screening for diagnostic purposes [ 5 ]. Per the approved protocol, tissue can be subjected to a variety of DNA/RNA/Protein technologies, and important diagnostic and therapeutic insight gained at the molecular level [ 6 ]. It should be noted, that wherever possible, adjacent uninvolved "normal" tissue free of disease should also be collected for comparison with the corresponding diseased tissue. The example in this review will focus on some published works [ 7 ] identifying diagnostic and prognostic biomarkers in the tumor tissue of patients with mesothelioma utilizing Affymetrix GeneChips for assessing gene expression within the tissue. In this case, we collected normal mesothelium of the peritoneum or the pleura in conjunction with mesothelioma tumor tissue. The precise flow of the clinical sample from the patient to the researcher will depend largely on the disease in question and pre-existing departmental procedures. For example, in a study of pancreatic cancer, urine is provided by the patient prior to the procedure, since we have found that intra-procedural catheterization can result in blood contamination that effects subsequent proteomic analysis. Blood is drawn from an IV access for collection of plasma and PBML, while surgically resected tissue is snap frozen on site (see below). However, a multiple myeloma research protocol requires the collection of bone marrow aspirates from both inpatient and outpatient clinics, which are placed into tissue culture media and rapidly transported on ice to the Flow Cytometry Molecular Diagnostic Laboratory for immunophenotyping and cytometric sorting of plasma cells. The sorted fraction is then archived in freezing media for subsequent DNA, RNA and/or protein analysis. Irrespective of procedure, it is important to maintain a log book that tracks the flow of a sample from the subject to freeze. Procurement time is particularly important for RNA and some protein analysis, since biomolecular degradation can be a significant factor. Prior to collection, our procurement team (lab personnel and research nurses) pre-label collection tubes and containers with anonymous identifier tags that are freezer safe. Several kits that include equipment for blood draws, urine collection and tissue procurement are preassembled and made available to the research staff to ensure rapid response for obtaining specimens. We have an on-call list of several members for the research team who can reach the pick-up point within 30 minutes of receiving a collection call. Through a coordinated multidisciplinary approach, consenting and scheduling is typically obtained well in advance in either the outpatient or inpatient setting. At the Karmanos Institute, we have found that the easiest way to insure proper specimen harvest from patients with solid tumors is for the surgeon to divide the specimen directly in the OR and supervise its collection and distribution to the laboratory for archiving. However at the Van Andel Research Institute, we require oversight from the Hospital Department of Pathology to release surgically resected samples within the OR to the research team. The critical issue is to ensure suitable material is harvested in the correct fashion to allow accurate histopathological diagnosis. In no way should the research effort impede this quality-of-care. As a safety margin, we hold all frozen tissue for a period of time until an official diagnosis has been reported. In addition, wherever possible we embed samples in OCT freezing media, and generate hematoxylin/eosin (H/E)-stained histopathology slides for each research sample. This not only provides high quality histopathology slides for possible diagnostic back-up, but also allows us to correlate our molecular findings with the histology of the same working sample. In addition to entering pertinent data regarding the official diagnostic pathology report, we have also developed standardized report templates within Dx2Tx that accompany each research sample, in which the pathological image is associated with the corresponding data addressing critical variables such as relative amount of each cell type, stage, morphology etc. As with all data entered into Dx2Tx, these metadata are directly amenable to subsequent statistical analysis (see below). At the time of resection, solid tissue samples are immediately frozen in either liquid nitrogen or, for sites with no supply, a dry ice-chilled isopentane bath. Samples are transported in aluminum foil, and stored at -80°C until further processing. It should also be pointed out that many of our protocols involve the collection of tissue biopsies in addition to surgical resections. Since it is physically difficult to split a biopsy (although this is done for example with bone marrow core biopsies), we typically acquire additional samples beyond what are required for accurate pathological diagnosis. If these additional samples are required by the research team, it is important to disclose the supplementary procedures to the patient in the informed consent. While obviously smaller in size compared to surgical resections, we procure these specimens as fresh tissue as described above. It is important to note, some of our molecular technologies allow for the use of as few as 500 cells for complete multiplex analysis. Since biopsy material will doubtless be the predominant source of tissue for future molecular-based diagnostics, it is important to develop protocols to address smaller sample size and possible mis-sampling issues. In addition, as a result of not restricting samples to excess surgical specimens, biopsies can rapidly increase sample accrual from larger cohorts of patients. Urine and blood samples are transported on wet ice to the research labs, while frozen tissue is transferred on dry ice. Upon arrival in the laboratory, urine is typically divided into suitably sized aliquots and frozen at -80°C. Blood is fractionated into plasma or serum (again aliquoted and frozen), while PBML are isolated by centrifugation and Ficoll/Hipaque density gradient centrifugation and cryogenically frozen. Fresh frozen tissue specimens are blocked in OCT if possible and stored at -80°C until further use. All specimens are associated with the donating subject within Dx2Tx, and archived in a simple grid system within dedicated and alarmed -80°C freezers. The location and use of each aliquot is tracked within Dx2Tx to readily allow for subsequent retrieval. Sample Procurement – The "Omics" The tissue collection protocols outlined above should be evaluated for compatibility for the molecular analysis to be performed. We typically perform SNP analysis on DNA isolated from the PBML fraction of whole blood, gene expression (mRNA) analysis on tissue, and proteomic analysis of blood plasma/serum, urine and tissue. The detailed discussion of each of these molecular protocols is beyond the scope of this review (see [ 6 ] for a good starting point of reference). There are important differences between the various platforms available that can result in some disparities in the results generated [ 8 - 10 ]. As such, it is essential to document and ideally database protocols, deviations from protocols (version control), and in general, all experimental variables that could potentially confound the results (for example see Minimum Information About a Microarray Experiment ). We enter these variables into Dx2Tx in association with the corresponding experimental procedure such that they can be analyzed together with the derived molecular and clinical data. In this fashion, important QC and QA parameters can be interrelated with molecular and clinical data. Experiments that do not meet certain QC/QA criteria can be excluded from the analysis if the data suggest that the variable(s) is a major confounder. In this fashion, Dx2Tx serves as an electronic notebook, the entries into which can be routinely recorded and statistically analyzed. This feature will be particularly important as potential applications advance towards clinical utility, for example through ensuring compliance with the FDA and accrediting agencies such as JACHO, CMS and CAP, and while maintaining the necessary state, federal, and insurance applications required to provide clinical diagnostic services. Data Analysis For so many investigators, analysis represents the "black box" of the translational pipeline. It is reasonable to state that statistical analysis of the data is one of the most important steps in the translational process. In addition to determining the sample size and inclusion/exclusion criteria necessary to test a hypothesis, it is essential to utilize the correct tests of statistical significance during both discovery and application phases of the translational pipeline. This review will not detail all possible permutations of biostatistics, but we will touch briefly on certain statistical concepts. The average gene expression profiling experiment can generate in excess of 30,000 individual data points per clinical sample, while current high-throughput SNP-chips can generate >100,000 discrete attributes per analysis. Similarly, proteomic experiments generate vast amounts of coupled data that includes both fractionation (such as 2-D gels or 2-D liquid chromatography) and detection (such as mass spectrometry) elements. Coupled with extensive clinical content, this collectively presents significant challenges for data storage and retrieval, as well as statistical analysis. The most frequent frustration of new users of the "omics" technologies remains "What are the data telling me?" Before proceeding with specific examples, it is important to grasp the concepts of normalization and data massaging or "filtering" [ 11 ]. To compare data across multiple samples, data is typically normalized or scaled such that results from one experiment are directly comparable to those of another. There are a number of methods to normalization across experiments including the use of a common reference sample or internal controls such as housekeeper genes/proteins that do not alter across experiments. Various forms of mathematical normalization can then be used (for example mean centering) to scale the results across experiments. As discussed above, through the tracking of protocols and minimizing experimental variation, one can reduce the degree of scaling required to directly compare data across different experiments. The normalization routines, as with pre-filtering of data, should be performed during an analysis session and not prior to data entry, since the different normalization and filtering routines can significantly effect the analytical results. As such, it is important to track the normalization and filtering criteria employed during analysis, such that results from sessions using different methods can be compared and contrasted. One problem in dealing with vast amounts of clinical and/or multiplexed molecular data from a relatively small sample population is that many of the observed correlations could (not necessarily do) occur by random chance. One method investigators use to minimize the probability of this so called "false discovery" is to reduce the complexity of the data through successive rounds of filtering. Depending on the application, we do use some filtering under certain circumstances. However, as with normalization, this is done during data analysis and not prior to database entry. In this fashion, we can filter and analyze the data "on the fly", allowing the user to evaluate the effect of filtering the raw data. For example, if we were looking for a specific set of genes that are found highly expressed in mesothelioma relative to other tumor types, we may eliminate genes in mesothelioma samples below a certain expression (intensity) threshold, and those above a certain level in other tumor types. This would maximize the likelihood of identifying sensitive and specific mesothelioma genes, and minimize the gene pool and hence probability of false discovery. We also use fold-change filters depending on the application. For example, we may only be interested in diagnostic biomarkers that display at least a 4-fold increase in mesothelioma tumor tissue relative to other tumor types and normal mesothelium. In addition, genes with certain characteristics (such as those that encode only secreted or transmembrane proteins [ 12 ] can be filtered (included or excluded) within Dx2Tx. Thus, while it is important to note that these filters are not tests of statistical significance, they nonetheless can be used in the context of biological logic, to maximize the likelihood of identifying clinically valuable data within multiplexed data. The logic flow depicted in Figure 4 shows the effect of successive rounds of filtering the data from mesothelioma samples with the intent of identifying possible biomarkers that may be detected in blood or pleural effusions. This logic-based informatics approach to identifying possible disease biomarkers in physiological fluids can lead to candidate proteins that can be specifically detected in the corresponding blood/urine samples by other methods. We have shown the utility of this predictive approach in parallel with proteomic analysis of plasma in some experimental models of cancer. While this approach is yet to be validated in the human disease, we believe that this represents an excellent supplement to existing biomarker discovery programs that is readily implemented through simple data filters. Figure 4 Sequential filtering of Affymetrix gene expression data to identify potential plasma biomarkers of mesothelioma. The key aspects of a disease biomarker include sensitivity and specificity. These can be partially addressed through logic-based filters within Dx2Tx. There are an abundant number of methods that can be employed during analysis of multiplex data to determine statistical significance. In the absence of becoming a quasi-expert in statistics, we highly recommend that the assistance of any number of statisticians is sort. The key is to understand why certain statistical tests are used under a variety of circumstances, and what the limitations of each test may be. We typically use similarity-based tests (such as hierarchical clustering, principle component analysis, multidimensional scaling, etc) to identify relationships between variables (clinical, experimental and molecular) within large datasets [ 13 , 14 ]. These metrics identify the degree of similarity (or difference) between various attributes, and are extremely powerful when attempting to discover inter-sample relationships based upon their molecular and/or clinical features. For example, unsupervised hierarchical clustering can be used to cluster the X attributes (such as mesothelioma samples) based upon the values of the Y-attributes (such as relative gene expression) as shown in Figure 5 . In addition, the Y attributes can be clustered, based upon their similarity across the X attributes, providing a 2-dimensional clustergram displaying overall relationships (Figure 6 ). The co-clustering of samples is essentially a raw form of a molecular diagnostic application since samples with similar genotypes cluster based upon biological similarity (phenotypes). The co-clustering of genes and/or clinical data is also a potentially powerful application. For example, genes/proteins with similar functionality are often co-regulated at the level of their expression, and hence typically "co-cluster" on a gene expression clustergram. This concept becomes particularly powerful when attempting to predict the function of unknown genes based upon their overall correlation with a gene of known functionality [ 15 - 17 ]. In the case of clinical data, inter-relating clinical and/or environmental events can also be performed. Hence, features that correlate (such as increased stage of disease and poor outcome) are typically adjacent on the clustergram (Figure 7 ). Coupled with the extensive collection of standardized clinical data highlighted throughout this review, this feature alone may have significant impact in the mining of clinical data in disciplines such as epidemiology. Figure 5 A two-color clustergram generated after hierarchical clustering of gene expression data (Affymetrix U95A) across 21 tumor samples collected from patients with mesothelioma. Clustering has been performed in only the X axis, such that samples are grouped based upon similarity in overall gene expression (the identified sample sub-groups are color coded). The gene expression data has been mean centered, such that degrees of red and green indicate relatively high and low expression of the corresponding gene respectively, while black represents the mean value across samples. In this fashion, the relative expression of many genes can be readily visualized across several samples simultaneously, and the relationships between samples observed. Figure 6 Hierarchical clustering of Affymetrix gene expression data as described in the legend to figure 5, with the exception that genes are also clustered based upon similarity in expression across the samples. In this fashion, correlations between samples and genes can be simultaneously observed. Figure 7 Clustering of mesothelioma tumor samples (X attribute) by clinical data (Y attribute) reveals possible epidemiological relationships between the various clinical features. Some well known relationships (such as stage of disease, lymph node status, death), as well as some less established patterns (such as a correlation between high platelet count and recurrence) are readily observed in this mode of operation. The degree of the clinical event is represented on a mean centered scale, such that red and green indicate relatively high and low extents respectively. A subset of samples has been selected based upon the extent of a single trait, in this case prolonged survival time (blue). In addition to correlating clinical and molecular data individually, these data types can be merged and viewed simultaneously within the same clustergram. In this fashion, possible associations between clinical, experimental and molecular features can be readily identified. For example, as shown in Figure 8 , there appears to be an association between T-stage and a number of genes known to be involved regulation of the cell-cycle. Classification of gene/protein function (so called "annotation") can provide important information with regards to the underlying molecular cause(s) of a clinical event. In addition to utilizing the publicly available annotations (for example see Gene Ontology ), we also map results to molecular pathways using detailed pathway mapping software now available. Taking this approach, the coordinated expression of genes/proteins can be seen to map to specific molecular networks, therefore providing important information as to which pathways maybe activated or in-activated in association with a clinical event. For example, when all the genes differentially expressed in recurrent mesotheliomas relative to non-recurring tumors at a defined statistical significance (p < 0.001) are mapped using the MetaCore™ software (MetaCore™ ), a clear signaling pathway associated with cell proliferation is identified that appears to be hyper-activated in aggressive mesothelioma tumors (Figure 9 ). As discussed below, in addition to providing clear diagnostic value, this information is particularly useful in the design of treatment strategies that may target key points within the identified molecular network. Figure 8 Clustering of samples based upon integrated clinical, experimental and molecular attributes. In this sense, molecular-clinical (genotype-phenotype) associations can be readily observed. As described in the legend to Figure 7, the extent of the clinical/molecular attribute is represented on the same normalized scale, such that red and green represent relatively high and low values respectively. Figure 9 Mapping molecular correlates of aggressive mesothelioma to highly curated molecular pathways can identify the underlying molecular mechanisms of the disease. This information could be used for diagnosis, as well identification of the key steps that may represent intervention points in the treatment of the disease. The red arrow indicates a predicted therapeutic target (EGFR). Pathway mapping was generated using MetaCore™ (GeneGo, Inc., St. Joseph, MI). For more information on this pathway mapping tool, see . Clinical Diagnostics With respect to identifying patterns (hypotheses) within the complex clinical and molecular datasets that could be translated into clinical diagnostic applications, we typically begin with unsupervised clustering techniques such as hierarchical clustering as shown above in Figures 5 , 6 , 7 . In this fashion, sample similarity with respect to clinical, experimental and/or molecular attributes can be assessed. Dx2Tx extends these analyses to identify clinical and/or experimental variables that statistically correlate with defined sample sub-groups. During this step of hypothesis generation, Dx2Tx runs back into the database housing all of the standardized clinical and experimental data and identifies correlates of the selected sub-groups. This is a highly powerful utility when operating in unsupervised mode, and requires an intricate link between data analysis and database content. Unsupervised clustering may for example identify the degree of molecular similarity across a cohort of patient samples, which could identify several clearly delineated groups at the genotype level. Running in hypothesis generation mode, Dx2Tx then identifies statistically significant correlates of these groups, and assigns clinical/experimental features to each. When Hypothesis Generator was executed on the 2 sample subgroups highlighted in Figure 5 , the clinical features time to recurrence (p = 0.003), T-stage (p = 0.004), survival time (p = 0.0002) and platelet count (p = 0.005) were returned as significant correlates of these sub-groups. Thus, while these samples may have been initially collected in the context of a different user-defined hypothesis, through the collection of standardized clinical data in addition to the generation of multiplexed molecular data, Dx2Tx was able to identify statistically significant patterns (hypotheses) within the data in an unbiased fashion. The user can of course decide which hypothesis to pursue. In this example, our ability to potentially utilize gene expression profiling to predict survival time of mesothelioma patients following surgery based upon gene expression within the tumor would have obvious prognostic value. Therefore, the next step would be to test this hypothesis and determine the accuracy of a possible diagnostic test. Once a hypothesis has been generated, Dx2Tx identifies samples against which the hypothesis can be tested. Certain inclusion and exclusion eligibility criteria can be considered and used to filter the content of the database to identify subjects/samples/experiments with certain characteristics. Dx2Tx also allows samples to be selected based upon the extent of any attribute(s) (Figure 7 ). For example, the investigator may be primarily interested in only a subset of the sample population that displayed the greatest and least extensive toxicity to a given drug. This is assisted through the selection of the trait of interest and setting the extent of the trait (i.e. by defining standard deviations from the population mean). This feature may be particularly important in retrospective analysis of large clinical trial cohorts, since the outliers for a given trait can be identified prior to sample procurement. Once the sample population is selected, we test the hypothesis across the series of selected samples in a 2-step process. A subset of the samples (typically defined as a training set) are selected (either logically or at random) from each subgroup (for example disease versus control) to develop a discrimination algorithm that identifies statistical correlates of the feature in question. It is worthwhile to note that Dx2Tx identifies clinical, experimental and molecular correlates of the selected feature(s), thereby integrating both clinical and molecular data into the potential diagnostic algorithm. The user can exclude any attribute from the input to the training algorithm. In a second cross-validation test, the trained algorithm is applied to the remainder of the samples (in retrospective mode of operation, with known outcome), to determine if the test could have accurately predicted the nature of the remaining samples. The outcome of the test is plotted using a receiver operator characteristic (ROC) curve to determine the accuracy of the test. The ROC curve is a way to visualize and quantify the effectiveness of a procedure by graphing the true positive rate (Y axis) against the false positive rate (X-axis). The area under the curve (AUC) provides an approximation of the accuracy of the test. A procedure with no effectiveness (AUC = 50%) would show a random 1:1 line, indicating that for every true positive, the procedure also generated a false positive. Generally, an AUC of 90–100% is considered excellent, while an AUC of 80–90% is good. The ideal diagnostic test would of course identify all true positives before encountering a false positive (AUC = 100%). In this working example, the hypothesis generated from analysis of unsupervised clustering of gene expression data from mesothelioma tumors is that survival time of patients can be predicted based upon the underlying genomic signatures of the tumor. Thus, patients with the shortest and longest survival time following surgery were placed into two groups. Each group was then randomly divided into 2 additional groups, the training set and the test set. The discriminating clinical and molecular features are first identified using a standard t-statistic for numerical data and chi squared for binary (including text) data. This test statistic is then used in a weighted voting metric [ 18 ]. Data are first converted to a respective z score in order to normalize data of different types to a similar scale. A more refined statistical package, which will more rigorously integrate the binary and non-binary data, is currently in the process of being implemented into the Dx2Tx solution. In this fashion, the experimental, molecular and clinical attributes that statistically correlate with survival time are first identified. In this example, no experimental variables (i.e. those which may denote a variation in experimental protocol or quality) were identified that correlated with patient survival time. The clinical parameters platelet count and T-stage were identified as clinical correlates of survival time and therefore included into the training algorithm. In addition, 157 genes were identified, the expression of which correlated with survival time (p < 0.05). Each attribute (platelet count, stage, and the 157 individual genes) was then weighted based upon the calculated t-statistic within the training group. A discrimination score (the sum of the t-statistic multiplied by the normalized z-score for each attribute) was then calculated for each sample within the training groups and a threshold decision point (the value at which a sample is classified as neither group 1 or 2) is set halfway between the means of the two test groups. Alternatively, a user can set the threshold in order to maximize either sensitivity or specificity of the assay, or set it to a value which would demarcate an acceptable test failure rate. In this fashion, the end-user can set the decision point of the classification algorithm on the side of false positives or false negatives based upon the clinical consequence of the test result. For example, if a positive test results in administration of a poorly tolerated treatment, the physician would typically error on the side of false negatives. At this time, a discrimination score is calculated for the remaining test samples, compared to the threshold decision point, and assigned a classification. The predicted classification is then compared to the actual outcome. While complicated, Dx2Tx performs this cross-validation metric in a matter of seconds. In this working example, the ROC plot generated from the prediction of the prognosis of patients with mesothelioma suggests that this particular diagnostic test is approximately 90% accurate at determining the 6 month survival of patients following surgery as determined by the area under the ROC curve (Figure 10 ). Once validated, the classification algorithm is stored within Dx2Tx, such that it can be applied to any future sample. Thus, through the capture of standardized clinical, experimental and molecular data, hypotheses can be rapidly generated and tested, and further developed into potentially useful diagnostic applications. At this point, the focus may shift from retrospective analysis to prospective studies. Figure 10 A Receiver Operating Characteristic (ROC) curve showing the performance of an integrated clinical and molecular diagnostic test for predicting prognosis of patients with mesothelioma. The Area under the curve (indicative of the tests accuracy) is approximately 90%. From Diagnosis to Treatment In addition to the identification of potential diagnostic applications, we have a major focus on identifying new treatment targets, and/or improving therapeutic strategies involving existing treatments. Patient variation, with respect to treatment response (efficacy and toxicity), is a well documented phenomenon [ 1 ]. Through the capturing of clinical data and pertinent samples across a large patient population that exhibits variable treatment response, retrospective statistical analysis of the integrated clinical, experimental and molecular data could reveal the underlying causes of this variation. For example, DNA polymorphisms in some isoforms of the cytochrome p450 enzymes have been associated with the variation in the rates of metabolism of many pharmaceutical drugs across a sample population [ 19 ]. As a result, a specific test now exists that could be used to better determine the optimal dose of some pharmaceutical treatments (Roche Release of AmpliChip ). These so called "companion diagnostics", which could accompany therapeutic agents and assist in treatment decisions, are also being developed for specific agents that display varying degrees of efficacy and toxicity across sample populations. With the accurate capture of longitudinal clinical data including toxicity and response assessment, associations between clinical response and molecular features of either the patient and/or the disease tissue should be readily identifiable within complex retrospective datasets. For example, we have identified a genomic signature within plasma cells isolated from multiple myeloma patients that correlates with tumor response to the drug melphalan. This genomic signature is currently being applied to additional patient samples using cross-validation statistics as outlined above, to determine the accuracy of this possible companion diagnostic application. As with most single agent treatment regimens, drug resistance in the area of oncology represents a significant problem in the treatment of the disease. Therefore, pre-treatment tests could conceivably identify the patients who would benefit the most and least from treatment. Our research also includes the discovery of possible early surrogate markers of therapeutic index. This ideally requires the collection of clinical specimens and data both pre and post treatment. In conjunction with the treatment of cell lines in culture and molecular analysis of livers and kidneys from treated mice, early biomarkers of efficacy and toxicity have been identified in association with several treatment regimens. If these biomarkers of response can be validated in retrospective patients, they could be written into future clinical studies to provide an early indication of therapeutic effect. These biomarkers may ultimately be used as surrogate markers to determine discontinuation or modification of protocols to maximize therapeutic index in prospective trials. There are currently a number of drugs in development, in clinical trials or that have recently received FDA approval that target specific molecular aberrations [ 20 - 23 ]. Unlike cytotoxic chemotherapies, molecularly targeted therapeutics often display a high degree of specificity against the selected target. Based upon the specificity of these drugs to defined proteins, it is envisioned that molecular-based diagnostics will naturally accompany these agents to identify the patient sub-population who will benefit from treatment. Because of the inherent genomic instability of cancer, combinations of these molecularly targeted drugs will almost certainly be required to ultimately treat the disease. Indeed, mathematical models of adaptive microevolution of the cancer cell suggest that a multi-modality treatment strategy that targets at least five individual molecular targets simultaneously will be required to minimize the chance of a single cell within the tumor acquiring resistance to each agent [ 24 ]. As part of our research effort, we are attempting to identify the optimal multi-modality targeting strategies to treat specific tumor types based upon their molecular makeup. For these reasons, we have incorporated a drug-target database that is regularly updated to include molecularly targeted agents as they are publicly disclosed. During analysis, we can filter the datasets to only include genes/proteins against which drugs have already been developed. For example, we can substitute the drug-target list as a filter in place of the known secreted or plasma membrane proteins discussed in relation to Figure 4 . When performing this function on the mesothelioma dataset, epidermal growth factor receptor (EGFR) inhibitors in combination with eniluracil and topoisomerase II inhibitors are identified as a possible combination treatment for mesothelioma based upon relatively high expression of their molecular targets EGFR, Dihydropyrimidine dehydrogenase and topoisomerase II respectively. Such hypotheses obviously require testing in a relevant preclinical model of the disease (see below). Nonetheless, because the corresponding drugs have already been developed, this is a readily testable hypothesis assuming the investigator can gain access to the therapeutic agent. As discussed below, however, we do not believe that expression levels of the molecular target alone are necessarily sufficient to predict drug efficacy. The FDA approval process of the EGFR inhibitor Iressa has received a great deal of attention [ 25 ]. Recently, it was shown that lung carcinomas from a subset of patients that possess activating mutations in the EGFR are most responsive to Iressa [ 26 , 27 ]. This raises an important concept we have been pursuing for some time; namely that diagnostic tests need to address activity of the target(s) rather than merely expression levels. Since an active molecule can often shut down its own expression, while conversely, hypo-active targets may consequently be over-expressed, this biological phenomenon of negative feedback often results in inverse expression-activity relationships (CPW, unpublished observations). Hence, in combination with gene silencing technologies (such as RNA interference) that mimic a selective drugs action, we are identifying the down-stream genomic and proteomic consequences of target gene disruption for the purpose of identifying biomarkers that could be used to assess target activity. We believe that rather than using expression levels of the target molecule alone as a rudimentary diagnostic test, this more sophisticated approach may yield greater success in attempting to predict the optimal treatment strategies based upon a molecular profile. In addition to pursuing existing drug targets, we are also attempting to identify novel targets that may warrant future drug discovery efforts. We typically attempt to validate only candidate genes/proteins that have or are predicted to have "drugable" characteristics. To determine whether a potential target is drugable, we are currently using some relatively simple criteria based upon the classes of drug targets that have been actively pursued by pharmaceutical/biotech companies to date. For example, we have annotated the publicly available drug targets described above using gene ontology and literature mining tools within Dx2Tx [ 9 ], and identified several recurring features of these targets (such as kinases, phosphatases, G-protein coupled receptors, etc). Any gene/protein identified with the same annotation is "drugable" by simple association. We also identify genes/proteins that co-cluster with genes/proteins with drugable features, since as described above co-expressed genes/proteins often share similar functionality. Future developments will include sequence, domain and structural-based predictions of drugable characteristics, but it remains to be seen if this selective approach will lead to accelerated clinical application in the future. Nonetheless, using the combined content and analytical power of Dx2Tx we can sequentially filter data in an attempt to identify specific targets for the disease in question and condense our target candidates further to identify those with the greatest potential for drug development. Preclinical Models of Disease The majority of the preceding discussion has focused on the identification of molecular correlates of disease, and identifying those that may represent diagnostic biomarkers and/or treatment targets for intervention. A large proportion of our translational research effort is dedicated to functional validation, where through various means, the expression and/or activity of the target are modified and the functional consequences addressed. While determining the function of a diagnostic biomarker is not necessary, the functional consequence of target gene/protein disruption is essential when establishing the true therapeutic value of a potential target. An ideal therapeutic target would be one that is causative of the disease, and inhibitors against which thus cause disease regression. The following discussion will focus on a high-throughput means by which we assess the functional significance of target gene/protein disruption in murine models of various human malignancies. There are a variety of approaches one can take to interfere with gene/protein expression and/or function with the purpose of demonstrating a definitive role in a biological process [ 28 ]. These include the use of pharmacological agents, antibodies and/or interfering mutants. However, these approaches require reagent access and/or some in-depth knowledge about the gene/protein in question. These approaches are also relatively low throughput and expensive, and can result in a significant bottleneck effect as potential targets are identified during the translational research effort. Antisense technologies represent an excellent approach to determine gene function, and are readily integrated into gene/protein discovery programs due to the wealth of gene/protein sequence information now available [ 29 , 30 ]. In this regard, the field of RNA interference has emerged as a highly specific and relatively simple way to disrupt genes [ 31 ]. RNA interference (RNAi) is a conserved phenomenon, whereby long double stranded RNA (dsRNA) is processed into small 21–23 nucleotide dsRNA fragments, termed short interfering RNA (siRNA) [ 32 ]. These fragments then target and degrade highly homologous RNA gene transcripts, thereby inhibiting gene expression in a sequence-specific manner. RNAi is thought to function to maintain genomic stability, regulate cellular gene expression, and defend cells against viral infection [ 33 , 34 ]. We have developed an avian retroviral vector that can deliver siRNA to cells expressing the viral receptor (TVA)
[35]. TVA expression can be directed to specific cells in vitro through exogenous transfection/infection, or in vivo through transgenic technology [ 36 ]. This system was developed to allow us to target the delivery of gene-specific siRNA to tumor and/or endothelial cells in vivo, such that the effect that target gene disruption has on tumor growth, metastasis and angiogenesis can be directly assessed. As discussed above, we typically target genes/proteins that we predict to have "drugable" characteristics, and in a sense this retroviral-siRNA approach mimics the optimal molecularly-targeted drug due to its targeted delivery and exceptional degree of gene specificity. In addition, these avian retroviruses do not replicate within mammalian cells and as such cells can be infected multiple times with vectors targeting multiple genes. This system therefore allows us to investigate the functional consequences of combinational targeting strategies. Because of the particular cloning strategies we have incorporated, it takes only six weeks from the discovery of a potential target gene to the point of assessing functional consequences of gene disruption in a relevant preclinical model of the disease. Because of the efficiency of the system, we are able to assess multiple targets simultaneously. For example, we are currently evaluating 19 targets in various combinations that have been predicted to display optimal efficacy in murine models of mesothelioma, pancreatic cancer, colorectal cancer and multiple myeloma. Thus, by introducing a systematic approach to target validation, we have limited the bottleneck between target discovery and functional validation. Improved murine models that depict specific features of the human disease in question are essential to validate the in vivo significance of experimental findings in a relevant preclinical setting. We particularly focus on the development of orthotopic xenograft models, in which human tumor cells are implanted into their corresponding site of origin within an immune compromised mouse. In this fashion, human tumors form in the tissue site of origin that more closely resemble the human counterpart with respect to biological behavior. These orthotopic models better recapitulate the various stages of tumor progression and more accurately reflect responsiveness to various therapeutic intervention strategies [37]. Coupled with in vitro and in vivo delivery of siRNA against multiple target genes, these preclinical models readily allow us to evaluate the requirement of proposed target genes in tumor progression. Through these studies, we are beginning to identify the optimal combination targeting strategies that may lead to the eventual treatment of aggressive cancers. Moving from Retrospective Analysis to Prospective Clinical Trials In the vast majority of cases in which molecular data is being correlated with clinical events, a hypothesis-driven inquiry is tested against archived clinically documented specimens. The majority of these trials are asking either a classification question (i.e. what profile sets this tumor apart from other tumors), a prognostication pattern (what group of genomic/proteomic patterns will predict time to progression or time to death), early detection (how is this tumor different from its "cell of origin" at the earliest time point recognizable such that a genomic/proteomic pattern could predict the development of malignancy in a high risk population), or response to therapy (is there a de novo set of genomic/proteomic parameters which predict either response or resistance to a given therapy). For classification phenomena, the investigator must currently rely on pathologic differences to stratify tumors into clusters which, upon genomic/proteomic analysis, have consistent and congruent distinguishing features. This is probably the easiest of analyses and relies mainly on established architectural and morphologic differences between tumors instead of clinical behavior and endpoints. Nevertheless, as with all discovery test-sets, validation prospectively must be performed in order to test the accuracy of the classification algorithms. This requires prospective application of the algorithm to blinded samples, and comparing the predicted outcome with the actual observed pathology. For the early detection, prognostication, and prediction of therapeutic response, one must link potential diagnostic tests to newly developing trials to allow validation of the established hypothesis. Before undertaking such an effort, however, the investigator must realize that the data were typically derived from a set of patients within a particular treatment regimen. For example, the patients presented in this article were characterized as having specimen procurement prior to definitive cytoreductive surgery followed by adjuvant therapy, and therefore a prospective trial which conforms to this treatment regimen should ideally be designed to avoid confounding variables. For the validation of diagnostics that predict time to recurrence or survival following treatment, a phase II trial of surgery and/or postoperative adjuvant therapy could be designed in which samples are harvested at the time of surgery. The diagnostic test would then be performed on the clinical specimen to obtain a prediction of patient outcome, and the patients' clinical course followed to see if there is prospective validation of the test. These analyses, however, must be performed in patient groups not weighted towards either high or low risk patients, and indeed, some stratification of clinical parameters should be specified at the outset in order to make reasonable comparisons. This is especially true if the molecular data is to be validated as part of a Phase III randomized trial comparing two treatment regimens after surgery. As the content of Dx2Tx expands to include clinical trial data, analysis can be performed on only those patients that conform to specified clinical parameters. Therefore, with the complete and standardized collection of clinical data from a large population of patients receiving various treatment regimens both on and off of protocol, established hypotheses can be further tested on additional retrospective subjects; in a sense a virtual trial that begins to more closely resemble the carefully controlled prospective clinical trial. The ideal clinical trials addressing the validity of molecular data to predict a clinical parameter are those involving patients with no prior diagnosis or treatments, who subsequently receive a single therapeutic regimen. Preferably, both pre- and post-treatment specimens are obtained from these patients. Such a pure trial could, for example, test whether a predefined genomic/proteomic pattern could predict a poor outcome despite favorable clinical parameters. Trials of patients receiving initial chemotherapy could also help to define the fidelity of the omics technologies for the prediction of chemosensitivity. These hypotheses may have been derived from a retrospective set of patients receiving the same chemotherapy or targeted therapy for which outcomes were known, or from the in vitro discovery of genomic/proteomic patterns defined in cell lines treated with the corresponding agent(s). Early detection clinical trials using genomic and/or proteomic technologies will typically take the longest to validate. After the discovery of proteins as candidate markers from analyses of serum, urine, or other fluids in patients with early stage malignancies compared to the appropriate high risk group, cohorts of high risk individuals (e.g. asbestos exposed or tobacco smokers) must be identified who are willing to have collections of the appropriate specimens longitudinally at given intervals. Moreover, in initial trials of these markers, the investigator must define at what time to disclose the value of the measured biomarker. One possibility would be to allow the prospective clinical trial to reach a certain endpoint (i.e. a certain number of cancers occur in the cohort which are detected by the standard of care radiographic or physical examination) and then reveal the results of the marker to see whether there is predictive value. Another way would be to define the value of the marker at follow-up intervals during the course of the trial and, in the face of radiographic or physical findings, initiate an invasive workup to "find" the predicted cancer. It is generally agreed that, although a two stage validation effort takes longer (i.e. blinded values for the marker in question until completion of the study followed by a study with invasive or semi-invasive investigations based on fluctuations or absolute levels of the marker), such a model is currently preferred. With respect to novel treatment strategies predicted from retrospective analysis and/or through the preclinical studies defined above, the next logical step is to attempt to validate the retrospective data which initially pointed the investigators to these potential treatment options. If the new therapy has never been used in humans before, it is important to define the maximum tolerated dose of the treatment. This is defined as the dose at which, in a Phase I clinical trial, a defined fraction (typically one third) of the patient cohort develop dose limiting toxicity. Once the dose or treatment strategy is found to be acceptable, a Phase II trial is designed to provide a measure of the activity of the agent(s) and to begin to define whether there is any benefit of the regimen in certain patient cohorts. The use of tools such as Dx2Tx could be invaluable in the future tracking and analysis of such trials, especially if samples are prospectively harvested at various times pre and post treatment. For example, in the Phase I/II trial, genomic/proteomic correlates of clinical toxicity could be readily identified by melding the clinical data with "omic" data from the patients who have undue toxicity. Molecular pathways that intuitively result in toxicity could be identified which could possibly be abrogated with another agent. With respect to determination of therapeutic efficacy, our present means of defining "clinical correlates of response" is essentially a "best guess" or subjective prediction of what clinical markers may indicate that the new agent is yielding a therapeutic effect. A more objective measure of downstream events resulting from an efficacious agent would result from analysis of the molecular data in tandem with clinical information for the responders compared to the non-responders. Dx2Tx would be able to then define the most important markers of response, both clinical and molecular, which could then be validated in a Phase III trial assessing agent efficacy. End Use It is unknown when an integrated clinical/molecular evaluation of the suspected or afflicted cancer patient will be of use to the end user, the practicing physician. Certainly, many of the same arguments that are used for and against "genetic testing" in other diseases may be used for such a global approach to oncology. There is no doubt however, the ability to define clinical behavior in a more efficacious and predictive manner, other than the archaic prognostic indicators which we use today, will help clinicians initiate and/or alter treatment course earlier and guide clinicians toward informed discussions with their patients in order to make treatment and/or surveillance decisions. If indeed the fidelity of combinatorial molecular and clinical medicine proves to be satisfactory, we may then be able to spare patients unnecessary treatment interventions which are currently doomed to failure. Moreover, the medical oncologist will be able to choose the correct cytotoxic and/or targeted therapy based on a global clinical-molecular snap-shot, which should translate into more favorable health economic policies and patient outcomes. Conclusions Throughout this review, we have highlighted the importance of designing research protocols involving human subjects that permit the collection of not only clinical specimens, but also extensive standardized clinical data in a longitudinal fashion. Through the merging of clinical and molecular data, non-biased patterns can be discovered that could translate into novel diagnostic and/or treatment opportunities. We have introduced a methodical approach for archiving and mining these seemingly disparate data sources that we believe can accelerate the translational research discovery pipeline. While there are clearly improvements to be made in the systematic collection of accurate and nationally standardized clinical data, we believe the integration of medical informatics and the molecular technologies can convert the once visionary concept of molecular-based medicine into a present reality. Competing Interests CW – A provisional patent has been filed on the Dx2Tx system described in this article. HP – This author declares that he has no competing interests. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC524522.xml |
516779 | The link between thyroid autoimmunity (antithyroid peroxidase autoantibodies) with anxiety and mood disorders in the community: a field of interest for public health in the future | Background To evaluate the association between mood and anxiety disorders and thyroid autoimmunity in a community sample. Methods: A community based sample of 222 subjects was examined. Psychiatric diagnoses were formulated using the International Composite Diagnostic Interview Simplified (CIDIS), according to DSM-IV criteria. All subjects underwent a complete thyroid evaluation including physical examination, thyroid echography and measure of serum free T4 (FT4), free T3 (FT3), thyroid-stimulating hormone (TSH) and anti-thyroid peroxidase autoantibodies (anti-TPO). Results 16.6% of the overall sample had an anti-TPO value above the normal cut-off. Subjects with at least one diagnosis of anxiety disorders (OR = 4.2, C.L. 95% 1.9–38.8) or mood disorders (OR = 2.9, Cl 95% 1.4–6.6, P < 0.011) were positive for serum anti-TPO more frequently than subjects without mood or anxiety disorders. A statistically significant association with anti-TPO+ was found in Anxiety Disorder Not Otherwise Specified (OR = 4.0, CL 95% 1.1–15.5), in Major Depressive Episode (OR = 2.7, CL 95% 1.1–6.7) and Depressive Disorder Not Otherwise Specified (OR = 4.4, S CL 95% 1–19.3). Conclusions The study seems to suggest that individuals in the community with thyroid autoimmunity may be at high risk for mood and anxiety disorders. The psychiatric disorders and the autoimmune reaction seem to be rooted in a same (and not easy correctable) aberrancy in the immuno-endocrine system. Should our results be confirmed, the findings may be of great interest for future preventive and case finding projects. | Background Autoimmune thyroid disease may be linked to depression [ 1 ] and anxiety [ 2 ]. Autoimmune disease and depression are not uncommon: the prevalence of autoimmune thyroid disease in the community ranged from 4 to 25% [ 3 ] and lifetime prevalence of Major Depressive Disorder ranged from 6 to 17% [ 4 ]. Thus the association may have a great relevance in terms of public health and prevention. The purpose of this investigation was to evaluate the relationship between mood and anxiety disorders and thyroid autoimmunity in a community survey. This research was carried out on the data base of two epidemiological studies aimed at defining the prevalence of psychiatric [ 5 ] and thyroid diseases [ 6 ] in Sardinia. On planning these surveys, researchers agreed to evaluate a representative sub-sample of a defined geographical area common to both endocrinological and psychiatric epidemiological surveys. This paper present the results of the cross psychiatric and endocrinological evaluation from the common areas of the two surveys. Methods The sample was extracted by randomization (1/10) subsequent to stratification according to age and sex, from the records of 2 Sardinian villages. Probands were interviewed face to face in their homes by specifically trained physicians. Two standardized forms were used to acquire information concerning: demographic data, state of health and use of social and health services. Psychiatric diagnosis was made using the Italian Simplified version of the Composite International Diagnostic Interview (CIDIS) [ 7 ]. The computer elaboration of data obtained enabled prevalence of psychiatric disorders according to DSM-IV [ 8 ] diagnostic criteria to be calculated. Anti-thyroid peroxidase autoantibodies (anti-TPO), considered as the most sensitive and specific marker of thyroid autoimmunity [ 9 ] was determined by RIA (Sorin Biomedica Diagnostics, Saluggia, Italy) with a cut-off value of 20 IU/ml. All subjects underwent a complete thyroid evaluation including physical examination, thyroid echography and measure of serum free T4 (FT4), free T3 (FT3), thyroid-stimulating hormone (TSH) and anti-thyroid peroxidase autoantibodies (anti-TPO). FT4 and FT3 were measured by means of a chromatographic method based on separation of free T4 on Lisophase columns (Technogenetics, Milan, Italy; normal values: FT4 6.6–16 pg/ml; FT3 2.8–5.6 pg/ml). TSH was measured by a chemiluminescent method (Ortho-Clinical Diagnostics Amersham, U.K.) with normal values ranging from 0.3–3.0 μU/ml. Thyroid echography was performed using a "real time" echograph (ALOKA Mod SSD 500 with a small parts 7.5). The association of anti- TPO+ with the main diagnoses deriving from CIDIS interview was calculated using Odds Ratio. Statistical significance was calculated using the X 2 test in 2 × 2 tables. Odds Ratio confidence intervals were calculated through application of the method of Miettinen [ 10 ]. Multivariate Logistic Regression was performed in order to evaluate the possible influences of gender and age on the association between anti-TPO+ and mood or anxiety disorders. The analysis was carried out considering mood (or anxiety) disorders as dependent variable, and anti-TPO+ (presence vs absence), gender (female vs male) and age (≤ 44 vs > 44) and their second order interactions as independent variables, by means of backward stepwise procedure; interactions lacking evidence of association (p > 0.20) were eliminated from the models. Results From a total of 261 subjects identified (age >18 years), 222 (85.1%), 127 females (57.2%), and 95 males (42.7%); over 44 years 127 (57.2%), 79 females (62.2%,) 48 males (37.7%), agreed to take part in the study whilst 20 (8,7%) refused to participate and 19 (7.3%) could not be traced. The final sample did not differ respect to the population of origin with reference to the variables applied in stratification. The lifetime prevalence of anxiety disorders in the sample was: Generalized Anxiety Disorder (GAD) 11.3%, Panic Disorder (PD) 2.7%, Anxiety Disorder Not Otherwise Specified (ADNOS) 5.4%, Social Phobia (SP) 5.4%; 18.5% had been diagnosed with at least one of these anxiety disorders. With regard to mood disorders, Major Depressive Episode (MDE) was present in 14.4%, Dysthymic Disorder (DD) in 2.7%, Depressive Disorder Not Otherwise Specified (DDNOS) in 4.0%; 18.9% had at least one of the above mentioned depressive disorders. 1.1% of the overall sample was affected by hypothyroidism, 16.6% had an anti-TPO value above the normal cut-off (anti-TPO+). Table 1 shows the association between anti-TPO+ and lifetime psychiatric diagnosis. Subjects with at least one lifetime diagnosis of anxiety disorders or one lifetime diagnosis of mood disorders presented anti-TPO+ more frequently than subjects without mood or anxiety disorders. An association with anti-TPO+ was found in ADNOS, in MDE and DDNOS. The attributable risk for ADNOS was 0.54, 0.37 for MDE and 0.34 for DDNOS, other conditions as PD and DD presented very high attributable risk. Table 1 Association between positivity anti-TPO (anti-TPO+), mood and anxiety diagnosis. Diagnosis Prev.nce N (%) Anti-TPO+ N (%) OR IC 95% χ2 P Att.able Risk One Anxiety Diagnosis GAD+PD+SP+ADNOS 41 (18.5) 15 (36.6) 4.2 1.9/38.8 12.6 0.001 0.54 One Mood Diagnosis MDE+DD+DDNOS 42 (18.9) 13 (30.9) 2.9 1.4/6.6 6.4 0.011 0.37 GAD 25 (11.3) 8 (32) 2.7 0.97/7.5 3.6 0.058 0.35 PD 6 (2.7) 3 (50) 5.4 0.7/37.3 2.8 0.096 0.68 SP 12 (5.4) 4 (40) 3.6 0.7/7.6 2.5 0.111 0.52 ADNOS 12 (5.4) 5 (41.7) 4.0 1.1/15.5 3.9 0.045 0.55 MDE 32 (14.4) 10 (31.2) 2.7 1.1/6.7 4.6 0.033 0.34 DD 6 (2.7) 2 (50) 5.2 0.3/16.8 1.3 0.250 0.67 DDNOS 9 (4.0) 4 (44.4) 4.4 1/19.3 3.8 0.049 0.60 GAD: Generalized Anxiety Disorder; PD: Panic Disorder; SP: Social Phobia; ADNOS: Anxiety Disorder Not Otherwise Specified; MDE: Major Depressive Episode; DD: Dysthymic Disorder; DDNOS: Depressive Disorder Not Otherwise Specified. The Multivariate Logistic Regression showed that gender and age do not interact with anti-TPO either in mood (interaction anti-TPO-gender p = 0.97, OR 1.03, CI 95%, 0.19–5.48; interaction anti-TPO-age p = 0.62, OR = 1.52, CI 95%, 0.30–7.77), or in anxiety disorders (interaction anti-TPO-gender p = 0.93, OR = 0.93, CI 95%, 0.18–4.80; interaction anti-TPO-age p = 0.67, OR = 0.70, CI 95%, 0.14–3.64). Interactions were therefore eliminated from the models. The final Logistic Regression models clearly indicated that gender and age do not influence the risk of one mood or anxiety diagnosis either as independent variables or as confounders (Table 2 , Table 3 and Table 4 ). Table 2 Frequency of anti-TPO+ in to the sample according sex and age and one mood diagnosis (OMD) and one anxiety diagnosis (OAD) diagnosis for multivariate logistic regression. Age Sex N anti-TPO+ (%) OMD (%) OMD (%) #anti-TPO+ OAD (%) OAD (%) #anti-TPO+ <45 Female 79 17 (21.5) 17 (21.5) 6 (35.2) 13 (16.4) 6 (35.2) <45 Male 48 8 (16.6) 7 (14.5) 2 (24) 7 (14.5) 3 (37.5) >44 Female 48 7 (14.5) 10 (20.8) 3 (42.8) 14 (29.1) 4 (57.1) >44 Male 47 5 (10.6) 8 (17.2) 2 (40) 7 (15.1) 2 (40) #TPO+ in the sub-group is the total Table 3 Multivariate logistic regression: effect of anti-TPO+ on risk of one mood diagnosis considering gender (female vs male) and age (≤ 44 vs > 44) effects. p OR CI 95% anti-TPO+ vs anti-TPO- 0.01 2.89 1.31–6.38 Gender (F vs M) 0.37 1.38 0.68–2.82 Age (≤ 44 vs > 44) 0.71 1.14 0.57–2.30 Table 4 Multivariate Logistic Regression: effect of anti-TPO+ on risk of one anxiety diagnosis considering gender (Female vs Male) and age (≤ 44 vs > 44) effects. p OR CI 95% anti-TPO+ vs anti-TPO- 0.001 4.50 2.02 – 10.04 Gender (F vs M) 0.23 1.58 0.75 – 3.31 Age (≤ 44 vs > 44) 0.08 1.91 0.92 – 3.96 Discussion The present study indicates an association between the presence of a lifetime diagnosis of mood or anxiety disorder and anti-TPO+ in a general population sample which had not been selected from medical or psychiatric health facilities. This association is independent by gender and age. Regarding specific diagnosis, MDE, DDNOS and ADNOS were associated with anti-TPO+. This finding is consistent with several previous clinical studies providing evidence for a significant association of mood disorders or post-partum depression and symptomless autoimmune thyroiditis with or without sub-clinical hypothyroidism [ 11 ]. Association between hypothyroidism and mood disorders is however controversial as other authors maintain that bipolar disorders rather than unipolar depression are characterized by an increased risk for the presence of anti-thyroid antibodies [ 11 ]. However, a study performed by Fountoulakis and collaborators [ 1 ] recently found a link between autoimmune thyroid disease and Unipolar Depression. In this study, compared to control patients all depressive subtypes had significantly higher thyroid binding inhibitory immunoglobulins, and atypical patients had significantly higher thyroid microsomal antibodies. Thyroid function markers Free Triiodothyronine (FT3), Free Thyroxine (FT4), and Thyroid Stimulating Hormone (TSH) were normal in all subjects suggesting that Unipolar Depression might be characterized by a "low-thyroid function syndrome". A sub-clinical dysfunction of axis Thyrotropin Releasing Hormone (TRH) – Thyroid Stimulating Hormone (TSH) with consequent alteration of circadian rhythms of TSH has been hypothesized in some depressive disorders. Indeed, this hypothesis may explain why some forms of mood disorders were associated with anti-TPO+ or thyroid autoimmunity without hypothyroidism, as defined by routine blood tests. A slight reduction in thyroid hormone secretion such as that found in sub-clinical hypothyroidism may affect cognition and mood [ 12 ]. At variance with other tissues which mainly rely on peripherally generated Triiodothyronine, the brain utilizes preferentially circulating thyroxine directly secreted by the thyroid gland and may become hypothyroid before other organs [ 13 ]. Moreover, a study carried out in a large community sample found no association between thyroid dysfunction, including hypothyroidism defined by thyroid blood test, and the presence of depression or anxiety symptoms [ 14 ]. This survey is rather limited due to the fact that the presence of depression or anxiety symptoms was defined using a self-rating scale, whilst in the present study depression and anxiety disorders were diagnosed by means of a structured psychiatric interview according to an international classifications. However, the link between thyroid autoimmunity and depression may involve other mechanisms related to the autoimmune pathogenesis of thyroid disease rather than hypothyroidism. Since several neuroendocrine secretory systems are involved in the control of immune reaction, a common neuroendocrine dysregulation involving cytokines might concur towards the pathogenesis of both affective disorders and autoimmune disease. Recent evidence suggests that thyroid autoimmunity may be affected by the Hypothalamic-Pituitary-Adrenal axis (HPA) through the balance of proinflammatory and antiinflammatory cytokines [ 15 ]. In line with this view, the increased frequency of post-partum depression, associated to the fact that pregnancy would seem to be a "protected" period, could explain at least in part the consequences on thyroid autoimmunity elicited by HPA-related modifications to the immunitary axis. Indeed, similar phenomena are observed in rheumatoid arthritis and multiple sclerosis [ 16 , 17 ]. With regard to the public heath aspects of this research, should these findings be confirmed, they would constitute a most important public health issue, due to the high attributable risk found. The attributable risk is a useful measure to document the burden of risk to a community. Attributable risk depends both on the magnitude of relative risk and on the prevalence of the risk factor in the population [ 4 ]. The high attributable risk of autoimmune thyroid for Major Depressive Disorder and anxiety and depressive sub-threshold syndromes may have implication for the development of preventive interventions. Particularly, further longitudinal studies will be required to confirm whether anxiety and depressive disorders are a consequence and not a cause of thyroid autoimmunity. Limitations The potential of the study is reduced by the small sample size, particularly regard to psychiatric diagnoses less frequently observed in the general population, such as Panic Disorder; the extension of the findings is therefore rather limited. Conclusions This study indicates an association between the presence of a lifetime diagnosis of mood or anxiety disorder and anti-TPO+. The psychiatric disorders and the autoimmune reaction seem to be rooted in a same (and not easy correctable) aberrancy in the immuno-endocrine system. If the findings are confirmed, they may prove to be of interest for future projects of case finding: a systematic screening for mood disorders in anti-TPO+ subjects and a systematic evaluation for thyroid diseases and thyroid autoimmunity in subjects with mood disorders may be advisable. Competing interests None declared. Authors' contributions MGC conceived the study, participated in the design of the study, performed the statistical analysis and drafted the manuscript. AL, SM and CS participated in the statistical analysis and drafted the manuscript. MCH, SM, MC, BC, LDO participated in its design and coordination. All authors have read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC516779.xml |
514608 | Epidemiological evidence of higher susceptibility to vCJD in the young | Background The strikingly young age of new variant Creutzfeldt-Jacob disease (vCJD) cases remains unexplained. Age dependent susceptibility to infection has been put forward, but differential dietary exposure to contaminated food products in the UK population according to age and sex during the bovine spongiform encephalopathy (BSE) epidemic may provide a simpler explanation. Methods Using recently published estimates of dietary exposure in mathematical models of the epidemiology of the new variant Creutzfeldt Jacob disease (vCJD), we examine whether the age characteristics of vCJD cases may be reproduced. Results The susceptibility/exposure risk function has likely peaked in adolescents and was followed by a sharp decrease with age, evocative of the profile of exposure to bovine material consumption according to age. However, assuming that the risk of contamination was proportional to exposure, with no age dependent susceptibility, the model failed to reproduce the observed age characteristics of the vCJD cases: The predicted cumulated proportion of cases over 40 years was 48%, in strong disagreement with the observed 10%. Incorporating age dependent susceptibility led to a cumulated proportion of cases over 40 years old of 12%. Conclusions This analysis provides evidence that differential dietary exposure alone fails to explain the pattern of age in vCJD cases. Decreasing age related susceptibility is required to reproduce the characteristics of the age distribution of vCJD cases. | Background Owing to the lack of data concerning age dependent dietary exposure to bovine material in food in the UK, mathematical models used in the study of new variant Creutzfeldt-Jacob disease (vCJD) have postulated an age risk function for contamination where the influence of age dependent dietary exposure to contaminated material could not be separated from intrinsic age related susceptibility [ 1 - 3 ]. For example, we modelled exposure/susceptibility by a plateau during the first 15 years of age followed by an exponential decrease afterwards [ 2 ]. Using this approach we estimated the duration of the incubation period for vCJD to circa 15 years; and predicted that the epidemic had likely peaked in 2001–2002 which is now consistent with the observed data [ 2 , 3 ]. The simple representation of the age risk function implied that few parameters were necessary to describe it, a desirable property when estimation is based on limited data. By June 2003, 139 cases of vCJD had been reported to the vCJD surveillance unit in the UK, as compared to the 79 cases used in our original paper [ 2 ]. This warrants the inclusion of more detail in the age risk function, especially to investigate whether the constant risk assumption in children holds, as it led to predict a bimodal age distribution for cases in the coming years [ 3 ]. At the same time, detailed estimates of the dietary bovine material consumption in the UK population according to age and sex have been made available [ 4 ]. This offers the opportunity to try to disentangle the role of exposure from that of age-specific susceptibility. Indeed, the estimates for dietary exposure show that it peaked during adolescence and decreased with age afterwards, a pattern which is consistent with the finding that most vCJD cases are young and were therefore at most teenagers during the years when the bovine spongiform encephalopathy (BSE) epidemic was at its maximum. We therefore set out to estimate the age risk function with a versatile description based on step functions, and to investigate whether dietary exposure alone could explain the age distribution of cases. Methods Data We obtained the age, sex and date of onset for the 137 vCJD cases reported to the UK vCJD unit as of June 2003, all of which had onset before October 2002. Delays in reporting may reach 18 months (R Will, personal communication) and bias downward the incidence curve. Therefore, we included in our analysis only the 129 cases with onset before November 2001, consisting in 71 men and 58 women. Mortality by age and sex for the UK was obtained from the Office for National Statistics [ 5 ]. Model The model extends our previously described method [ 3 ]. The instantaneous risk of infection for vCJD, λ(a,t) , was assumed to depend on date t and age a in a multiplicative manner, with λ(a,t) = f(a) g(t) , where parallels the entry of infected cows in the human food chain estimated from back-calculation results in the UK [ 6 ] and r corresponds to the impact of the specified risk material ban in 1989; f(a) is a function varying with age only (see following paragraph). Occurrence of cases of vCJD of sex i (i = Male or Female) is then modelled by a Poisson process in the (age, time) plane, with intensity , so that the expected number of onsets of sex i in the period [ t,t +d t ) and age [ a , a +d a ) is π i ( a , t ) d a d t [ 2 , 7 ]. In the equation above, h is the distribution of the incubation time, S i ( a,t ) is the probability of survival at time t for individuals of sex i aged a calculated from census information, and β ι is the intensity of a homogeneous Poisson process. The incubation period distribution h was taken to be lognormal, as the choice of a particular shape is not critical for the estimates [ 3 ]. This distribution was also assumed to be independent of sex, and of age at contamination. Estimates of the parameters were obtained by numerical maximization of the log-likelihood, defined as , where the first sum is on all observed cases where s i is the sex of individual i , and the second is over both sexes ( s ) and requires integration of the intensity over the domain (in time and age) D = [0,100] × [1 Jan 1980,31 Nov 2001]. Custom FORTRAN code, using SLATEC and TOMS library for numerical routines, was used in this step. In all instances, our model allowed estimation of the mean and standard deviation of the incubation period, the shape of f(a) , and β . The model predicted distribution of age for cases with onset before November 2001 was determined by integrating ( a , t ) over the period [1/1/1980, 1/11/2001] for consecutive age class of width 10 years. The χ 2 distance was computed to measure agreement between the observed and the model calculated age distributions, but no formal test was carried out. Age Risk function Model "Susceptibility & Diet" To allow a versatile shape in the age risk function, we chose a non parametric description based on step functions rather than a mathematical function with few parameters. We wrote so that the shape of f(a) was unconstrained for a<25 and decreased exponentially after age 25. We then estimated f i , i = 1,...,6 and α at maximum likelihood. With this particular choice, it is possible to visually check the progression in risk with age during infancy ( f 1 , f 2 , f 3 ), the presence of a peak in risk in teenagers ( f 2 , f 3 , f 4 ), and whether the drop in risk after 20 is strong enough to be exponential ( f 4 , f 5 , f 6 ). Model "Diet Alone" To investigate whether dietary exposure to meat was sufficient to explain the young age of vCJD cases, we first determined the exposure to meat according to age and sex from UK data, using consumption of carcass meat [ 4 ]. In these data, it is seen that the quantity consumed by week, as well as the percentage of consumers, changes with age. We therefore defined e(a) as the product of mean consumption by the corresponding percentage of consumers in age class a , and wrote f(a) = f 0 e(a) , where f 0 was estimated at maximum likelihood. Model "Susceptibility | Diet" Finally, to estimate the influence of age standardised on dietary exposure, we wrote and likewise estimated f i from the data. In model " Susceptibility & Diet " and " Susceptibility | Diet ", we normalised the f i by the largest of the values in the graphical presentation of results, yielding risks relative to the highest risk age class. Confidence intervals Maximum likelihood estimators have a limiting normal distribution around the true parameters, with limiting variance/covariance matrix given by the inverse of the Fisher information [ 8 ], therefore the quadratic form , where θ stands for a vector of k parameters, for the estimates, and i ( θ ) for the Fisher information matrix, has a limiting chi-squared distribution with k degrees of freedom. We therefore determined 95% confidence intervals by finding the values of parameters so that Q ( θ ) was less than the 95 th quantile of the corresponding chi-squared distribution. Results The estimated age risk function in model " Susceptibility & Diet " is presented in Figure 1 , and shows that an increase during childhood, peak during adolescence and sharp decrease afterwards provided the best fit. In this model, the average incubation period was estimated at 13.2 years (CI95% [11.2, 15.8]), with standard deviation 2.0 years (CI95% [1.1,3.7]). The model predicted age for the cases showed good agreement with the observed distribution ( χ 2 = 2.45; Figure 2 ). The shape of the estimated age risk function in Figure 1 is evocative of the age profile of dietary exposure to bovine carcass meat in the 1980s in the UK, as shown in Figure 3 , where an increase in consumption was noted during childhood and adolescence, and decreased afterwards. However, in model " Diet Alone ", while the average incubation period was estimated at 12.1 (CI 95% [10.2, 14.2]) years with a standard deviation of 2.4 (CI95% [1.2, 4.1])years, close to that of the first model, the predicted age distribution of the cases was at odds with the observed distribution ( χ 2 = 58.4), as apparent in Figure 2 . More precisely, the predicted percentage of cases aged over 40 was 48% with the " Diet Alone " assumption, when the observed percentage was only 10%; it was 12% with the " Susceptibility & Diet " model. When estimating the residual influence of age, once exposure has been taken into account, model " Susceptibility | Diet " led to a profile for the fi as shown in Figure 4 . This profile retained the major characteristics of the age risk function obtained in the " Susceptibility & Diet " model, although with an earlier maximum susceptibility. With this model, the average incubation period established at 12.6 years (CI 95% [10.5, 14.7]) with standard error 1.8 years (CI95% [1.2, 3.5]), and the predicted age distribution of cases showed agreement similar to that of the " Susceptibility & Diet " model ( χ 2 = 2.36). Discussion Incorporating differential dietary exposure to BSE infected products according to age and sex, in a flexible age risk function for vCJD contamination, we found that exposure alone could not explain the young age of vCJD cases seen in the UK. Decreasing age related susceptibility had to be assumed to reproduce the characteristics of the age distribution of these cases. In all instances, the estimates for the epidemiological characteristics of vCJD were in line with those previously reported from comparable number of cases [ 3 , 9 , 10 ], and pointed to an epidemic of moderate size. We obtained a smaller point estimate for the mean incubation period when the age risk was allowed to change during childhood rather than assumed constant (13.2 yrs CI95% [11.2, 15.8] vs. 16.4 yrs CI95% [11, 24] [ 3 ]). All models considered here predicted that, by 2010, the epidemic should have ended, provided it is limited to individuals with the observed susceptible genotype (as of today, all vCJD cases are methionine homozygous at codon 129 of the PrP gene[ 11 ]). The average number of cases to come is predicted at 43, leading to a total size of 172 cases. From the model " Susceptibility & Diet ", it appears that a previous assumption of a constant age risk in children and adolescents [ 2 ] has likely led to overestimate the risk of infection in young children. The age-risk profile estimated here leads to a smaller risk than previously found in children born after 1980, to fewer cases among the young and an overall smaller total size for the epidemic than reported before. Furthermore, since few young cases are expected in the future, the bimodality of the predicted age distribution of cases is not found anymore. The estimated profile of the susceptibility/exposure age risk function agreed with that selected in scenario analysis [ 9 ], although our maximal risk is among the 15–20 years old rather than in the 10–15 years old. However, even after adjustment for dietary exposure, susceptibility remains rapidly decreasing in adults, as found in the " Susceptibility | Diet " model. This finding confirms that obtained in a recent analysis based on scenario analysis, where it was found that exponential decrease in susceptibility in the oldest cohorts was desirable [ 10 ]. This is also found here by direct estimation, and moreover, the age susceptibility appears to increase among children up to age 5–10, be almost constant among teenagers and rapidly decreasing afterwards. Three hypotheses have been put forward to explain the young age of the vCJD cases: age dependent incubation period, age dependent exposure, and age dependent susceptibility. Age dependent incubation period has recently been revived by Cooper et al., because scenarios including longer incubation periods in old cohorts provided a better fit of the data[ 10 ]. However, the increase of the mean age of cases with time which should occur with age dependent incubation periods [ 2 ], is not supported by the data. The correlation between age at onset and calendar time remains indeed extremely small (Spearman correlation coefficient r = -0.025, n = 129). The potential role of differential dietary exposure was proposed early in the epidemic [ 12 ], because substantial differences in exposure existed in adolescents and adults. A disequilibrium in sex ratio towards males was for example predicted on this basis [ 13 ], but is not statistically established today ( χ 2 test, P = 0.49) although the proportion of males is 55%. Our analysis shows that, while the estimated age risk function for infection with vCJD exhibits a peak in the young that was also found in dietary exposure to potentially BSE infected products, differential dietary exposure alone does not account for the observed pattern in the age of the vCJD cases. Indeed, the relative exposure does not decrease rapidly enough with age to reduce the risk in older adults. Therefore, an additional effect of age is required to fully account for the age of the vCJD cases. Once exposure is taken into account, this effect appears to be peaking in children less than 10 years old, and decreasing afterwards. Contamination through BSE infected food is today the most plausible explanation for the occurrence of vCJD; this is the rationale for correcting the risk of infection by the level of dietary exposure. However, this explanation remains today hypothetical, because even if epidemiology and biochemistry favour a link between BSE and vCJD, this is not regarded as ultimately conclusive [ 14 ]. In this work, we examined the further possibility that, provided dietary exposure was the culprit, differences in age-related exposure could explain the age distribution of cases. This assumes that the risk of vCJD infection may have been larger in those consuming more bovine products; however population based models linking bovine products consumption to vCJD incidence were not conclusive in this respect [ 15 ]. Conclusions The reasons for an increased susceptibility in teenagers compared to young children and adults are today speculative. In many infectious diseases, a decrease in susceptibility with age is mediated through the immunological responses to repeated exposure since birth. However, for vCJD, most cohorts in the population have had the same duration of exposure to BSE infected material. Due to the age range considered, biological processes involved in the maturation of the immune system; in response to hormonal changes may be incriminated. For example, changes in the permeability of the intestinal barrier occurs with age with decreasing Peyer's patches [ 16 ]. Further experimental research is required to provide explanations for what remains a very unusual characteristic in an infectious disease. List of abbreviations vCJD: variant Creutzfeldt-Jacob Disease BSE : Bovine Spongiform Encephalopathy Competing interests None declared. Authors' contributions PYB designed the study, analysed the results and drafted the manuscript. JYC participated in the design and writing. AJV participated in the design and writing. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC514608.xml |
549080 | A new example of viral intein in Mimivirus | Background Inteins are "protein introns" that remove themselves from their host proteins through an autocatalytic protein-splicing. After their discovery, inteins have been quickly identified in all domains of life, but only once to date in the genome of a eukaryote-infecting virus. Results Here we report the identification and bioinformatics characterization of an intein in the DNA polymerase PolB gene of amoeba infecting Mimivirus, the largest known double-stranded DNA virus, the origin of which has been proposed to predate the emergence of eukaryotes. Mimivirus intein exhibits canonical sequence motifs and clearly belongs to a subclass of archaeal inteins always found in the same location of PolB genes. On the other hand, the Mimivirus PolB is most similar to eukaryotic Polδ sequences. Conclusions The intriguing association of an extremophilic archaeal-type intein with a mesophilic eukaryotic-like PolB in Mimivirus is consistent with the hypothesis that DNA viruses might have been the central reservoir of inteins throughout the course of evolution. | Background Mimivirus is the largest known virus, both in particle size (>0.4 μm in diameter) and genome length, recently discovered in amoeba, following the inspection of a hospital cooling tower prompted by a pneumonia outbreak [ 1 ]. Recently, its entire 1.2-Mbp genome sequence was determined [ 2 ]. Extensive phylogenetic studies and gene content analyses defined Mimivirus as a new family of nucleocytoplasmic large DNA viruses (NCLDV) besides Poxviridae , Iridoviridae , Phycodnaviridae and Asfarviridae , and suggested its early origin, probably before the individualization of the three domains of life [ 2 ]. While analyzing Mimivirus genome sequence, we noticed the unusual length of its putative DNA polymerase. A detailed analysis identified an intein in this gene. After the recent discovery of an intein in Chilo iridescent virus [ 3 ], an insect-infecting NCLDV of Iridoviridae , this is the second report of an intein sequence in a eukaryote-infecting virus. Inteins are "protein introns" that catalyze self-splicing at the protein level. The splicing is defined by the self-catalytic excision of an intervening sequence ("intein") from a precursor host protein where it is located, and the concomitant ligation of the flanking amino- and carboxy-terminal fragments ("exteins") of the precursor. Inteins often possess a homing endonuclease domain, and are considered as mobile elements. Since their first discovery in 1990 [ 4 , 5 ], inteins have been identified in a wide variety of organisms, including bacteria, archaea, and unicellular eukaryotes, albeit with sporadic distribution (see for a comprehensive list). For instance, they are relatively abundant in some hyperthermophilic archaea species (such as Methanococcus jannaschii possessing nineteen inteins), but absent in closely related species such as Methanococcus maripaludis [ 6 ]. Similarly, they are observed in many unrelated bacterial clades, but appear often limited to several species within each clade. It was suggested that viruses were potential "vectors" of inteins across species and responsible for the sporadic distribution of inteins [ 3 ]. Accordingly, inteins have been identified in many bacteriophages and prophages [ 7 - 10 ]. To our knowledge, the sole published account of eukaryote-infecting viruses harboring an intein concerns iridoviruses [ 3 ]. Results Eukaryotic Polδ-like Mimivirus PolB Mimivirus genome sequence exhibits a putative ORF (R322, 1740 amino acid long) corresponding to a family B DNA polymerase PolB. This ORF R322 exhibits high scoring sequence homology (BLAST E-value<10 -24 ) against eukaryotic PolBs in the public database. However, this Mimivirus PolB is much larger than its eukaryotic and viral homologues (about 1000 aa), and its optimal alignment with the other PolB sequences reveals four unmatched extraneous segments (Fig. 1A , Fig. S1 ). Focusing on these extra segments, we identified a 351-aa intein (position 1053 to 1403) in the Mimivirus PolB sequence. Figure 1 (A) Locations of inteins found in different DNA polymerases of the family B (PolB) (I, II, III; filled triangles) and other extra segments identified in the Mimivirus PolB (i1, i2, i3; open triangles). Nanoarchaeum equitans PolI is encoded in two pieces of genes (NEQ068, NEQ528), the break point of which corresponds to the position III intein integration site. Full intein motifs are comprised of the C-terminal part of NEQ068 and N-terminal part of NEQ528. (B) A phylogenetic tree of the family B DNA polymerases (PolBs) from diverse organisms, including Mimivirus (R322; GenBank AY653733), Paramecium bursaria Chlorella virus 1 (PBCV), Ectocarpus siliculosus virus (ESV), Invertebrate iridescent virus 6 (IIV), Lymphocystis disease virus 1 (LDV), Amsacta moorei entomopoxvirus (AME), Variola virus, Asfarvirus, eukaryotic DNA polymerase α and δ catalytic subunits, and archaeal DNA polymerase I. Intein containing genes are indicated by bold letters in the figure. Numbers in parentheses on the right of species name designate the numbering of paralogs. Sequences corresponding to inteins or Mimivirus extra segments (i1, i2, i3) were removed for the tree reconstruction. N. equitans PolI split genes were concatenated. (C) A phylogenetic tree based on the intein sequences found in PolBs. Numbers (I, II, and III) in parentheses on the right of species names indicate the intein integration sites. In (B) and (C), trees were built using a neighbor joining method, and rooted by the mid-point method. Bootstrap values larger than 70% are indicated along the branches. After removing those four Mimivirus specific insertions, the Mimivirus PolB sequence exhibited the highest BLAST scores (E-value = 10 -125 , 32% identity) against a soybean DNA polymerase Polδ (SWISS-PROT: O48901) with an alignment covering both the entire Mimivirus and the target sequence. Near equivalent matches are observed with a variety of eukaryotic (from yeast to human) family B DNA polymerase sequences. The best viral homologues were found in phycodnaviruses (E-value = 10 -116 ). Conserved carboxylate residues (aspartate and glutamate) at the exonuclease and polymerase active sites [ 11 , 12 ] were all identified in the Mimivirus PolB (Fig. S1 ). There was no other ORF encoding a putative PolB in the genome. These suggest that R322 encodes a functional PolB. Consistent with the homology search result, a phylogenetic analysis places the Mimivirus PolB near the root of eukaryotic Polδs (Fig. 1B ). A similar branching position is obtained for the seven universally conserved Mimivirus genes [ 2 ]. Despite low bootstrap values for some of the deep branches in the Fig. 1B , this tree clearly indicates the lack of any specific affinity between the Mimivirus PolB and the archaeal PolB sequences containing inteins (bold letters in the Fig. 1B ). It should also be noted that several other large DNA viruses are known to possess PolBs with a similar phylogenetic pattern [ 13 ]. Canonical/archaeal type Mimivirus intein The Mimivirus intein sequence (351 aa) exhibits significant sequence similarities to several known inteins (E-value<10 -4 ), all of which are from thermophilic/halophilic archaea. The best matching intein (E-value = 3 × 10 -8 ) is the second intein of the Thermococcus sp . PolB (InBase: Tsp-GE8 Pol-2) with 24% amino acid sequence identity. The Mimivirus sequence exhibits all the expected features required for an active intein (Fig. 2 ). Sequence motifs [ 14 ] characterizing the splicing domain (N1-4, C2, C1) and the dodecapeptide LAGLIDADG homing-endonuclease domain (EN1-4) were all identified in the Mimivirus sequence except N4 motif. N4 motif is occasionally absent in the previously characterized active inteins [ 14 ]. Amino acid residues providing nucleophilic groups in self-splicing reactions are all present: the first serine and the last asparagine residues of the intein, and the first threonine residue of the downstream extein. Accordingly the Mimivirus intein is a canonical "asparagine-type" intein, of which the close homologues have previously been observed only in archaea species. In contrast, the previously reported Chilo iridescent virus intein is a non-canonical "glutamine-type" exhibiting a glutamine residue at the C-terminus [ 3 , 15 ]. The threonine and histidine residues in the N3 motif assisting in the initial acyl rearrangement at the N-terminal splice junction are also conserved. Thus, we predict that the Mimivirus intein is an active intein capable of self-splicing. The presence of a homing endonuclease domain suggests that this intein also retained its capacity to spread to other sites of the genome or to other organisms. Figure 2 The Mimivirus DNA polymerase PolB intein. The 351 amino acid residues intein sequence is shown with, respectively, the last and the first three amino acid residues of the N-extein and the C-extein. Bold letters represent amino acid residues essential for protein splicing. Conserved intein sequence motifs are indicated by underlines (N1, N2, N3, EN1, EN2, EN3, EN4, C2 and C1). The sequence part matching to the Pfam LAGLIDADG endonuclease domain (PF00961, E-value = 0.16) is indicated by italic letters. The intein/extein boundaries are shown by '|'. Other three inserts that we identified in the Mimivirus PolB are rather short. Those inserts are unique to Mimivirus, being not found in other PolB sequences. One of the extra segments of 197 aa found at the position 'i3' (Fig. 1A ) exhibits a marginal sequence similarity to an intein within the replication factor C of Methanococcus jannaschii (E-value = 0.002, Fig. S2 ). However, it also exhibits a comparable level of sequence similarities to several unrelated database sequences, apparently containing low complexity sequences. The i3-insert lacks sequence features required for an active intein. The remaining two extra segments (88 and 121 aa at the position 'i1' and 'i2', respectively) did not exhibit any significant similarity to known protein sequences. The biological properties of those three Mimivirus specific inserts remain to be characterized. Mimivirus intein belongs to a specific allele type Inteins have been identified in different types of DNA polymerases [ 16 ]. DNA polymerase catalytic subunits known to contain inteins are archaeal PolI, archaeal DNA polymerase II (PolII), bacterial DNA polymerase III α subunit (DnaE) and bacteriophage DNA polymerase I. Among these, archaeal PolI belongs to the family B DNA polymerase. Archaeal PolI contains up to three intein alleles, the insertion of which always occurs at one of three strictly conserved positions (I, II and III in Fig. 1A ). Interestingly, the location of the bipartite inteins that separate the two PolI gene pieces of Nanoarchaeum equitans [ 17 ] coincides with position III. Remarkably, Mimivirus intein is exactly located at the position III (Fig. 1A ). The sequence around the insertion site is highly conserved among different PolBs from evolutionary distant organisms such as Escherichia coli and human (Fig. 3 ). The crystal structure of Pyrococcus kodakaraensis PolI [ 11 ] reveals that those three distinct sites are in close spatial proximity, in the middle of the DNA binding domain and active site. Figure 3 Sequence alignment of Family B DNA polymerases from the Archaea, Bacteria and Eukarya domains. The Mimivirus PolB sequence was used without its intein sequence. Only the region of the alignment around Mimivirus intein insertion site ("YGD|TDS") is shown. The insertion site precisely coincides with the most conserved positions in the sequences, as indicated by bold letters. This is the sole region in the entire sequence exhibiting 6 consecutive identical residues among PolB of the Archaea, Bacteria and Eukarya domains. SWISS-PROT/TrEMBL IDs are DPOL_ARCFU ( Archaeoglobus fulgidus ), Q8TWJ5 ( Methanopyrus kandleri ), DPO2_ECOLI ( Escherichia coli ), Q87NC2 ( Vibrio parahaemolyticus ), Q8SQP5 ( Encephalitozoon cuniculi ), and DPOD_HUMAN (Human). Perler et al . observed that inteins present in the same location within homologous genes ("intein alleles") tend to be more similar with each other than with inteins in different locations of the same gene or in different genes [ 18 ]. This phenomenon appears not only the simple consequence of regular vertical transmission of inteins, but also the result of lateral acquisitions through "homing" [ 19 ] at the same site of highly similar genes (i.e. "alleles") by the mechanism involving gene conversion [ 18 ]. Remarkably, the Mimivirus PolB intein holds this rule. The Mimivirus intein exhibits higher sequence homology scores to inteins at the position III of archaeal PolI (designated as "pol-c allele") than to inteins in the other PolI locations (I, II) or inteins in other genes. A phylogenetic analysis of the Mimivirus intein and other PolI inteins also supports the classification of the Mimivirus intein in this specific "intein allele"-type (Fig. 1C ). This underlines the presence of intein subclasses ("intein alleles") each exhibiting its own preference of harboring site, even in such distantly related homologous genes such as Mimivirus PolB and archaeal PolI. It is implausible that the intein homing mechanism involving gene conversion have led to the direct transfer of an intein between such distantly related homologous genes. Nucleotide sequences (18 bp) around the pol-c allele insertion site do not exhibit unexpectedly high level of sequence similarities between Mimivirus (TATGGAGAC/ACGGACTCA for the amino acid sequence YGD/TDS) and archaeal sequences. For instance, the sequences from M. jannaschii and Pyrococcus horikoshii exhibit 7-missmaches (TAT ATT GAC/AC T GA TGG A; MJ0885) and 5 mismatches (TAT AT AGAC/ACGGA TGG A; PH1947), respectively. To the best of our knowledge, no evidence has been reported for a homing endonuclease recognizing such different sequences, although homing endonucleases are known to be rather tolerant of single-base-pair changes in their lengthy DNA recognition sequences [ 19 ]. A similar observation has been reported for DnaB inteins of Rhodothermus marinus and Synechocystis sp. PCC6803 [ 20 ]. A shift in the base compositions between intein and extein coding sequences is considered as indicating a recent acquisition of inteins [ 20 ]. Mimivirus PolB extein/intein DNA sequence compositions do not show a significant difference. Both exhibit similar G+C-contents (29%) and codon usages. In contrast, Thermococcus fumicolans PolI coding DNA (GenBank: Z69882) exhibits a G+C-content of 57% for the extein regions, compared to G+C-contents of 47% and 49% for its two inteins. Discussion Archaeal PolI inteins have been described only in extremophiles, growing under conditions of temperature over 80°C (hyperthermophiles) or of high salinity (10 times that of sea water; halophiles). Mimivirus is mesophilic, growing in amoeba under the temprature of 37°C. The association of an archaeal-seqeunce-like intein with a eukaryotic-like PolB in Mimivirus thus suggests an indirect interaction between mesophilic eukaryotic viruses and extremophilic archaeabacteria. Mesophilic euryarchaea species similar to the methanogens associated with rumen [ 21 , 22 ] or related species found in human beings [ 23 ] might have mediated the transition of inteins between extreme environment and moderate one in the course of evolution. However, no data are available yet on the presence of inteins in the PolB genes of such mesophilic archaebacteria. Lateral transfer (homing) might be responsible for the phylogenetic incongruence between inteins and exteins, and the same intein locations within homologues of distantly related organisms such as Mimivirus and archaea. However, given the specificity of homing endonucleases to long recognition sequences (12–40 bp) and the low level DNA sequence similarity between viral and archeal PolB homologues, a single recent homing event appears quite unlikely. The spread of inteins is better explained by a series of transfers, where inteins progressively accommodated small changes in their homing recognition sequences while retaining their gene position specificity. Such a cascade of transfers could have been mediated by DNA viruses [ 3 ]. Consistent results now start to accumulate including recent identification of several inteins in different iridoviruses (S. Pietrokovski pers. comm.), and an intein in a golden brown alga-infecting virus HaV of the Phycodnaviridae [ 24 ]. Given the similar base compositions of Mimivirus intein and extein, the low level of intein homology between Mimivirus and archaea, and the likely early origin of the Mimivirus/NCLDV lineage [ 2 ], it is tempting to speculate that these DNA viruses might have acquired inteins very early on, and acted as their central reservoir disseminating inteins across different domains of life in the long course of evolution. Conclusions We have characterized a new viral intein found in the eukaryotic-type putative DNA polymerase PolB of Mimivirus by binformatics methods. The conservation of the active site motifs for splicing as well as its insertion at a catalytically important site of the PolB sequence suggests that the intein is most likely to be functional. Our phylogenetic analyses revealed that the intein sequence is closest to extremophilic archaeal inteins. The intriguing association of an extremophilic archaeal-type intein with a mesophilic eukaryotic-like PolB in Mimivirus is consistent with the hypothesis that DNA viruses might have been the central reservoir of inteins throughout the course of evolution. Methods Sequence homology searches were carried out with the use of the BLAST programs [ 25 ] against the SWISS-PROT/TrEMBL database [ 26 ] and the New England Biolabs Intein Database [InBase, ; [Perler, 2002 #1380]]. Pfam [ 27 ] searches were carried out with the use of its web site . Multiple sequence alignments were generated with the use of T-Coffee [ 28 ]. Intein sequence motifs were identified through the inspection of a multiple intein sequence alignment. Neighbor joining tree analyses were conducted with the use of MEGA version 2.1 [ 29 ]. All the gap containing columns in multiple sequence alignments were removed before phylogenetic tree analyses. The gamma distance was applied to compute evolutionary distances. The gamma shape parameter (alpha) was estimated using the GZ-GAMMA program [ 30 ]. The sequence and annotation data for the Mimivirus PolB and intein was deposited to GenBank (accession number: AY606804). The complete genome sequence of Mimivirus is also available at GenBank (accession number: NC_006450). For a comprehensive description of the Mimivirus complete genome sequence and preliminary characterizations of the viral particle, see [ 2 ]. Competing interests The author(s) declare that they have no competing interests. Authors' contribution HO carried out most of the sequence analysis, contributed to the interpretation of the results, and drafted the manuscript. DR contributed to the interpretation of the results. JMC contributed to the construction of the sequence alignment, participated in the interpretation of the results and finalized the manuscript. Supplementary Material Additional File 1 Supplementary figure S1 Sequence alignment of Mimivirus PolB and eukaryotic Polδs. The Mimivirus intein sequence is removed, and its insertion site is highlighted by amino acid residues in red corresponding to the left three and right three resides around the insertion site. Three Mimivirus specific inserts (i1, i2 i3) were highlighted by blue letters. Conserved carboxylate residues in the exonuclease and polymerase active sites are highlighted by green background. Eukaryotic sequences were Encephalitozoon cuniculi (TrEMBL/SWISS-PROT: Q8SQP5), Schizosaccharomyces pombe (P30316) and Glycine max (soybean, O48901). Sequence alignment was obtained with the use of T-Coffee. Click here for file Additional File 2 Supplementary figure S2 Sequence alignment of Mimivirus insert i3 and known intein sequences. Intein sequences are from Methanococcus jannaschii replication factor C (Mja RFC-3) and Pyrococcus abyssi replication factor C (Pab RFC-2). Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549080.xml |
549533 | From wealth to health: modelling the distribution of income per capita at the sub-national level using night-time light imagery | Background Sub-national figures providing information about the wealth of the population are useful in defining the spatial distribution of both economic activity and poverty within any given country. Furthermore, since several health indicators such as life expectancy are highly correlated with household welfare, sub-national figures allow for the estimation of the distribution of these health indicators within countries when direct measurement is difficult. We have developed methods that utilize spatially distributed information, including night-time light imagery and population to model the distribution of income per capita, as a proxy for wealth, at the country and sub-national level to support the estimation of the distribution of correlated health indicators. Results A first set of analysis are performed in order to propose a new global model for the prediction of income per capita at the country level. A second set of analysis is then confirming the possibility to transfer the country level approach to the sub-national level on a country by country basis before underlining the difficulties to create a global or regional models for the extrapolation of sub-national figures when no country data set exists. Conclusions The methods described provide promising results for the extrapolation of national and sub-national income per capita figures. These results are then discussed in order to evaluate if the proposed methods could not represent an alternative approach for the generation of consistent country specific and/or global poverty maps disaggregated to some sub-national level. | Background Economy, income and poverty do affect and are affected by population's health in many ways. At broad scale, the macro relationship between life expectancy and the gross national product (GNP) is well known and has been presented in different publications [ 1 ]. At a smaller scale a very robust relationship exist between an adult individual's income and that individual's health. This has been confirmed in the review done by Benzeaval and Judge [ 2 ] of sixteen studies coming from four different countries and for which the authors conclude by saying that: "All of the studies that include measures of income level find that it is significantly related to health outcomes." The conclusion of another study performed in Tanzania [ 3 ] shows that the poorest tercile of the households in this country are the ones presenting the poorest health status indicators confirming the relationship between poverty and health status. The same study also confirms the effect of the geographic distribution of poverty on the health status of the population. In return, high level of poverty also becomes an important factor of vulnerability for the population which becomes more exposed to diseases, especially infectious ones. By identifying the poorest area within a country it becomes possible to plan more effective intervention aimed at improving the health status of the population and therefore potentially reducing their level of poverty. As poverty tends to be clustered in specific places it is important to have access to disaggregated data. In addition to that, aggregated, national-level poverty data tends to mask sub-national variations [ 4 ]. The development of variables that can be used as indicators of economic status is not straightforward. Even the measurement of the most basic of economic variables – such as national income levels – is fraught with problems. This reliable measurement of income is particularly problematic for low-income countries, given the lack of well-developed national income accounting methods and the large size of the "informal" sector in these economies. These problems are compounded when information is sought on the spatial and temporal changes in economic activity [ 5 ]. Recent advances in measurement and estimation techniques, though, have helped substantially. So much so that it is now routine for national statistical offices of almost all countries to report on national economic activity numbers, as well as for international organizations such as the World Bank and the International Monetary Fund to report their own versions of "adjusted" income numbers (both in local currency units as well as in purchasing-power parity terms). However, the use of self-reported income for measurement of economic status is widely regarded to be problematic [ 6 ]. In a cross-section, income for any given household tends to be a relatively noisy indicator of its underlying longer-term economic status. From an accounting point of view, income numbers for subsistence-farming and self-employed households are particularly troublesome. In addition, respondents often perceive income-related items as being invasive and this can lead to non response bias. For these and other reasons, survey income has tended to be significantly under-reported and inconsistent with income estimated using national accounts statistics. Survey-based estimates of income are often lower than those of consumption for the same household, even though national accounts data show aggregate positive savings rates [ 7 ]. The degree of under-reporting in income has been found to vary by income deciles: lower-income households tend to be more likely to under-report than higher-income households. In several instances, poorer household have been found to report expenditure levels that far exceed reported income levels – possibly because of greater underreporting of income than of expenditure – indicating the implausible implication that the poor are chronic dissavers [ 8 ]. For all these reasons, most of the national household surveys, such as the World Bank's Living Standards Measurement Study (LSMS) [ 9 ] and many national surveys prefer to measure consumption and not income as the indicator of household welfare. If sub-national level consumption figures are therefore available for most low income countries, through the use of these tools, the data they are producing is not comparable over countries making it impossible, for the moment, to build a consistent global map of poverty, or food insecurity, disaggregated to some sub-national level. In this context the growing uses of Geographic Information Systems (GIS) as well as the generation of new geocoded data sets might offer new perspectives in order to produce globally consistent poverty maps and help predicting the distribution of welfare at the sub-national level when reliable sub-national data are still not available. As the consumption indicators produced are not comparable over countries this attempt is done using income per capita expecting that this indicator might be comparable. In terms of data, satellite imagery is offering great potential for global data sets depicting weather patterns plus the physical and biological environment. If most of the data sensed concerns bio-physical parameters (e.g. clouds and vegetation...) there is one parameter, sensed by some satellites that can be used in the socio-economic context: night-time light. In their publications on the use of this parameter, Elvidge et al. [[ 10 - 12 ] and [ 13 ]] compares country level surface area with detected lighting at night (area lit) with population, energy usage, and economic activity. They found a strong correlation between area lit and Gross Domestic Product (GDP) for 21 countries as illustrated in Figure 1 . Figure 1 Area lit (km2) versus 1994 Gross Domestic Product. GDP estimates for 21 countries on a log-log plot (extracted from Elvidge et al., 12) More recently, an attempt to obtain a global map of socio-economic parameters at the sub-national level has been made by Doll et al. [ 14 ]. In their approach, the country-level relationship found between area lit and Gross Domestic Product (GDP) was applied to a 1° × 1° resolution grid. This research showed a potential solution for obtaining sub-national distribution map for this parameter. The methodology presented in this paper follows a different approach, using other parameters in combination with light in order to directly predict GDP per capita and adjust the results to specific conditions. It also demonstrates the role played by other environmental and socio-economic factors on this prediction. Results and discussion The analysis performed at both the country and sub-national levels as well as the results obtained are presented in the coming sections. The country level analysis The objective of the country level analysis is to go beyond the observations done so far [[ 10 - 13 ] and [ 14 ]] and to see if an other component or combination of the light information (number of cells with light, total frequency of observation, mean frequency of observation) with other parameters could provide a model for the prediction of income per capita at the country level. The relationship between area lit and GDP shown by Elvidge et al. in 1997 [ 10 ] has already been confirmed for a larger number of countries by Doll et al. in 2000 [ 14 ] and Elvidge et al. one year later [ 13 ]. Starting from this result, a new set of analysis is performed using the different parameters described in the methods section. For the parameters stored in grids (light and surface area) the GIS tool is used in order to extract the figures based on the country delimitation. The first set of result are reported in Table 1 and present the correlation factors existing between the different parameters expressed in log for 171 countries with: Table 1 Correlation factor between the parameters used for the country level analysis Logdp Logdppc Lopopun Losurfli Lotofre Lonbrpix lomeanf Logdppc 0.4295 Lopopun 0.8708 -0.0700 Losurfli 0.9340 0.3963 0.8159 Lotofre 0.9374 0.4679 0.7808 0.9906 Lonbrpix 0.9318 0.4249 0.7980 0.9975 0.9927 Lomeanf 0.4413 0.5134 0.2079 0.3728 0.48 0.3709 Losurfco 0.6723 -0.1513 0.8250 0.7442 0.6842 0.7265 -0.0164 - Logdp: log of the GDP figure (expressed in ppp) - Logdppc: log of the GDP per capita figures (expressed in ppp) - Lopopun: log of the UN population figures - Losurfli: log of the area lit (km 2 ) - Lotofre: log of the total frequency of light observation - Lonbrpix: log of the number of cells being highlighted - Lomeanf: log of the mean frequency of light observation - Losurfco: log of the surface area of the country (km 2 ) The logarithm of all the parameters has been used in this analysis for the following reason: a) the relationship is expected to be linear in logs, b) if there is heteroskedasticity, the log form is one way to remove the problem, c) the parameters can be interpreted as elasticities if both the dependent and the independent variables are in logs The following observation can then be extracted from Table 1 : 1) Three of the light parameters are correlated to each other (lotofre, lonbrpix and losurfli). The forth one (lomeanf), not correlated to the previous ones, seems to contain a different information content connected to light, 2) Apart from the mean frequency of light observation all the parameters are more highly correlated with GDP rather than with GDP per capita, 3) There is no significant difference in the values obtained for expressing the correlation between the first three light parameters (lotofre, lonbrpix and losurfli) and GDP. Lomeanf shows a lower correlation with GDP, 4) Population presents a high correlation with GDP but none with GDP per capita. The same is observed for the surface area of the country, 5) The correlation between the first three light parameter mentioned under point 1) and population is good which is not the case with the log of the mean frequency of light observation, 6) A high correlation exists between the surface area of the country and the population. The above-mentioned observations seem to indicate that the high correlation coefficient observed between the 3 light parameters (lotofre, lonbrpix and losurfli) and GDP is explained by the strong correlation between these parameters and population and the one between population and GDP. On the contrary the mean frequency of light observation shows a stronger correlation with GDP per capita. By predicting GDP per capita instead of GDP we therefore avoid any circularity in the model. Based on all the observation made it is decided to consider the use of the following variable in order to model GDP per capita at the country level: - the mean frequency of light observation as this variable present the highest correlation with GDP per capita, - the total frequency of light observation as this parameter provides the highest correlation with GDP per capita among the other three light parameters found to be correlated to each other (lotofre, lonbrpix and losurfli). This parameter also contains more variability than the number of cells or the area lit which is an advantage when working with small areas that could be completely highlighted. - the total population and surface area of the country. Even if these parameters do not provide a good correlation with GDP per capita they are a necessary adjustment factor for the light and population variable (density). As the correlation between GDP per capita and the different variables may not be linear the correlation existing between the log of GDP per capita and the square of the log of these variables is analysed. This analysis shows that the square improves the prediction only for the two selected light parameters (total frequency and mean frequency of light observation) which are used as additional variables, the square of the log of population and the surface area of the country being not used. From that point, 63 combination of the 6 variables kept for the analysis are tested in order to find a regression for modelling the log of GDP per capita at the country level. This is firstly done using the full data set, then trying to analyse the role of the climate and the one of the GDP composition by sector. Table 2 contains the information regarding the regression based on the best combination of significant variables (P > | t | < 50) using the full data set (171 countries). Table 2 Information about the best regression obtained for the country level data set with F (5, 165) = 156.46, Prob > F = 0.0000, R-squared = 0.8258, Adj R-squared = 0.8205 and Root MSE = 0.20552 logdppc Coef. Std. Err. t P > |t| [95% Conf. Interval] lopopun -0.4665182 0.038419 -12.14 0.000 -0.5423734 -0.39066 lotofre2 0.0574057 0.002592 22.15 0.000 0.0522876 0.062524 lomeanfr -2.677929 1.115337 -2.40 0.017 -4.880102 -0.47576 lomeanf2 0.9731717 0.364591 2.67 0.008 0.2533075 1.693036 losurfun -0.1320768 0.028785 -4.59 0.000 -0.1889105 -0.07524 _cons 7.465969 0.855281 8.73 0.000 5.77263 9.154675 The plot of the observed log of GDP per capita versus the predicted figures obtained with this regression as well as the plot presenting the residuals is reported in Figure 2 . Figure 2 Prediction of income per capita at the country level (in log). a) Plot of the observed log of GDP per capita versus the predicted ones obtained with the regression in Table 2 b) Plot presenting the residuals versus the log of GDP per capita for the same regression The application of this model results in a significant over estimation of GDP per capita for 10 countries (United republic of Tanzania, Malawi, Zambia, Sao Tome and Principle, Tajikistan, Azerbaijan, Uzbekistan, Kyrgyzstan, Yugoslavia and Egypt) and an under estimation for most of the high income countries presenting a GDP per capita figure higher than 12'500 US$. This can be better visualized by transforming the log of GDP per capita into GDP per capita figures for the 171 countries, creating a new graph (Figure 3 ) similar to the one shown in Figure 2a . Figure 3 Prediction of income per capita at the country level. Plot report the predicted versus the observed GDP per capita for the 171 country level data set using the model reported in Table 2 Figure 3 shows that the development of the lightning infrastructure follows the economic development within a country until reaching a certain level of development after which the model is underestimating income per capita for many countries. This is explained by the fact that from this level, an increment of income is not necessarily reflected in the spatial extension of the outdoor lightning system or the creation of infrastructure that requires specific lightning at night (highways, factories, etc.). There are however some countries for which the model gives a good estimate of GDP per capita, such as Qatar (QAT), United Arab Emirates (ARE); Finland (FIN), Sweden (SWE), Monaco (MCO), Canada (CAN), Norway (NOR). Finally, the model gives a clear overestimation of GDP per capita for the United States of America (USA). The following explanation can be given for these countries: - "over lighting" due to an above-average wealth of the country (e.g. for Qatar, the United Arab Emirates and maybe Monaco), - Two explanations are possible regarding the Nordic European countries that appears in this list (Norway, Sweden and Finland) connected to the fact that the night-time light grid that is used in the context of this work is based on data collected during winter: a) specific climatic conditions during winter requiring excessive lighting of the infrastructure which may be at the origin of their location in the graph (the same observation is likely to be done for Canada). b) snow may be at the origin of an over estimation of income in these countries as snow makes lights look bigger and brighter than they appear in the no-snow condition. - The position of the United States in the graph is related to the fact that the square of the log of the total frequency of light observation is the highest observed in the entire sample. Several hypothesis are proposed in order to explain this situation. This includes the fact that light is more easily spread due to the big habitable surface, electricity is cheap and the road network very wide. The Root Mean Square Error (Root MSE) observed when applying this model is of 4000 US$. 27 countries are presenting an error higher than this value. The 5 countries with the highest error are: Luxembourg, Switzerland, Austria, Australia and Germany. The 5 lowest error are observed for: The Federal State of Micronesia, Burkina Faso, Haiti, Mali and Rwanda. Trying to improve the prediction of GDP per capita at the country level by including other variables we can observe that grouping the countries according to their agricultural level is giving better results than grouping them by climatic type. The figures given by the World Bank [ 15 ] or by the CIA world factbook 2001 [ 16 ] allows a decomposition of the economy into 3 sectors: agriculture, industry and services, the last two being the sectors that are the source of most of the light production. Thus the type of economy within a country has an important effect on the development of the infrastructure and then indirectly on their level of lighting. When the percentage of GPD due to the agricultural sector (Figure 4 ) is used to group the countries there is an improvement in the specific regression for each group. Figure 4 Repartition of the GDP per capita in function of the percentage of GDP due to agriculture In Figure 4 , we can observe a continuous distribution of the points with breaks indicated by the vertical and horizontal lines. These breaks (5, 10 and 25 %) are therefore used for grouping the countries (170 countries used for this analysis as 1 variable is missing for 1 country part of the initial sample) and the same approach than the one described earlier is applied in order to find the regression giving the best prediction for each group as follow: - below 5 % (38 countries): lopopun, lotofre, losurfun (Adj R-squared: 0.5785) - between 5 and 10 % (28 countries): lopopun, lotofre, losurfun (Adj R-squared: 0.6289) - between 10 and 25 % (55 countries): lopopun lotofre2 losurfun (Adj R-squared: 0.5463) - above 25 % (49 countries): lopopun lotofre losurfun (Adj R-squared: 0.5512) We can observe the same combination of variables (lopopun lotofre losurfun) in 3 of the 4 grouping but the coefficients for each of these variables are significantly different. Figure 5 shows the plot of the observed log of GDP per capita versus the predicted figures obtained as well as the plot presenting the residuals when applying these regressions to the respective group while Figure 6 shows the same graph than the one reported in Figure 5a without the logarithmic function. Figure 5 Prediction of income per capita at the country level (grouping by agricultural level in log). a) Plot of the observed versus the predicted log of GDP per capita figures obtained when applying the best model by agricultural level b) Plot presenting the residuals versus the log of GDP per capita for the same regression Figure 6 Prediction of income per capita at the country level (grouping by agricultural level). Plot of the predicted GDP per capita versus the observed one for the 170 countries of the data set when applying the best model found for each of the GDP agricultural contribution group. From the 170 countries used in this analysis, 33 countries are presenting an error higher than the Root MSE (2877 US$). The 5 countries presenting the highest error are: Luxembourg, Australia, Switzerland, New Zealand and Singapore. The 5 lowest error are observed for: Chad, Somalia, Bangladesh, Iran and the Republic of Moldova. The sub-national level analysis The sub-national level analysis has two objectives: to examine the possibility of applying the country level approach to the sub-national level on a country-by-country basis and to explore the possibility of generating global or regional models to be used for countries where sub-national data are either missing or are deemed to be unreliable. This analysis is based on the log of the same parameters successfully used at the country level. The population, total frequency and mean frequency of light observation as well as the surface area figures are extracted from the grids described in the methods section using the boundaries of corresponding sub-national administrative or statistical units. In order to keep the consistency with the values used for country level analysis, the sub-national figures for the population and the total frequency of light observation are adjusted by applying an homogeneous factor and rounding the resulting figures to the closest integer number. Transferring the country level approach to the sub-national level The analysis of the correlation factor existing at the sub-national level between each variables and GDP per capita (expressed in log) shows significant heterogeneity from one country to another. For few countries, some variables even appear as being not significant for the prediction of GDP per capita at the sub-national level. The list of variables observed in the country specific regression giving the best prediction vary therefore also from one country to another. In this case we do not look for models that were based on significant variables but take the one that is presenting the highest Adj R-squared value as the number of observation is generally low. The Adj R-squared obtained for these regressions varies from 0.2492 for The Netherlands to 1.00 for Portugal. This analysis indicates that it may be difficult to find a universal global or regional model that could be applied at the sub-national level. It is nevertheless important to underline the fact that at least one of the light variables is present in all the regression which is not the case for the population or the surface area of the sub-national units. For 3 countries (Italy, the Republic of Korea and the United Kingdom) the high correlation existing between GDP and population seems to indicate that these sub-national figures have been generated using a linear model based on population only. This finding demonstrates that not all reported sub-national data are reliable and emphasizes the need for independent methods for generating sub-national estimates. The sub-national level data for these 3 countries are therefore taken out from the sample for all the analysis reported in this publication. Figure 7 summaries, in a unique graph, the result obtained when applying the country specific regression and converting the results into GDP per capita figures. Figure 7 Prediction of income per capita at the sub-national level (country specific model). Plot of the predicted GDP per capita versus the observed one when applying the sub-national country specific model. Table 3 list the number of units with an error bigger than the country specific Root MSE (also listed) as well as the percentage of units this correspond to. This represents a total of 97 sub-national units (mean value of 12.8 % of unit per country). All the countries presenting a national GDP per capita higher than 15,000 US$ as well as the 3 Latin American countries part of the sample (Argentina, Brazil and Mexico) are at the top of this list. The countries for which the prediction obtained is bellow the Root MSE for all the units are: Bangladesh, Greece, India, Mozambique, Portugal and South Africa. Table 3 Root MSE, number of units outside the 1*RMSE range and percentage of the total number of unit this represents when applying the country specific model Country Root MSE Number of unit outside 1*RMSE % of the number of units DEU 4615 17 48 USA 3546 24 47 ARG 3240 7 29 ESP 2312 4 25 BEL 2384 2 18 BRA 1405 5 18 SVN 1427 2 16 MEX 1777 5 15 IDN 1651 4 14 RUS 2693 11 14 AUT 1405 1 11 THA 2074 9 11 FRA 1605 2 9 NLD 2669 1 8 FIN 1437 1 5 SWE 1264 1 4 CHN 981 1 3 Among these 97 units, 31 are containing at least one city with a population larger than 1,000,000 inhabitants (including some capital cities). Their presence in the list can be explained by the high concentration of buildings which represents vertical highlighted structures for which the satellite sensor is not able to capture the total intensity of the light being produced. This also confirms that in some countries the capital cities are more highly lit than their in country counterparts. Five other units contains oil and/or gas production infrastructure (2 in Indonesia, 2 in the USA and 1 in the Russian Federation). The separation of the lights produced by gas flares from the city lights could be the explanation for the under prediction of income per capita in these units. The difficulty in these cases is to know if the income produced by this activity remains within the concerned unit or goes directly to the government or even outside the country. This observation is to be extended to the offshore infrastructures that should also be considered in the model. Even if no particular characteristics are identified for the remaining 61 units, these results illustrate the possibility to transfer the approach developed at the country level to the sub-national level on a country-by-country basis. This analysis also demonstrate the need to have income figures for some sub-national units in order to find the regression that provides the extrapolated figures for the remaining ones. This concerns units containing big cities (more than 1,000,000 inhabitants) and units containing oil and/or gas production. Additional analysis have to be performed in order to make sure that these are the only cases or if other specificity also have to be taken into account. An out of sample analysis should also be performed in order to determine what is the smallest sample (% of all the units) necessary to obtain a regression providing prediction of acceptable quality. Generating a global or regional model for the prediction of sub-national figures As the availability of disaggregated income data is very poor, and despite the observation made in the previous section regarding the heterogeneity of the country specific sub-national model, tests and analysis are done in order to see if it is possible to generate a model that would allow the generation of sub-national income per capita series for countries where no sub-national figures are available. Four approaches are used: - application of the country level model described in Table 2 - generation of a model based on all the sub-national data (with and without country dummy) - generation of a model grouping the countries by climatic types - generation of a model grouping the countries by agricultural level using the GDP composition by sector The best results being obtained with the last approach, only this one is described in details in this section. For the other approaches only the major findings are reported. Applying the country level regression to the sub-national data set and presenting the results on a summary graph (all the countries on the same graph) gives an acceptable prediction for sub-national units presenting a GDP per capita figure bellow 5,000 US$. Above this limit the application of this approach under estimate GDP per capita. When we generate country specific graphs we can observe that the important variety of income per capita figures at disposal in the sample is in fact at the origin of the observation done earlier on the summary graph and not the result of a good prediction for each country considered in the sample. This emphasizes the need to make country specific graph for analysing the results. This analysis also demonstrates that applying the country level regression to the sub-national level may introduce an important quantitative error. In this regards, the attempt done by Doll et al. in their publication [ 14 ] should only be considered as a qualitative result. When using the entire sub-national sample for generating one unique regression, the combination of variables giving the best results (Adj R-squared = 0.7692) is the same set found for the country level model (Table 2 ). The coefficients observed for each of the variables are also very close to the ones observed for the country level regression (-0.4955732*Lopopun + 0.059074*lotofre2 - 2.959219*lomeanfr + 1.176556*lomeanf2 - 0.1686316*losurfun + 7.774052). These small differences have an impact on the prediction of the income per capita figures for the units presenting a value higher than 10,000 US$ (ppp). The sub-national model gives better result than the country level model in only a few countries: Austria, Brazil, Greece, Portugal, Spain, South Africa, Thailand and the Netherlands. If we integrate a dummy variable in the same regression we can observe one more time the same combination of variables than the one obtained for the country level model (Table 2 ) but the coefficients are slightly different. This alternate approach improves the prediction of the sub-national GDP per capita figures (Adj R-squared = 0.8971) reducing the dispersion of the estimates obtained for the middle and high income countries. Another advantage of this approach is the fact that a good correlation exist between the country specific residual observed when applying the regression in Table 2 at the country level and the country specific constant obtained when applying the sub-national level model including the dummy variable (Figure 8 ). Figure 8 Correlation between the country and the sub-national level prediction. Illustration of the correlation existing between the residual observed at the country level and the country specific constant observed at the sub-national level In theory we could then apply the regression mentioned in Figure 8 on the residual found for a particular country during the country level analysis in order to find the constant to be used for predicting the sub-national GDP per capita figure for the same country. Even if the Adj R-squared for this regression is quite high (0.50) we can observed that Mozambique already represents an outliner indicating that using this correlation may unfortunately also generate important errors in the estimation of the country specific constant. In conclusion, even if this approach is giving better results than the previous ones (application of the country level regression and generation of a unique regression without country dummy) and is presenting an important advantage due to the existing correlation with the country level model we can not consider that the results obtained are of sufficient quality for applying it to other countries where we would not have any sub-national figures. Grouping the countries part of the sub-national data set by climatic types improves the prediction of GDP per capita for Bangladesh, France, India, The Netherlands, The Russian Federation, Sweden and Spain. This improvement mainly concerns sub-national units presenting an observed GDP per capita higher than 10,000 US$ and is not related to a particular climatic type. When the sub-national sample is grouped according to the percentage of GDP associated with agriculture in each country, using the same cut off point than for the country level analysis (Figure 4 ), the following set of countries are obtained: - Below 5 % of GDP due to agriculture: Austria, Belgium, Finland, France, Germany, Portugal, Slovenia, Spain, Sweden, the Netherlands and USA, - Between 5 and 10 %: Argentina, Brazil, Greece, Mexico, Russian Federation and South Africa, - Between 10 and 25 %: China, Indonesia, and Thailand, - More than 25 %: Bangladesh, India and Mozambique. The regression found for each group (based on significant variables except for the country specific constant and the country dummy) are presenting an Adj R-squared varying from 0.5731, for the countries presenting a percentage of GDP due to agriculture between 5 and 10 %, to 0.8013 for the countries with a percentage higher than 25 %. Like for the grouping by climatic type, the list of variables found is different for each group. Figure 9 shows the plot of the observed versus the predicted log of GDP per capita as well as the plot presenting the residuals obtained when applying these regressions. Figure 10 contains the same graph than the Figure 9a when taking out the logarithmic function. Figure 9 Prediction of income per capita at the sub-national level (grouping by agricultural level in log). a) Plot of the observed versus the predicted log of GDP per capita obtained when applying the regression giving the best prediction with the sub-national data set separated into groups based on the percentage of GDP due to agriculture and with a country specific dummy variable b) Plot presenting the residuals versus the log of GDP per capita for the same regression Figure 10 Prediction of income per capita at the sub-national level (grouping by agricultural level). Plot of the predicted versus the observed GDP per capita for the sub-national data set when applying the models grouping the countries using the percentage of GDP due to agriculture including a country dummy. Using this approach, 108 sub-national units are presenting a prediction error bigger than the Root MSE, which represent a mean percentage of units by country of 21 %. The countries for which this grouping induces a significant improvement of the prediction, compare to the models used previously, are: Bangladesh, India, Belgium, Spain, France, the Netherlands, Portugal, Slovenia, Sweden, USA, Finland and the Russian Federation which represents 5 more countries than when using the grouping by climatic types. We can also observe that this improvement does mainly concerns countries were agriculture represent less than 5 % of the country GDP. Only exceptions: Bangladesh, India and the Russian Federation. Even so, 3 of the countries for which the unique regression with dummy variable was not capturing the variability of GDP per capita before (Mexico, Argentina and Greece) are still presenting the same type of results. Due to the small number of observation for each grouping it is not possible to identify if a correlation exists between the residuals observed at the country level and the country specific sub-national constant making it difficult to generate a regional model for these groupings using the same approach than the one described previously (Figure 8 ). Conclusions This study demonstrates that night-time lights data are useful in generating estimates of both national and sub-national GDP per capita figures. Because night-time lights are produced with zero reliance on national reporting data, they provide an independent measure of economic activity. The country level results reported here confirm the conclusion given by Doll et al. [ 14 ] regarding the possibility of expanding the relationship between area lit and GDP to a larger number of countries. But these results raise the question whether the high correlation observed between these two parameters is in fact not coming from the good correlation existing between population and area lit and the one between population and GDP. By introducing parameters other than area lit in the regression we have demonstrated the possibility of independent estimation of GDP per capita at the country level with a high level of confidence. This offers an interesting possibility for completing country level data sets for which data are missing or are in error. Special attention should nevertheless be paid to small densely populated territories with high level of lighting (e.g. Singapore and Monaco). In these areas, the models tend to underestimate GDP per capita as light produced by vertical infrastructures like high rise buildings fails to expand the area of lighting. This effect may be reduced if brightness information on the lights is available. The graph reported in Figure 3 demonstrate that satellite observed area lit and percent frequency of lighting can be successfully used for the prediction of GDP per capita at the country level until a certain limit of economic development above which the relationship breaks down. GDP per capita estimates can be improved by developing models for groups of countries having similar climate or having similar proportions of the country GDP associated with agriculture. Even if the Root MSE observed when using these two models are close to each other our preference goes to the model grouping the countries by agricultural level as it produces fewer outliers. The disadvantage of adding more sub-groupings than the one proposed is that the number of countries in each group becomes small, reducing the strength of the model. The result demonstrate that the approach developed for the country level can also be apply at the sub-national level but only through the generation of country specific models. Figure 11 illustrate the result obtained by the application of the country specific model found for South Africa as well as the distribution of the prediction error expressed in percentages. Figure 11 South Africa, application of the country specific model. a) Distribution of the predicted GPD per capita figure when applying the country specific model b) Distribution of the prediction error when applying the same model In this regard the possibility to generate country specific model for the extrapolation of sub-national income per capita figure is offering an interesting solution for countries where sub-national data on welfare are not available or where the application of other methods (e.g. small area estimation) would be difficult. Nevertheless, the fact that income per capita figures are necessary for some sub-national units in order to generate the country specific model represent a limitation to the application of this approach. If the analysis done in the context of this work already gives an indication of the type of units for which it would be necessary to have a good estimation of the income per capita figures additional work would be necessary in order to confirm this list, maybe add other type of units and also have a better indication regarding the minimum number of units for which input data are needed to insure a good prediction. In addition to that, further analysis are also needed in order to define the level of desegregation to which it would be possible to go based on the developed approach. It will finally be necessary to consider including the amount of light due to gas or oil production infrastructure in the model if the income they are producing stays within the sub-national units where they are located. Despite the caveats, the results obtained clearly demonstrate an interesting potential for making independent estimates of GDP per capita at the sub-national level especially for low income countries where the prediction obtained are of good quality and the need for them definitively the most important. Improving the data sets required to operationalize this approach may be far easier than improving the national reporting of sub-national economic data. It would nevertheless be important to compare the results obtained by this approach with other methods also producing sub national estimated using consumption indicators in the context of poverty mapping exercises [[ 4 ] and [ 17 ]]. In addition to that, the possibility to maybe apply this approach to consumption indicators instead of income per capita figure should be explored as this would then represent an additional method, presenting the advantage of being less sophisticated, for the generation of desegregated country specific poverty maps. An other advantage would be that this approach would then not only be applicable to data collected in the context of the World Bank Living Standards Measurement Study (LSMS) [ 9 ] but also to the ones collected for example in the context of the WHO World Health Survey [ 18 ]. Even if this new instrument is presenting an additional advantage, by also collecting health indicators at the household level, work has to be done in order to confirm that the data collected are sub-nationally representative. Trying to generate a consistent global or regional poverty map desegregated to some sub-national level, the grouping of the sample at disposal by climate or percentage of GPD due to agricultural level is definitively improving the prediction but does not provide us with a model generating consistent estimates for all the countries. Between these two grouping the preference again goes to the second one which improves the estimation for the middle and high income countries. The result obtained when applying this model to South Africa as well as the distribution of the prediction error expressed in percentages are reported in Figure 12 as an example. Despite being the best regional model analysed in the context of this research the grouping by agricultural level is still producing significant error when applying it on a country by country basis (Figure 12b ) compare to the country specific model itself (Figure 11b ). Figure 12 South Africa, application of the model grouping the countries by agricultural level. a) Distribution of the predicted GPD per capita figure when applying the model grouping the countries by agricultural level. b) Distribution of the prediction error when applying the same model (see Figure 11 for the legend). In this regards, the analysis performed is not offering us the possibility to generate consistent global map showing the sub-national distribution of income per capita figures. These type of exercises are sorely needed to improve our knowledge regarding the health and well being of people in the poorest areas of the world. In this context night-time lights remain a useful data set for the evaluation of the impact of international efforts to improve the economic and therefore health conditions of these populations. It would therefore be important to pursue the type of work described in this paper and to see if the results obtained could not be improved. For both the country and the sub-national level model, the combination of the two groupings used in the context of this present work could for example improve the proposed models but this would have to be confirmed through additional analysis. The more recent global night-time light mosaic that NOAA has compiled is another element which might improve the results presented in the context of this work. This new global mosaic has the advantage of covering a longer period of observation (full years instead of the six month composite used in this study) which would improve the homogeneity of the distribution of the number of observation per pixel and also reduce the effect of the snow observed on images collected during winter. These new products were processed with major improvements in the exclusion of all but the highest quality segments of the individual orbits. They also include one information that was missing in the data set used for the context of this work: the digital number brightness of the lights. It has been found that adding in brightness of the lighting greatly improves the relationship to variables such as electric power consumption and GDP. Even if some difficulties to make products with brightness values remains, due to saturation in urban centers and the lack of on-board calibration, this new generation of grids could allow a significant improvement of the models described in the present paper. Another advantage of this data set is that the NGDC is producing a full global composite for each year from 1992 through 2004 allowing therefore for trend analysis. Before being able to test such new source of spatially distributed information it would be important to correctly address the question of the lack of documentation regarding the methods used for generating sub-national income estimates reported by individual countries in case this indicator would provide better results than consumption when trying to generate a consistent global poverty map. Inclusion of erroneous data may lead to misleading interpretations (see the case of the data for the United Kingdom, Italy and the Republic of Korea in the sub-national level analysis section). Such discrepancies also underline the value of high quality geospatial data for use in making independent estimates of economic activity. This for example includes standardization of the basic vector GIS layers (national and sub national borders for example) in order to insure a proper use of the information collected at the sub national level or stored in different raster layers (night-time light and population for the present work). Among the different initiatives that are trying to answer the need for standardization in this area we can mention the Second Administrative Level Boundaries data set project (SALB). The first objective of this project is to create a redistributable Second Administrative Level Boundaries global data set (SALB) representative of January 2000 to be used with the GIS technology. The information finally collected has extended the period of representativity of the database for finally covering the period 1990-present. A process has also been put in place in order to insure the updates of the database in the future. You can find all the relevant information about this project as well as the data already available on the project web site [ 19 ]. The growing use of this database and more specifically of its specific coding scheme should improve the availability and comparability of sub national income figures in the future. Finally, the shifts observed between all these layers of geographical distributed information is emphasizing the need for defining a "master" that could be used as a ground reference when generating or working with them. If the precision of its georeferencing is confirmed, the global mosaic of satellite images that are now publicly available (e.g the Landsat mosaic) could for example constitute this master. Methods In order to perform the analysis at the country and at the sub-national level it was necessary to compile existing data sets on income as well as geographically distributed parameters that would be used to model these figures using a GIS. These concerns: - night-time light imagery - population - international and sub-national boundaries - surface area - climate The data sets used for the context of this paper are presented now. In order to insure the consistency of the analysis, all the data sets compiled or created were adjusted to 1995, which is the year of representativity of the only night-time light grid available at the beginning of this research. We utilized GDP per capita data that have been collected at both the country and sub-national level. For the country level, GDP in International Dollars (GDP I$) for 171 countries have been calculated for the year 1996 from GDP figures expressed in local currency unit using price level data [ 20 ] and were adjusted to be representative of 1995. For the sub-national level, GDP figures representative of the first or second administrative or statistical level have been collected for 26 countries (653 units) from the 5 continents (Table 4 ). These figures were homogenised in order to obtain a final data set expressed in Power Purchasing Parity (PPP) representative of 1995. A country specific adjustment factor has been applied to these figures in order to keep the consistency with the country figures before dividing them by their corresponding population. Table 4 Source of the income figures for the sub-national analysis Country ISO 3 Code Admin-Stat./Level/Nb unit Representativity Source of the data Income level Argentina ARG Adm/1/24 1995 National statistics agency, Argentina Middle Austria AUT Adm/1/9 1994 Sozialstatistic Austria High Bangladesh BGD Admin/2/19 1993 Statistical Yearbook of Bangladesh, 1994 Low Belgium BEL Admin/2/11 1994 MaconUSA High Brazil BRA Admin/1/27 1997 IBGE; Brazil Middle China CHN Admin/1/30 1998 Statistical Yearbook of China, 1998 Low Finland FIN Stat/3/19 1997 Statistics Finland High France FRA Stat/2/22 1994 MaconUSA High Germany DEU Stat/2/35 1994 MaconUSA High Greece GRC Admin/1/13 1994 Statistical Office of Greece High India IND Admin/1/25 1991 Statistical Outline of India 1999–2000 Low Indonesia IDN Admin/1/27 1994 Statistical Information Services, Indonesia Middle Italy ITA Stat/2/20 1994 MaconUSA High Mexico MEX Admin/1/32 1995 INEGI, Mexico Middle Mozambique MOZ Admin/1/10 1997 Instituto Nacional de estatistica, Mozambique Low Netherlands NLD Stat/2/12 1994 MaconUSA High Portugal PRT Stat/2/7 1994 MaconUSA High Russian Federation RUS Admin/1/77 1996 State Com. Of the Russian Federation on Statistics Middle Slovenia SVN Stat/2/12 1996 Statistics Slovenia Middle South Africa ZAF Admin/1/9 1994 Statistics South Africa Middle Republic of Korea KOR Admin/1/14 1995 National Statistical Office of Korea High Spain ESP Stat/2/16 1994 MaconUSA High Sweden SWE Stat/3/21 1996 Statistics Sweden High Thailand THA Admin/1/76 1995 Chulalongkorn University, Bangkok Middle United Kingdom GBR Stat/2/35 1994 MaconUSA High United States of America USA Admin/1/51 1997 Harvard University, USA High The night-time light grid used has been provided by NOAA's National Geophysical Data Center (NGDC). This grid data set is the result of a 6-month 1 km resolution composite based on images collected between October 1994 and March 1995 by the U.S. Air Force Defence Meteorological Satellite Program (DMSP) Operational Linescan System (OLS) [ 10 ]. Only the grid with the distribution of lights associated with human settlements has been used in the context of the present research. Due to improvements in the algorithms used during the processing, this grid is different than the one used by Doll et al. for their publication [ 14 ]. By comparing these two grids it has been possible to identify some of these differences and to take advantage of their respective specificity for generating the grid used here. From this grid it possible to extract 4 parameters connected to light observation at night which are capturing a different information making it possible to extract the figures at the country or sub-national level for the analysis: - the number of cells highlighted at night - the area lit (surface area being highlighted at night) which is giving an indication of the extension of the highlighted surfaces, - the total frequency of light observation (obtained by adding the percents frequency of light detection observed in each cell on a given surface) which is giving an indication of the total intensity of the highlighting - finally, the mean frequency of light observation, in the highlighted areas, which gives us an indication of the dispersion of this intensity. For example: one pixel with a 100 % of light observation or 100 pixels with 1 % of light observation are giving the same value for the total frequency of light observation but a totally different figure for the mean frequency of light observation and area lit. These information being stored in a grid it is possible to extract them at the country or sub-national level for the analysis. Regarding population, the country level data that have been used are the UN population figures for the year 1995 [ 21 ]. For the sub-national level the Gridded Population of the World (GPW) version 2 has been selected [ 22 ] as being consistent with the UN country level data set. In addition to that, this data set is offering the possibility to use GIS in order to make spatial analysis at the sub-national level. Climate is known to have an important influence on many elements on the earth surface, including human behaviour and well being [ 23 ], and also on the need for specific lighting of infrastructure. The Köppen climate classification distribution grid derived by the Food and Agriculture Organization (FAO) from the International Institute for Applied Systems Analysis (IIASA) data sets has therefore been used for the context of this project [ 24 ]. Using this grid it is possible to determine the general climate of any geographical entity (national or sub-national). Köppen categories are based on the annual and monthly averages of temperatures and precipitation. Five major climatic types are recognized in this system, each type being designated by a capital letter (each of them being divided into sub types): A – Tropical moist climates: all months have an average temperature above 18 degrees celsius. B – Dry climates: with deficient precipitation during most of the year C – Moist mid-latitude climates with mild winters D – Moist mid-latitude climates with cold winters E – Polar climates: with extremely cold winters and summers The delimitation of the units of analysis (countries and sub-national units), corresponding to the GDP figures collected for the context of this study, have also been prepared in a format that could be used in the GIS tool. In order to insure the consistency from one country to another the delimitation of the international borders has been adjusted to 1995. Finally, the surface area of the units of analysis has also been included in the data set. Two types of software have been used in analysing national and sub-national income: a Geographic Information System (GIS) and a statistical package. In order to take advantage of the specificity of the two modes of representation observed in the compiled data set (vector and raster) we have worked with the ArcView GIS software (version 3.2) for the vector part complemented by the Spatial Analyst extension (version 1.1) for the raster part. As no statistical analysis could be done directly in the GIS software the statistical support was provided by the STATA software (Version 5). | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549533.xml |
549527 | Correlation of sperm penetration assay score with polyspermy rate in in-vitro fertilization | Background The sperm penetration assay (SPA) is used to predict the fertilizing capacity of sperm. Thus, some programs rely on SPA scores to formulate insemination plans in conjunction with in-vitro fertilization (IVF) cycles. The purpose of this study was to evaluate if a relationship exists between SPA scores and polyspermy rates during conventional IVF cycles. Methods A total of 1350 consecutive IVF patients using conventional IVF insemination were evaluated in the study. Oocytes were inseminated three hours post-retrieval by the addition of 150,000 to 300,000 progressively motile sperm. Approximately 18 hours after insemination, the oocytes were evaluated for fertilization by the visualization of pronuclei. The presence of three or more pronuclei was indicative of polyspermy. Polyspermy rates, fertilization success, embryo quality, and pregnancy rates were analyzed retrospectively to evaluate their relationship with SPA score, count, motility, number of progressively motile sperm inseminated, oocyte pre-insemination incubation time, patient age, and diagnosis. Results A significant positive relationship was observed between SPA score and polyspermy rate (r s = 0.10, p < 0.05). Patients with a normal SPA score had significantly higher polyspermy rates than those with abnormal SPA scores (6.3% ± 1.5% vs. 2.0% ± 0.7%, p < 0.05). Fertilization percentage was significantly lower in the group with severely abnormal SPA scores versus all other SPA groups (57.5% ± 2.1% vs. 70.2% ± 1.3%, p < 0.005). Although embryo quality was not affected, both clinical pregnancy and implantation rates improved slightly as SPA score increased. In addition, there was a decrease in the rate of spontaneous abortion as SPA score increased. Conclusions These data indicate SPA score is positively correlated with polyspermy rates and IVF fertilization percentage. Additionally, there is a slight increase in clinical pregnancy rates, and embryo implantation rates with increased SPA. Furthermore, there is a slight decrease in spontaneous abortions rates related to increased SPA. | Findings The sperm penetration assay (SPA) is used to evaluate the fertilizing capacity of human spermatozoa [ 1 ]. Some in-vitro fertilization (IVF) programs rely on the SPA to formulate their insemination plans in conjunction with IVF cycles [ 1 , 2 ]. Couples with normal SPA scores normally undergo conventional IVF insemination in contrast with intracytoplasmic sperm injection (ICSI), which is used for those with a diminished SPA. Over the past decades success rates have steadily improved for human IVF embryo transfer programs. Recent advances in ovulation induction, micromanipulation, culturing conditions, and media formulations have fostered a technological revolution for IVF leading into the 21'st century (Pool, 2002). However, many common problems still remain. One common problem associated with conventional IVF insemination is polyspermic fertilization when more than one sperm successfully penetrates and fertilizes the oocyte. These pre-embryos are considered abnormal and typically discarded. Thus, increased rates of polyspermy ultimately result in a reduced number of embryos. Studies have been conducted that evaluate the factors associated with human IVF polyspermy [ 3 - 5 ]. However, most of the reports are based on data over a decade old. Recent improvements in IVF techniques may alter these relationships. Furthermore, no study has evaluated the relationship between sperm SPA score and incidence of polyspermy in human IVF. The purpose of this study was to evaluate if a relationship exists between SPA scores and polyspermy rates during IVF cycles utilizing conventional insemination. We also evaluated the relationship between SPA score and fertilization rate, embryo quality, and pregnancy rate. After Institutional Review Board approval, we conducted a retrospective study of 1350 consecutive IVF patients using conventional IVF insemination. Intracytoplasmic sperm injection (ICSI) patients were excluded from the study. The fertilizing capacity of each patient was assessed with the SPA. The SPA was performed on patients undergoing IVF using techniques previously described [ 6 ]. The SPA was performed on patients less than sixty days prior to the IVF cycle. The SPA score reflects the percentage of eggs successfully penetrated by the patient's sperm. SPA patients were stratified into 4 groups according to SPA score (Group 1: severely abnormal SPA (< 10% penetration), n = 182; Group 2: abnormal SPA (10–19% penetration), n = 368; Group 3: normal SPA (20 – 29% penetration), n = 404; Group 4: high-normal SPA (> 30% penetration), n = 396). During the IVF cycle ovarian stimulation was performed using standard techniques of gonadotrophin-releasing hormone (GnRH) agonist down-regulation combined with controlled ovarian stimulation using a combination of recombinant follicle stimulating hormone (rFSH) and urinary-derived gonadotropin stimulation. Ovarian follicles were aspirated using a trans-vaginal ultrasound-guided needle. Oocytes were inseminated three-hours post-retrieval by the addition of 150,000 to 300,000 progressively motile sperm. Approximately 18-hours post-insemination, the oocytes were evaluated for fertilization by visualization of pronuclei. The presence of three or more pronuclei was indicative of polyspermy in which case the fertilized oocyte was discarded. Embryos were cultured in HTF medium and transferred 72-hours post-retrieval. Embryo quality was assessed using a previously reported embryo scoring system that took into account the number of cells present and the level of cellular fragmentation [ 7 ]. Polyspermy rates, fertilization success, embryo quality, and pregnancy rates were analyzed retrospectively with respect to their relationship with SPA score. The correlation between SPA score and polyspermy was conducted using Spearman's correlation coefficient. Additionally, polyspermy rates, fertilization percentage, and embryo quality for the different SPA groups were analyzed for statistical difference using Kruskal-Wallis analysis. Lastly, pregnancy rates in the different groups were evaluated with a Chi-square analysis. A significant positive relationship was observed between SPA score and polyspermy rates in 1350 IVF patients undergoing conventional IVF insemination (r s = 0.10, p < 0.05). Patients with normal SPA scores (Group 3) had significantly elevated polyspermy rates over those with severely abnormal SPA scores (6.3 ± 1.5 – 113/1,791 vs. 2.0 ± 0.7 – 20/1,020, p < 0.05, Table 1, please see file "Table1-Aoki et al.doc", Additional file 1 ). Conventional IVF fertilization rate was significantly lower in patients with severely abnormal SPA (57.5 ± 2.1 – 1,135/1,974) vs. all other patients (70.2 ± 1.3 – 8,519/12,135; p < 0.005, Table 1, please see file "Table1-Aoki et al.doc", Additional file 1 ). Embryo quality showed no relationship to the SPA (Table 1, please see file "Table1-Aoki et al.doc" Additional file 1 ). Clinical pregnancy and implantation rates were increased in Group 4 (50.8% – 202/396 and 27.4% – 343/1255, respectively) versus Group 1 (31.6% – 58/182 and 15.7% – 96/612, respectively, p < 0.05) but no significant relationship between SPA and IVF outcome was observed throughout the entire study population (Table 1, please see file "Table1-Aoki et al.doc" Additional file 1 ). Similarly, there was a decrease in the rate of spontaneous abortion in Group 4 (14.7% – 30/202) versus Group 1 (31.3% – 18/58), although no significant relationship was observed within the entire population (Table 1, please see file "Table1-Aoki et al.doc" Additional file 1 ). No relationships were observed between IVF polyspermy rates and maternal age, maternal diagnosis, amount of progressively motile sperm added, sperm morphology, ovarian stimulation protocol, or post-oocyte retrieval pre-insemination incubation time. Additionally, confounder analysis indicated the SPA groups were similar with respect to paternal age, maternal age and IVF diagnosis. Therefore, no standardization of the data with respect to these variables was required. This is the first report to correlate SPA score with IVF polyspermy rates. Significantly elevated IVF polyspermy rates were observed in patients with normal SPA scores. Previous reports have established a relationship between IVF polyspermy and oocyte age/maturation, state of the zona pellucida, number of progressively motile sperm used for insemination, oviductal cytokine expression, and in-vitro culture conditions such as pH, temperature, and media supplementation [ 5 ]. Unlike other reports, we did not find a relationship between the number of progressively motile sperm used for insemination and IVF polyspermy rates. This discrepancy is most likely due to the narrow range of sperm number we used for insemination (150,000 to 300,000) compared with the studies of van der Ven et al. who used 500,000 to 1.5 *10 6 sperm [ 4 ]. Our results are consistent with Ho et al. who found no relationship between these variables [ 3 ]. However, the data still suggest minimizing the number of sperm added for IVF insemination in patients with increased SPA scores. Poly-pronuclear formation in in-vitro fertilized eggs usually arises from polyspermic fertilization but may also be a product of second polar body retention [ 8 ]. Thus, one potential pitfall of associating poly-pronuclear fertilized oocytes with polyspermy is the uncertainty related to second polar body retention. However, a recent report showed only a small percentage of fertilized oocytes with three-pronuclei (2.5%) arise from second polar body retention. The findings of this study validate our study design of categorizing poly-pronuclear fertilized oocytes as polyspermic. Moreover, we would expect the contribution of second polar body retention to be consistent throughout the SPA groups negating any confounding effects. Consistent with other reports, the SPA was correlated with fertilization percentage [ 1 , 6 ]. Furthermore, clinical pregnancy and implantation rates did show improvement in patients with a high-normal SPA versus those with a severely abnormal SPA. In addition, spontaneous abortion rates were significantly different between these two groups. This data strongly suggests ICSI may be an attractive alternative for patients with severely affected penetration ability. Done correctly, the SPA provides a reliable assessment of the fertilizing ability of human spermatozoa with very low false negative rates (< 0.03%) and serves a valuable tool for clinicians to treat infertility patients with the appropriate modality [ 9 ]. The SPA is particularly useful in light of recent concerns about ICSI and imprinting diseases [ 10 - 12 ]. These concerns have led to the recommendation that IVF clinicians should be careful to employ the technique only when necessary and the SPA provides a prognostic tool to appropriately make that decision. Based on these findings, it may be valuable for individual IVF programs to re-evaluate the clinical utility of SPA testing. The SPA may prove valuable for clinicians to avoid increased amounts of polyspermy and unnecessary wasting of oocytes. Meanwhile, the SPA does not appear to be a reliable indicator of IVF embryo quality or pregnancy rates. Authors' Contributions VWA designed the study, carried out statistical evaluation, and was the primary author of the manuscript. CMP, KPJ, HHH, MG, and IH were involved in management of the IVF cases and collection of the clinical data. DTC was responsible for the implementation of the study, administrative requirements including IRB approval, statistical evaluation of the data, and preparation of the manuscript. Supplementary Material Additional File 1 Table 1. Relationship between SPA and IVF outcome measures Click here for file | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC549527.xml |
553966 | Expression analysis of the mouse S100A7/psoriasin gene in skin inflammation and mammary tumorigenesis | Background The human psoriasin (S100A7) gene has been implicated in inflammation and tumor progression. Implementation of a mouse model would facilitate further investigation of its function, however little is known of the murine psoriasin gene. In this study we have cloned the cDNA and characterized the expression of the potential murine ortholog of human S100A7/psoriasin in skin inflammation and mammary tumorigenesis. Methods On the basis of chromosomal location, phylogenetic analysis, amino acid sequence similarity, conservation of a putative Jab1-binding motif, and similarities of the patterns of mouse S100A7/psoriasin gene expression (measured by RT-PCR and in-situ hybridization) with those of human S100A7/psoriasin, we propose that mouse S100A7/psoriasin is the murine ortholog of human psoriasin/S100A7. Results Although mouse S100A7/psoriasin is poorly conserved relative to other S100 family members, its pattern of expression parallels that of the human psoriasin gene. In murine skin S100A7/psoriasin was significantly upregulated in relation to inflammation. In murine mammary gland expression is also upregulated in mammary tumors, where it is localized to areas of squamous differentiation. This mirrors the context of expression in human tumor types where both squamous and glandular differentiation occur, including cervical and lung carcinomas. Additionally, mouse S100A7/psoriasin possesses a putative Jab1 binding motif that mediates many downstream functions of the human S100A7 gene. Conclusion These observations and results support the hypothesis that the mouse S100A7 gene is structurally and functionally similar to human S100A7 and may offer a relevant model system for studying its normal biological function and putative role in tumor progression. | Background Human S100A7 or psoriasin was first identified as an over-expressed secreted protein in psoriatic skin[ 1 ]. More recently its expression in both pre-invasive (ductal carcinoma in situ , DCIS) and invasive human breast cancer was demonstrated [ 2 ]. Although highly expressed in DCIS and generally down-regulated in invasive breast cancer, the expression of psoriasin/S100A7 in both in-situ and invasive breast cancer is correlated with markers of poor prognosis [ 3 , 4 ] and in invasive carcinoma also with poor clinical outcome [ 5 ]. Support for psoriasin/S100A7 having a functional role in this aggressive phenotype is shown by the observation of increased growth and tumorigenesis when breast cancer cells over-expressing psoriasin/S100A7 are grown as xenografts in nude mice [ 6 ]. This activity may in part be mediated by the ability of psoriasin/S100A7 to interact with c-Jun activation domain-binding protein 1 (Jab1) [ 6 ] and enhance pro-survival pathways [ 7 ] and protect against anoikis in human breast and head and neck squamous cancer cells [ 8 ]. However, further investigation of the physiological and pathophysiological function of psoriasin/S100A7 can only be effectively undertaken using genetic manipulation of an animal model. To undertake such studies we need to characterize and clone the mouse psoriasin gene. Interestingly, during the preparation of this manuscript Marenholz etal. , [ 9 ] have argued that the possible ancestral homolog of the human S100A7 paralogs [ 10 ] has been identified in the mouse on the basis of genome sequence analysis and gene orientation, but has not been characterized fully as yet. The following study provides evidence of the cloning of a cDNA of the putative mouse ortholog of human psoriasin/S100A7, investigates its expression under conditions of mouse mammary tumorigenesis and skin inflammation, and confirms that the factors involved in its regulation are comparable to those involved in regulation of the human counterpart. Methods Tissues Murine studies were conducted in accordance with the principles and procedures recommended and approved by the University of Manitoba Animal Care Review Board. Mammary tumors in CD1 mice were generated chemically using 7, 12-dimethylbenz anthracine (DMBA) as previously described [ 11 ]. Freshly dissected mouse tissues were either stored frozen at -70°C, or processed to generate formalin fixed paraffin-embedded tissue blocks. Acute dermatitis in C57/B6 mice was induced by the topical application of 20% croton oil (dissolved in dimethyl sulfoxide) to a 1 cm length mid-tail portion [ 12 ]. The tail skins were stimulated continuously with croton oil every 4 hours for 24 hours. Mice were divided into five time-groups (each containing 3 mice) that were designated 0, 4, 8, 16, and 24 hours. At each time-point, the mice were sacrificed. The tail-skins were harvested and paired specimens of normal and inflamed skin tissue were fixed in 3.7% formaldehyde in 0.1 M phosphate buffered saline for 16–18 hours, followed by paraffin embedding. Sections from the above blocks were used for preparation of haematoxylin and eosin stained sections for light microscopic examination, in situ hybridization (ISH), and immunohistochemistry (IHC). Human tumor tissues were obtained from the Department of Pathology, University of Manitoba. All cases were coded and therefore anonymous and prior approval was obtained from the University of Manitoba Research Ethics Board and the Pathology Access to Tissue Committee. Different cervical tumor pathologies (total n = 39 cases) including in-situ adenocarcinoma (n = 10) and squamous cell carcinoma (n = 9), as well as invasive adenocarcinoma (n = 10) and squamous carcinoma (n = 10) were selected for one cohort study. In addition, different invasive lung tumor types (total n = 78 cases) including mesothelioma (n = 10), small cell carcinoma (n = 15), adenocarcinoma (n = 28), and squamous carcinoma (n = 25) were selected to form another study cohort. Squamous differentiation within both murine and human tumors was determined by standard morphological criteria including cytoplasmic keratinization and cellular stratification relative to keratin pearls. In-situ hybridization Paraffin embedded 5 μm tissue sections were analyzed by in-situ hybridization according to a previously described protocol [ 13 ]. The plasmid pCR4-TOPO-mPsor-ORF, consisted of pCR4-TOPO plasmid (Invitrogen Canada Inc, Burlington, ON) containing a 344 base pair insert of the mouse psoriasin cDNA, also known as mouse S100A15 (from nucleotide 94 to 437 as numbered in AY465109, and from 1 to 344 as numbered in AY582964). One microgram of linearized template DNA was used to generate 35 S-UTP-labeled sense and antisense cRNA probes using the Riboprobe System (Promega, Madison, WI) according to the manufacturer's instructions. Sense probes were used as controls. In-situ hybridization and washing conditions were as previously described[ 2 ]. Sections were developed using Kodak NTB-2 photographic emulsion and counter-stained with Lee's stain after 2–6 weeks. Levels of mouse S100A7/psoriasin RNA expression were assessed by microscopic examination at low power magnification and with reference to the negative sense control. This was done by scoring the estimated average signal intensity (on a scale of 0 to 3), where 0 is no expression and 3 is a high proportion of strong focal expression. Immunohistochemistry Immunohistochemistry (IHC) was performed on serial 5 μm sections from a representative, formalin fixed paraffin embedded tissue block from each tumor. For human tumor blocks, psoriasin/S100A7 IHC was performed essentially as described [ 5 ] and human psoriasin/S100A7 was detected using a previsouly characterized rabbit polyclonal antibody [ 3 , 5 ]. For murine tumors estrogen receptor-alpha (ERα) was detected using an affinity purified rabbit polyclonal antibody, MC-20, raised against a C-terminal peptide of mouse ERα (#sc-542, Santa Cruz Biotechnology Inc, CA). Antibodies were applied using an automated tissue immunostainer (Discovery module, Ventana Medical System), 3, 3-diaminobenzidine IHC kit and bulk reagents were supplied by the manufacturer. Briefly, the Discovery staining protocol was set to "Standard Cell Conditioning", followed by 60 minutes incubation at 42°C with primary antibody and 30 minutes incubation at 42°C with secondary antibody (goat anti-rabbit-IgG-HRP, Jackson Immuno Research Labs Inc). Primary antibody concentrations initially applied to the Ventana instrument were 1:200 for ERα and 1:200 for the secondary antibody translating into final dilutions of 1:600 after 1:3 dilution with buffer dispensed onto the slide with the primary antibody. Slides were counterstained with hematoxylin. Levels of psoriasin/S100A7 and ERα expression were scored semi-quantitatively in tissue sections, under the light microscope. Scores were obtained by estimating average signal intensity (scale of 0 to 3) and the proportion of epithelial cells showing a positive signal (0–100%). The intensity and proportion scores were then multiplied to give an overall IHC-score. RNA extraction and reverse transcription Total mouse RNA was extracted using Trizol™ reagent (Invitrogen) according to the manufacturer's instructions, and the integrity of the RNA was confirmed by denaturing gel electrophoresis as previously described [ 14 ]. RNAs from the various frozen tissues were reverse transcribed. One μg of total RNA was reverse transcribed in a final volume of 30 μl composed of 50 mM Tris-HC1 (pH 8.3), 75 mM KC1, 3 mM MgC1 2 , 0.5 μM random hexamers (Invitrogen) 0.5 mM dNTPs, 0.01 mM DTT in the presence of 300 units of MMLV-RT (Invitrogen), and 4 units RNase inhibitor at 37°C for 1 hour, followed by 5 minutes at 95°C and kept at -20°C until used. PCR conditions The primer pairs used were as follows; Mouse S100A7/psoriasin C-terminus 5'-ATG CCA GAC ACA CCA GTG GAG-3' (sense; nucleotides 111–131 in GenBank acc. AY465109) and 5'-GGT AGT CCT TCA CCA GCT TGC-3' (antisense; nucleotides 358–378). Mouse S100A7/psoriasin open reading frame 5'-TGA AGG GTC CAT CAG TCA-3' (sense; nucleotides 94–111 in GenBank acc. AY465109) and 5'-CTA GTA GAG GCT GTG CT-3' (antisense; nucleotides 421–437). Mouse β-actin Primers were designed according to the mRNA sequence (GenBank acc. NM_007393): 5'-TCT ACG AGG GCT ATG CTC TCC-3' (sense; nucleotides 574–594) and 5'-GGA TGC CAC AGG ATT CCA TAC-3' (antisense; nucleotides 883–903). According to the chromosome 5 genomic contig sequence (GenBank acc. NT_039324), these primers span an 87-bp intron with the antisense primer binding across the intron-exon boundary. PCR reactions were performed essentially as previously described [ 3 ]. To amplify cDNA corresponding to mouse S100A7/psoriasin, an initial 2 minutes at 94°C was followed by 36 cycles (30 seconds at 94°C, 30 seconds at 56°C, 30 seconds at 72°C). Twenty six cycles were used to amplify β-actin cDNA (30 seconds at 94°C, 30 seconds at 58°C, 30 seconds at 72°C). PCR products were separated on 1.5% agarose gels containing ethidium bromide (0.1 μg/ml) as previously described [ 13 ]. Identity of the 344 bp product corresponding to mouse S100A7/psoriasin and the 330 bp product corresponding to β-actin were confirmed by subcloning and sequencing as described previously [ 13 ]. Semi-quantitative PCR analyses were performed using three independent PCRs for each sample for both mouse S100A7/psoriasin and β-actin. Signals visualized with UV irradiation on a GelDoc2000/ChemiDoc System (Bio-Rad), were quantified by densitometry using the Quantity One software (version 4.2; Bio-Rad). Mouse S100A7/psoriasin expression was standardized to β-actin expression assessed from the same cDNA in separate PCR reactions and run in parallel on separate gels. The standardized mean of each triplicate PCR was then expressed relative to the levels in a "moderately expressing" sample selected for each batch of cDNAs to be analyzed. Phylogenetic analysis Amino acid sequences of S100 family genes were aligned using Clustal X [ 15 ]. This alignment was used to construct a phylogenetic tree based on a Poisson corrected neighbour-joining distance method [ 16 ] available in the computer software package MEGA v3.0 [ 17 ]. The reliability of the phylogeny's interior branches was tested by a bootstrap test with 1000 replications [ 18 ]. The human sequences used, were GenBank Accession numbers: AAH05019 (S100A14), NP_789793 (5100A15), NP_006262 (S100A1), NP_005969 (S100A2), NP_002951 (S100A3), AAH00838 (S100A4), NP_002953 (S100A5), AAH09017 (S100A6), AAH34687 (S100A7), XP_060509 (S100A7L-2), AAH05928 (S100A8), AAH47681 (S100A9), AAH15973 (S100A10), AAH14354 (S100A11), NP_005612 (S100A12), NP_002952 (S100A13). The mouse sequences used, were GenBank Accession numbers: NP_035439 (S100A1), NP_035440 (S100A3), NP_035441 (S100A4), NP_035442 (S100A5), AAH03832 (S100A6), NP_955454 (S100A7/15), NP_038678 (S100A8), AAH27635 (S100A9), AAH25044 (S100A10), AAH21916 (S100A11), NP_033139 (S100A13), AAH25607 (S100A14). Results Identification and cloning of mouse S100A7/psoriasin cDNA A BLASTP search of mouse sequence databases was performed using the human psoriasin (S100A7) amino acid sequence and identified a mouse sequence highly similar to human S100A7, which had been previously predicted by automated computational analysis (GenBank acc. XM_143311, later replaced by AY465109). Using this predicted sequence in a BLASTN search, a mouse skin expressed sequence tag (EST) (GenBank acc. AA792680) containing a highly similar sequence was identified. A pair of mouse "psoriasin"-specific PCR primers was designed that would only amplify a sequence of mouse psoriasin contained within the EST. According to the recently posted genomic sequence (GenBank acc. AY465110) these primers span a 1992-bp intron, thus the potential amplification of contaminating genomic DNA is minimized (Figure 1 ). An additional set of mouse psoriasin primers ("psoriasin-ORF") was later designed to amplify the entire predicted open reading frame (Figure 1 ). These primers also span the 1992-bp intron. Both sets of putative mouse psoriasin primers were used for RT-PCR and semi-quantitatively measure the potential expression of the putative mouse psoriasin in mouse skin, mammary gland and mouse mammary tumors. These tissues were chosen because the human psoriasin/S100A7 was described as differentially expressed in skin and breast. The results, using primers that amplified the entire predicted ORF of the putative mouse psoriasin, are shown in Figure 2 , with β-actin as a control in parallel. A 344 bp PCR product was amplified from cDNA derived from normal mouse skin and mammary gland using mouse psoriasin primers (see above for description). Based on our PCR results, mouse psoriasin mRNA is expressed in skin as well as in mammary gland where it is more highly expressed in mammary tumors than in normal mammary gland (Figure 2a ). The psoriasin RT-PCR product and the 330 bp product generated using the mouse β-actin primers were both cloned and sequenced to confirm their identities. The nucleotide sequence of the putative mouse psoriasin RT-PCR product (GenBank acc. AY582964) was 100% identical to the coding sequence of the putative mRNA in the GenBank sequence database (GenBank acc. XM_143311). Start/stop codons, PCR primer binding sites, and intron/exon boundaries are indicated in Figure 1 . However, during the course of this study the predicted mRNA sequence of GenBank acc. XM_143311, named as 'similar to the S100 calcium-binding protein A7 (psoriasin)' was removed and apparently replaced by another entry GenBank acc. XM_356221 for a predicted mRNA that was named as 'similar to the S100 calcium-binding proteins A15'. GenBank acc. AY465109 then appeared as a cDNA cloned from mouse skin and was named as 'Mus musculus S100 calcium-binding protein A15 mRNA'. The amino acid sequences of the various Genbank entries believed to describe the mouse psoriasin are aligned in Figure 3 to illustrate that the amino acid sequences of the ORF from each entry are identical. From here on, we will refer to this gene as mouse S100A7/ psoriasin. In addition, according to the recently posted genomic sequence (GenBank acc. AY465110 and NT_078386), the mouse S100A7/psoriasin gene maps within the murine S100 cluster on chromosome 3 (LOC381493, NCBI Map Viewer Link), in a region similar to the human S100 cluster on chromosome 1q21 [ 10 , 19 ]. Our sequence database searches and RT-PCR experiments demonstrate a putative mouse ortholog of the human psoriasin gene exists and like human psoriasin, it is expressed in skin and mammary tissue. Phylogenetic analysis The putative mouse psoriasin gene (originally "Similar to S100A7") has been renamed "mouse S100A15" in Genbank (AY465110) [ 20 ]. This name implies higher homology to human S100A15 than to human psoriasin/S100A7. However, it has been shown that human S100A7 and S100A15 diverged sometime during primate evolution [ 10 ]. A phylogenetic tree was constructed to examine the evolutionary relationships between the mouse and human S100 genes (Figure 4 ). The results are consistent with the divergence of human S100A7 and S100A15 occurring by duplication after the human/mouse split in evolution, and thus the mouse gene is likely ancestral to both human paralogs. Presence of putative Jab1 binding motif in mouse S100A7/psoriasin A notable structural feature of the human psoriasin/S100A7 gene is a consensus Jab1-binding domain sequence that we have speculated may confer the ability of human psoriasin/S100A7 to interact with Jab1 and so play an important role in the mechanism by which human psoriasin/S100A7 medicates its action [ 6 ]. Jab1 was originally identified as a factor influencing c-Jun transcription of AP-1 regulated genes [ 21 ]. It is now known that Jab1 is a component of a multimeric protein complex (the CSN/COP9 signalosome) and that Jab1 interacts with many components of cell signalling pathways [ 22 ]. This may be in the context of either phosphorylation or proteasomal activities as Jab1 is also the only known deneddylating protein active in control of the SCF-cullin ubiquitin ligases [ 23 ]. A putative Jab1-binding motif is also present in the murine S100A7/psoriasin sequence but is absent from the human S100A15 sequence (Figure 5 ). Of the three amino acid residues known to be conserved in the Jab-1 binding motif of Jab-1 binding proteins (see highlighted residues in Figure 5 ), one is different in the mouse S100A7/psoriasin. However, the change from aspartic acid in the human S100A7 to a glutamic acid in mouse S100A7/psoriasin is a conservative change and the residue remains acidic. In contrast, the equivalent residue in human S100A15 is threonine, a neutral residue containing an hydroxyl group in its aliphatic side chain, and also having the potential to be post-translationally modified by phosphorylation. These observations support further our hypothesis that the so-called mouse S100A15/psoriasin gene encodes a product that is functionally more similar to human psoriasin/S100A7 than human S100A15. Since the human and mouse Jab1 proteins are 99% identical (Genbank accession number, mouse Jab1 AAC33900; human Jab1 NP_00364) we speculate that conservation of the Jab1-binding domain in the mouse protein results in conservation of the interaction as well. Expression of mouse S100A7/psoriasin during mouse mammary tumorigenesis Our initial results shown in Figure 2 suggest that the expression of mouse S100A7/psoriasin may be upregulated during mammary tumorigenesis. This was of interest since we and others have previously found that human psoriasin/S100A7 expression is highly up-regulated in human breast cancer compared to normal breast tissue [ 2 , 24 ]. To confirm and extend this observation, further analysis of murine mammary tumors was conducted by both semi-quantitative RT-PCR and by in situ hybridization. cDNAs were generated from matched normal mammary gland and mammary gland tumors from DMBA-treated CD1 wild-type mice (n = 6, same as in Figure 2A ). Semi-quantitative RT-PCR analysis showed significantly increased expression of mouse S100A7/psoriasin mRNA in mammary gland tumors relative to matched normal mammary gland (p < 0.05) by Wilcoxon signed-rank test, one-sided (Figure 2B ). To confirm this differential expression and to determine which cell type within the mammary tumors was expressing mouse S100A7/psoriasin, in situ hybridization was performed on sections from paraffin-embedded tissue blocks corresponding to the fresh frozen samples from which RNA was extracted to generate cDNAs used for RT-PCR analysis. A strong but focal signal was observed in all six mammary gland tumors while the signal was weak or undetectable in all of the matched normal mammary gland tissues (Table 1 , Figure 6 ). Assessment of the sections at high magnification and correlation with serial sections stained only by H&E, revealed that the S100A7/psoriasin expression was usually confined to a subset of epithelial tumor cells that showed morphological features of squamous differentiation (Figure 6 ). Within these regions of squamous differentiation expression was usually decreased with proximity to the squamous surface. Semi-quantitative RT-PCR and in situ hybridization results showed a significant positive correlation (Spearman's rank correlation coefficient r = 0.63, p < 0.02) (Table 1 ). Although up-regulation of mouse S100A7/psoriasin expression occurred in DMBA induced tumors compared to adjacent normal mammary gland tissue, expression levels showed a wide-scatter amongst the mammary tumors (Figure 2 ). This was of interest since ERα has been reported to have variable expression in chemical carcinogen induced tumors [ 25 - 27 ] and we have previously shown an inverse association of human psoriasin/S100A7 expression with ERα in breast cancer [ 3 , 5 ]. We therefore compared the expression of mouse S100A7/psoriasin as determined by in situ hydridization with ERα expression assessed by IHC in adjacent sections from the same tumors. An inverse association between mouse S100A7/psoriasin and ERα expression was clearly evident but categorical contingency analysis did not reach statistical significance (Fisher's exact test, p = 0.073), likely due to small numbers. However detailed comparison within tumors showed that in tumors that expressed both genes, mouse S100A7/psoriasin expression was restricted to areas within heterogeneously positive ERα tumors that lacked ERα expression. Interestingly, mouse mammary tumors that develop due to targeting of the neu oncogene to the mouse mammary gland (MMTV- neu ) did not express mouse S100A7/psoriasin RNA nor ERα (data not shown). Expression of mouse S100A7/psoriasin in a model of skin inflammation Human psoriasin/S100A7 was originally cloned from psoriatic skin, an inflammatory skin disease [ 1 ]. Mouse S100A7/psoriasin was however first identified in an expression library constructed from normal mouse skin (Genbank Acc AA792680 and AY465109) and our data have confirmed mouse S100A7/psoriasin RNA expression in normal mouse skin (see Figure 2A ). To determine if mouse S100A7/psoriasin expression is upregulated in inflammation, ISH was performed on sections from normal mouse tissues (see Figure 2A ) as well as tissue from a model of mouse skin inflammation. As seen in mammary tumor tissues, a strong but focal signal was observed in some samples and assessment of the sections at high magnification revealed that the mouse S100A7/psoriasin expression was confined to a subset of epithelial cells surrounding hair shafts (Figure 7A ) consistent with the localization of human psoriasin/S100A7 in normal human skin [ 3 ]. No signal was observed using a mouse S100A7/psoriasin sense probe. In addition we have speculated that mouse S100A7/psoriasin expression would mimic its human counterpart and be upregulated under conditions of skin inflammation. A significant increase in mouse S100A7/psoriasin RNA expression is seen at 24 hrs after croton oil application correlating with the development of inflammation characterized by leucocytic inflammation, (Figure 7B , Table 2 ). Expression of human psoriasin/S100A7 in relation to differentiation in human tumors The pattern of mouse S100A7/psoriasin expression observed in mouse tumors and skin suggested an association of expression with the process of squamous differentiation. Studies of human psoriasin /S100A7 support a similar association in skin and bladder cancer [ 28 , 29 ]. To determine further the relation between psoriasin and different pathways of epithelial differentiation we therefore examined human psoriasin expression relative to different human tumor types in the cervix and lung, since overt squamous differentiation is rare in human breast cancer. In tumors in both cervix and lung, psoriasin expression was significantly correlated with squamous differentiation (p < 0.0001). Expression was observed in over 90% of squamous carcinomas in both organs, but was a rare occurrence (<10%) in adenocarcinomas and absent in other tumor types in the lung (Figure 8 ). As we have previously observed in both breast and skin, psoriasin expression in the cervix was also most highly expressed in squamous carcinoma-in-situ with lower levels in invasive squamous carcinoma (p = 0.0013). Discussion We have used structural and expression analysis to show that a gene currently classified as mouse S100A15 (GenBank acc. AY465110) which maps within the murine S100 cluster on chromosome 3 (see acc. NT_078386, NCBI Map View Link and LOC381493) [ 10 ] should be reclassified as mouse S100A17/psoriasin, as suggested by Marenholz et al. [ 9 ]. This gene is not highly conserved relative to other members of the S100 family (~40% amino acid sequence similarity of mouse S100A7 with either human S100A7 and/or human S100A15, compared to >60% amino acid sequence similarity of most other known mouse S100A proteins to their human counterparts, unpublished data) [ 10 ]. However, phylogenetic analysis shows that while mouse retained an ancestral S100A7/psoriasin gene, the human ortholog underwent duplication and functional differentiation making the assignment of orthology/paralogy only on the basis of phyogenetic information uncertain. In the human, the closely related psoriasin/S100A7 and S100A15 appear to be functionally distinct. Although both were isolated as overexpressed genes in human psoriatic skin [ 1 , 20 ], we have previously identified S100A7 to be differentially expressed in neoplastic mammary gland [ 30 ]. Neither gene is substantially detectable in database analysis of the available normal human mammary gland libraries, and only psoriasin/S100A7 is present at high copy numbers (up to 2,300 tags per 200,000) in several mammary gland neoplasia libraries, whereas S100A15 is absent or rarely detectable (<10 tags per 200,000) [ 30 ]. Additionally, psoriasin/S100A7 is expressed in several SAGE database libraries representing human skin neoplasia, while S100A15 is not detected (data not shown). Thus, despite some similarities in their occurrence and expression in association with psoriasis [ 20 ] these closely related human genes show dissimilar expression patterns in mammary tumors suggesting distinct functional roles. In contrast, our studies show that the pattern of gene expression of mouse S100A7/psoriasin during mammary tumorigenesis and skin inflammation occurs in a similar pattern to that seen with human psoriasin/S100A7 [ 2 , 3 ]. Furthermore, while mouse S100A7/psoriasin shows approximately equivalent amino acid sequence similarity to both human S100A7 and human S100A15 proteins, both mouse S100A7/psoriasin and human psoriasin/S100A7 contain a putative Jab1-binding motif that is functionally significant in the human at least [ 6 , 7 ] but this motif is not found in human S100A15. We conclude on the basis of chromosomal location, phylogenetic analysis, amino acid sequence similarity, conservation of a putative Jab1-binding motif, and similarities in patterns of expression, that mouse S100A7/ psoriasin is the murine ortholog of human psoriasin/S100A7. Expression of human psoriasin/S100A7 was originally attributed to skin pathologies where abnormal squamous differentiation occurs [ 1 ]. In breast tumors where psoriasin/S100A7 is also expressed, overt squamous differentiation is rare. However it has been detected and suggested as a marker of squamous differentiation in a subtype of bladder cancer [ 29 ]. Extending the latter finding our data here shows that in cervix and lung, two tissues where squamous and glandular differentiation are both common differentiation pathways for carcinomas, psoriasin/S100A7 is commonly expressed and almost exclusively associated with squamous tumor subtypes. The parallel finding that mouse S100A7/psoriasin was distinctively associated with areas of squamous differentiation within murine breast adenocarcinomas suggests that similar factors are involved in the regulation of both psoriasin/S100A7 and murine S100A7/psoriasin genes, and is in keeping with a similar role for the murine gene. A role for human psoriasin/S100A7 in inflammation has been suggested, since psoriasin/S100A7 can be secreted and was shown to be chemotactic for neutrophils and CD4+ T-cells in vitro [ 31 ]. Our current data from a model of mouse skin inflammation is also consistent with such a role, since a significant upregulation in mouse S100A7/psoriasin expression occurs simultaneously with the acute phase of skin inflammation after application of croton oil to the mouse tail skin. The role for psoriasin/S100A7 in tumorigenesis may be related to the activity of pro-survival pathways [ 7 ] and acquisition of apoptosis resistance [ 8 ]. Our results here also provide indirect support for this hypothesis, as mouse S100A7/psoriasin RNA was expressed in keratinocytes at the margin of squamous differentiation, where resistance to apoptotic stimuli may be an important component of the sequence of differentiation and to allow time for production of large amounts of keratin before finally undergoing desquamation, denucleation and cell death [ 32 ]. Finally, psoriasin has been implicated in human breast cancer progression. Specifically, psoriasin/S100A7 has been associated with the pre-invasive DCIS phenotype [ 2 ], augmentation of several characteristics of malignancy in vitro and in vivo [ 4 , 6 ] and with poor outcome in invasive estrogen receptor-negative tumors [ 3 , 5 ]. Our current results also support the involvement of S100A7/psoriasin in murine mammary tumorigenesis. Additionally, differential expression of mouse S100A7/psoriasin was observed between DMBA and MMTV- neu induced mammary tumors, and a strong trend towards a negative correlation between mouse S100A7/psoriasin and estrogen receptor alpha expression emerged in the DMBA induced mammary tumors. Such data are consistent with the pattern of expression of human psoriasin/S100A7 in human breast tumors, and are consistent with the view that mouse S100A7/psoriasin subserves similar roles to human psoriasin/S100A7 in mammary tumorigenesis and breast cancer progression. It is nevertheless possible that human psoriasin/S100A7 and mouse S100A7/psoriasin have several functions, depending on cellular context. This is reflected by differences already observed in localization of expression in different cell types within a tissue or subcellular localization, since psoriasin has been found in the nucleus, in the cytoplasm, at the cell periphery/plasma membrane and it can also be secreted [ 33 , 34 ]. Conclusion S100 proteins have been known to be differentially expressed during tumorigenesis as well as other disease states for some time. However, the functional roles they may play in disease processes are poorly understood [ 33 , 34 ]. In breast cancer, psoriasin/S100A7 is associated with important biological and clinical aspects of the disease [ 30 ] and the identification of its potential ortholog in the mouse is an important step to facilitate understanding of its function and mechanism of action. The results presented in this study strongly support the hypothesis that mouse S100A7 is the murine ortholog of human psoriasin/S100A7, and provide a rationale for the manipulation of mouse S100A7/psoriasin in mice to gain important insights into the function of human psoriasin/S100A7. Competing interests The author(s) declare that they have no competing interests. Authors' contributions MW carried out the database searches, cloning, sequencing, ISH, RT-PCR of mouse S100A7/psoriasin, phylogenetic analysis and co-ordinating frozen tissue and tissue block collection from other authors. She was involved in data analysis and writing of the manuscript and critically revising it. EE helped MW in database searches and motif analyses. He was involved in writing and revising it critically. ML carried out the mouse skin inflammation experiments and prepared the tissue blocks. He was involved in writing and revising it critically. SA, GQ and AA were involved in carrying out the tissue analysis and data acquisition and analysis in Figure 8 . LS-C supervised and carried out some of the ISH. YN carried out the IHC. AC supervised and helped carry out the phylogenetic analysis. He was involved in writing the manuscript and revising it critically. YM and RS provided the mouse mammary tumor tissues, other mouse tissues and respective blocks for analysis. They were both involved in writing the manuscript and revising it critically. LCM is the Principal Investigator of the project and was responsible for the overall supervision and co-ordination of the project. She is the corresponding author and contributed to the writing, critical revision together with draft preparation, final preparation for manuscript submission, preparation and submission of all revisions. PHW is the co-P.I. of the project. He supervised and was involved in the interpretation of all histopathology, ISH and IHC. He contributed to all data analysis, to the writing of the manuscript and its critical revision. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC553966.xml |
544956 | Is immunotherapy an effective treatment for Alzheimer's disease? | Immunotherapy in patients with Alzheimer's disease (AD) is rapidly becoming a hot topic of modern geriatric and clinical gerontology. Current views see immunization with Aβ peptide, the amyloidogenic protein found in senile plaque of AD patient's brains, or the infusion of preformed antibody specific for human Aβ, as possible therapeutic approaches to improve the cognitive status in the disease. Animal models of the disease have provided positive results from both approaches. Thus, an initial clinical trial using immunization with human Aβ in AD patients was started, but then shortly halted because of an unusually high incidence (6%) of meningoencephalitis. A long and currently ongoing debate in the scientific community about the pro or contra of vaccination or passive immunization with Aβ in AD is thereafter started. Here, the authors would like to stress few points of concern regarding these approaches in clinical practice. | Immunotherapy in patients with Alzheimer's disease (AD) is rapidly becoming a hot topic of modern geriatric and clinical gerontology. M.E. Weksler [ 1 ] in the article entitled "The immunotherapy of Alzheimer's disease" published in Immunity and Ageing discussed this theme. Current views see immunization with Aβ peptide, the amyloidogenic protein found in senile plaque of AD patient's brains, or the infusion of preformed antibody specific for human Aβ, as possible therapeutic approaches to improve the cognitive status in the disease. Animal models of the disease have provided positive results from both approaches, since either vaccination with human Aβ or infusion of preformed antibodies specific for Aβ, have resulted in a decrease of amyloid plaques density in the brain of amyloid precursor protein (APP) transgenic mice [ 2 , 3 ]. Improved memory performances after Aβ vaccine on APP transgenic mice have also been claimed [ 4 ]. Several other studies from transgenic mice have there after reinforced the suggestion that vaccination or passive immunotherapy might result in a relevant clinical effect in human patients (see the Weksler article [ 1 ]). An initial clinical trial using immunization with human Aβ in AD patients was started and then shortly halted because of an unusually high incidence (6%) of meningoencephalitis (see the Weksler article [ 1 ]). A long and currently ongoing debate in the scientific community about the pro or contra of vaccination or passive immunization with Aβ in AD is thereafter started. Here, we would like to stress few points of concern regarding these approaches in clinical practice. 1) The claimed animal model for AD is unfortunately incomplete even if useful model for the human disease. 2) Vaccination with human Aβ in mice induces an immune response against a foreign protein, i.e. human Aβ, and the mouse Aβ homolog does not appear to be involved. On the contrary, Aβ vaccination in man may potentially induce an autoimmune like disease in the brain and other peripheral tissues of susceptible patients. At the moment we do not have the ability to predict which patients will suffer destructive immune responses. 3) The effects of both vaccination or passive immune therapy in AD brains might be non-specific, as already suggested by a recent report [ 5 ] and by the original histopathological investigation from AD patients deceased after meningoencephalitis [ 6 ]. This notion was then reinforced by the report of these authors [ 7 ]. 4) Vaccine therapy is not always effective in the elderly, since immune defects of variable degree are often present in old persons [ 8 ]. In old clinically ill AD patients this immune activation might fail or activate noxious auto-aggressive immune responses. Once again we cannot predict which patient will experience one or the other condition after the vaccination. 5) Historically vaccination has been successful in the prevention of the diseases much and much less in the therapy of ongoing diseases. A significant gain would be achieved by Aβ vaccination or passive immune therapy, if these manipulations will work in the very early stages of the disease. At the moment, no clinical data on this topic are available and those from animal models have not extensively addressed this topic. 6) A general consideration on AD is also mandatory. In fact, clinical signs of dementia show up after extensive synapse and neuron loss have already occurred in the brain. How can an immune response restore an already compromised nervous circuit or revive dead neurons? The very modest decrement of cognitive deterioration rate claimed in a proportion of AD patients receiving the vaccine is a marginal therapeutic goal. In fact, the disease has not been cured, and the modest clinical slow down in the AD progression takes big prices, i.e. patients and care givers will continue to suffer the catastrophic effects of the disease for a longer time. We feel that Aβ vaccination and in a less extent passive immune therapy are now exciting experimental protocols for animal research and experimental neurology investigations, probably premature for clinical applications. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544956.xml |
516037 | Upper limb neuropathy in computer operators? A clinical case study of 21 patients | Background The character of upper limb disorder in computer operators remains obscure and their treatment and prevention have had limited success. Symptoms tend to be mostly perceived as relating to pathology in muscles, tendons or insertions. However, the conception of a neuropathic disorder would be supported by objective findings reflecting the common complaints of pain, subjective weakness, and numbness/tingling. By examining characteristics in terms of symptoms, signs, and course, this study aimed at forming a hypothesis concerning the nature and consequences of the disorder. Methods I have studied a consecutive series of 21 heavily exposed and severely handicapped computer-aided designers. Their history was recorded and questionnaire information was collected, encompassing their status 1/2 – 1 1/2 years after the initial clinical contact. The physical examination included an assessment of the following items: Isometric strength in ten upper limb muscles; sensibility in five homonymously innervated territories; and the presence of abnormal tenderness along nerve trunks at 14 locations. Results Rather uniform physical findings in all patients suggested a brachial plexus neuropathy combined with median and posterior interosseous neuropathy at elbow level. In spite of reduced symptoms at follow-up, the prognosis was serious in terms of work-status and persisting pain. Conclusions This small-scale study of a clinical case series suggests the association of symptoms to focal neuropathy with specific locations. The inclusion of a detailed neurological examination would appear to be advantageous with upper limb symptoms in computer operators. | Background Upper limb pain and dysfunction are frequent complaints associated with computer work. However, the responsible pathology and the pathophysiological mechanisms are insufficiently understood. In addition, there is no consensus with regard to physical findings that may reflect symptoms. The involvement of the nerves in "non-specific" upper limb disorder, e.g. in computer operators, is suggested by various observations: The demonstration of an elevated threshold to vibratory stimulation [ 1 - 3 ]; abnormal upper limb tension tests [ 4 , 5 ]; reduced nerve mobility [ 6 , 7 ]; abnormal nerve tenderness (mechanical allodynia) [ 8 ]; changed axonal flare reaction [ 9 ]; allodynic response to supra-threshold vibration [ 2 ]; reduced muscle strength [ 10 , 11 ] and sympathetic reflexes [ 12 ]; and thermographic changes [ 13 ]. Still, clinical practice and epidemiological studies tend to attribute upper limb symptoms in computer operators to a disorder in muscle, tendon, or insertion [ 14 ]. Focal neuropathy including carpal tunnel syndrome is infrequently reported [ 15 , 16 ]. Upper limb pain in computer operators shares the features of a neuropathic pain: Common analgesics tend to be ineffective. Pain may be evoked spontaneously or may appear to constitute an abnormal response to stimuli with frequent occurrence of allodynia. In addition, there are often non-painful abnormal spontaneous or evoked sensory phenomena such as numbness/tingling. The common experience of weakness which may further deteriorate on use would also be compatible with an upper limb nerve affliction. A precise and accurate diagnosis is crucial for effective management and rehabilitation, and also for epidemiological studies concerning causation. In order to get a better understanding of the pathophysiological mechanisms, the injured tissue should be precisely located. This might not necessarily be where symptoms predominate. I have aimed at studying a clinical series of computer operators with upper limb complaints and dysfunction in terms of • exposure characteristics; • symptoms and past treatment; • physical findings which may reflect an affliction of the peripheral nerves; • prognosis with regard to symptoms and work-status. Methods Patients This study comprises a consecutive series of 21 computer-aided designers with pain and functional limitations in the dominant upper limb. All patients were referred to a department of occupational medicine for diagnostic and aetiological assessment and management. Three patients were males of median age 27 years (range 25–41) and 19 were females of median age 35 years (range 25–55). Clinical examination and interpretation Interview Patients were interviewed about the character, distribution, initial presentation and development of their symptoms. Special attention was given to the presence of upper limb pain, subjective weakness and numbness/tingling, and to other symptoms included in a standard protocol for work-related upper limb disorders [ 17 ]. Physical examination A subsequent physical examination included extracts of diagnostic criteria for selected clinical disorders (tension neck syndrome, cervical syndrome, supra-and infraspinous tendinitis, bicipital tendinitis, frozen shoulder, acromioclavicular arthrosis, epicondylitis, tenosynovitis, and wrist and forearm peritendinitis) [ 17 ]. Upper limb nerve afflictions were defined from an additional neurological examination consisting of the following components: • Manual assessment of the isometric strength in a selection of ten upper limb muscles (Figure 1 ). Any reduction of strength was registered as weakness [ 18 , 19 ]. Patients were encouraged to provide maximal muscle effort on both sides for each muscle tested, despite any potential discomfort. • The sensibility (algesia by pinprick, aestesia by moving touch [ 19 ]) was assessed in five homonymous innervation territories (Figure 2 ): • The axillary nerve (the deltoid area); • The musculocutaneous nerve (the dorsal forearm); • The radial nerve (the first dorsal web); • The median nerve (the tip of the second finger); • The ulnar nerves (the tip of the fifth finger). • The perception of vibration (tuning fork 256 Hz [ 20 ]) was additionally estimated in the ulnar and median territories (tips of second and fifth fingers). Any sensory deviation from normal was registered as abnormal. • Assessment of tenderness with slight pressure at 14 locations along the course of nerves [ 8 ]. Any mechanical allodynia was registered as abnormal: • The brachial plexus (scalene triangle, passage behind the pectoralis minor muscle); • The suprascapular nerve (suprascapular notch); • The axillary nerve (quadrilateral space); • The musculocutaneous nerve (passage through the coracobrachial muscle); • The median nerve (just proximal to the elbow, at the passage between the two heads of the pronator teres muscle, at the passage below the arcade of the common superficial flexor muscle, and at the carpal tunnel); • The radial nerve (triceps and brachioradial arcades); • The posterior interosseous nerve at the arcade of Frohse (supinator tunnel); • The ulnar nerve (sulcus of the ulnar nerve and Guyon's canal in the hypothenar). Assessments in patients with unilateral disorder were based on comparison to contra-lateral findings defined as normal. In patients with bilateral disorder, test-results were related to other findings in the same limb assumed to be normal, e.g., strength in adjacent muscles or sensibility in adjacent innervation territories [ 21 ]. The definition and location of a nerve affliction ("neuropathy") was based on a traditional approach with a focus on the topography and innervation patterns of the upper limb nerves. Special consideration was given to the presence of normal strength in certain muscles and of reduced strength in others [ 18 , 19 ], and to localized mechanical allodynia at the appropriate location(s) along the nerve trunk [ 8 ]. I have operated with two sets of criteria for the definition of focal neuropathy assuming the second criterion to be more convincing: • Criterion 1: The presence of a pattern of muscle-weakness suggesting a focal neuropathy at a defined location, at which mechanical allodynia with slight pressure at the nerve is present. • Criterion 2: Criterion 1 plus sensory deviations from normal in one or several sensory territories located peripherally to focal neuropathy. In addition, double crush [ 22 ] at the appropriate location(s) was arbitrarily defined when strength reductions and/or mechanical allodynia were either equivalent or more prominent distally. The double crush theory refers to the phenomenon that a focal neuropathy increases the vulnerability of the nerve as a whole, resulting in a tendency of focal neuropathy to occur at several locations along the course of a nerve. Location of neuropathy Brachial plexus neuropathy at chord level was defined with reduced strength in the deltoid, biceps, and radial flexor of the wrist muscles, when weaknesses were accompanied by brachial plexus tenderness at its passage behind the pectoral muscle. Depending on the extent of brachial plexus involvement, additional muscles may be weak and mechanical allodynia may extend in the proximal or medial direction. Median neuropathy at elbow level was defined with reduced strength in the radial flexor of the wrist muscle along with mechanical allodynia involving the median nerve at elbow level (at the passage proximal to the elbow, between the two heads of the pronator teres muscle, and/or at the arcade of the superficial flexor of digits muscle). With an isolated median neuropathy, the deltoid, biceps, and ulnar extensor of the wrist muscles must be intact. Double crush involving the brachial plexus and the median nerve was defined in the following situations: • Strength in the radial flexor of the wrist muscle was reduced as much as / more than it was in the deltoid or biceps muscles. • Mechanical allodynia was either the same or more conspicuous at the median nerve at elbow level than it was at plexus level. Posterior interosseous neuropathy was defined with reduced strength in the ulnar extensor of the wrist muscle along with tenderness at the nerve-passage below the arcade of Frohse in the dorsal proximal forearm. With an isolated posterior interosseous neuropathy, the deltoid, biceps, short radial extensor of the wrist, and radial flexor of the wrist muscles must be intact. Double crush involving the brachial plexus and the posterior interosseous nerve was defined in the following situations: • Strength in the ulnar extensor of the wrist muscle was reduced as much as / more than in the deltoid, biceps, or radial flexor of wrist muscles. • Mechanical allodynia was either the same or more conspicuous at the arcade of Frohse than it was at plexus level. Other potential focal neuropathy was defined according to similar criteria, e.g., an isolated carpal tunnel syndrome would require reduced strength in the short abductor of the wrist muscle but preserved strength in the radial flexor of the wrist muscle. An isolated ulnar neuropathy at elbow or wrist level would require reduced strength in the abductor of the fifth digit and intact proximal muscles. In addition, mechanical allodynia should be present at the appropriate locations along nerve trunks. Management Patients were recommended to freely move and use the symptomatic upper limb within the limits of immediate and subsequent pain aggravation. All patients were offered physiotherapy based on the concept of adverse neural tension [ 23 , 24 ] and encouraged to return to computer work after optimizing the work station ergonomics and work organization. Patients unable to return to work were advised concerning rehabilitation: Maximizing variation during future work; keeping the upper limbs close to the body; and avoiding repetition and static postures. Questionnaire 1/2 – 1 1/2 years after the initial examination the patients responded to a questionnaire: The exposure characteristics; symptoms (pain, weakness/fatiguability, numbness/tingling); past treatment; pain intensity at the first encounter and at follow-up to be quantified on a VAS-scale from 0 (no pain) to 10 (extreme pain); and the present status with regard to functional limitations and work. Statistics The change of the level of reported pain between the first consultation at the department and at follow-up was assessed by Friedman's test. Results Exposure characteristics All 21 patients returned the questionnaire. The mean duration of work with computer-aided design was 95 months (16–260 months). The self-reported daily mean of time spent with computer work constituted 81% (50–100%) of the total working time. 86% of the respondents reported aggravating factors during the months prior to the onset of symptoms, including high work intensity, overwork or other work conditions causing an unusual strain. Symptoms and past treatment Pain in the dominant upper limb was common to all patients. It had a mean duration of 24 months (1–60 months) and was the main symptom in 13 patients. All but one patient had a subjective feeling of weakness/fatiguability. Five patients reported this to be the most disturbing symptom. 19 patients experienced numbness/tingling which constituted the main symptom in three of them. Five patients had bilateral symptoms. All patients had received treatment prior to admission: A limited and transitory effect of past physiotherapy was reported in four out of 17, of pain killers in one out of 10, and of local steroid injections in two out of three patients. For the remaining patients the past treatment had no effect. Physical examination According to the defined criteria for work-related upper limb disorders [ 17 ], non-neuropathic disorders were not identified. In all 21 patients reduced strength was demonstrated in the following muscles: Deltoid, biceps, triceps, and the radial flexor, short radial extensor and ulnar extensor of the wrist. In a smaller number of patients there were additional strength-reductions in the pectoral, infraspinatus, latissimus and abductor of the fifth digit muscles (Figure 1 ). Sensory abnormalities were identified in 19 out of 21 patients. The median nerve territory was most frequently involved. However, most patients had additional sensory deviations in territories innervated by the radial, musculocutaneous, axillary, or ulnar nerves (Figure 2 ). In all patients, mechanical allodynia was present at the brachial plexus at chord level, i.e., on its passage behind the pectoral muscle. In two patients it was also present at trunk level, i.e., at the scalene triangle. Furthermore, mechanical allodynia was present in all patients at the posterior interosseous nerve at the arcade of Frohse, and at the median nerve at one or several locations around the elbow. No mechanical allodynia was observed at the suprascapular, axillary, musculocutaneous, ulnar or radial nerves, nor at the median nerve at the volar wrist (Figure 3 ). Contra-lateral mechanical allodynia was present at the brachial plexus in five patients, and additionally at the posterior interosseous and median nerves in four and three of these patients, respectively (Figure 3 ). The five patients with contra-lateral mechanical allodynia had bilateral complaints. According to the defined criteria, the patterns of physical findings in all 21 dominant limbs suggested the presence of a brachial plexus neuropathy in combination with a median and posterior interosseous neuropathy at elbow level (Figure 4 ). No other nerve entrapments were defined in this sample. Prognosis At follow-up after 1/2 – 1 1/2 years, only two out of the 21 patients remained in computer work. Three were employed in other jobs while the majority was training for other jobs (eight patients) or unemployed (eight patients). Eleven patients reported a beneficial effect of the proposed physiotherapy, five experienced no effect, and five did not pursue the suggested treatment. On a group basis, the mean of pain when worst was reduced from 8.0 at enrolment to 6.1 at follow-up. The corresponding figures when pain was least were 4.3 and 2.7, respectably. This reduction was significant (χ 2 = 8.0 and 9.0, p < 0.005 and 0.003, respectively). However, the pain persisted on a disturbing level in the majority of patients (Figure 5 ). The severity of pain at follow-up was unrelated to the present occupational status and to the severity of pain at enrolment. Neither of the two parameters was related to sex, age, or the duration of exposure. Discussion Working in computer-aided design involves an almost continuous operation of the pointing device for extended periods of time. Consequently, the dominant upper limb strain is more pronounced than in other computer work. Following this exposure, all the patients were severely handicapped in this very limb. In most patients, the unilateral disorder enabled the examiner to conveniently compare the outcomes of the physical examination in the dominant limb with the contra-lateral findings. There is a general consensus that reduced muscle strength, sensory deviations from normal, and localized mechanical allodynia are related to afflicted peripheral nerves [ 19 ]. The relation to an underlying neurological change is indicated by the occurrence of these abnormalities in patterns, in accordance with anatomical facts. The complaints of pain in all dominant upper limbs, and subjective complaints of weakness and numbness/tingling in most of them, were reflected by a rather uniform pattern of strength-reductions, mechanical allodynia, and sensory deviations from normal suggesting the involvement of the brachial plexus at chord level and of the posterior interosseous and median nerves at elbow level. In this sample, I found no indication of carpal tunnel syndrome, ulnar neuropathy at elbow or wrist level, radial neuropathy before division of the posterior interosseous nerve, nerve root compression, or any other neuropathic or non-neuropathic upper limb disorder. Abnormalities were detected by tests which are included in the classical neurological examination. To enable the examiner to assess a single muscle at a time, the limb position during testing should aim at limiting disturbing interference from other muscles [ 21 ]. The mostly minor strength reductions demanded the quantification of strength-reductions to include Grade 4+ (Contraction against gravity and strong resistance [ 18 ]). The identification of such minor weakness demands simultaneous testing on the right and the left side. The absence of facial expressions, withdrawal, or complaints from the patients suggested that the voluntary contraction during strength testing was not influenced by simultaneous pain. The physical examination in this clinical case study was not blinded concerning patient-related information, such as the presence and location of symptoms. However, in a recent validation of the physical tests applied we have found that blinded examiners could reliably assess the individual items (individual muscle strength [ 21 ], sensory qualities, mechanical allodynia), as well as the occurrence of findings in patterns in accordance with the course of nerves and the innervated tissue. Findings were also reflected by the presence of symptoms. Clinical experiences have led to hypotheses suggesting upper limb symptoms in computer operators to be related to prolonged non-neutral and predominantly static positions including a flexion of the shoulder and a submaximally pronated forearm. This may result in a muscular imbalance from some muscles being successively shortened and their antagonists passively stretched and weakened [ 25 ]. Pain and functional limitations may result from limited available space for the nerves especially at locations close to joints or adjacent to bony prominences, fibrous bands or tunnels. This may cause tension, friction, and compression [ 24 , 25 ]. Reduced axoplasmatic flow at a proximal site may lessen the ability of nerves to withstand adverse forces at a more distal site (or the reverse), as described in the double crush phenomenon [ 22 ], and the mobility of the entire nerve may be impaired by such external affliction [ 6 , 24 , 26 ]. From the topography of the brachial plexus, the lateral chords would appear to be most at risk behind the pectoralis minor muscle. The muscles supplied from this part of the plexus (deltoid, biceps, radial flexor of wrist, triceps, short radial extensor and ulnar extensor of the wrist) were invariably involved. In a few limbs there was an additional involvement of the pectoral, small abductor of the fifth digit, latissimus dorsi, and infraspinatus muscles (Figure 1 ). This is concurrent with a medial or proximal extension of a brachial plexus affliction. Caution should be exercised when drawing a comparison between the outcome of this study of patients, referred with a serious disorder, and studies of "healthy" computer operators in occupation. Still, it would be relevant to compare with upper limb findings in computer operators described by others. A study of 533 visual display terminal workers has suggested an array of upper limb disorders in 22%, dominated by tendon related conditions in 15% and probable nerve entrapment in 4% [ 15 ]. In a study of 632 newly hired computer operators, the one year incidence of neck and shoulder symptoms was 58% and of hand/arm symptoms 39%. Symptoms were explained by physical findings in the neck/shoulder region in 35% of participants ("somatic shoulder/neck syndrome" in 33%) and in the hands/arms in 21% (de Quervain's syndrome in 15%) [ 27 ]. In a recent major cross-sectional study of almost 7000 computer operators, 20% complained of moderate to severe pain. The physical examination, however, was only able to disclose a limited number of upper limb disorders, similar to what would be expected in the general population [ 28 ]. Self-reported numbness/tingling in 10.9% of the computer operators was attributed to carpal tunnel syndrome in a minority (numbness/tingling in the median nerve territory in 4.8%, and symptoms at night in 1.4%) and unexplained in the remaining subjects [ 16 ]. Based on the same material, nerve entrapment was only diagnosed in 12 subjects (supinator syndrome and pronator syndrome defined by localized palpation tenderness with withdrawal, and pain with provocative maneuvers). No new cases of nerve entrapment occurred during a one year follow-up [ 29 ]. However, the diagnoses depend on the choice and validity of the clinical tests employed and on the diagnostic criteria applied. The "somatic shoulder/neck syndrome" [ 27 ] is characterized by nonspecific signs and may well be a neuropathic condition. Discomfort with the Finkelstein maneuver [ 15 , 27 ] is not specific for de Quervain's syndrome [ 30 ]. If not associated with first dorsal compartment tenderness and swelling, this diagnosis would seem to be unjustified. The common occurrence of de Quervain's syndrome in computer operators who hardly move their thumb would seem unlikely. My findings are more in accordance with those of Pascarelli, who studied 485 upper limb patients out of which 70% were computer operators. A detailed and comprehensive physical examination demonstrated protracted shoulders in 78% and head forward position in 71%. This was also frequent in my study-patients but not systematically registered. A neurogenic thoracic outlet syndrome in 70% was suggested by tests stressing the brachial plexus and by the demonstration of mechanical allodynia [ 31 ]. In a former study of 53 computer operators with severe upper limb disorders Pascarelli has found a high prevalence of reduced muscle strength and impaired passive wrist deviation associated with an increase in forearm pain. These findings were attributed to myofascial shortening and found to be useful clinical indicators of injury [ 32 ]. Conclusions The limited success of the prevention and management of computer-related upper limb disorders demands new approaches to practice and research in the field. The inclusion in future studies of the presented systematic examination of the upper limb nerves may provide additional diagnostic information. This may lead to future improvement of the prevention and management of computer-related upper limb disorders. Competing interests None declared. Authors' contributions JRJ designed the study and conducted all the clinical examinations, and wrote the manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC516037.xml |
520750 | The efficacy of duloxetine: A comprehensive summary of results from MMRM and LOCF_ANCOVA in eight clinical trials | Background A mixed-effects model repeated measures approach (MMRM) was specified as the primary analysis in the Phase III clinical trials of duloxetine for the treatment of major depressive disorder (MDD). Analysis of covariance using the last observation carried forward approach to impute missing values (LOCF_ANCOVA) was specified as a secondary analysis. Previous research has shown that MMRM and LOCF_ANCOVA yield identical endpoint results when no data are missing, while MMRM is more robust to biases from missing data and thereby provides superior control of Type I and Type II error compared with LOCF_ANCOVA. We compared results from MMRM and LOCF_ANCOVA analyses across eight clinical trials of duloxetine in order to investigate how the choice of primary analysis may influence interpretations of efficacy. Methods Results were obtained from the eight acute-phase clinical trials that formed the basis of duloxetine's New Drug Application for the treatment of MDD. All 202 mean change analyses from the 20 rating scale total scores and subscales specified a priori in the various protocols were included in the comparisons. Results In 166/202 comparisons (82.2%), MMRM and LOCF_ANCOVA agreed with regard to the statistical significance of the differences between duloxetine and placebo. In 25/202 cases (12.4%), MMRM yielded a significant difference when LOCF_ANCOVA did not, while in 11/202 cases (5.4%), LOCF_ANCOVA produced a significant difference when MMRM did not. In 110/202 comparisons (54.4%) the p-value from MMRM was lower than that from LOCF_ANCOVA, while in 69/202 comparisons (34.2%), the p-value from LOCF_ANCOVA was lower than that from MMRM. In the remaining 23 comparisons (11.4%), the p-values from LOCF_ANCOVA and MMRM were equal when rounded to the 3 rd decimal place (usually as a result of both p-values being < .001). For the HAMD 17 total score, the primary outcome in all studies, MMRM yielded 9/12 (75%) significant contrasts, compared with 6/12 (50%) for LOCF_ANCOVA. The expected success rate was 80%. Conclusions Important differences exist between MMRM and LOCF_ANCOVA. Empirical research has clearly demonstrated the theoretical advantages of MMRM over LOCF_ANCOVA. However, interpretations regarding the efficacy of duloxetine in MDD were unaffected by the choice of analytical technique. | Background Treatment effects are often evaluated by comparing change over time in outcome measures. However, valid analyses of longitudinal data can be problematic, particularly if some data are missing for reasons related to the outcome measure [ 1 , 2 ]. Since the problem of missing data is almost ever-present in clinical trials, numerous methods for handling missingness have been proposed, examined, and implemented [ 3 ]. A common method of analyzing clinical trial data is to use analysis of variance or analysis of covariance (ANOVA or ANCOVA) with missing data imputed by the last observation carried forward approach (LOCF_ANCOVA). The popularity of LOCF_ANCOVA may be due to its simplicity, and also the belief that violations of the restrictive assumptions inherent to LOCF_ANCOVA lead to a conservative analysis [ 4 ]. Considerable advances in statistical methodology, and in our ability to implement these methods, have been made in recent years. Thus, methods that require less restrictive assumptions than LOCF_ANCOVA are now readily implemented. For example, likelihood-based repeated measures approaches have a number of theoretical and practical advantages for analysis of longitudinal data with dropout [ 4 ]. One such method, termed MMRM (Mixed Model Repeated Measures [ 5 ]), has been studied extensively in the context of neuropsychiatric clinical trials [ 6 - 9 ]. In these studies, MMRM was found to be more robust to biases from missing data than LOCF_ANCOVA, and thereby provided superior control of Type I and Type II errors. The LOCF_ANCOVA method was shown to underestimate treatment group differences in some scenarios, while overestimating differences in others. When no data were missing, the two methods yielded identical results. The MMRM approach was specified as the primary analysis in the Phase III clinical trials of duloxetine for the treatment of major depressive disorder (MDD), while LOCF_ANCOVA was specified as a secondary analysis. In the present investigation, we provide a comprehensive summary of results from MMRM and LOCF_ANCOVA in the eight acute-phase clinical trials that formed the basis for duloxetine's New Drug Application (NDA) for MDD. The primary objective of this investigation was to determine whether differences in results between MMRM and LOCF_ANCOVA influenced conclusions regarding the efficacy of duloxetine. Methods Data The data source for this investigation was the eight acute-phase clinical trials in which duloxetine was compared with placebo in the treatment of MDD. Relevant details of these studies are highlighted in Table 1 . Table 1 Summary of studies included in the comparisons between MMRM and LOCF_ANCOVA Study Treatment Duration Drug Number of Patients Drug Dose & Design 1 9 weeks Placebo 122 - Duloxetine 123 60 mg/d (QD) 2 9 weeks Placebo 139 - Duloxetine 128 60 mg/d (QD) 3 8 weeks Placebo 70 - Duloxetine 70 40–120 mg/d (20 mg-60 mg BID) Fluoxetine 33 20 mg/d (QD) 4 8 weeks Placebo 75 - Duloxetine 82 40–120 mg/d (20 mg-60 mg BID) Fluoxetine 37 20 mg/d (QD) 5 8 weeks Placebo 90 - Duloxetine 91 40 mg/d (20 mg BID) Duloxetine 84 80 mg/d (40 mg BID) Paroxetine 89 20 mg/d (QD) 6 8 weeks Placebo 89 - Duloxetine 86 40 mg/d (20 mg BID) Duloxetine 91 80 mg/d (40 mg BID) Paroxetine 87 20 mg/d (QD) 7 8 weeks Placebo 93 - Duloxetine 95 80 mg/d (40 mg BID) Duloxetine 93 120 mg/d (60 mg BID) Paroxetine 86 20 mg/d (QD) 8 8 weeks Placebo 99 - Duloxetine 93 80 mg/d (40 mg BID) Duloxetine 103 120 mg/d (60 mg BID) Paroxetine 97 20 mg/d (QD) Results are summarized from all rating scale total scores, subscales, and global assessments that were specified a priori in the various protocols to be analyzed for mean change from baseline to endpoint, and were collected at more than one postbaseline time point (Table 2 ). Efficacy measures that were assessed only at baseline and endpoint were not included in this summary because repeated measures analyses were not possible for these outcomes. Thus, the present investigation included every rating scale total score and subscale from every clinical trial relevant to duloxetine's NDA for an indication in major depression. In total, 20 efficacy and health outcome variables were included in the summary of MMRM and LOCF_ANCOVA. Some of the eight trials included multiple dose arms; therefore, some outcomes were assessed in as many as 12 comparisons with placebo. Table 2 Outcomes included in the summary of results. 17-item Hamilton Rating Scale for Depression (HAMD 17 ) Total score [24] Core subscale (items 1, 2, 3, 7, 8) (not published) Maier subscale (items 1, 2, 7, 8, 9, 10) [25] Anxiety/Somatization subscale (items 10, 11, 12, 13, 15, 17) [26] Retardation subscale (items 1, 7, 8, 14) [27] Sleep subscale (items 4, 5, 6) [27] Montgomery-Asberg Depression Rating scale [28] Hamilton Anxiety Rating Scale total score [29] Clinical Global Impression of Severity [26] Patient Global Impression of Improvement [26] Visual Analog Scale for pain [30] Overall pain severity Headaches Back pain Shoulder pain Time in pain while awake Interference with daily activities Somatic Symptom Inventory [31] 26 Item total score 28 Item total score (includes 2 additional questions on painful physical symptoms) Quality of Life in Depression Scale total score [32] Comparisons of MMRM and LOCF_ANCOVA focused on contrasts between duloxetine and placebo. However, six of the studies also included known effective antidepressants approved for marketing in the United States and other countries. Contrasts between duloxetine and the active comparators are not included in this summary since these results may draw attention to the drug versus drug results and detract from the primary focus of comparing MMRM with LOCF_ANCOVA. Statistical analysis This summary makes no attempt to provide formal statistical comparisons of results from MMRM and LOCF_ANCOVA. Previous research has demonstrated conclusively that in the absence of missing data the two methods yield identical endpoint contrasts, while differences do exist in the presence of subject dropout [ 6 - 9 ]. Furthermore, formal statistical comparisons are typically applied to random samples obtained from larger populations in order to assess the uncertainty associated with the sampling. However, the eight studies included in this summary are not a sample, but rather represent all of the acute-phase, double-blind, placebo-controlled trials of duloxetine. Thus, there is no uncertainty associated with sampling. Consequently, results from the two methods need only be summarized in order to assess how differences between the methods may influence overall conclusions regarding the efficacy of duloxetine. Three overall summary measures were used to compare results from the two analytic techniques: 1) With regard to statistical significance of the difference between duloxetine and placebo, the proportion of outcomes showing agreement between MMRM and LOCF_ANCOVA was compared with the proportion of outcomes for which MMRM and LOCF_ANCOVA yielded disparate results; 2) The proportion of outcomes for which MMRM yielded the lowest p-value was compared with the corresponding proportion for LOCF_ANCOVA; 3) The number of outcomes for which "substantial evidence of efficacy" was demonstrated. In regulatory settings, the criterion for substantial evidence of efficacy is frequently the demonstration of a statistically significant advantage over placebo in two or more studies. This criterion was utilized here to define substantial evidence of efficacy for a particular outcome. The frequency of lower p-values provides a "fine-tuned" measure of sensitivity of the two analytic methods. However, in certain cases such an assessment may actually be misleading. For example, to distinguish between p= .800 and p = .810, or between p =. 002 and p =. 003, implies that the methods yielded different results when in fact the similarities far outweigh the differences. Hence, it is equally appropriate to simply categorize based upon the presence or absence of a significant difference. Furthermore, given the large number of outcomes assessed across the eight studies, it would not be surprising to see the two methods disagree with regard to statistical significance on at least a small number of outcomes. Therefore, perhaps the most clinically meaningful summary measure is the number of outcomes for which substantial evidence of efficacy was demonstrated. Three outcomes were selected for more detailed presentation of results: the 17-item Hamilton Rating Scale for Depression (HAMD 17 ) total score, the HAMD 17 Maier subscale, and the Visual Analog Scale (VAS) for overall pain severity. The HAMD 17 total score was an obvious choice as it was the primary outcome in all studies. The other outcomes were selected since they are frequently focal points in manuscripts and presentations regarding duloxetine's efficacy. Finally, we provide case studies to help explain how and why results from MMRM and LOCF_ANCOVA may differ. MMRM and LOCF_ANCOVA analyses were specified in the Phase III duloxetine protocols as follows. In LOCF_ANCOVA analyses, change from baseline to the last observation was the dependent variable. Treatment and investigative site were included as categorical independent variables, and baseline severity was included as a covariate. In MMRM analyses, change from baseline to all postbaseline times was the dependent variable. Independent variables included the fixed, categorical effects of investigative site, treatment, time, and the treatment-by-time interaction, along with the continuous covariates of baseline severity and the baseline severity by time interaction. Parameters were estimated using Restricted Maximum Likelihood with the Newton-Raphson algorithm. The protocols specified an algorithm for choosing the best fitting covariance structure. In all cases an unstructured matrix provided the best fit. Hence, within-patient errors were modeled using an unstructured covariance matrix. Results The protocols for the eight studies in the duloxetine NDA specified a priori a total of 202 mean change analyses for the 20 rating scale total scores or subscales. The frequency of significant outcomes and the frequency of higher/lower p-values for each analytic technique are summarized in Table 3 . MMRM and LOCF_ANCOVA agreed with regard to substantial evidence of efficacy for 18 of the 20 outcomes, with each analysis yielding substantial evidence for 15 outcomes. That is, MMRM and LOCF_ANCOVA both found substantial evidence of efficacy for 14 outcomes; both methods did not find substantial evidence for four outcomes; and each method found substantial evidence when the other did not for one outcome (Table 3 ). Table 3 Summary of results from MMRM and LOCF_ANCOVA in duloxetine clinical trials. Outcome Frequency of significant outcomes 1 MMRM LOCF_ANCOVA P-value lower with MMRM LOCF_ANCOVA Outcome significant with MMRM but not LOCF_ANCOVA Outcome significant with LOCF_ANCOVA but not MMRM HAMD 17 Total score 9\12 6\12 9\12 1\12 3\12 0\12 Core subscale 9\12 8\12 5\12 3\12 2\12 1\12 Maier subscale 10\12 9\12 6\12 2\12 2\12 1\12 Anxiety subscale 6\12 5\12 7\12 3\12 2\12 1\12 Retardation subscale 8\12 6\12 6\12 3\12 4\12 2\12 Sleep subscale 1\12 1\12 10\12 2\12 0\12 0\12 MADRS Total score 5\10 3\10 7\10 2\10 2\10 0\10 HAMA Total score 3\10 3\10 4\10 5\10 0\10 0\10 CGI-Severity 6\12 6\12 8\12 2\12 0\12 0\12 CGI-Improvement 1\2 1\2 2\2 0\2 0\2 0\2 PGI-Improvement 9\12 7\12 8\12 2\12 2\12 0\12 Somatic Symptom Inventory 26-item Average score 2\10 2\10 7\10 3\10 0\10 0\10 28-item Average score 3\10 2\10 7\10 3\10 1\10 0\10 VAS Pain Severity Overall pain 3\10 4\10 2\10 8\10 1\10 2\10 Headaches 1\10 0\10 5\10 5\10 1\10 0\10 Back pain 2\10 2\10 6\10 4\10 1\10 1\10 Shoulder pain 2\10 1\10 3\10 6\10 2\10 1\10 Daily activities 1\10 0\10 2\10 8\10 1\10 0\10 Pain while awake 2\10 2\10 5\10 5\10 1\10 1\10 QLDS Total score 1\4 2/4 1\4 2\4 0\4 1\4 Totals 84\202 70\202 110\202 69\202 25\202 11\202 1. Some of the eight trials included more than one dose arm. Therefore, an individual outcome could be assessed in as many as 12 comparisons with placebo. In 166/202 outcomes (82.2%), MMRM and LOCF_ANCOVA agreed with regard to the statistical significance of the difference between duloxetine and placebo. In 25 cases (12.4%) MMRM yielded a significant difference whereas LOCF_ANCOVA did not, while in 11 cases (5.4%) LOCF_ANCOVA yielded a significant difference when MMRM did not. Both methods tended to yield significance more frequently in depression rating scales and subscales than in outcomes related to somatic and painful physical symptoms. The studies were generally underpowered for these secondary somatic and pain outcomes owing to the greater variance in changes score for these outcomes. For example, the variance in VAS overall pain severity was approximately nine-fold greater than the variance in HAMD 17 total scores, leading to a three-fold greater standard error. In 110 of the 202 outcomes (54.4%) the p-value from MMRM was lower than that from LOCF_ANCOVA, while in 69 cases (34.2%) the p-value from LOCF_ANCOVA was lower than that from MMRM. In the remaining 23 cases (11.4%) the p-values from LOCF_ANCOVA and MMRM were equal when rounded to the 3 rd decimal place (usually as a result of both p-values being < .001). More detailed results from the three focus outcomes (HAMD 17 total score, HAMD 17 Maier subscale, and VAS overall pain) are presented in Table 4 . In the case of the HAMD 17 total score, the advantage of duloxetine over placebo in mean change from baseline to endpoint from MMRM analyses was greater than the corresponding advantage from LOCF_ANCOVA in 9/12 comparisons (Table 4 ). In 9/12 comparisons the p-value from MMRM was lower than that from LOCF_ANCOVA, while LOCF_ANCOVA yielded a smaller p-value in one case, and p-values were identical in the two remaining cases. In 3/12 comparisons, MMRM yielded a significant difference when LOCF_ANCOVA did not, but in no instance did LOCF_ANCOVA produce a significant difference when MMRM did not. When averaging results across all eight studies, the advantage of duloxetine over placebo in HAMD 17 total score was 2.4 from MMRM analyses compared with 2.0 from LOCF_ANCOVA. Thus, the advantage of duloxetine over placebo, based on LOCF_ANCOVA results, was approximately 83% as large as the advantage from MMRM. Table 4 MMRM and LOCF_ANCOVA analysis of three focus outcomes from duloxetine clinical trials. Study 1 Study 2 Study 3 Study 4 Study 5 Study 6 Study 7 Study 8 Dose 60 mg QD 60 mg QD 60 mg BID 60 mg BID 20 mg BID 40 mg BID 20 mg BID 40 mg BID 40 mg BID 60 mg BID 40 mg BID 60 mg BID Δ p Δ p Δ p Δ p Δ p Δ p Δ p Δ p Δ p Δ p Δ p Δ p HAMD 17 Total Score MMRM 4.9 < .001 2.2 .024 3.1 .009 0.9 .415 1.4 .143 1.5 .116 2.4 .034 3.6 .002 2.2 .001 3.3 < .001 1.4 .045 1.6 .014 LOCF 3.8 < .001 1.7 .048 2.1 .066 0.4 .681 1.2 .222 1.5 .138 2.4 .022 3.1 .003 2.2 .007 3.0 < .001 0.9 .253 1.5 .054 HAMD 17 Maier Subscale MMRM 2.8 < .001 1.6 .003 2.0 .005 0.7 .282 1.2 .037 1.4 .012 1.1 .068 1.7 .005 1.4 < .001 2.0 < .001 0.9 .022 1.2 .002 LOCF 2.3 < .001 1.4 .007 1.0 .030 0.6 .481 1.0 .058 1.3 .012 1.5 .028 1.8 .004 1.4 .001 1.9 < .001 0.7 .090 1.0 .014 VAS Overall Pain Severity MMRM 5.9 .055 4.4 .135 NA NA 0.3 .931 1.2 .731 1.0 .771 7.4 .035 5.5 .085 6.3 .050 5.6 .044 1.3 .625 LOCF 6.9 .019 5.2 .037 NA NA 3.5 .573 3.5 .647 3.2 .710 7.1 .048 6.1 .063 5.6 .086 7.8 .014 5.5 .066 NA = not assessed Similar results were obtained for the Maier subscale. Thus, the advantage of duloxetine over placebo in mean change from baseline to endpoint from MMRM was greater than that from LOCF_ANCOVA in 9/12 comparisons (Table 4 ). The p-value from MMRM was lower than that from LOCF_ANCOVA in 6/12 comparisons, while LOCF_ANCOVA produced a lower p-value in 2 cases, and p-values were identical in the remaining four cases. In 2/12 comparisons MMRM yielded a significant difference when LOCF_ANCOVA did not, while there was one instance in which LOCF_ANCOVA yielded a significant difference when MMRM did not. Averaged over all eight studies, the advantage of duloxetine over placebo in mean Maier subscale score was 1.5 from MMRM analyses compared with 1.3 from LOCF_ANCOVA. Thus, the advantage of duloxetine over placebo based on LOCF_ANCOVA results was approximately 87% as large as the advantage from MMRM. For VAS overall pain severity, the advantage of duloxetine over placebo from MMRM analyses was greater than the corresponding advantage from LOCF_ANCOVA in 2/10 comparisons (Table 4 ). The p-value from MMRM was lower than that from LOCF_ANCOVA in 2/10 comparisons, while in the remaining 8 comparisons the p-value from LOCF_ANCOVA was lower than that from MMRM. In 1 comparison MMRM yielded a significant difference when LOCF_ANCOVA did not, while in 2 comparisons LOCF_ANCOVA yielded a significant difference when MMRM did not. Over all eight studies, the average advantage of duloxetine over placebo in VAS overall pain severity was 3.9 from MMRM analyses compared with 5.4 from LOCF_ANCOVA. Thus, the advantage of duloxetine based on LOCF_ANCOVA results was approximately 138% as large as the advantage from MMRM. Case Studies Mean changes in HAMD 17 total score and VAS overall pain severity from two studies (Studies 1 and 2) are used to further illustrate MMRM and LOCF_ANCOVA (analysis of variance with missing data imputed via last observation carried forward) results. Results from these studies were originally reported by Detke et al [ 10 , 11 ]. In both studies the advantage of duloxetine over placebo in HAMD 17 total score tended to increase over time whereas duloxetine's advantage in VAS overall pain was greatest at intermediate visits (Tables 5 and 6 ). Table 5 MMRM and LOCF_ANCOVA analyses of HAMD 17 total score in Studies 1 and 2. STUDY 1 STUDY 2 MMRM THERAPY Week N Mean change Std.error p-value N Mean change Std.error p-value DULOX 1 121 -2.89 0.36 .435 123 -2.89 0.38 .601 PLACEBO 1 115 -2.50 0.37 136 -2.64 0.36 DULOX 2 112 -5.72 0.49 < .001 109 -5.54 0.48 .071 PLACEBO 2 110 -3.39 0.50 129 -4.43 0.45 DULOX 3 105 -7.37 0.53 < .001 108 -6.82 0.55 .287 PLACEBO 3 103 -4.58 0.54 122 -6.06 0.52 DULOX 5 100 -8.76 0.60 < .001 98 -8.58 0.66 .116 PLACEBO 5 101 -5.74 0.60 111 -7.20 0.62 DULOX 7 91 -9.93 0.64 < .001 89 -10.14 0.69 .008 PLACEBO 7 93 -5.82 0.65 97 -7.69 0.65 DULOX 9 84 -10.91 0.70 < .001 81 -10.46 0.71 .024 PLACEBO 9 89 -6.05 0.69 90 -8.29 0.67 LOCF_ANCOVA DULOX LOCF_ANCOVA 121 -9.47 0.63 < .001 123 -8.75 0.71 .048 PLACEBO LOCF_ANCOVA 115 -5.67 0.66 136 -7.02 0.68 Table 6 MMRM and LOCF_ANCOVA analyses of VAS overall pain severity in Studies 1 and 2. STUDY 1 STUDY 2 MMRM THERAPY Week N Mean change Std.error p-value N Mean change Std.error p-value DULOX 1 120 -2.69 1.97 .181 121 -2.02 1.89 .157 PLACEBO 1 113 1.04 2.04 134 1.38 1.79 DULOX 2 113 -8.20 1.87 .003 108 -6.95 2.04 .003 PLACEBO 2 108 -0.42 1.93 127 0.91 1.90 DULOX 3 105 -10.46 1.95 .005 106 -9.81 1.90 < .001 PLACEBO 3 101 -2.56 2.01 119 -1.59 1.78 DULOX 5 99 -8.68 2.20 .028 95 -9.70 2.05 .011 PLACEBO 5 99 -1.85 2.24 108 -2.92 1.92 DULOX 7 91 -10.25 2.34 .016 87 -8.91 2.12 .254 PLACEBO 7 92 -2.26 2.38 94 -5.76 2.01 DULOX 9 83 -8.68 2.18 .055 80 -10.05 2.24 .135 PLACEBO 9 88 -2.80 2.20 88 -5.65 2.12 LOCF_ANCOVA DULOX LOCF_ANCOVA 121 -8.31 2.09 .019 121 -10.39 2.05 .037 PLACEBO LOCF_ANCOVA 114 -1.40 2.18 134 -5.22 1.94 In the case of the HAMD 17 total score, advantages for duloxetine over placebo at endpoint (Week 9) from MMRM in Studies 1 and 2 were 4.86 (p < .001) and 2.17 (p = .024), respectively. The corresponding advantages from LOCF_ANCOVA were 3.80 (p < .001) and 1.73 (p = .048). Although the differences were significant for both methods in both studies, MMRM yielded treatment contrasts that were approximately 25% greater than LOCF_ANCOVA. For VAS overall pain, the advantage of duloxetine over placebo at endpoint from MMRM in Studies 1 and 2 were 5.88 (p = .055) and 4.40 (p = .135), respectively. The corresponding advantages from LOCF_ANCOVA were 6.91 (p = .019) and 5.17 (p = .037). In both studies, the endpoint differences were significant from LOCF_ANCOVA, but not from MMRM. The LOCF_ANCOVA treatment contrasts were approximately 15% greater than those from MMRM. Standard errors from LOCF_ANCOVA were approximately 5% smaller than the Week 9 standard errors from MMRM for both the HAMD 17 total score and VAS overall pain. Discussion In many areas of clinical research, the impact of missing data can be profound [ 2 , 12 - 14 ]. Traditional approaches to analyses of data from clinical trials with dropouts, such as LOCF_ANCOVA, have focused on ease of implementation and interpretation. However, simple methods rely upon assumptions that are often unrealistic. For example, LOCF_ANCOVA assumes that patient dropout is completely random, i.e. it is unrelated to the outcome being analyzed. Hence, in an analysis of efficacy data, LOCF_ANCOVA assumes that patients do not drop out due to lack of efficacy. The LOCF_ANCOVA approach also assumes that, for those patients who drop out, their observations would not have changed had they stayed in the trial. When these assumptions do not hold true, estimates of treatment effects and associated standard errors may be biased [ 2 - 4 , 6 , 7 , 15 - 17 ]. Considerable advances in statistical methodology, and in our ability to implement these methods, have been made in recent years. Methods such as MMRM, which require less restrictive assumptions regarding missing data, may now be easily implemented with standard software [ 4 , 5 , 18 , 19 ]. No universally superior approach to analysis of longitudinal data exists. However, a series of studies [ 6 - 9 ] demonstrated empirically what may have been anticipated from statistical theory – namely that the MMRM approach, while providing no guarantee of immunity from bias due to subject dropout, was a sensible analytic choice for many clinical trial scenarios. MMRM has repeatedly been shown to provide adequate control of Type I (false positive) and Type II (false negative) errors in a wide variety of situations modeled after neuropsychiatric clinical trials. In these head-to-head comparisons involving 456,000 data sets, the LOCF_ANCOVA approach did not perform as well as MMRM. We therefore specified MMRM as the primary analysis and LOCF_ANCOVA as a secondary analysis in the Phase III clinical trials of duloxetine in the treatment of MDD. Similar results regarding control of Type I and Type II error for LOCF_ANCOVA and mixed-effects model analyses have been obtained independently [ 16 , 20 - 22 ]. Furthermore, following an independent investigation of data from two placebo- and active-comparator controlled duloxetine trials, in which treatments were coded A, B, C, etc. to blind analysts to the treatment names, Molenberghs et al [ 4 ] concluded that MMRM analysis was a sensible choice for those data. The theoretical differences between MMRM and LOCF_ANCOVA have been summarized [ 5 , 18 ], established empirically [ 6 - 9 ], and proven mathematically [ 4 ]. However, we are unaware of any previous investigations of how these differences manifest themselves in efficacy assessments of a new medicinal product. The VAS pain results highlight a limitation of endpoint analyses of any type, namely that they provide only a snapshot view of the response profile. From LOCF_ANCOVA analysis, one can only conclude that drug was superior to placebo at endpoint. However, MMRM analysis reveals that drug had a significant effect early in the trials, but that advantage was somewhat transitory as the placebo group tended to "catch up" over time. In order to understand the response profile of a drug, the entire longitudinal profile should be considered [ 2 ]. From MMRM, the entire profile can be assessed from the same analysis that provided the primary result (the contrast at endpoint). In the duloxetine database, results from MMRM and LOCF_ANCOVA were in general agreement regarding substantial evidence of efficacy and frequency of significant differences. However, MMRM tended to be more sensitive to drug-placebo differences for outcomes related to overall depressive symptoms and core emotional symptoms of depression, with mean advantages over placebo that were 10% to 20% greater than LOCF_ANCOVA. However, MMRM did not universally increase duloxetine's advantage over placebo in comparison to results from LOCF_ANCOVA. For example, in somatic and painful physical symptom outcomes, results from LOCF_ANCOVA showed mean advantages over placebo that were approximately 40% greater than that from MMRM. Therefore, while the overall conclusions regarding the efficacy of duloxetine were unaffected by the choice of analytic method, this should not mask the important differences between MMRM and LOCF_ANCOVA. The advantages of MMRM and similar methods over LOCF_ANCOVA have been conclusively demonstrated in many studies and are evident in the duloxetine data. Khan et al [ 23 ] compiled a database from FDA summaries of efficacy for all antidepressants approved between 1985 and 2000. Less than half of the studies – which were analyzed using LOCF_ANCOVA as the primary analysis – found significant advantages for drug over placebo. These studies were generally anticipated to have at least 80% power and, if the analysis worked as expected, the success rate would be 80%. Therefore, LOCF_ANCOVA was less sensitive to the drug effects than anticipated. In the duloxetine database, MMRM yielded a 75% success rate for the primary outcome measure (HAMD 17 total score), while LOCF_ANCOVA produced a 50% success rate, in comparison to the expected rate of 80%. While many factors may reduce the success rate in Phase III clinical trials, the use of a statistical method with known inflation of Type II error (false negative results) is an obvious suspect. An unduly high rate of false negative results could be especially problematic in early phases of drug development where only one or two chances exist to make the correct decision regarding the efficacy of a drug. It is noteworthy that one of the instances when MMMRM yielded a significant difference on the primary outcome (HAMD 17 total score) when LOCF did not was in the Phase II study (Study 3). Also consider that across all therapeutic areas only about 50% of the molecules that enter Phase III testing receive regulatory approval. Many factors may reduce the success rate of Phase III development. However, the use in Phase II of a statistical method with known inflation of Type I error (false positive results) is an obvious suspect. Thus, the unexpectedly low success rate in Phase III is consistent with the conclusion that LOCF_ANCOVA inflates Type I error (as a result of an unduly high rate of false positive results in Phase II). Hence, results from the duloxetine NDA are consistent with research suggesting a move away from LOCF_ANCOVA and other simple analytic techniques to methods such as MMRM that are more robust to the biases from missing data. Conclusion Important differences exist between MMRM and LOCF_ANCOVA. Research has clearly demonstrated the advantages of MMRM over LOCF_ANCOVA. However, interpretations regarding the efficacy of duloxetine in MDD were unaffected by the choice of analytical technique. Competing Interests Drs. Mallinckrodt, Raskin, Wohlreich, Watkin, and Detke are employees of Eli Lilly and Company. Author's Contributions CHM designed the statistical analyses, and participated in interpretation of data and drafting of the manuscript. JR, MMW, JGW, and MJD participated in interpretation of data and drafting of the manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC520750.xml |
516792 | Mohmmar Qadaffi, Open Access, and Retrovirology | Retrovirology has been publishing as an Open Access online journal for approximately six months. In this editorial, I review the reasons for and the advantages of Open Access publishing, update our progress to date, and summarize where we intend to go with this journal. | Qadaffi's lesson Thirty-five years ago, on September 1 st , 1969, I was an eleven year old boy living with my family in Benghazi, Libya. That morning, I awoke to an eerie silence. The normally bustling traffic outside of our apartment in a busy suburb of Benghazi was not to be heard. In fact, there was to be little or no traffic for the next several days. History will recount that day as the first day of Mohmmar Qadaffi's bloodless coup d'etat, seizing power in Libya. What I learned that September week in 1969 was that access to and distribution of information tipped the fate of a state. On that September day, there were in fact two groups of army officers vying to overthrow the then enfeebled Libyan King Idris. Qadaffi, at the time only twenty-seven years old, headed a small group of very junior officers. He, however, had a bold and swift strategy. His first targets were not the dams, the power stations, the police stations...; but the radio stations (in 1969 television had a very minor penetrance in Libya) and newspapers. By dint of controlling the airwaves and printing presses and by freely distributing his information during time of chaos, Qadaffi held sway over the minds and the sentiments of the populace. He was thus able to pre-empt the efforts of a larger competing group of more senior army officers. What does Qadaffi's story have to do with Open Access publishing? Imagine how the outcome might have been different in 1969 if broadcast radio was not freely and openly accessible by the masses but was an expensive subscription service limited to a few. In that setting, and perhaps in all settings, information that reaches a little audience at a steep cost is information with little effect. Retrovirology, freely and openly accessible by all Curiously, the force that drives a coup d'etat is the same that motivates scientists to publish (although I am not suggesting that scientists are like Qadaffi). As scientists, we want everyone who is interested to read what we have done; and ideally we want that information to get to the widest audience in the fastest, most informative, and least expensive manner. Until now, ideal and reality were two entirely different matters. Traditional publishing in print journals reaches a limited audience, is slow, bulky, and expensive. By its very nature (i.e. pages of paper are heavy to mail and expensive to print), there is a real impetus on the part of print publishers to restrict the space available to authors. We are all too familiar with the results: abbreviated presentation of methods and text, the necessity to "inadvertently" not cite many colleagues' publications, and the ubiquitous use of microscopic figure panels that provoke blurred vision and headaches. I daresay that many after having read a short "letter" or a "brief definitive" report have been left scratching heads and wondering how the experiments were done, where did the vectors come from, and why many controls were definitely missing. For the retrovirology community, Retrovirology aims to change such "reality". Retrovirology is published by BioMed Central, an independent publisher committed to ensuring immediate free access to peer-reviewed biomedical research. Retrovirology is Open Access which means that it is universally and freely available online to everyone. You as the author retain copyright, and your paper is guaranteed to be archived in at least one internationally recognised repository [ 1 ]. More importantly, publishing in Retrovirology is rapid. For example, a full research article was reviewed and published in less than 20 days [ 2 ]; and a review article was similarly processed in 8 days [ 3 ]. Retrovirology will also give you all the space and all the colored prints (at no extra charge) to present your findings informatively and attractively. If we don't publish your paper, we will tell you honestly what is scientifically deficient about your work. I promise not to obfuscate myself using annoyingly belittling mantra that my editorial decision was guided by "intense competition for space". Judge a book by its cover? This is my second editorial for Retrovirology readers [ 4 ]. Retrovirology has been publishing on-line now for 6 months. While a first editorial can speak of dreams and aspirations, the burden of a second editorial is to demonstrate results and progress. Aided by an international cohort of extremely capable editors (Monsef Benkirane, Ben Berkhout, Masahiro Fujii, Michael Lairmore, Andrew Lever, and Mark Wainberg) and sixty editorial board members [ 5 ], I am indeed encouraged and gratified by Retrovirology's initial achievements. As of this writing, we have published 22 papers. These are good papers; and importantly, these papers are being widely-read. I can say with some pride that half (11 out of 22; see Table 1 ) of our published papers have been read more than 1,000 times each. The remaining papers are not far behind; our most recently published paper, barely 1 week on-line, has already had more than 200 readers. Given that retrovirology is a relatively small community of researchers, my personal experience of twenty years in this field tells me that these are very respectable numbers. Table 1 Access statistics for the top 11 Retrovirology articles. Article Access Count 1 Early steps of retrovirus replicative cycle Sébastien Nisole, Ali Saïb Retrovirology 2004, 1 : 9 1890 2 Establishment of a novel CCR5 and CXCR4 expressing CD4+ cell line which is highly sensitive to HIV and suitable for high-throughput evaluation of CCR5 and CXCR4 antagonists Katrien Princen, Sigrid Hatse, Kurt Vermeire, Erik De Clercq, Dominique Schols Retrovirology 2004, 1 : 2 1602 3 Evolution of the HIV-1 envelope glycoproteins with a disulfide bond between gp120 and gp41 Rogier W Sanders, Martijn M Dankers, Els Busser, Michael Caffrey, John P Moore, Ben Berkhout Retrovirology 2004, 1 : 3 1332 4 Use of a multi-virus array for the study of human viral and retroviral pathogens: gene expression studies and ChIP-chip analysis Elodie Ghedin, Anne Pumfery, Cynthia de la Fuente, Karen Yao, Naomi Miller, Vincent Lacoste, John Quackenbush, Steven Jacobson, Fatah Kashanchi Retrovirology 2004, 1 : 10 1313 5 Increased mortality associated with HTLV-II infection in blood donors: a prospective cohort study Jennie R Orland, Baoguang Wang, David J Wright, Catharie C Nass, George Garratty, James W Smith, Bruce Newman, Donna M Smith, Edward L Murphy, Retrovirology 2004, 1 : 4 1209 6 Retrovirology and young Turks... Kuan-Teh Jeang Retrovirology 2004, 1 : 1 1140 7 HIV-1 gene expression: lessons from provirus and non-integrated DNA Yuntao Wu Retrovirology 2004, 1 : 13 1103 8 Multi-faceted, multi-versatile microarray: simultaneous detection of many viruses and their expression profiles Biehuoy Shieh, Ching Li Retrovirology 2004, 1 : 11 1097 9 Two discrete events, human T-cell leukemia virus type I Tax oncoprotein expression and a separate stress stimulus, are required for induction of apoptosis in T-cells Takefumi Kasai, Kuan-Teh Jeang Retrovirology 2004, 1 : 7 1088 10 Apoptosis of uninfected cells induced by HIV envelope glycoproteins Barbara Ahr, Véronique Robert-Hebmann, Christian Devaux, Martine Biard-Piechaczyk Retrovirology 2004, 1 : 12 1076 11 HIV CTL escape: at what cost? Stephen M Smith Retrovirology 2004, 1 : 8 1022 Despite that many of you would argue to me that you would rather publish in Cell , Science , and Nature rather than Retrovirology . I wouldn't dispute you on that point. On the other hand, I would ask you to pause and think about a refrain that I am sure all of you have taught your children, "Don't judge a book by its cover!" Please, don't sell your own work short; and please don't belittle the objectivity of your colleagues. If you have a wonderful piece of work, do you truly need the cover of Cell to establish that point? On the other hand, if you have a horrible study, do you think that hiding it behind the cover of Nature would fool many of your colleagues for long? There was a time in the print age when display space on reading room shelves was limited and perhaps justifiably dominated by "the big three". Today, Retrovirology will get your paper listed in PubMed within 48 hours of acceptance for publication. Anyone in this electronic era whose reading habits are not guided by PubMed in one way or another is probably someone who you may not care to read your study. In short, Retrovirology will promise your immediacy of publication and visibility for your work. Your papers will be well-read by your colleagues. Beyond that, I believe, and I think you do too, the impact of your work should rest on its merits and not on our or any other journal's cover. Open Access is not free lunch Besides teaching about book covers, I think the other concept that we teach our children is that "There is no free lunch!" At the end of the day, anything worth doing has a price tag. I, my colleague editors, and all the editorial board members at Retrovirology work for free for the journal. We can do this because we collect a salary elsewhere doing research. On the other hand, we do have a dedicated staff based at the publisher in London. No one would question that the salaries for these professionals as well as other publication costs must be borne. To fund these costs, from October 1, 2004, authors of articles accepted for publication will be asked to pay a single all-inclusive article-processing charge (APC) of US$600; there are no additional charges beyond this sum. Retrovirology will not levy additional page or colour charges on top of this fee. You are free to publish any number of colour figures and photographs, at no extra cost. I support the need for APC, because without this Retrovirology cannot be published. I think most of you will agree based on your experience with other publishers that US$600 is more than modest and reasonable. As a point of comparison, the Public Library of Science has set up two new Open Access journals, and has set their APCs at US$1500, fully two and one-half times our charge, for each accepted article [ 6 ]. Nonetheless, I realize that any APC fee can present an unacceptable financial hardship for some authors. These authors should feel free to contact me in confidence, and I will consider fee-waiver requests on a case-by-case basis. In closing, let me thank those of you who have read, submitted and reviewed papers for Retrovirology . More than anything else, Retrovirology is not my journal as much as it is your journal. I welcome your advice, and I look forward to hearing from you. Abbreviations APC = article-processing charge. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC516792.xml |
340957 | Engineering Bacteria to Make “Unnatural” Natural Drugs | null | Faced with new and ongoing threats to public health, researchers are becoming increasingly resourceful in their quest to discover new drugs. Drug researchers have long looked to living organisms for inspiration, either mimicking or extracting chemical formulas from naturally occurring compounds. Bacteria and fungi, for example, produce a wide range of compounds—some of which give them a selective advantage in their own environments—that provide important pharmaceutical activities. One class of these natural compounds are the polyketides, which make up a large portion of the antibiotics (including erythromycin and tetracycline) and antitumor drugs (such as doxorubicin and epothilone) that have been isolated from various microorganisms. Polyketides are synthesized by bacteria and fungi by the appropriately named polyketide synthases (PKSs). PKSs can be thought of as large molecular factories containing a series of enzymes working on an assembly line: each enzyme in the line adds molecules to a primer, or starter, unit—which is usually an acetate molecule—and then hands off the growing chain to the next enzyme. The specific enzymes set all the characteristics of the polyketide, including the chain length, the building blocks used, and the branching pattern of the molecules. Although microorganisms generate polyketides with a variety of characteristics, one goal of drug discovery research is to increase this diversity even further—a larger pool of polyketides promises more drugs with enhanced pharmaceutical applications. Early attempts at creating artificial polyketides focused on altering the functional characteristics of naturally occurring polyketides—the length of the chain, the building blocks, and the patterns of the branches. Chaitan Khosla and colleagues have taken this approach one very large step further. Rather than changing the machinery to modify the growing structure of a polyketide, they engineered bacteria to use an alternative, nonacetate primer molecule. This has important practical implications because some medicinally significant compounds do not use the usual acetate primer unit. By dissecting out the specificities of the “starter” and longer, multiunit “elongation” PKS enzymes and by mixing and matching modules, they have produced novel polyketide analogs (in this case, of anthraquinone) with more effective medically relevant properties. One of the compounds they engineered shows enhanced efficacy in blocking the growth of breast cancer cells that depend on the activity of the estrogen receptor, while a second polyketide inhibits an enzyme linked to adult-onset diabetes, demonstrating just two possible new therapeutic applications for synthesized polyketides. But, as the authors propose, this method promises to reveal new pharmaceutical agents that haven't even been discovered yet. A synthetic polyketide | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC340957.xml |
340943 | Cracking the Polyketide Code | Polyketides, natural products from microorganisms, have been a main source of antibiotics. Understanding the 'programming' of the enzymes that produce these complex molecules has opened a new field of drug discovery | For half a century, natural products from microorganisms have been the main source of medicines for treating infectious disease. The most important chemical class of these antibiotics, apart from the penicillins, is the polyketides. They are made by the stepwise building of long carbon chains, two atoms at a time, by multifunctional enzymes that determine the chain length, oxidation state, and pattern of branching, cyclisation, and stereochemistry of the molecules in a combinatorial fashion to produce an enormous variety of structures. Recent elucidation of the genetic ‘programming’ of the enzymes has opened a new field of drug discovery based on rationally engineering the enzymes to produce ‘unnatural natural products’ with novel properties. Following the development of penicillin for the treatment of septicemia in the early 1940s, numerous antibiotics were discovered and introduced into medicine. While a fungus makes penicillin, semisynthetic derivatives of which have been a mainstay of antibacterial therapy for decades, most natural antibacterial antibiotics come from a group of soil-dwelling, filamentous bacteria called the actinomycetes, of which Streptomyces is the best-known genus. These organisms make an amazing array of so-called secondary metabolites that have evolved to give their producers a competitive advantage in the complex soil environment, where they are exposed to stresses of all kinds ( Challis and Hopwood 2003 ). The compounds have many functions, but those with antibiotic activity are the most important from the human perspective. Actinomycete antibiotics include such antibacterial compounds as tetracycline and erythromycin, antifungal agents like candicidin and amphotericin, anticancer drugs such as doxorubicin, and the antiparasitic avermectin ( Walsh 2003 ). While many different chemical classes are represented amongst actinomycete antibiotics, one class accounts for an extraordinary proportion of the important compounds, including all those mentioned above. This chemical family is made up of the polyketides. They are synthesized by multifunctional enzymes called polyketide synthases (PKSs), which are related to the fatty acid synthases that make the lipids essential for the integrity of cell membranes, but they carry out much more complex biosynthetic routines. Repeated rounds of carbon chain building and modification use a series of independently variable reactions selected according to a ‘program’ characteristic of each PKS ( Reeves 2003 ). Recent research has focused on determining this program so as to be able to modify it in rational ways by genetic engineering and thus generate novel drug candidates. The resulting field of ‘combinatorial biosynthesis’ of ‘unnatural natural products’ has been given added urgency by the rise of multidrug-resistant pathogens, of which MRSA (methicillin-resistant Staphylococcus aureus ) is simply the most discussed of a series of threats ( Walsh 2003 ). How do PKSs work and how can we make new ones? Molecular Diversity The heart of PKS function is the synthesis of long chains of carbon atoms by joining (condensing) together small organic acids, such as acetic and malonic acid, by a so-called ketosynthase function. This uses the building units in the form of activated derivatives, called coenzyme A (CoA) esters, so we speak of acetyl-CoA and malonyl-CoA, for example. The special form of condensation that joins them is driven by loss of carbon dioxide. Thus, when acetyl-CoA, with two carbon atoms, joins with malonyl-CoA, with three carbons, one of the latter is lost and a chain of four carbon atoms results ( Figure 1A ). Further rounds of condensation extend the chain by two carbons at each step. If the chain-extender unit, instead of being malonyl-CoA, is methylmalonyl-CoA, which has four carbon atoms, the linear carbon chain is still extended by two carbons, and the ‘extra’ carbon forms a methyl side branch. More complex extender units produce more complex side branches. Figure 1 The Chemistry of Polyketide Chain Assembly (A) Acetic acid and malonic acid are converted to their coenzyme A (CoA) esters and then attached, by specific acyl transferases, to components of the polyketide synthase (PKS): acetyl-CoA is attached to the active site of the ketosynthase, and malonyl-CoA to a structural component of the PKS called the acyl carrier protein (ACP). Condensation of the two units by the ketosynthase, with loss of one carbon from malonyl-CoA as carbon dioxide, produces a four-carbon chain attached to the ACP. This is transferred back to the ketosynthase, and further rounds of condensation with malonyl-CoA (as shown) or other chain extender units produce a polyketide chain. (B) The three-step reductive cycle that converts a keto group to a hydroxyl, then to a double bond, and finally to a fully saturated carbon. (C) A complex polyketide in which keto groups, hydroxyl groups, double bonds, and fully saturated carbons occur at different positions along the chain, depending on the operation of the reductive cycle after each condensation. Choices of the number and type of the building units are variables in determining polyketide structure. Another concerns the keto groups (C=O) that appear at every alternate carbon atom in the growing chain as a result of the condensation process (accounting for the name polyketide). They may remain intact. Alternatively, some may be modified or removed by a series of three steps ( Figure 1B ), any of which may be omitted. This results in keto groups remaining at some points in the chain; hydroxyl groups (–OH), formed by reduction of a keto group, at others; double bonds between some adjacent carbon atoms, resulting from removal of the hydroxyl by loss of water (dehydration); or full saturation with hydrogen atoms elsewhere, arising by ‘enoyl’ reduction of the double bond ( Figure 1C ). A further variable concerns the stereochemistry of the hydroxyl groups and methyl or other carbon branches, each of which can exist in two possible configurations. Finally, the nascent carbon chain adopts different folding patterns after it leaves the PKS, and ‘tailoring’ enzymes can then add sugars or other chemical groups to it at many alternative positions, enabled by the pattern of chemical reactivity built into the polyketide by the PKS. The challenge has been to understand the programming of the PKS that accounts for this gamut of structural variation. During the 1990s, the ability to manipulate actinomycete genes, developed over previous decades, mainly using the model species Streptomyces coelicolor ( Hopwood 1999 ), was combined with chemical and biochemical experiments to begin to crack this ‘polyketide code’. The first studies were on organisms making antibiotics of the ‘aromatic’ family, which includes tetracycline and doxorubicin, as well as the model compounds actinorhodin (made by S. coelicolor itself) and tetracenomycin. The main variable in their structure is carbon chain length, with few choices of different building units or keto group modification, so the programming would (in principle) be simple. The DNA sequences responsible for such PKSs revealed sets of genes encoding proteins, including ketosynthases, ketoreductases, and acyl carrier proteins (ACPs) (the unit of the PKS on which the growing carbon chain is tethered; see Figure 1A ), that would come together to form a multicomponent PKS resembling a typical bacterial fatty acid synthase. In contrast, the DNA sequence of the gene set for the complex polyketide erythromycin, made by a relative of Streptomyces called Saccharopolyspora erythraea , which has more involved programming, revealed multifunctional proteins with the various enzymic functions carried out by active sites on the same polypeptide chain, as in a mammalian fatty acid synthase. The big surprise, though, was the finding of six sets, or modules, of such active sites, corresponding to the six rounds of condensation needed to build the carbon chain ( Cortes et al. 1990 ; Donadio et al. 1991 ). The modules each contain an acyl transferase (to load the extender unit onto the enzyme), as well as a ketosynthase and an ACP domain, together with exactly those reductive activities needed to generate the required pattern of modification of the chain at each step of elongation. Thus was born an ‘assembly line’ model in which the program for the PKS is hardwired into the DNA and expressed in a linear array of active sites (domains) along the giant protein. This consists of the six chain-building modules, preceded by a short module for loading the starter unit and ending in a domain for releasing the completed carbon chain from the PKS. The carbon chain of the polyketide would be assembled and modified progressively as the molecule moved along the protein, interacting with each domain in turn, which would select extender units, make carbon–carbon bonds, and modify keto groups as appropriate, depending on the presence or absence of domains for the three steps in the reductive cycle. The model arose from the gene sequence, but was rapidly tested by mutating individual domains or adding or deleting whole modules and by observing predicted changes in the polyketide product. Soon, dozens of engineered compounds had been made, and the field mushroomed with the isolation of more and more clusters of genes for complex polyketides that both proved the generality of the model (with minor variations) and filled the need for spare parts for the engineering of countless new polyketides ( Shen 2003 ). Several biotech companies were founded to exploit the potential for drug discovery. Aromatic PKS Programming Meanwhile, the programming of the aromatic PKSs was harder to understand. They had been found to contain only a single ketosynthase, which has to operate a specific number of times to build a carbon chain of the correct length, so how is this determined? How does a single reductive enzyme know which keto groups to modify? And how is the starter unit for building the carbon chain selected (the extender units are normally all malonyl-CoA, so no choice is involved)? Considerable progress had been made in constructing novel compounds by mixing and matching PKS subunits, but this was largely based on empirical knowledge about which components to put together ( McDaniel et al. 1995 ). A specific subunit of the PKS, named the chain length factor (CLF), was deduced to have a major influence on carbon chain length ( McDaniel et al. 1993 ), but this conclusion was not universally accepted in the absence of experimental evidence on its mode of action. Two recent publications by the Khosla laboratory at Stanford University describe significant advances in understanding aromatic PKS programming and promise to turn the spotlight back onto engineered members of this class of compounds as potential drug candidates by allowing rational manipulation of the two key variables: carbon chain length and choice of starter unit. In the first paper ( Tang et al. 2003 ), the authors explore the hypothesis that the CLF exerts control over carbon chain length by associating closely with the ketosynthase, a protein with which it shares considerable amino acid sequence similarity, giving rise to a channel of a certain size at the interface between the two proteins. By systematically changing amino acids at four key positions in the CLF, the size of the channel was altered. Thus, large amino acid residues in the CLF of a PKS that makes a 16-carbon chain were replaced by less bulky residues found in one that builds a 20-carbon chain, and the chain length of the product increased as expected. The authors propose that the length of the channel is the main factor in controlling the number of chain-extension steps that can take place to fill it. While protein–protein interactions with other PKS subunits may modulate this chain length control, the work represents a major step in understanding and manipulating the chain length of aromatic polyketides. What about the choice of starter unit? Most aromatic polyketides start with acetyl-CoA. An important earlier publication by Leadlay and colleagues ( Bisang et al. 1999 ) had shown that this is not loaded directly onto the PKS, as had been assumed, but is derived by loss of carbon dioxide from a molecule of malonyl-CoA previously loaded onto the enzyme. This decarboxylation is catalysed by the CLF as an activity independent of its role in influencing carbon chain length. There are, however, certain aromatic polyketides, including the anticancer drug doxorubicin, an antiparasitic agent called frenolicin, and the estrogen receptor agonist R1128, that have different starters. What Tang et al. (2004) have deduced, as described in this issue of PloS Biology , is that the PKSs for these compounds consist of two modules of active sites. The components of each module are not activities carried on the same protein, as in the PKSs for the complex polyketides, but are all separate proteins. They form functional modules nevertheless. The newly recognized modules in the producers of compounds that start with nonacetate units have a dedicated ACP and a special ketosynthase that carries out a first condensation, joining the unusual starter unit to the first malonyl-CoA extender unit. The starter module then hands the resulting ‘diketide’ on to the second module (first reducing it, if appropriate, using reductive enzymes ‘borrowed’ from fatty acid biosynthesis) for typical extension by successive condensation with malonyl-CoA units to complete the chain. If the starter module is not present, the second module defaults to its typical habit of decarboxylating malonyl-CoA to acetyl-CoA and starts the chain with that. The excitement of the work for biotechnology is that it offers the prospect of engineering promising drug candidates by making novel combinations of starter and extender modules and perhaps of feeding the starter modules with a whole range of unnatural substrates ( Kalaitzis et al. 2003 ). It is encouraging that already, in the proof-of-principle studies reported by Tang et al. (2004) , some products with improved in vitro antitumor activity were obtained. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC340943.xml |
524250 | Unanticipated Antigens: Translation Initiation at CUG with Leucine | Major histocompatibility class I molecules display tens of thousands of peptides on the cell surface for immune surveillance by T cells. The peptide repertoire represents virtually all cellular translation products, and can thus reveal a foreign presence inside the cell. These peptides are derived from not only conventional but also cryptic translational reading frames, including some without conventional AUG codons. To define the mechanism that generates these cryptic peptides, we used T cells as probes to analyze the peptides generated in transfected cells. We found that when CUG acts as an alternate initiation codon, it can be decoded as leucine rather than the expected methionine residue. The leucine start does not depend on an internal ribosome entry site–like mRNA structure, and its efficiency is enhanced by the Kozak nucleotide context. Furthermore, ribosomes scan 5′ to 3′ specifically for the CUG initiation codon in a eukaryotic translation initiation factor 2–independent manner. Because eukaryotic translation initiation factor 2 is frequently targeted to inhibit protein synthesis, this novel translation mechanism allows stressed cells to display antigenic peptides. This initiation mechanism could also be used at non-AUG initiation codons often found in viral transcripts as well as in a growing list of cellular genes. | Introduction Immune surveillance by cytotoxic T cells (CTLs) is a key mechanism for detecting and eliminating abnormal cells. These include cells infected with viruses or bacteria, and those that have suffered tumorigenic transformations ( Townsend and Bodmer 1989 ). The antigen receptors of CTLs probe the repertoire of peptide/major histocompatibility complex (MHC) class I complexes on the target cell surface. Target cells display peptides derived from virtually all cellular translation products; the presence of foreign peptides, such as those derived from viral proteins, can trigger a T cell response. Each cell presents tens of thousands of distinct peptides as potential ligands for the CTL antigen receptor ( Rammensee et al. 1993 ; Engelhard 1994 ). Most peptides are represented at fewer than ten copies per cell, and by some estimates only three copies of the antigenic peptide are sufficient for target cell lysis ( Purbhoo et al. 2004 ). CTLs are thus a very sensitive probe for the peptides displayed by MHC class I. The antigen-presenting cells (APCs), which include almost all nucleated cells, are also very efficient in generating peptides for display by MHC class I ( Pamer and Cresswell 1998 ; Princiotta et al. 2003 ). In fact, they are so efficient that the complex mixture of peptides on the cell surface includes some peptides that should in theory never have been translated in the first place ( Shastri et al. 2002 ). These peptides, referred to as “cryptic,” are derived from the 5′ and 3′ “untranslated” regions of the RNA or from alternate translational reading frames. Cryptic peptides have been identified as targets for CTLs specific for tumors as well as virus-infected cells ( Mayrand and Green 1998 ; Cardinaud et al. 2004 ). Several studies have shown that these peptides can arise in tumor cells and cultured cell lines despite the absence of conventional AUG codons ( Malarkannan et al. 1995a ; Dolstra et al. 1999 ; Malarkannan et al. 1999 ; Ronsin et al. 1999 ). Recently, using a transgenic approach we demonstrated that such peptides can also be expressed in a variety of normal cells and can elicit CTL responses ( Schwab et al. 2003 ). Remarkably, a distinct translation mechanism appeared to be responsible for their generation, because it was capable of decoding the CUG initiation codon as leucine rather than the expected methionine residue. How APCs generate peptides using non-AUG codons remains obscure. It is believed that cells express only one class of initiator tRNA, RNA i Met , which is specific for AUG and is always charged with the methionine residue ( Peabody 1989 ; Rajbhandary and Chow 1995 ). 40S ribosomes are preloaded with RNA i Met and other initiation factors even before they approach the mRNA to be translated. Initiation at non-AUG codons is therefore thought to be caused by “wobble” in the pairing of the non-AUG codon with the anticodon of the RNA i Met ( Peabody 1989 ). This mispairing results in incorporation of the methionine residue at the non-AUG initiation codon. How cells initiate translation with a nonmethionine residue is thus not explained by current translational theory. It is nevertheless important to understand this mechanism not only because antigenic peptides can arise from non-AUG initiated translation, but also because expression of a growing number of genes appears to be mediated by translation initiated at non-AUG codons. In this study, we used T cells as probes to analyze the translation mechanism that allows the generation of CUG-initiated antigenic peptides and decodes CUG as the leucine rather than the methionine residue. We found some similarities, but also key differences, between the translation mechanisms mediating initiation at conventional AUG versus CUG codons. Results/Discussion Decoding of the CUG Initiation Codon as Leucine Does Not Depend upon the mRNA Sequence We had previously shown both in transfected cell lines and in a transgenic mouse model that CUG can be decoded as leucine when it serves as the initiation codon for the (CTG)-TFNYRNL peptide (the initiation codon is in parentheses and the remaining amino acids are in single-letter code) ( Malarkannan et al. 1999 ; Schwab et al. 2003 ). Although there are numerous examples of non-AUG initiation codons, this is to our knowledge the only known instance in which the identity of the first residue was investigated and found not to be methionine, with the exception of initiation directed by cricket paralysis virus-like internal ribosome entry sites (CPV-IRES). Normally the RNA i Met occupies the P site of the ribosome, where translation initiation begins. In contrast, the CPV-IRES binds the ribosome and positions it precisely so that initiation begins in the A site, without requiring RNA i Met ( Jan and Sarnow 2002 ). In the cricket paralysis virus, initiation begins with alanine encoded by GCU, and in the Plautia stali intestine virus initiation begins with glutamine encoded by CAA ( Sasaki and Nakashima 2000 ; Wilson et al. 2000 ). Because there was no obvious IRES-like sequence in the transgene we had used, and non-CUG leucine codons failed to direct efficient translation, we suspected that a different mechanism might account for initiation at the CUG codon ( Schwab et al. 2003 ). Therefore we first asked whether the leucine start was dependent upon the mRNA sequence. To address this question, we identified the first amino acid of CUG-initiated peptides that were presented by MHC class I molecules and thus detectable by appropriate T cells. In this assay, we cotransfected cells with cDNA constructs encoding the peptide as well as the MHC molecule that binds and presents the peptide on the cell surface. Binding to the MHC protects the initiation residue from being removed by cellular proteases, which can trim antigenic peptides in the cytoplasm or the endoplasmic reticulum ( Kloetzel 2004 ; Rock et al. 2004 ). We then extracted the peptides translated in the cells and separated them by reverse-phase high performance liquid chromatography (RP-HPLC), using conditions in which the methionine-initiated peptide elutes at a different time from the leucine-initiated peptide. Finally, we assayed the HPLC fractions for the presence of the peptide. We added to each fraction a T cell hybridoma that cross-reacts with the methionine and the leucine-initiated forms of the peptide, as well as appropriate APCs that will display the peptide on their MHC molecules. We measured the T cell response to each fraction. By comparing the retention time of the T cell-stimulating fractions from the cell extract with the retention time of synthetic peptides, we can determine the identity of the peptide made by the cells. This assay has several advantages over conventional assays used to detect initiating amino acids: MHC molecules protect the first amino acid from cleavage, T cells can detect nonmethionine residues with high sensitivity, and we can analyze the products of translation initiation in vivo. To determine if decoding of the CUG initiation codon as leucine was influenced by 5′ or 3′ untranslated regions (UTRs) of the mRNA, we transfected COS-7 cells with cDNA encoding the K b MHC molecule and the *(CTG)-TFNYRNL* peptide ([CTG]YL8). We refer to this as the xYL8 model. We placed the peptide in three different contexts: first, in the pcDNA1 vector; second, upstream of the IRES in the pIRES2-eGFP vector; and, finally, in the 5′ UTR of green fluorescent protein (GFP) in the pcDNA1 vector. The peptides translated in the transfected cells were extracted and fractionated by HPLC. Each fraction was tested for the presence of the leucine-initiated LTFNYRNL (LYL8) peptide and the methionine-initiated MTFNYRNL (MYL8) peptide using BCZ103 T cells and K b -expressing L cells (a fibroblast cell line) as APCs. With each construct we found a single peak of antigenic activity that eluted in the same fraction as the synthetic LYL8 peptide ( Figure 1 A– 1 C). In contrast, when the cells were transfected with a construct encoding the ATG-initiated peptide ([ATG]-TFNYRNL), the single peak of antigenic activity eluted in fractions identical to those containing the synthetic MYL8 peptide ( Figure 1 A and 1 D). This result rules out the possibility that the apparent use of leucine was due to post-translational modification of the MYL8 peptide, which caused it to coelute with the LYL8 peptide. We conclude that the cells were capable of decoding the initiating CTG codon as leucine regardless of the 5′ or 3′ sequences flanking the LYL8 coding sequence. Figure 1 The CUG Initiation Codon Is Decoded as Leucine Independent of RNA Sequence (A) The indicated synthetic peptides mixed with extracts from untransfected COS-7 cells were separated by RP-HPLC. Each fraction was tested for BCZ103 T cell-stimulating activity with K b+ B7.2 + L cells as APCs. After overnight incubation, the β-galactosidase induced in activated T cells was measured using the substrate chlorophenol red-β-pyranoside, which yields a colored product with absorbance at 595 nm. The arrows indicate the reproducible peak elution times for the MYL8 and the LYL8 peptides. Injections of buffer alone (Buffer) were carried out under identical conditions, and the fractions were assayed in parallel to ensure absence of cross-contamination between runs. (B–D) Extracts from COS-7 cells transfected with cDNA encoding K b and the indicated constructs were separated by RP-HPLC. Fractions were tested for BCZ103 T cell-stimulating activity as in (A). (E) The indicated synthetic peptides mixed with extracts from untransfected COS-7 cells were separated by RP-HPLC. Each fraction was tested for the specific DBFZ T cell-stimulating activity with D b+ B7.2 + L cells as APCs as in (A). The arrows indicate the reproducible peak elution times for the MM9 and the LM9 peptides. (F and G) Extracts from COS-7 cells transfected with cDNA encoding D b and the indicated constructs were separated by RP-HPLC. Fractions were tested for DBFZ T cell-stimulating activity as in (E). (H) A range of concentrations of the indicated synthetic peptides was tested for BCZ103 T cell-stimulating activity with K b+ B7.2 + L cells as APCs. (I) A range of concentrations of the indicated synthetic peptides was tested for DBFZ T cell-stimulating activity with D b+ B7.2 + L cells as APCs. We next tested whether the LYL8 coding sequence itself enabled the leucine start. We examined the initiating amino acid used for a different peptide presented by the D b MHC class I molecule that satisfied the conditions required for our assay: that an MHC molecule present the peptide, that HPLC allow distinction between the leucine- and methionine-initiated forms, and that a T cell cross-react with the peptides with leucine or methionine residues at the first position. When COS-7 cells were transfected with cDNA constructs encoding the D b MHC molecule and the *(CTG)-SNEN-METM peptide derived from the influenza nucleoprotein, only the leucine-initiated LM9 peptide was detected in the HPLC-fractionated cell extracts ( Figure 1 E and 1 F). We again assessed whether the apparent use of leucine could be due to post-translational modification of the methionine-initiated peptide. We transfected cells with a construct containing the ATG initiation codon in place of the CTG codon. Analysis of cell extracts after HPLC fractionation now revealed a single peak of antigenic activity that eluted in the same fractions as the synthetic MM9 peptide ( Figure 1 E and 1 G). Similar results were also obtained with the (CTG)-SHL8 peptide that was presented by the D b MHC class I molecule ( Malarkannan et al. 1995b ) (unpublished data). We conclude that the peptide sequence itself was irrelevant to the decoding of the CTG initiation codon as leucine. At present it is difficult to quantify the fraction of the total translated material that is initiated with leucine, because the different peptides may have different stabilities in the cell. Furthermore, the T cell hybridomas could respond to the methionine- and leucine-initiated peptides with differing sensitivities. Indeed, the BCZ103 T cell hybridoma responds to LYL8 approximately 30-fold better than to MYL8 ( Figure 1 H). In contrast, the DBFZ T cell hybridoma recognizes its leucine- or methionine-initiated cognate peptides with comparable sensitivity ( Figure 1 I). Nonetheless, translational initiation with the leucine residue is readily detected. In previous studies, we explored the possibility that leucine was used as the first amino acid because the ribosome may have begun at an upstream alternate initiation codon (there are no upstream ATGs in these constructs) and read through the stop codon before the CTG initiation codon in the LYL8 coding sequence. We showed that increasing the number of stop codons upstream of CTG from one to six had no effect on LYL8 expression, and that substituting the CTG codon with other leucine-encoding triplets essentially ablated peptide expression ( Malarkannan et al. 1999 ; Schwab et al. 2003 ). We assessed the possibility of translational read-through in the different model systems used here. If read-through were responsible for LYL8 expression, one would predict that expression of the downstream LYL8 would be proportional to expression of the upstream sequence. To test this hypothesis, we made DNA constructs in which the LYL8 coding sequence was placed directly after the stop codon terminating translation of another peptide, SSVVGVWYL (SVL9) ( Mendoza et al. 1997 ). We then placed the [SVL9*LYL8] cassette in-frame (R0) with and out-of-frame (R1) with an ATG initiation codon ( Figure 2 A). These constructs were transfected into cells along with the appropriate MHC molecules, and peptide expression was assayed with the 30NX/B10Z T cell hybridoma specific for the SVL9/D b complex ( Figure 2 B) and the BCZ103 T cell hybridoma specific for the LYL8/K b complex ( Figure 2 C). T cells were added directly to the transfected cells without an intervening extraction step. Cells expressing the R0 cassette were highly active as APCs. Shifting the cassette out-of-frame with the initiating ATG in the R1 construct dramatically reduced expression of the SVL9 peptide. Remarkably, the cells transfected with either the R0 or the R1 constructs were equivalent in their ability to stimulate the LYL8-specific T cells ( Figure 2 C). To confirm that the CTG initiation codon was decoded as leucine, we analyzed HPLC fractionated cell extracts ( Figure 2 D). Again, a single peak of activity was found in the fractions that matched the leucine-initiated LYL8 synthetic peptide. Thus, translation of the LYL8 peptide was independent of upstream conventional translation initiation events, and translational read-through events were not detected. Figure 2 Expression of the Cryptic LYL8 Peptide Is Independent of the Efficiency of Upstream Translation Initiation Events (A) The nucleotide sequences of R0 and R1 constructs encode the SVL9 peptide followed by a termination codon and the LYL8 peptide. In the R0 construct, the SVL9 peptide is in frame with an ATG initiation codon. In the R1 construct, a single nucleotide is inserted after the ATG codon and causes the SVL9*LYL8 coding sequence to be out-of-frame with the ATG. The in-frame translation products are underlined and arrows indicate the potential initiation codons. (B and C) The R0 and R1 constructs were transfected into Lmtk – cells along with the appropriate MHC molecule. They were tested for SVL9/D b expression using the 30NX/B10Z hybridoma (B) and LYL8/K b expression using the BCZ103 T cell hybridoma (C). (D) Extracts from COS-7 cells transfected with cDNA encoding K b and the indicated constructs were separated by RP-HPLC. Fractions were tested for BCZ103 T cell stimulating activity as in Figure 1 . We also considered the possibility that leucine was used as the first amino acid because of an RNA modification that introduced an AUG immediately upstream of the peptide-coding sequence. However, the experiments below show that the 5′ UTR influences the leucine start because it is affected by the Kozak context, by the presence of an upstream hairpin, and by the presence of upstream initiation codons. Together, these findings demonstrate that the mRNA remained intact. The Kozak Context Affects the Efficiency of Initiation at CUG The efficiency of initiation at a given AUG codon depends on the identity of the surrounding nucleotides. These nucleotides, commonly referred to as the “Kozak context,” have a substantial influence on protein synthesis. Kozak found that G CC A CC AUG G is optimal, and that the nucleotides at positions –3 and +4 are the most influential, while the nucleotide at the –6 position exerts a smaller effect. Changing these nucleotides can change protein expression by 20-fold ( Kozak 1987a , 1987b ). To determine whether the leucine start was affected by the nucleotides surrounding the CUG initiation codon, and to allow better prediction of probable CUG initiation codons, we varied the nucleotides at positions –6, –3, and +4 relative to the CUG codon. We inserted synthetic oligonucleotides *( N CC N CC CTG N CC)SEL8* into the pcDNAI vector, where N represents the degenerate nucleotides flanking the CTG initiation codon for the SLVELTSL (SEL8) peptide presented by the K b MHC molecule to bm1BZ19.4 T cells ( Malarkannan et al. 1996 ). We transfected COS-7 cells with plasmid DNA from 96 individual randomly picked bacterial colonies together with the cDNA for K b . When the transfected cells were tested for their ability to stimulate bm1BZ19.4 T cells, we noticed a substantial variation in the T cell response, suggesting that the plasmids differed in their ability to express the SEL8 peptide ( Figure 3 A). As a negative control, none of the cells transfected with each of the 96 plasmid DNAs encoding the minigene *( N CC N CC CCC N CC)SEL8* stimulated the T cells, demonstrating that initiation activity was restricted to the CTG codon ( Figure 3 B). To identify which nucleotides were associated with the variation in peptide expression, we determined the nucleotide sequences of three sets of 18 CUG-initiated constructs that yielded high, intermediate, or low T cell responses ( Figure 3 C). As summarized in the panels in Figure 3 D, among the plasmids that yielded high responses, 70% had T at position –6, 75% had A at position –3, and 80% had G at +4. Conversely, among the plasmids that yielded low responses, these nucleotides were infrequent. Thus, the optimal context for the CUG mediated initiation is T CC A CC CUG G , which is in close agreement with Kozak's consensus sequence ( G CC A CC AUG G ) for the AUG start and extends earlier findings that showed that an A at +5 and U at +6 can also enhance initiation at the CUG codon ( Boeck and Kolakofsky 1994 ; Grunert and Jackson 1994 ). Furthermore, the fact that CUG initiation activity was influenced by the Kozak context provides additional evidence that the CUG codon was decoded during translation initiation rather than translation elongation. Figure 3 The Optimal Nucleotide Context for the CUG Initiation Codon (A and B) The indicated degenerate oligonucleotides were cloned into the pcDNA1 vector. “ N ” represents any one of T, A, C, and G nucleotides. The CTG or CCC initiation codons are boxed and the peptide coding sequence is indicated by [SEL8]. 96 randomly picked plasmids for the CTG-initiated peptide and an equal number for the CCC-initiated peptide were purified. The plasmids were transfected into COS-7 cells along with the K b MHC class I molecule, and the T cell response was measured. Each bar represents the T cell response to cells transfected with an individual plasmid. (C) Three sets of 18 representative plasmids, each yielding high, intermediate, and low responses (as shown) were selected for nucleotide sequencing. (D) Summary of the nucleotide sequences of plasmids yielding high, intermediate, and low responses. The left, middle, and right panels, respectively, correspond to the plasmids shown in (C). Each panel shows the percent of each nucleotide found at the –6, –3, and +4 degenerate positions indicated by the “ N ” in A. For example, the upper left square shows that, of the high T cell-stimulating plasmids, the –6 position was T for 67%, A for 6.7%, C for 20%, G for 6.7%, a pyrimidine (T or C) for 87%, and a purine (A or G) for 13%. An Excellent Kozak Context Enhances Both the CUG/Leucine and the CUG/Methionine Starts In the above model, because the initiation codon was not included within the final SEL8 antigenic peptide product protected by the MHC molecule, the identity of the amino acid residue specified by the CUG initiation codon could not be determined. Thus, we could not distinguish whether the Kozak context affected the leucine start, the methionine start, or both. To resolve this question, we turned to the (CTG)YL8 model, in which the predominant T cell-stimulating activity is the leucine-initiated LYL8 peptide (see Figure 1 A– 1 C). At the –6 and –3 positions, we placed the best (T, A) and the worst (G, T) nucleotides. We were unable to vary the +4 position from the original A, because that would alter the peptide's second amino acid and likely affect its detectability by the BCZ103 T cell hybridoma. We first transfected cells with the two constructs as well as the appropriate K b MHC cDNA. After 2 d the transfected cells were assayed for their ability to stimulate the BCZ103 T cell. Cells expressing the LYL8 peptide with its CTG initiation codon in the “Excellent Kozak” context (T at –6, A at –3) were superior to those with the CTG codon in a “Poor Kozak” context (G at –6, T at –3) in stimulating the T cell response ( Figure 4 A). Next, we extracted peptides from the cells expressing the two constructs and analyzed the antigenic activities after HPLC fractionation. Again, a single peak of activity corresponding to the leucine start was detected in the extract of cells expressing the “Poor Kozak” construct ( Figure 4 B). In contrast, in the extract from cells expressing the “Excellent Kozak” construct not only was the total amount of the LYL8 peptide higher, but a new activity peak corresponding to the methionine-initiated MYL8 peptide was also clearly detected. By comparing the T cell response to the cell extracts with the response to known quantities of synthetic peptides, we determined that the construct with the CTG initiation codon in an “Excellent Kozak” context yielded approximately 6-fold more LYL8 peptide than the construct with CTG in a poor context. The amount of MYL8 peptide increased as well, but the change cannot be quantified, because MYL8 was undetectable in extracts of cells transfected with the construct containing the CTG initiation codon in the “Poor Kozak” context ( Figure 4 B). We conclude that the Kozak context enhanced not only the overall efficiency of the CTG/leucine start, but also the ability of the CTG codon to be decoded in the conventional “wobble” mode as the methionine residue. Figure 4 The Kozak Context Enhances the Leucine as well as the “Wobble” Methionine Start (A) Lmtk – cells were transfected with K b cDNA and the indicated “Excellent” and “Poor” constructs encoding the (CTG)YL8 peptide. They were tested for LYL8/K b expression using the BCZ103 T cell hybridoma. (B) Extracts from COS-7 cells transfected with cDNA encoding K b and the indicated constructs were separated by RP-HPLC. Fractions were tested for BCZ103 T cell-stimulating activity with K b+ B7.2 + L cells as APCs. The arrows indicate the peak elution positions for the MYL8 and the LYL8 peptides. T cell responses to fractions collected after injecting sample buffer alone (Buffer) are also shown to indicate absence of cross-contamination between runs. Ribosomes Scan 5′ to 3′ for the CUG Initiation Codon In most cases, ribosomes bind mRNA at the 5′ cap and scan in the 3′ direction for the first AUG in an appropriate Kozak context. Thus, for approximately 90% of mRNA transcripts, the 5′-most AUG initiates protein synthesis ( Kozak 1991 ). In order to develop predictive algorithms for the CUG initiation codon, we asked whether ribosomes similarly scanned for the CUG/leucine start. We took advantage of the fact that a heat-stable hairpin in the mRNA blocks 40S ribosomal scanning ( Kozak 1986 , 1989a ). We designed a hairpin based on the one Kozak used to trap scanning 40S ribosomes ( Kozak 1989b ). It extends Kozak's original sequence in order to maintain stability at 37 °C, and includes a bulge to prevent the longer hairpin from targeting the mRNA as a potential substrate for RNA interference. We placed the hairpin 42 nucleotides upstream of the initiation codon. To control for nonspecific effects, such as RNA stability, we also included GFP under the control of an IRES element downstream of the peptide-coding sequence. We first transfected COS-7 cells with constructs encoding the ATG-initiated MYL8 peptide and another encoding MYL8 downstream of the heat-stable hairpin. We titrated the transfected cells and assayed the T cell response to the peptides presented on the cell surface. As expected, the presence of the hairpin inhibited MYL8 expression ( Figure 5 A). This effect was not due to the hairpin destabilizing the mRNA, because GFP expression as measured by flow cytometry of the same cells was not decreased by the presence of the hairpin ( Figure 5 B). Next, we performed the same experiment with COS-7 cells transfected with the CTG-initiated peptide with and without an upstream hairpin sequence and found that the hairpin inhibited translation of LYL8 as well ( Figure 5 C). Again, the level of GFP expression was similar in the cells transfected with constructs with or without the hairpin ( Figure 5 D). Similar inhibition of both ATG- and CTG-initiated translation was also observed with a second set of constructs in which the hairpin was placed 68 nucleotides upstream of the initiation codon, making it less likely that the hairpin interfered with a potential IRES-like ribosomal landing pad (unpublished data). Both conventional, ATG-mediated and cryptic, CTG-mediated translation events were disrupted by upstream hairpins, suggesting that both require ribosomal scanning in the 5′-to-3′ direction. Figure 5 The CUG Start Is Blocked by a Heat-Stable Hairpin in the 5′ UTR COS-7 cells were transfected with cDNA encoding K b and the indicated constructs. (A and C) The cells were titrated and peptide expression was tested with BCZ103 T cells. (B and D) GFP expression in the transfected cells was assayed by fluorescence-activated cell sorting. GFP fluorescence (shaded histograms) is not observed in untransfected cells (or in cells transfected with a vector not encoding GFP [unpublished data]). A Set of Ribosomes Is Scanning Specifically for the CUG/Leucine Start We next asked whether ribosomes responsible for the CUG/leucine start were scanning specifically for CUG initiation codons, or whether they were able to start at conventional AUG initiation codons as well. To address this question, we placed “decoy” ATG and CTG codons upstream of and out of frame with the CTG codon initiating expression of the peptide, and asked whether their presence affected peptide translation. As a positive control, we transfected cells with a construct encoding the ATG-initiated MYL8 and another encoding the same MYL8 peptide but with three ATGs upstream of and out of frame with the peptide (ATG) 3 ATG. The control constructs had CAGs instead of ATGs, because the CAG codon does not possess initiation activity. As expected, the presence of upstream out-of-frame ATGs dramatically reduced ATG-initiated MYL8 peptide expression. The reduction in MYL8 peptide was seen both when the transfected cells were used directly to stimulate T cells and when peptides from the transfected cells were extracted, separated by HPLC, and then assayed with T cells ( Figure 6 A and 6 B). This result is in complete agreement with the scanning model of translation initiation. The marked reduction in peptide expression occurs because ribosomes initiate translation at the first AUG and traverse the downstream peptide-coding region in the wrong reading frame; the ribosomes that missed the first AUG would start at the second or the third AUG and also traverse the peptide coding region in the wrong reading frames ( Bullock and Eisenlohr 1996 ; Bullock et al. 1997 ). Figure 6 Ribosomes Are Scanning Specifically for the CUG/Leucine Start (A, C, and E) Lmtk – cells were transfected with the indicated constructs and K b cDNA. After 2 d they were tested for MYL8/K b or LYL8/K b expression using the BCZ103 T cell hybridoma. Error bars represent the standard deviation of three replicate wells. (ATG) 3 ATG (solid circles, A) denotes the ATG-initiated peptide preceded by three ATGs upstream of and out of frame with the peptide; (CAG) 3 ATG (open circles, A) is the identical DNA construct but the upstream ATGs were replaced with CAG. (ATG) 3 CTG (solid circles, C) denotes the CTG-initiated peptide preceded by three ATGs upstream of and out of frame with the peptide; (CAG) 3 CTG (open circles, C) is the identical DNA construct but the upstream ATGs were replaced with CAG. (CTG) 3 CTG (solid circles, E) denotes the CTG-initiated peptide preceded by three CTGs upstream of and out of frame with the peptide; (CAG) 3 CTG (open circles, E) is the identical DNA construct but the upstream CTGs were replaced with CAG. (B, D, and F) Extracts from COS-7 cells transfected with cDNA encoding K b and the indicated constructs were separated by RP-HPLC. Fractions were tested for BCZ103 T cell-stimulating activity with K b+ B7.2 + L cells as APCs. Arrows indicate the peak elution positions of the MYL8 and the LYL8 peptides. Points on graphs correspond to those in (A), (C), and (E). We then transfected cells with a construct encoding (CTG)YL8 and another encoding (CTG)YL8 with three ATGs upstream of and out of frame with the peptide (ATG) 3 CTG. When the transfected cells were used directly to stimulate T cells, the upstream ATGs had little effect ( Figure 6 C). However, HPLC analysis of the peptides produced in transfected cells revealed that the upstream ATGs virtually abolished expression of the MYL8 peptide, whereas LYL8 expression was not affected ( Figure 6 D). Thus, upstream out-of-frame ATG initiation codons inhibited the conventional “wobble” CUG/methionine start, but not the CUG/leucine start. Finally, we transfected COS-7 cells with constructs encoding (CTG)YL8 with three CTGs upstream of and out of frame with the peptide (CTG) 3 CTG. When the transfected cells were used directly to stimulate T cells, we saw a small but consistent inhibition ( Figure 6 E). Remarkably, HPLC analysis of the transfected cell extracts showed that only LYL8 expression was decreased without any change in the expression of the MYL8 peptide ( Figure 6 F). The observation that upstream ATGs inhibit the wobble start but not the leucine start, and that upstream CTGs inhibit the leucine start but not the wobble start, suggests that a separate set of ribosomes is scanning for the CUG/leucine start, distinct from the ribosomes used for the methionine start. Note, however, that the upstream CUGs, despite an “Excellent Kozak” context, inhibited the leucine start weakly. This effect contrasts with the inhibition caused by the upstream AUGs on the AUG/methionine or the CUG/methionine starts and suggests that other features are required for an efficient CUG/leucine start. Interestingly, one form of the ASCT2 amino acid transporter is initiated with multiple CUG and GUG codons in close proximity ( Tailor et al. 2001 ). This redundancy may be required if any given non-AUG codon is used inefficiently. Furthermore, many mRNAs with CUG starts have GC-rich regions immediately downstream from the initiation codon and have been hypothesized to form hairpins that cause the ribosome to pause long enough to recognize the CUG codon ( Kozak 1990 ). The Leucine Start Is Enhanced in the Presence of Phosphorylated Eukaryotic Translation Initiation Factor 2α Finally, we were interested to know whether the leucine start requires eukaryotic translation initiation factor 2 (eIF2), which is responsible for loading the RNA i Met onto the 40S ribosome. Cells target eIF2 by phosphorylating its α subunit (eIF2α) to inhibit protein synthesis in response to a number of stress signals, including viral infection, starvation, and the accumulation of unfolded proteins. Ribosomes release eIF2-guanosine diphosphate (GDP) after the AUG initiation codon is reached, and GDP is exchanged for guanosine triphosphate (GTP) with the assistance of another protein, eIF2B, before eIF2 can be used for another round of translation initiation. When eIF2α is phosphorylated, it binds eIF2B with unusually high affinity and thus prevents subsequent nucleotide exchange. Because eIF2B is limiting in the cell, phosphorylation of only a fraction of eIF2α can substantially inhibit translation globally ( Hershey and Merrick 2000 ). To approach this question, we transfected HeLa cells with (CTG)YL8 or (ATG)YL8 constructs. We then assayed peptide expression in cells that had or had not been treated to induce phosphorylation of eIF2α. It was a challenge to induce phosphorylation of eIF2α for long enough to see an effect on peptide expression without causing substantial toxicity. Furthermore, we could not disrupt peptide/MHC assembly in the endoplasmic reticulum, a requirement that ruled out standard reagents such as dithiothreitol, thapsigargin, and tunicamycin. We also did not want to inhibit peptide elongation, which ruled out amino acid starvation. The optimal treatment for our purposes was sodium arsenite (NaAs). Arsenite reacts with sulfhydryl groups and causes phosphorylation of eIF2α presumably by inducing an unfolded protein response, although the precise mechanism remains unknown ( Brostrom and Brostrom 1998 ). We transfected HeLa cells with DNA encoding the CTG and ATG-initiated peptides. After 12 h, we treated the cells with 50 μM NaAs for 4 h before assaying them for expression of eIF2α and peptides. Western blot analysis confirmed that the amount of phosphorylated eIF2α in NaAs-treated cells was enhanced ( Figure 7 A). As a control, the amount of tubulin in the cells remained unchanged. Figure 7 The Leucine Start Is Enhanced in the Presence of Phosphorylated eIF2α HeLa cells transfected with cDNA encoding K b together with cDNA encoding either the ATG- or CTG- initiated peptides were treated for 4 h with 50 μM NaAs, with brefeldin A (BfA), or left untreated (UT). (A) Transfected cells treated with NaAs or without (UT) were lysed and tested for phosphorylation of eIF2α by Western blot and for tubulin as a loading control. (B) The transfected cells were titrated and tested for their ability to stimulate BCZ103 T cells. We assayed the HeLa cells for peptide expression using the BCZ103 T cell hybridoma ( Figure 7 B). We found that expression of the ATG-initiated MYL8 peptide was reduced by NaAs, as expected. The reduction was dramatic because the effect of NaAs was equivalent to the effect of brefeldin A, which stops protein transport from the endoplasmic reticulum to the Golgi and thus prevents any new peptides from trafficking to the cell surface. Surprisingly, expression of the CTG-initiated LYL8 peptide increased upon NaAs treatment, although it too was inhibited by brefeldin A treatment. This result was consistent in six independent experiments. The difference in ATG versus CTG initiation is not likely due to a unique effect of NaAs that somehow stabilizes LYL8 and not MYL8, because an alternative method of inducing phosphorylation of eIF2α using β-interferon and polyIC ( Savinova and Jagus 1997 ; Kaufman 2000 ), gave similar reproducible results (unpublished data). We conclude that, in contrast to initiation at conventional AUG codons, initiation at the CUG/leucine codon was eIF2α-independent. The effect of eIF2α phosphorylation on the leucine start strikingly mirrors the effect of eIF2α phosphorylation on proteins whose synthesis is directed by the CPV-IRES, which does not require eIF2 ( Wilson et al. 2000 ). Intriguingly, eIF2-independence of the CUG start is consistent with the observation of Donahue et al. (1988) that mutations in a Zn(II) finger domain of eIF2β permit initiation at a UUG codon. The data implicates the nucleic acid-binding function of eIF2β in start-site selection ( Donahue et al. 1988 ; Huang et al. 1997 ). Identification of the factors required for the CUG start will illuminate not only leucine initiation, but also the role of these components in the methionine start. The CPV-IRESs are to date the only known sequences that allow eIF2-independent initiation in eukaryotic cells. Viruses employ a host of creative strategies to prevent phosphorylation of eIF2α ( Kaufman 2000 ). Initiation at non-AUG codons, which is relatively common in viral transcripts, may provide a way to continue translation despite the lack of eIF2. Similarly, despite general translational inhibition during viral infection, cells need to continue generating antigenic peptides to flag down T cells. In addition to viral proteins and antigenic peptides, a number of regulatory cellular proteins have non-AUG initiation codons. For example, c-Myc has two distinct isoforms, the longer of which is initiated with CUG and may inhibit proliferation, as it is absent in a number of tumor-derived cell lines. Intriguingly, synthesis of the CUG-initiated form increases when cells reach high density, specifically when methionine is limiting ( Hann et al. 1988 ; Hann et al. 1992 ). It should be interesting to test whether this and other CUG-initiated proteins have a leucine start. In addition, whether other non-AUG initiation codons such as GUG or ACG are decoded in a manner similar to the CUG codon described here remains to be determined. In summary, we found that when CUG acts as an alternate initiation codon, it can be decoded as leucine as well as methionine. The leucine start does not depend on mRNA structure or sequence, but its efficiency can be enhanced by the Kozak context. A set of ribosomes is scanning 5′ to 3′ specifically for the CUG initiation codon. While the methionine start is inhibited when cells are treated with NaAs, the leucine start is enhanced, suggesting that leucine initiation is independent of eIF2. This novel translation initiation mechanism provides cells not only antigenic peptides but also a potential tool for translational control. Materials and Methods Cell lines Lmtk – , COS-7, K b+ B7.2 + L, D b+ B7.2 + L, BCZ103, 30NX/B10Z, and DBFZ cells have been described ( Mendoza et al. 1997 ; Malarkannan et al. 1999 ). Cell lines were maintained at 37 °C with 5% CO 2 in RPMI 1640 with 10% fetal bovine serum, 1 mM sodium pyruvate, 50 μM β-mercaptoethanol, 0.3 mg/ml glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin (normal medium). HeLa cells were obtained from ATCC (#CCL-2; Manassas, Virginia, United States), and cultured in Eagle's minimal essential medium modified with Earle's balanced salt solution, nonessential amino acids, 2 mM L-glutamine, 1 mM sodium pyruvate, 1500 mg/l sodium bicarbonate (ATCC #30–2003), 10% fetal bovine serum, and 10 μg/ml ciprofloxacin. Plasmid construct sequences Sequences for constructs depicted in Figure 1 are as follows. CTG-YL8 in the BstXI/XbaI sites of the pcDNA1 vector (Invitrogen, Carlsbad, California, United States),
5′-TGTGTAGCTGACCTTCAACTACCGGAATCTATAGCTAG-3′; CTG-YL8 in the EcoRI/BamHI sites of the pIRES2-eGFP vector (Clontech, Palo Alto, California, United States),
5′-AATTAGACGAAGGTCTAGCTGACCTTCAACTACCGTAACCTGTAGATC-3′; CTG-YL8 in the HindIII/BamHI sites of the pcDNA1 vector with GFP in the EcoRI site,
5′-AGCTAGCTGACCTTCAACTACCGGAATCTATAGATCGATC-3′; ATG-YL8 in the EcoRI/BamHI sites of the pIRES2-eGFP vector,
5′-AATTAGACGAAGGTCTAGATGACCTTCAACTACCGTAACCTGTAGATC-3′; CTG-M9 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGCTGAGCAACGAGAACATGGAGACCATGTAGTGCACTAG-3′; and ATG-M9 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGATGAGCAACGAGAACATGGAGACCATGTAGTGCACTAG-3′. Sequences for constructs depicted in Figure 2 are as follows. R0 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGATGAGCAGCGTCGTCGGCGTTTGGTACCTCTAGCTGACCTTCAACTACCGGAATCTCTAG-3′; and R1 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGATGGAGCAGCGTCGTCGGCGTTTGGNACCTCTAGCTGACCTTCAACTACCGGAATCTCTAG-3′. Sequences for constructs depicted in Figure 3 are as follows. CTG-initiated peptide in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGNCCNCCCTGNCCAGTGTTGTTGAATTCTCCAGCCTCTAG-3′; and CCC-initiated peptide in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGNCCNCCCCCNCCAGTGTTGTTGAATTCTCCAGCCTCTAG-3′. Sequences for constructs depicted in Figure 4 are as follows. “Excellent Kozak” CTG-YL8 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGTCGACCCTGACCTTCAACTACCGGAATCTCTAG-3′; and “Poor Kozak” CTG-YL8 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGGCGTCCCTGACCTTCAACTACCGGAATCTCTAG-3′. Sequences for constructs depicted in Figure 5 are as follows. CTG-YL8 or ATG-YL8 in the pIRES2-eGFP vector, as in Figure 1 hairpin CTG-YL8 or ATG-YL8 in the pIRES2-eGFP vector (sequence from 5′ cap to start of peptide, whose sequence is as in Figure 1 ):
5′-GCTAGCGCTACCGGACTCAGATCGTGTCCGGATTTGGGGCGCGTGGTGGCGGCTTTTCGCGCGCGCGACGCGTCGCGCGCGCGTTTTGCCGCCACCACGCGCCCCTTTAGTACTTGAGCTCAAGCTTCGAATTAGACGAAGGTCTAG-3′. Sequences for constructs depicted in Figure 6 A and 6 B are as follows. CAG 3 ATG-YL8 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGAACCCAGGGTCGACCCAGGACCCAGGTAGTCGACCATGACCTTCAACTACCGGAATCTCTAG-3′; and ATG 3 ATG-YL8 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGAACCATGGGTCGACCATGGACCATGGTAGTCGACCATGACCTTCAACTACCGGAATCTCTAG-3′. Sequences for constructs depicted in Figure 6 C and 6 D are as follows. CAG 3 CTG-YL8 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGAACCCAGGGTCGACCCAGGACCCAGGTAGTCGACCCTGACCTTCAACTACCGGAATCTCTAG-3′; and ATG 3 CTG-YL8 in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGAACCATGGGTCGACCATGCACCATGGTAGTCGACCCTGACCTTCAACTACCGGAATCTCTAG-3′. Sequences for constructs depicted in Figure 6 E are as follows. CAG 3 CTG-YL8 in the HindIII/XbaI sites of the pcDNA1 vector mutated to alter the two CTGs in the 5′ UTR to CAG,
5′-AGCTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTTAGAAACTGACCTTCAACTACCGGAATCTCTAG-3′; and CTG 3 CTG-YL8 in the HindIII/XbaI sites of the pcDNA1 vector mutated to alter the two CTGs in the 5′ UTR to CAG,
5′-AGCTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACTGGATTACAAACTGGATTACAAACTGGATTTAGAAACTGACCTTCAACTACCGGAATCTCTAG-3′ Sequences for constructs depicted in Figure 6 F are as follows. CAG 3 CTG-YL8 in the HindIII/XbaI sites of the pcDNA1 vector,
5′-AGCTACAAACAGGNTTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTTAGAAACTGACCTTCAACTACCGGAATCTCTAG-3′; and CTG 3 CTG-YL8 in the HindIII/XbaI sites of the pcDNA1 vector,
5′-AGCTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACAGGATTACAAACTGGATTACAAACTGGATTACAAACTGGATTTAGAAACTGACCTTCAACTACCGGAATCTCTAG-3′. Sequences for constructs depicted in Figure 7 are as follows. CTG-YL8 in the in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGTAGCTGACCTTCAACTACCGGAATCTGTAGCTAG-3′; ATG-YL8 in the in the BstXI/XbaI sites of the pcDNA1 vector,
5′-TGTGAAAAACAGGCCAAACAGGCCAAACAGGAAGATGACCTTCAACTACCGGAATCTCTAG-3′. Transfections The DEAE-dextran transfection method was used in Figures 2 B and 2 C, 3 , 4 A, 6 A, 6 C, and 6 E. It has been described previously ( Serwold et al. 2001 ). Briefly, Lmtk – or COS-7 cells were transfected in 96-well plates (1 × 10 4 cells per well) with the indicated concentration of plasmid DNA, 0.1 mg/ml DEAE-dextran, 0.1 mM chloroquine, 10 ng/ml MHC class I cDNA, 5 ng/ml B7.2 cDNA, and 10% NuSerum (Collaborative Biomedical Products, Becton Dickinson, Bedford, Massachusetts, United States) in RPMI 1640. After 90 min, the cells were shocked with 10% DMSO in PBS for 2 min. After 2 d incubation in normal medium, T cells (1 × 10 5 cells per well) were added to assay peptide expression. GeneJuice (Novagen, Madison, Wisconsin, United States) transfection reagent was used in Figures 1 , 2 D, 4 B, 5 , 6 B, 6 D, 6 F, and 7 . It was used according to the manufacturer's instructions, except that we used 8 ml of medium, 8 μl of GeneJuice, 1.33 μg of MHC class I DNA, and 1.33 μg of peptide DNA per 10-cm dish. Peptide expression in the transfected cells was assayed after 2 d by titrating intact cells or after fractionating the cell extracts by HPLC. T cell assay The T cell assay has been described previously ( Sanderson and Shastri 1994 ). Briefly, Lac Z-inducible T hybridomas (1 × 10 5 cells) were incubated with APCs in 96-well plates for 6–24 h. The response, accumulation of intracellular β-galactosidase, was measured with the substrate chlorophenol red-β-D-galactopyranoside (CPRG). The product was measured with a 96-well plate reader at 595 nm and 655 nm as the reference wavelength. HPLC The HPLC assay has been described previously ( Serwold et al. 2001 ). Briefly, 2 d after transfection, cells were lysed in 10% formic acid, and the lysate was passed through a 10 kDa cutoff filter (Millipore, Bedford, Massachusetts, United States). The <10 kDa fraction was injected onto a 2.1 × 250 mm C18 column (Vydac, Hesperia, California, United States) and separated by RP-HPLC, with 0.1% TFA in water as the polar buffer and 0.1% TFA in acetonitrile as the nonpolar buffer. Three-drop fractions (except analysis of xM9, with five-drop fractions) were collected in 96-well plates, dried in a vacuum centrifuge, and analyzed by the addition of APCs (5 × 10 4 per well) and T cell hybridomas (1 × 10 5 cells per well). eIF2α phosphorylation Cells were transfected with GeneJuice (Novagen) as described above, but the transfection medium was left on for only 4 h before the cells were lifted and split into multiple dishes with fresh medium. Cells were allowed to rest for 12 h before sodium arsenite treatment. They were treated with 50 μM NaAs (Sigma, St. Louis, Missouri, United States), 1× GolgiPlug containing brefeldin A (PharMingen, San Diego, California, United States), or left untreated for 4 h. The cells were then lifted and counted. For the T cell assay, they were titrated in a 96-well plate. The assay was as described above, except that 1× GolgiPlug was added in order to “freeze” the cells in their state at the end of treatment. For the Western blot, they were incubated on ice for 5–10 min in lysis buffer (20 mM HEPES [pH 7.5], 150 mM NaCl, 1% Triton X-100, 10% glycerol, 1 mM EDTA, 10 mM tetrasodium pyrophosphate, 100 mM NaF, 17.5 mM β-glycerophosphate, 0.4 U/ml aprotinin, 10 μg/ml leupeptin, 1 mM PMSF, 0.1 mM pepstatin A, and complete protease inhibitor cocktail [Roche, Basel, Switzerland]). The lysate was spun for 15 min at 4 °C, and the supernatant was transferred to a tube containing an equal volume of 2× SDS-PAGE sample buffer (100 mM Tris-HCl [pH 6.8], 20% glycerol, 4% SDS, bromophenol blue, and 5% β-mercaptoethanol). The sample was then heated in water just off the boil for 5 min, separated on a 10% SDS-PAGE gel, and transferred to a nitrocellulose membrane. The membrane was blocked for 1 h at room temperature in TBS–0.1% Tween 20–5% bovine serum albumin (TBS, 0.02 M Tris-HCl [pH 7.6] with 0.137 M NaCl), incubated for 1 h with primary antibody (#44–728, at a 1:2000 dilution; Biosource, Camarillo, California, United States) in TBS–0.1% Tween 20–5% bovine serum albumin, washed four times for 5 min with TBS–0.1% Tween 20, incubated for 40 min with secondary antibody (anti-rabbit-HRP #NA934V, at a 1:30,000 dilution; Amersham, Little Chalfont, United Kingdom), washed four times for 5 min with TBS–0.1% Tween 20, incubated for 5 min in substrate (SuperSignal West Femto Maximum Sensitivity Substrate, #34095; Pierce Biotechnology, Rockford, Illinois, United States), and exposed to film. An antibody to α-tubulin (#sc-5546; Santa Cruz Biotechnology, Santa Cruz, California, United States) was used as a loading control. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC524250.xml |
517498 | Liver, spleen, pancreas and kidney involvement by human fascioliasis: imaging findings | Background Fasciola hepatica primarily involves the liver, however in some exceptional situations other organs have been reported to be involved. The ectopic involvement is either a result of Parasite migration or perhaps eosinophilic reaction. Case presentation Here we report a known case of multiple myeloma who was under treatment with prednisolone and melphalan. He was infected by Fasciola hepatica, which involved many organs and the lesions were mistaken with metastatic ones. Discussion Presented here is a very unusual case of the disease, likely the first case involving the pancreas, spleen, and kidney, as well as the liver. | Background Human fascioliasis, a commonplace infection caused by a leaf-shaped Trematode Fasciolahepatica affects a human host by chance [ 1 - 4 ]. It seems that the rate of infection is increasing in many countries worldwide [ 1 , 5 ]. Human fascioliasis has to be differentially diagnosed from such hepatic and biliary diseases as acute hepatitis, neoplasm, visceral toxocariasis, biliary tract diseases, hepatic amebiasis, and infection with other liver flukes like schistosomiasis [ 2 , 5 ]. Diagnosis of the disease is achieved by locating the ova either in feces or duodendal drainage or by [ 7 - 9 ]. Imaging techniques proved to be the most useful method for confirming the diagnosis and also the follow-up of fascioliasis [ 6 , 8 , 10 ]. Small, often peripheral, nonenhancing, hypo dense nodules with tortuous, linear, branching tracts in CT scans, which decrease in size after successful therapy, are highly suggestive of the disease [ 8 - 12 ]. Immature flukes can produce ectopic masses or abscesses in various locations and during the acute phase of the disease, other structures such as subcutaneous tissue, heart, lungs, pleura, abdominal wall, brain, cecum, epididymis, and stomach can be involved [ 2 , 8 , 9 ]. In a particularly unusual report, direct peritoneal involvement with granuloma formation has been reported [ 1 ]. Eosinophilic reactions can be exhibited in the body as pleuritis and pericarditis [ 7 , 9 ]. Here, we present an extremely unusual radiologic incident of fascioliasis involving multiple organs. Case presentation In Dec 2001, a 52 year old man, who was a known case of Multiple Myeloma (MM), was presented to one of our affiliated hospitals with persistent right upper quadrant and epigastric pain, and anorexia for a period of 1 month. At the time of admission, the patient had been receiving prednisolone and melphalan for his MM, which was currently in remission. His recent condition began with tongue and facial edema two weeks before appearance of the abdominal pain. Upon physical examination, mild epigastric tenderness and a palpable liver were found. Neither icterus nor any positive sign of cardiopulmonary abnormalities were noted. Additionally, the patient did not have a fever and his peripheral lymph nodes were not enlarged. Initial laboratory findings were as follows: a hemoglobin of 10.6 g/dL, white blood cell count of 10,800/mm3 with 18% eosinophils, and a sedimentation rate of 90 mm at the end of the first hour. The total bilirubin was 0.5 mg/dL (0.2–0.8 mg/dl), alanine aminotranferase (ALT) 48 IU/L (0–40 IU/L), aspartate aminotransferase (AST) 46 IU/L (0–40 IU/L), and alkaline phosphatase (Alk P) 651 IU/L (60–140 IU/L). The eosinophilia fluctuated between 12 to 55 percent in various tests performed during the time period in question, with no unique patterns noted. Serum electrophoresis showed a monoclonal spike in the gamma region. Specific enzyme-linked immunosorbant assay (ELISA) produced a positive result for Fasciola hepatica, while the test was negative for Toxocara canis. Serologic tests for the presence of hepatitis A, B, and C viruses were negative. Blood and urine cultures were found to be sterile. Other laboratory studies, including repeated stool examinations for ova and parasites, showed no abnormalities. Chest x-rays did not demonstrate any parenchymal or pleural abnormality. Abdominal ultrasonography showed a mild hepatomegaly with multiple hypoechoic lesions in the liver. A CT scan revealed multiple but poorly defined, hypodense lesions in the liver, and a completely enlarged pancreas with mild bilateral pleural reaction, suggesting metastatic cancer (Fig. 1). In the search for a potential malignancy, diagnostic laprascopy was performed, which revealed the presence of white-colored lesions ranging from 1 to 3 cm in diameter on the surface of both lobes of the liver with mild ascites. Multiple liver and peritoneal biopsy specimens revealed fibrinoid necrosis, associated with granulomatous reaction and a high concentration of eosinophils in the liver, (Fig. 2) accompanied by markedly inflamed peritoneal tissue with eosniphilic infiltration. No malignant cells were identified and no evidence of extramedulary plasmocytoma was found. Specific staining for fungal organisms and acid-fast bacilli were negative. The ascitic fluid also had a high level of eosinophils. Endoscopic retrograde cholangiopancreatography (ERCP) failed to show any filling defect within the biliary tree. Furthermore, the patient underwent bone marrow aspiration that only indicated high eosniophilic infiltration. The patient was placed on albendazole (400 mg twice daily for 1 week). The treatment was well tolerated and the abdominal pain was improved rapidly. At the time of discharge, the patient was in good clinical condition. During a follow-up visit two months later, a second CT unexpectedly showed not only an increase in the number and size of the hypodense lesions in the liver, but also the extension of lesions into the pancreas, the spleen and both kidneys (Fig. 3). No evidence of peripheral enhancement of the hepatic lesions or ascites was documented. The patient was still experiencing upper quadrant pain on the right side. Laboratory investigations produced a white blood cell count of 7200/mm3 with 16% eosinophilia. Repeated stool examinations failed to identify ova and parasites. The patient was given triclabendazole (10 mg/kg, bid for two days). As recommended, the patient had another follow-up CT scan three months later. At that time all of his symptoms were resolved. Follow-up CT scans revealed a considerable improvement in the number and size of the lesions. At the time of the CT scan, all of his symptoms were resolved (Fig. 4A, B). At this time the WBC was 6000/mm3 with 6% eosinophils. After 5 months and in the last CT, the lesions had almost disappeared completely (Fig. 4C, D). Discussion While fascioliasis is a well-known human parasite, it sometimes produces unusual characteristics that may influence a clinician to misdiagnose the condition. In the vast majority of cases, the diagnosis is difficult in both acute and chronic phases and some important conditions such as liver abscesses and metastasis cannot be easily differentiated from fascioliasis [ 1 , 13 ]. Interestingly, the larvae is able to migrate to a number of ectopic locations such as subcutaneous areas, intestines, pleura, lungs, abdominal wall, brain, cecum, epididymis, stomach, pericardial, or cerebral sites, producing very unique clinical manifestations [ 8 , 9 ]. Serologic testing, when performed by an enzyme linked immunoabsorbent assay (ELISA) for the detection of antibodies specific to the parasite is nearly one hundred percent sensitive and specific. Thus ELISA may be used to confirm the diagnosis in acute and chronic phases [ 7 - 9 ]. Among imaging tools, ultrasonography is of little diagnostic value during the acute phase, while a contrast-enhanced CT scan can be very useful for diagnosis [ 9 , 17 ]. In CT images one can find two distinct kinds of lesions: single or multiple hypodense nodular areas caused by deposition of the parasite (abscess-like lesions) and tunnel-like branching or tortuous hypodensity which is created as the result of parasite migration through the liver and is highly suggestive of the disease [ 9 , 17 ]. If peripheral tortuous lesions are present, hepatic fascioliasis should be the primary diagnostic consideration [ 14 ]. As clinical and laboratory findings of fascioliasis may easily be confused with many other conditions, a high index of suspicion is required to establish a correct diagnosis [ 9 , 17 ]. Both CT scan and ultrasonography can be helpful in evaluating the response to treatment [ 15 ]. The ingested metacercariae of Fasciola hepatica penetrate the intestinal wall and migrate through the peritoneal cavity to reach the liver. However, ectopic migration to other locations is one of the strangest manifestations of the infection [ 8 , 9 ]. The precise route of migration toward ectopic sites is unclear but most often occurs in the acute stage of the disease [ 9 ]. In addition, a syndrome of eosinophilic reaction without direct parasitic involvement may accompany acute fascioliasis [ 9 ]. Ascites is not a common finding in fascioliasis, but it is not unheard of, as the peritoneum is the usual route of migration of the parasite towards the liver. Kabaalioglu A. et al reported mild splenomegaly and enlargement of the left rectus abdominalis muscle in hepatic fascioliasis [ 17 ]. In a study the authors found that among pleuro-pulmonary diseases, parenchymal infiltrates resembling the Loeffler syndrome and pleural effusion were the most common radiological features [ 8 ]. Radiologic manifestations of fascioliasis in a case with peritoneal involvement have been described as low-density lesions in the mesentery [ 18 ]. While the radiologic presentations of the disease are contradictory, reporting any new set of images related to the unusual organ involvement is of paramount importance. The case presented here is probably the first report of fascioliasis with spleen, pancreas, and kidney involvement. Migration of Fasciola hepatica to the spleen and pancreas and perhaps the kidneys was considered as the cause of the imaging findings in this patient, however hypersensitivity reaction to Fasciola antigens might also be implicated. Pathology examination to confirm the ectopic locations of the involvement was not performed, however improvement of the lesions on follow-up imaging studies after treatment defended the involvement of those regions by the parasite The compromised immunity of the patient linked to the use of immunosuppressive medicines could partly explain the extensive extra-hepatic involvement of the disease. The patient responded fully to Triclabendazole and is now healthy. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC517498.xml |
518967 | Frequency of myasthenic crisis in relation to thymectomy in generalized myasthenia gravis: A 17-year experience | Background Myasthenic crisis is the most serious life-threatening event in generalized myasthenia gravis (MG) patients. The objective of this study was to assess the long-term impact of thymectomy on rate and severity of these attacks in Iranian patients. Methods We reviewed the clinical records from 272 myasthenic patients diagnosed and treated in our neurology clinic during 1985 to 2002. Fifty-three patients were excluded because of unconfirmed diagnosis, ocular form of MG, contraindication to surgery, concomitant diseases and loss to follow-up. The Osserman classification was used to assess the initial severity of the disease. Frequency and severity of the attacks were compared between two groups with appropriate statistical tests according to the nature of variables. Multivariate logistic regression analysis was used to assess the predictors of myasthenic crisis in the group of patients without thymoma. Results 110 patients were in thymectomy group and the other 109 patients were on medical therapy. These two groups had no significant differences with respect to age at onset, gender, Osserman score in baseline and follow up period. 62 patients (28.3% of all 219 patients) had reported 89 attacks of myasthenic crisis. 20 patients of 62 (32%) were in thymectomy group and 42 (68%) were in the other group. There was significant difference between the two groups in number of patients with crisis (P = 0.001; odds ratio = 2.8 with 95% CI of 1.5 to 5.2). In addition, these attacks were more severe in group of non-thymectomized patients as the duration of ICU admission was longer and they needed more ventilatory support during their attacks. Regression model showed thymectomy and lower age at onset as two predictors of decrement in myasthenic crisis rate in non-thymomatous MG patients. Conclusions It is suggested that frequency and severity of myasthenic attacks as important endpoints in evaluation of MG patients. Thymectomy seems to have a preventive role on rate and severity of these attacks. | Background Myasthenia Gravis (MG) is the mostly known autoimmune disease in human, mediated by autoantibodies against the nicotinic receptor of acetylcholine in neuromuscular junctions [ 1 ]. Two cardinal features of this disease are muscle weakness and fatigability. Although periorbital area contains the only affected muscles in some patients, generalized weakness develops in approximately 85% of patients, affecting bulbar and limb muscles as well as the neck extensors and the diaphragm. Myasthenic crisis refers to a rapid deterioration in neuromuscular function with respiratory compromise due to ventilatory muscle insufficiency or weakness of upper airway musculature or both [ 1 ]. These attacks might be triggered by multiple factors, including infections, physical or emotional stresses, aspiration, electrolyte disturbances, changes in medications and inadvertent administration of non-depolarizing muscle relaxants or other drugs. In this state, respiratory function should be monitored closely for evidence of respiratory failure and ventilatory support should be initiated in the setting of emerging respiratory failure [ 2 , 3 ]. In general, four methods of treatment are currently in use [ 4 ]: anticholinesterase agents, immunosuppressives, surgical thymectomy, and short-term immunotherapies, including plasma exchange and intravenous immune globulin (IVIG). Since the first operation for MG in 1911 [ 5 ], thymectomy has become an increasingly accepted procedure for treatment of MG, as it can help to achieve complete clinical remission rate between 18% and 50% [ 6 - 9 ] and clinical improvement in majority of patients [ 10 ]. The rationale for thymectomy is that about 75% of MG patients have thymic abnormalities; of these, 85% have hyperplasia and 15% have thymoma [ 11 ]. Although there is no convincing evidence regarding the use of thymectomy in non-thymomatous MG patients [ 12 , 13 ], most authorities recommend this procedure in all patients of ages between puberty and 60's [ 14 , 15 ]. However, this is not a general idea. Thymectomy effects on occurrence and intensity of respiratory crisis have not been studied yet. Concerning value of this life-threatening event in MG patients, we studied the role of thymectomy on long-term frequency and severity of these attacks. Methods We reviewed the clinical records from 272 myasthenic patients diagnosed and treated in the neurology clinic of Shariati Hospital, Tehran, during 1985 to 2002. All included patients had positive history and physical examination, confirmed with positive Tensilon test and electromyography (with repetitive nerve stimulation test). Searching for thymomas, all patients had a chest computed tomography (CT) scan. The following patients were excluded from the study: 3 patients with doubtful test results and unconfirmed diagnosis; 25 patients with ocular form of MG (as these patients have no presentation of crisis); 2 patients with thymoma (and two episodes of crisis in the previous year), but contraindicated for surgery because of coronary artery disease; 6 patients with thyroid function abnormalities (because of their known effects on induction and reduction of myasthenic crises); 2 patients with concomitant rheumatoid arthritis (and consumption of high doses of corticosteroids and immunosuppressives); and 15 patients lost from follow-up with incomplete data. 219 MG patients enrolled in the study. Myasthenic crisis was defined as an attack that compels physicians to hospitalize the patient and carefully supervise patient's respiratory capacity [ 3 ]. In most of these conditions, ventilatory support was needed. However, sometimes patient could be managed without intubation as a result of spontaneous improvement or use of plasmapheresis or IVIG. All included patients were eligible for surgery. Patients with CT confirmed thymoma underwent elective thymectomy as soon as possible. In the first visit after diagnosis confirmation in non-thymomatous patients, the advantages and drawbacks of operation (including risks of anesthesia in MG patients) had been explained to them and deciding between the two choices of thymectomy or conservative therapy had been left to the agreement between physicians and patients. No particular protocol had been established for these patients and physicians had only a consultant role for patients with no force on them. For clinical assessment of the initial severity of the disease, the classification of Osserman [ 16 ] had been applied to all patients in the first visit and inserted into patient's clinical profile. In following visits, patients' condition has been referred to this initial assessment. Data were collected on demographic factors, disease course at the first year of onset (or the year before surgery in thymectomized patients), time of surgery, pathology report, number of myasthenic crises (post-op in thymectomized patients), precipitating factors for attacks, ICU admission and the need for respiratory support, plasmapheresis or IVIG in each attack. 12 patients in thymectomy group had history of myasthenic crisis before surgery that were not considered. The participants' baseline characteristics and follow-up data were analyzed by Student's t test and analysis of variances (ANOVA) as parametric tests, Mann-Whitney U test as non-parametric test, and Pearson's chi-square test for qualitative differences (SPSS software, version 10.0). P values less than 0.05 were taken to indicate statistical significance. Continuous variables are shown as the mean ± standard deviation (SD) and non-parametric variables with their median and range. Odds ratios and relative risks with 95% confidence intervals (95% CI) were calculated to assess the proportional risk of crisis between two groups. In the subgroup of patients without thymoma, multivariate logistic regression analysis was used to assess the relationship between frequency of myasthenic crisis and several risk factors such as sex, start age of the disease, Osserman score and thymectomy. According to its percentile, start age was divided into three categories as bottom one-third (under percentile 33), middle one-third (between percentiles 34 and 66), and top one-third (over percentile 67). Results 110 patients underwent thymectomy (76 by trans-sternal and 34 by trans-cervical approaches) and 109 patients were on conservative therapy. Baseline characteristics and follow-up periods are shown in Table 1 . Time interval between onset of myasthenic symptoms and thymectomy ranged from 1 to 192 months with median of 12. Mean age at thymectomy was 30.3 ± 12.8 years. During follow-up period, each patient had about 13 visits on average. Table 1 Characteristics of thymectomized patients just before thymectomy and non-thymectomized patients at the first year of the diagnosis Thymectomized patients (n = 110) Non-thymectomized patients (n = 109) P value Age at diagnosis (years) 29.2 ± 13.7 33.0 ± 15.9 0.059 Gender (male/female) 41/69 53/56 0.089 Osserman Score* = I 0 0 Osserman Score = IIa 46 48 Osserman Score = IIb 37 50 Osserman Score = III 25 11 Osserman Score = IV 2 0 Follow up period (years) 6.4 ± 4.3 7.9 ± 5.6 0.134 * Mann-Whitney U test showed no significant difference between two groups (Z = |1.421| & P = 0.155) Totally, 62 patients in both groups experienced 89 attacks of respiratory failure during their follow-up, which accounts for 28.3% of the study population. As it is shown in Table 2 , 18.2% of thymectomized patients (20 of 110) and 38.5% of non-thymectomized patients (42 of 109) had history of these attacks. The difference between two groups was significant in this regard. Additionally, Odds ratio for being affected twice or more among patients on conservative therapy was 4.2 (with 95% CI of 1.1 to 16.7). Three patients in thymectomy group had history of crisis during post-op period in the hospital. Main triggering factors for myasthenic crises were lack of compliance to the drugs, pneumonia, and unknown causes Table 3 . Table 2 relation between myasthenic crisis frequency and thymectomy Thymectomized patients (n = 110) Non-thymectomized patients (n = 109) P value Patients with myasthenic crisis (n = 62) 20 42* 0.001 Episodes of myasthenic crisis (n = 89) 25 64 † 0.016 * This group had more probability of having crisis with an odds ratio of 2.8 (with 95% confidence interval of 1.5 to 5.2) † Myasthenic crises were more prevalent in this group with a relative risk of 2.6 (with 95% confidence interval of 1.8 to 3.8) Table 3 Characteristics of patients with myasthenic crisis in thymectomized and non-thymectomized patients* Thymectomized patients (n = 20) Non-thymectomized patients (n = 42) Age at diagnosis (years) 31.4 ± 10.6 30.5 ± 12.7 Age at first crisis (years) 33.5 ± 11.2 34.1 ± 11.3 Gender (male/female) 6/14 20/22 Triggering factors Pneumonia 3 (15%) 13 (30%) Other infections 2 (10%) 4 (10%) Aspiration 2 (10%) 3 (7%) Stresses † 1 (5%) 2 (5%) Drug intolerance 8 (40%) 10 (24%) No obvious cause 4 (20%) 10 (24%) * There were no statistical differences between two groups † Refers to physical and emotional stresses as well as menstruation To assess the severity of each attack, we considered duration of ICU admission (days), need to respiratory support and need to plasmapheresis or IVIG during crisis as indicating variables. As it is shown in Table 4 myasthenic crises were almost more severe in patients under conservative therapy. Two patients (both on conservative therapy) had passed away during respiratory failure (mortality rate = 2/89 = 2.2%) and these conditions were so protracted in some of patients that they were confined to ICU beds for more than one month. Table 4 Severity parameters for each attack in patients with myasthenic crisis Thymectomized patients (n = 25) Non-thymectomized patients (n = 64) P value* Median Days in ICU (Range) 7(2–38) 12(2–55) 0.044 Need to Ventilatory support 19 (76%) 59 (92%) 0.037 Need to Plasmapheresis or IVIG 7 (28%) 23 (36%) 0.476 * P values from chi-square test for assessment of difference between two groups Pathology reports in thymectomized patients revealed 16 cases of thymoma. Fifty percent of these patients (8 of 16) had experienced crisis, which accounts for 40% (8 of 20) of thymectomized patients with history of crisis. Mann-Whitney U test revealed significant difference in this regard (P < 0.001). Severity indexes did not differ significantly between thymomatous and non-thymomatous patients Table 5 . Table 5 Characteristics of thymectomized patients with respect to pathology of thymus Thymus Pathology N (% of 110) Age at diagnosis* (years) Sex (m/f) Patients with Myasthenic crisis (%) Number of Myasthenic crisis Days in ICU (Median & Range) Ventilatory support Normal 51 (46.4%) 29.7 ± 13.9 18/33 3 (5.9%) 3 3 & 2–4 3 Hyperplasia 43 (39.1%) 27.4 ± 15.2 14/29 9 (20.9%) 10 7 & 2–30 7 Thymoma 16 (14.5%) 37.3 ± 8.9 9/7 8 (50.0%) 12 9 & 3–38 9 * Analysis of variances indicated that age at diagnosis in thymomatous patients was significantly higher than others (P = 0.007). There was no statistically significant difference regarding other variables. Totally, there were 203 patients without thymoma. Regression analysis in this subgroup of patients revealed that group of non-thymectomized patients as well as two third of patients with higher ages at the beginning of the disease were more prone to respiratory attacks Table 6 . Table 6 logistic regression model for predictors of myasthenic crisis among patients without thymoma (n = 203). Predictors Odds Ratio Lower 95% CI Upper 95% CI P value Start age (years) 1.016 0.996 1.038 0.085 Sex (male) 0.926 0.482 1.778 0.612 Thymectomy (performed) 0.234 0.113 0.484 <0.001 Osserman classification (score III) 1.269 0.410 3.928 0.529 Start age category (one third of younger patients) 0.272 0.097 0.759 0.021 Discussion Persuasive evidence for thymectomy in non-thymomatous generalized myasthenic patients is not on hand. In an attempt to establish standards in this regard, American Academy of Neurology advised thymectomy as only an option to increase the probability of remission or improvement in these patients [ 13 ]. In this review, thymectomy was associated with a median relative rate of medication free remission of 2.1 and a relative rate of improvement of 1.7. However, these improvements were significant in only a small number of the studies reviewed and the majority did not show a significant benefit with thymectomy. In addition, none of the studies was randomized or used blinded outcome assessments, and in most thymectomy was performed in younger patients. Surprisingly, none of these studies has considered myasthenic crisis frequency as their study endpoint. For instance, in the study of Cohen et al. in 1981, 15 of 28 thymectomized patients experienced 21 crises during follow-up period [ 17 ]. In another study, 13 of 27 patients with crisis had previous thymectomy, six with thymoma [ 18 ]. This study was the first one that sought thymectomy impact on occurrence of crisis. In our experience, persistence on thymectomy for all myasthenic patients would not be so wisely. Response to thymectomy is highly variable among these patients. Method of gaining agreement between physicians and patients would bring some benefits. As it is shown in Table 3 , rate of intolerance to drugs was lower among the patients who preferred medical therapy. However, it is notable that there is a general tendency among both physicians and patients towards operation in younger ages and in patients with more severe course. In this study, thymectomized patients were about 4 years younger on average and had higher Osserman scores at the beginning (Table 1 ); but the difference was not statistically significant because of large study population and dispersion of baseline characteristics. When myasthenic crisis occurrence is considered as the main endpoint of treatment in myasthenic patients, thymectomy has a great impact. Patients on conservative therapy had risk of about 3-fold of thymectomized patients to encounter respiratory attacks. However, the overall frequency of myasthenic crisis among our study population could be considered more than similar reports [ 2 , 4 ], perhaps reflecting some unidentified racial differences or lower compliance of Iranian patients to continuous medications. Some previous studies have shown the benefit of thymectomy in the control of symptoms of MG. It has been revealed that delay in surgery could worsen the prognosis of MG [ 19 ] and the chance of benefiting from thymectomy increases when the history of MG is short and the stage of the disease is early [ 20 ]. Concerning differences of thymectomy effects in thymomatous and non-thymomatous patients, several studies have been conducted previously, showing some discrepancies in the results [ 21 - 23 ]. According to this study, thymomatous patients are more prone to respiratory attacks compared with all other MG patients. However, their attacks are not drastically different in severity of complications. During myasthenic crisis, Weakness of the respiratory muscles may be out of proportion to that of other skeletal muscles. In rare cases, ventilatory failure is the only clinically apparent manifestation of the disease [ 24 , 25 ]. In addition, sometimes prediction of incoming attack is very difficult. In the present study, a 27-year-old woman experienced an attack after 7 months of complete remission on no drug. Generally, infection is the most common trigger for myasthenic crisis [ 1 ]. Higher rate of infection in non-thymectomized patients in this study could be logically due to higher amounts of immunosuppressives in the medication regimen of these patients that we could not evaluate in this study. However, concerning frequency of myasthenic crises, this reduction of immunosuppressives in the regimen of thymectomized patients could be interpreted as an important benefit of thymectomy in MG patients. The main limitation of this study was its retrospective design and the lack of randomization. Apparently, within a period of 17 years, many changes have occurred to diagnostic and therapeutic facilities as well as knowledge and attitude of physicians. These changes could have significant consequences on the outcome of myasthenic patients. Although various operative approaches seem not to differ drastically [ 26 ], various types of immunosuppressives and dosage of them have important role in the prognosis of patients, which we did not assess them in this study. We also could not gather information about acetylcholine receptor antibodies, which play an important role in the pathophysiology of MG. Further studies concerning the role of these antibodies in the occurrence and severity of myasthenic crises are required. Conclusions In conclusion, we suggest evaluation of various aspects of myasthenic crisis as an important endpoint in long-term assessment of generalized MG patients. Moreover, we showed that thymectomy has a preventive role on rate and severity of myasthenic attacks in age and disease severity matched groups of patients. However, this effect needs further evaluation in upcoming prospective studies. Competing interests None declared. Authors' contributions In advance, suggestion of the design of the study was from our professor AkS. Data extraction and initial analysis were done by AlS and HT. SA participated in the design and implementation of the study. AM performed additional analyses and wrote the first draft of the paper. MS and SA both had helpful and valuable comments in revising the paper. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC518967.xml |
544942 | The role of the applied epidemiologist in armed conflict | Background Applied epidemiologists are increasingly working in areas of insecurity and active conflict to define the health risks, suggest feasible means to reduce these risks and, monitor the capacity and reconstruction of the public health system. In 2001, The Carter Center and the United States Institute for Peace sponsored a conference within which "Violence and Health" was discussed and a working group on applied epidemiology formed. The group was tasked to describe the skills that are essential to effective functioning in these settings and thereby provide guidance to the applied epidemiology training programs. Methods We conducted a literature review and consultation of a convenience sample of practitioners of applied epidemiology with experience in conflict areas. Results and conclusions The health programs designed to prevent and mitigate conflict are in their early stages of implementation and the evaluation measures for success are still being defined. The practice of epidemiology in conflict must occur within a larger humanitarian and political context to be effective. The skills required extend beyond the normal epidemiological training that focuses on the valid collection and interpretation of data and fall into two general categories: (1) Conducting a thorough assessment of the conflict setting in order to design more effective public health action in conflict settings, and (2) Communicating effectively to guide health program implementation, to advocate for needed policy changes and to facilitate interagency coordination. These are described and illustrated using examples from different countries. | Introduction In 2004 it is estimated that there are 95 violent conflicts worldwide [ 1 , 2 ]. The profound consequences to the well-being of communities from these conflicts are disproportionately distributed. Ninety percent of those who die in war are civilians, half are female, and more children will die or be disabled than soldiers [ 2 - 9 ]. Prior to the 1990s, humanitarian assistance in the context of active violence was the domain of emergency medical services; public health and epidemiology focused on refugees and displaced populations [ 3 , 6 , 10 , 11 ]. As war became endemic in certain areas, primarily civil and geographically less demarcated, the international public health community was pressured to provide prevention and primary health services to the indigenous population as well [ 12 , 13 ]. The pervasiveness of war and the magnitude of its effects have led public health experts to advocate for directed strategies to prevent and mitigate its effects on communities [ 2 , 4 , 5 , 12 - 14 ]. These strategies are described in the WHO program called Health as a Bridge for Peace (HBP) and have included developing reference and training materials to support health workers in war settings [ 15 - 17 ]. In February 2001 the Carter Center and the United States Institute for Peace (USIP), in collaboration with CARE, Emory University and the Centers for Disease Control and Prevention (CDC), sponsored a meeting on "Violence and Health". The goals of the meeting were to determine the impact of violent conflict on public health and to advise public health training programs on means to enhance the work of public health professionals in settings of violent conflict. During the meeting a specialty workgroup for "training public health care professionals" was formed and asked to focus on applied epidemiologists by describing their role in conflict prevention, mitigation, and documentation. Applied (or field) epidemiology was initially created by the US Centers for Disease Control and Prevention in the 1950s as a post-doctoral training program [ 18 ]. Over time this program has been adopted by more than 35 governments, the World Health Organization (WHO), and numerous schools of public health and non-governmental organizations (NGOs) [ 19 ]. The applied epidemiologist is trained to conduct high quality scientific studies and to translate findings into practical, effective public health programs. Increasingly, applied epidemiologists are recruited to areas of insecurity and conflict to define the health risks, suggest feasible means to reduce these risks, and monitor the capacity and reconstruction of the public health system [ 20 - 22 ]. This trend reflects a growing demand from donors, governments, military, and humanitarian groups for credible information to support the planning and evaluation of health inputs in war. These realities have led to an emerging professional interest within public health on the causes and effects of conflicts [ 2 , 13 , 14 , 23 - 25 ]. Methods Because of the increasing demand for applied epidemiologists in conflict, the workgroup leaders reviewed the relevant literature and sought further input from applied epidemiologists and public health practitioners experienced with conflict settings to better understand training needs in this setting. The literature review and discussions focused on the role of applied epidemiologists in violent conflict and what skills are needed to function effectively. The literature review included MEDLINE and Google keyword searches for "war", "conflict", "complex emergencies" and "epidemiology" combined with either "conflict", "disaster" and "refugees". The examination included specific training and reference materials developed to support Health as a Bridge for Peace [ 16 , 17 ]. The leaders expanded their workgroup to include other experts representing the Training in Epidemiology and Public Heal th Network (TEPHINET- a consortium of applied epidemiology training programs in 40 countries), the US Centers for Disease Control and Prevention (CDC), the World Health Organization (WHO), foundations, schools of public health, and numerous private organizations [ 20 , 21 ]. This paper presents the results of the literature review, as well as the answers provided by members of the group at the meeting and subsequently by members of the expanded workgroup via phone interviews about the implications of these recommendations for training field epidemiologists. Examples from the personal experience of the authors are presented and are referenced by giving the individual's initials. Results Applied epidemiologists have an important role in conflict prevention, mitigation, and reconstruction, but to be effective in these circumstances their training must emphasize knowledge about international law, human rights, and complex emergencies [ 16 , 17 ]. In addition, the skills needed must emphasize not only valid collection and interpretation of data but also highlight skills that fall into two general categories: (1) Conducting a thorough assessment of the conflict setting and players, which requires the collection, analysis and interpretation of qualitative and quantitative data to design effective public health action; and (2) Communicating effectively to guide health program implementation, to advocate for needed policy changes and to facilitate interagency coordination. We discuss each of these categories and illustrate each using examples from the experiences of workgroup members in actual conflict settings. Conducting a thorough assessment of the conflict setting in order to design more effective public health action in conflict settings The first contribution of the applied epidemiologist in conflict is to insist that a thorough assessment of the situation be done as soon as possible. Without such insistence the rush to begin programs without adapting interventions to local circumstances will prevail with predictably poor results [ 2 , 4 , 10 ]. The literature is replete with examples of quick humanitarian action with dire consequences due to poor assessment [ 10 , 24 , 25 ]. The need to precede the bulk of interventions with valid assessment data might best be summarized in the phrase "Don't just do something, stand there" (and first assess the situation) [ 20 , 26 , 27 ]. The use of validated qualitative research methods is essential to the processes of building trust and performing high quality assessments. Many workgroup experts suggested including social and behavioural scientists with practical field skills from the beginning; if this is not possible then their techniques for collecting data in the field should be adopted. Using ethnographic skills increases the accuracy and validity of assessments and allows the development of a common vocabulary of key terms for communicating with local people, even when using a translator [ 28 , 29 ]. Despite the inevitable concerns about time, qualitative assessment methods create only a minimal delay in quantitative investigations or services as they require relatively few respondents and can be done in the early stages of response and on a part-time basis. Information gathered in this way makes it less likely that data collected will be misused for political reasons and more likely that samples will be representative and services well targeted [ 10 , 26 , 27 ]. Once there is agreement that an assessment will be done- either prior to, or (more usually) simultaneously with emergency interventions- the role of the applied epidemiologist is to plan and coordinate the collection of valid and useful data. [ 19 , 20 , 26 ] The applied epidemiologist brings knowledge of how to achieve this in difficult and resource-poor circumstances by knowing whether and how to make compromises in study design, time, cost, logistics, and security while maintaining acceptable validity. Adequate preparation for the assessment, such as learning the priorities of communities and the language they use to describe health and illness, is invaluable. A good assessment defines the specific context for the health work and assures that the information obtained has a utility that justifies the cost and the risk to health workers and communities. Professionalism and impartiality during the assessment process can set the tone for and facilitate future work. If the role of the epidemiologist in collecting valid and useful data is understood by the community and other relief workers, there is a greater likelihood that the data will be appreciated as being impartial (even if certain groups have an interest in denying this). The description of previous health programs is one source of useful information for assessments that can be obtained before fieldwork commences. Did previous programs engender trust or suspicion? Did they work, and if so, why? This information can indicate a community's ability to work together or with outside agencies. In addition, if a new program resembles a past program, it will inherit the community's view of the past program, for good or ill. For example, in planning a maternal and child health program in Afghanistan with WHO, we determined that communities that previously had good experiences with girls' schools were much more willing to consider other projects that educated or benefited girls (SMM). Thus, maps from the previous ministries of education and health showing the location of these schools provided initial leads that were followed up by interviews with community members. Rapid survey methods are most often used to describe the current needs of a population in conflict [ 3 , 20 ]. They aim to describe: • baseline health status, risks and determinants (e.g. mortality, morbidity, water purity, sanitation, vectors, food, and shelter) • availability, quality, and use of health services [ 22 ] • access to health care • the security of a situation [ 21 ], and • changes in the population (e.g. migration) and the conflict (i.e. location of the front lines) [ 3 , 21 ]. Field experience in previous conflicts enables the epidemiologist to know what data are most important to assist in determining risks and designing and assessing interventions in a specific situation. For an assessment to be useful, it must have clear objectives and resist the constant pressure to add more questions [ 21 ]. Negotiating questionnaire content among partners can be greatly assisted by agreeing on the final report format, its length and especially, the content of tables and graphs. Despite the potential shortcomings of rapid surveys, systematically collected data that are appropriately interpreted are superior to anecdote. Additionally, by emphasizing and clearly documenting what sources of data are used to build a conclusion it can be re-evaluated over time. In 1990, patient logs compiled from multiple NGO-supported clinics within Afghanistan revealed that the majority of patient visits were for routine primary care services and not war-related injuries as was assumed. The result was a significant revision of the training curriculum for Afghan medics and physicians in more than 15 non-governmental health service programs (SMM). Many of the challenges in generating useful, high-quality data can be addressed with innovative data collection methods and judicious interpretation [ 21 ]. In conflict settings, a rapid survey of a defined area should take about 1–2 days, and a report of its findings should be provided immediately to decision makers and service providers. For example, an immunization survey in Uganda determined that people in the safest and most easily served areas were un-immunized; it was concluded therefore, that people in hard-to-reach areas would also need immunization. This is an example of the application of the "best case" survey strategy [ 30 ]. There are many examples of innovative ways to select a sampling frame in the conflict setting [ 21 ]. Global Positioning Systems are increasingly used to identify areas for sampling. Alternatively, satellite imagery can be used to estimate population density and select a geographic area for sampling. The most common approaches for selecting participants in rapid population surveys during conflict are simple random sampling and cluster sampling. The former is easier to analyze and communicate but takes more time in the field. Cluster surveys are more complex to analyze and communicate but involve less field time. More important than which method for sampling is used is to recognize the methods potential biases and how these might limit its applicability [ 16 , 29 ]. In war, obtaining a denominator is very challenging [ 22 , 23 , 26 , 31 , 32 ]. Available population data vary widely in quality, and the movement of persons during conflict can result in inaccurate estimates. The population may be ideologically divided by the conflict, as well as widely dispersed, highly mobile, or in refugee camps [ 5 ]. The socio-political circumstances of a population can result in over- or underestimation of its size [ 11 , 24 , 29 , 31 ]. For example, women and children may be hidden for protection, or their mobility and access to health and social services severely restricted, resulting in a great underestimation of their numbers. During the civil conflict in Ethiopia in the 1990s, families sequestered young boys to prevent their recruitment as soldiers (SMM). Alternatively, population overestimates may occur when food, drugs, or other resources are rationed [ 21 , 29 ]. Multiple registrations of the same person and not reporting losses through death or departure are ways that families can increase or at least not reduce their resources for use or sale [ 31 ]. On the Thai-Khmer border in the 1980s the Khmer Rouge recruited members of other communities to increase their numbers for a census. Addressing these challenges requires an awareness of the forces on the population and the use of measures to accommodate to these forces. In Thailand the UN Border Relief Organization resorted to arriving without notice, calling the community in for registration, and marking those already registered with an indelible stamp to reduce over-counting (PB). In Afghanistan, discussions with community religious leaders revealed that the local mosque recorded households' numbers so as to equitably share food at Islamic holidays. Thus, religious leaders were able to provide an excellent estimate of the number of households and their size (SMM). Accurate counts of births, deaths, and disease are also affected by social and political instability [ 31 ]. Conflict amplifies the risk of chronic and acute health problems while reducing the chance that the affected population will have access to health care [ 2 , 13 , 28 ]. These increases may go unnoticed, however, because of underreporting from disrupted health systems; demand might even appear to decline [ 26 ]. On the other hand, if local authorities believe that reporting disease will increase the flow of resources, such as drugs and equipment, they may inflate these numbers [ 26 ]. Rural physicians in the Philippines in 2001 described exaggerating reports of malaria cases in the dry season with the aim to stockpile anti-malarial medications to cover the needs during the malaria season (SMM). Similarly, individuals might feign illness to stock up on medicine if they believe that existing services will be unreliable. For these reasons, data from health services is frequently inaccurate during conflict and should be evaluated in the context of the pressures on the community to survive [ 10 , 11 , 13 , 21 , 23 , 24 , 29 ]. The epidemiologist will have to assess the level of access to health-care services for military and non-military personnel, particularly women, children, the elderly, and those with chronic medical conditions [ 2 , 11 , 26 ]. Defining age-, sex- and cause-specific mortality rates will help assess specific vulnerable groups and determine whether they need special help in accessing services [ 26 , 29 ]. Populations in conflict might suffer overcrowding, poor water and sanitation, and inadequate rations as well as being targeted for violence [ 2 , 5 , 11 , 13 ]. All of these can be assessed epidemiologically and determine what interventions should be provided and to whom they should be targeted. Assessing the data from disease reporting systems is a core function of epidemiologists. If these systems are intact and can describe trends of priority diseases or conditions over time, maintaining them might be worth the effort [ 21 , 22 ]. Typically, however, conflict will interrupt health surveillance activities, including data collection, analysis, interpretation and dissemination, resulting in disease underestimates [ 2 ]. As a result, to answer high-priority questions, epidemiologists may need to employ surveillance strategies with less emphasis on routine reporting and more on surveys, sentinel sites, or sentinel populations [ 21 , 23 , 26 ]. Applied epidemiologists believe that the value of epidemiologic science must also be measured by whether and how it is used to create effective public health action [ 19 ]. Applying a public health approach to violence, war, and the factors that initiate and promote them may help identify effective prevention and response programs. For example, knowing the types of weapons used in a conflict makes it possible to predict the types of injuries that will occur [ 33 , 34 ]. At a larger scale, developing systematic predictors of violent conflict may allow earlier intervention, similar to food security and famine early warning programs that monitor the risk of malnutrition. In these programs the emergence of selected behaviours serves as a warning of population risk [ 35 ]. The long-term "health" effects on the public from social disruption, violence, poverty, oppression, and torture, are poorly understood, and even less is known about the effectiveness of our programs to address them [ 36 ]. Practicing epidemiology in conflict involves working in rapidly changing circumstances and uncertain security. Formal epidemiologic assessments and public health actions in war can only proceed where the environment is "permissive". The security situation for health workers, civilians and programs, as well as community participation and ownership of health programs, must be continuously monitored. Selected parameters that reflect the level of security, participation, demographics and disease should be discussed between partners and security agencies and re -evaluated frequently [ 2 ]. Health agencies must demand a level of security to function; however, their conduct on the ground may influence the extent to which they are afforded protection by local communities. Acute and chronic war settings, while superficially similar, represent different challenges to the epidemiologist and can significantly affect the type of intervention selected [ 23 ]. In wars of short duration – such as Kuwait after the first Gulf war – the coping mechanisms of the population (for example, how they maintain their health and where they go for health services) are still based on peacetime conditions [ 32 , 37 ]. The epidemiologist may be able to focus on the pre-existing public health and surveillance system, which is likely to be partially intact for at least some parts of the population [ 26 ]. The epidemiologist can work with those local health workers and officials who are still in place in order to understand the local situation, define and prioritize health problems, and reestablish basic services in a form acceptable to local people. A chronic war setting (e.g. Afghanistan) is more challenging [ 13 , 23 ]. Knowledge of the pre-war situation is still useful, but often little is left of the previous public health system, and people will have developed new coping mechanisms to deal with prolonged war. The epidemiologist must spend more time investigating the current situation and assist appropriate agencies to plan and build new systems that meet current needs. Because of the lack of predictability, security and centralized authority, all plans and programs will need to be tested at a smaller scale before expansion to a larger population. Building systems will take longer due to the lack of pre-existing resources and trained personnel. In Iraq, although the current conflict is "acute", it also follows ten years of sanctions after a previous war and a major change in leadership. Although much of the pre-2003 health infrastructure remains, the goal is not to replicate the old system with its poor supplies and limited access. Communicating effectively to guide health program implementation, advocate for needed policy changes and facilitate interagency coordination All aspects of communication are more difficult during conflict [ 10 , 26 ]. Pre-existing infrastructure may have been destroyed and, if not, it will be part of the struggle for power and control. Practical solutions to communication needs require flexibility and a higher proportion of the budget than is needed in more stable and secure situations. Radio, satellite, and cellular phones have advantages that make them very useful, but because of their cost these technologies are often out of reach for many programs. Aid organizations should collaborate to solve communication problems and share resources; donors need to encourage and reward this by allowing local coordination of grant funds to reduce redundancy. Included in communication skills are the abilities to facilitate inter-agency coordination and collaboration and to work with multiple constituencies. Coordination among program and donors offers an opportunity to eliminate waste and to encourage equity. In Uganda in 2001, local police allowed public health workers in isolated locations to use their radios to communicate surveillance information on outbreaks to the Ministry of Health (Personal communication P. Nsubuga 2001). Ongoing consultation with communities is necessary for the integrity of programs and to monitor security. Epidemiologists can improve the quality of the information they receive and disseminate by developing alliances with partners who have a communication infrastructure in place- and people travelling into areas that are restricted or insecure. For example, contacting agencies that work in agriculture, health care, education, land-mine removal, and food provision about sharing or coordinating field staff that monitor and supervise projects can be helpful. Jointly planning the efficient use of field staff from all types of programs using checklists, and other tools to help the non-technical visitor be able to bring back useful information and support local communities can help these communities feel less isolated, reduce waste and increase the data sources about health programs and problems. Workgroup members suggested that journalists may be potential sources of useful information, insofar as they have access to restricted areas. They may be able to: • report on health-related activities or risks (e.g., immunizations, outbreaks, and unexpected behaviour in selected populations), • provide reliable communication with all sides and help establish ceasefires for the provision of health services [ 38 ], and • help control rumours by defining means to check information early and responding to misinformation quickly [ 22 , 26 ]. The different goals of the media and public health and the rapid turnover of journalists, however, may limit useful cooperation. Epidemiologists in Colombia's Centro de Referencia Nacional sobre Violencia (CRNV) used national media coverage to highlight political homicides. They published information regarding deaths along one particular river in Colombia to raise awareness of the level of violence occurring there. This data-based approach drew less political resistance while focusing public attention on the deaths [ 39 , 40 ]. Effective communication strategies should target policy-makers from local to international levels. Epidemiologists can be an effective voice for public health if they maintain their credibility with policy-makers. Sometimes the most important role for the epidemiologist is to act as a witness to describe the local situation to an international audience of policy-makers and to advocate for action [ 2 , 10 , 27 , 41 , 42 ]. Using a public health approach the epidemiologist can describe the realities of war, its effects on individuals and communities, and the consequences of certain types of weapons, warfare and humanitarian approaches. We must think of war not as a natural consequence of life but as a preventable tragedy with multiple long-lasting implications. The decision to use violence must be examined openly and not sanitized [ 43 ]. A less rapid but possibly more respected method for bringing awareness of local health and social issues to national and international levels is publishing research in credible journals. Peer-reviewed work can be a tool to facilitate negotiations between governments, NGOs and communities, and to influence international and national level policy-makers. Effective advocacy for public health includes the ability to use communication skills such as advocacy, consensus development, and negotiation to promote population health to policy-makers [ 23 , 26 , 28 , 42 , 44 ]. As much as possible, messages should be communicated directly to the local population rather than through political leaders or another entity, as the messages may be tainted or obscured by differing agendas. For example, in the Thai border refugee camps epidemiological data were interpreted and then translated into health messages that were communicated by program staff via megaphones while travelling around the community (PB). In 2003 the US military personnel and reservists in Iraq were responsible for providing humanitarian aid and direct interactions with civil authorities; a role for which they had no training (EJN). In addition, because of the military association, the information gathered was considered inherently biased and dismissed. During conflict, health and information services that were once unified under a single health ministry can become fragmented with parts falling under the aegis of government, those resisting the government, or nongovernmental organizations [ 23 , 31 ]. Access to health care and humanitarian resources can be important in political battles at the local level [ 26 ]. Each party in a conflict wants control of hospitals, health workers, food, and medications so as to be able to offer them to the community, thus enhancing their credibility, and to aid its military efforts (SMM). In Colombia, some communities have lived in conflict for > 40 years despite local, national, and international efforts to find solutions to the chronic violence. Epidemiologists there described attempts to ameliorate animosities by involving local constituencies in health activities: the Church, NGOs, elected officials, and other leading organizations or personalities in the community (Personal communication Jorge Jara 2002). Local rather than national health authorities chose priorities, from outbreak investigations and interventions to data analysis to surveillance of violence and influenza. Through this process trust was built between the Colombia National applied epidemiology program and the local health authorities, thereby assuring access to the population that transcended partisanship. Parties on all sides of a political/military issue collaborated on a community health project, an unprecedented step. Many workgroup members felt that negotiation skills were essential when working with multiple aid groups or the main parties in the armed conflict. For example, consensus building is useful when the health problems are not well defined and diverse groups must collectively identify and solve them. At the simplest level this includes the skill of running meetings and enhancing a group's utility and the understanding that disagreement might not necessarily represent failure. The goal is to help all parties understand that improving health meets the interests of all groups and to develop complimentary strategies. Too often decisions are rushed and participants pushed aside, leading to unsuccessful outcomes and future difficulties working together [ 44 ]. When disagreements are so deep that one or both sides will not accept any solution that benefits the other side, negotiation will be fruitless. Currently, in Iraq, addressing the emergency and reconstruction needs of the Iraqi people through a collaborative process that involves the full range of stakeholders – Iraqi, Coalition, UN, NGOs, Civil Affairs, Donors – remains out of reach (EJN). To facilitate the coordination of health services, the epidemiologist needs to be familiar with the many agencies working in the target area, including governmental, nongovernmental, and United Nations agencies. The diversity of groups may make it difficult to establish a shared view of the situation, much less a collaborative plan of action. Additional obstacles to coordination may stem from competition for resources, lack of information or true differences in their philosophy about humanitarian aid and their ability to deliver it. Donors in particular can create a powerful force for coordination if they agree on priorities and processes. International donors can influence policy-makers by providing resources and technical support conditional on their commitment to the resolution of violent conflict [ 4 ]. It is the job of the applied epidemiologist to supply compelling credible information for these decisions. By providing credible information and establishing it as the foundation for action the epidemiologist can greatly assist coordination. Numerous workgroup members mentioned that being able to provide published medical and public health literature to government officials, donors and NGOs facilitated their efforts to coordinate agencies and to build consensus by providing an impartial standard based on scientific data [ 32 ]. One implication is that the epidemiologist needs rapid access to international literature, possibly via the Internet, even in war zones. Summary Applied epidemiologists can use their skills and position to promote positive population health policies and programs to address inequities that exacerbate conflict and violence [ 1 , 2 , 4 ]. In addition, credible on-site information can reduce the waste and harm of poorly planned humanitarian assistance [ 9 , 10 , 23 - 25 ]. Epidemiologists' work among policy-makers and the public uniquely positions them to communicate the effects of war, advocate for the population and assist in the reconstruction of health systems [ 9 , 15 , 23 ]. As impartial agents, epidemiologists can promote dialogue between conflicting parties, influence public opinion, facilitate projects that require cooperation, and coordinate multilateral responses to health and humanitarian needs [ 13 , 22 , 23 ]. Banning landmines, creating days of tranquillity for immunization, and surveillance of homicide resulting in limiting firearms in Colombia are examples of these ideas at work [ 38 - 40 , 44 , 45 ]. Providing health resources, including epidemiological support, should be part of a much larger diplomatic and humanitarian strategy. In fact, when a larger framework and commitment are missing, the health activities are unlikely to result in lasting changes and only increase the risks to the entire population. [ 26 , 47 ]. Epidemiologists are not diplomats nor should they be gratuitously sent into conflict situations as humanitarian band-aids. In addition, there are real risks that scientific information, meant to improve health, might be usurped for political aims rather than humanitarian purposes [ 14 , 27 ]. For example, in Colombia the data from health surveys undertaken with the best of intentions were used to locate populations for military action (Personal communication G. Suarez, 2002). Epidemiologists need to advocate for focused field research on conflict resolution and violence prevention, and to evaluate the success of health programs in conflict [ 2 , 25 , 36 ]. It is no longer practical to conceptualize complex humanitarian emergencies like war into distinct phases [ 2 ]. In reality the settings for disasters, refugees, active conflict, and anxious peace are in constant transition. Defining and evaluating social and individual risk markers for violence may facilitate setting up early warning systems [ 46 ]. The reluctance to evaluate health programs in war stems from the assumption that doing anything is better than doing nothing and perhaps that good evaluation is too difficult. These flawed assumptions preclude progress and condemns us to rigid approaches resulting in the continued use of interventions of unknown effectiveness [ 13 ]. To paraphrase a military saying, "we are always treating the health problems of the last war". Well-documented epidemiologic methods to assess conflict settings combined with ongoing re-assessment of the assumptions for interventions will greatly facilitate health program evaluation. In addition, measures of success for health programs need to reach much further into cultural competence, economic development, and social well-being. Based on the results of these studies more effective and cost-effective methods for public health response could be implemented in the future. With the number and intensity of current armed conflicts there is even greater urgency to begin this work. The knowledge and skills described here are not typically part of epidemiology training. Yet they do fit with the epidemiologist's mission of gathering data and using it to improve the lot of populations. The materials developed to date to train health personnel working in war zones need more specificity and case examples. However, whether epidemiologists are able to respond to war largely depends on the availability of appropriate additional training and modelling by teachers and practicing epidemiologists [ 2 , 23 ]. Education programs designed for these goals would need to expand knowledge of human rights and international law, qualitative research methods, innovative ways to gather reliable population information during conflict, and effective methods to communicate this information. The working group strongly emphasized that these skills cannot be imparted solely in didactic courses, and training must include simulations with supervised field practice. However well-intentioned the science and programs of epidemiology, we must be vigilant that epidemiology benefits public health, that it is not used to contribute to the prolongation of conflict and that it does not become part of it [ 13 , 22 - 25 , 33 ]. Insofar as epidemiologists influence decision-makers, we must strive to do so by adapting our work to the goals of peace through every means at our disposal. [ 33 , 42 ]. We should all strive for a time when, through the efforts of public health workers and others, war too will be eliminated. - Jimmy Carter [ 2 ] | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC544942.xml |
516023 | Analysis of HIV-1 Vpr determinants responsible for cell growth arrest in Saccharomyces cerevisiae | Background The HIV-1 genome encodes a well-conserved accessory gene product, Vpr, that serves multiple functions in the retroviral life cycle, including the enhancement of viral replication in nondividing macrophages, the induction of G2 cell-cycle arrest, and the modulation of HIV-1-induced apoptosis. We previously reported the genetic selection of a panel of di-tryptophan (W)-containing peptides capable of interacting with HIV-1 Vpr and inhibiting its cytostatic activity in Saccharomyces cerevisiae (Yao, X.-J., J. Lemay, N. Rougeau, M. Clément, S. Kurtz, P. Belhumeur, and E. A. Cohen, J. Biol. Chem. v. 277, p. 48816–48826, 2002). In this study, we performed a mutagenic analysis of Vpr to identify sequence and/or structural determinants implicated in the interaction with di-W-containing peptides and assessed the effect of mutations on Vpr-induced cytostatic activity in S. cerevisiae . Results Our data clearly shows that integrity of N-terminal α-helix I (17–33) and α-helix III (53–83) is crucial for Vpr interaction with di-W-containing peptides as well as for the protein-induced cytostatic effect in budding yeast. Interestingly, several Vpr mutants, mainly in the N- and C-terminal domains, which were previously reported to be defective for cell-cycle arrest or apoptosis in human cells, still displayed a cytostatic activity in S. cerevisiae and remained sensitive to the inhibitory effect of di-W-containing peptides. Conclusions Vpr-induced growth arrest in budding yeast can be effectively inhibited by GST-fused di-W peptide through a specific interaction of di-W peptide with Vpr functional domain, which includes α-helix I (17–33) and α-helix III (53–83). Furthermore, the mechanism(s) underlying Vpr-induced cytostatic effect in budding yeast are likely to be distinct from those implicated in cell-cycle alteration and apoptosis in human cells. | Background Human immunodeficiency virus 1 (HIV-1) Vpr is a small virion-associated protein that is incorporated into virions through a specific interaction with the p6 domain of the p55 gag precursor protein [ 1 , 2 ]. Increasing evidence suggests that Vpr plays important roles during HIV-1 replication and pathogenesis. First, virion-associated Vpr has been shown to act early in viral infection as a facilitator of HIV-1 preintegration complex (PIC) entry through the limiting nuclear pore. This activity of Vpr is thought to be responsible for Vpr's ability to enhance HIV-1 replication in nondividing cells, most notably in terminally differentiated macrophages [ 3 - 5 ]. Second, expression of Vpr induces a G2 cell cycle arrest, which is thought to indirectly enhance viral replication by increasing transcription from the HIV-1 long terminal repeat (LTR) [ 6 , 7 ]. Even though the molecular mechanism of Vpr-mediated cell-cycle G2 arrest is still obscure, it has been known that Vpr expression leads to inactivation of the mitotic p34cdc2/cyclinB complex in human cells [ 8 , 9 ] as well as in fission yeast Schizosaccharomyces pombe (Sc. Pombe) [ 10 - 14 ]. Involvement of protein phosphatase 2A (PP2A), Wee1, Cdc25C, and 14-3-3 proteins has also been implicated [ 8 - 12 , 14 ] but the host cell proteins directly engaged by Vpr are not yet identified. Noteworthy, HIV-1 Vpr expression induces also a growth arrest in Saccharomyces (S.) cerevisiae [ 15 - 17 ]. Deletion mapping studies showed that the C-terminal 33 amino acids, including the H(S/F)RIG motif, contributed to this cytostatic effect [ 15 , 18 ]. Although this region has also been implicated in Vpr-mediated cell-cycle dysregulation in mammalian and S. Pombe cells [ 19 - 25 ], the molecular mechanism of Vpr-growth arrest in budding yeast is thought to be distinct since growth arrest occurs independently of any evident block at the G2/M transition [ 16 ]. Accordingly, it has been reported that the G2/mitosis transition in budding yeast is regulated differently than in mammalian cells and fission yeast [ 26 , 27 ]. Indeed, Vpr cytostatic effect observed in S. cerevisiae has been proposed to result from gross mitochondrial dysfunction [ 17 ] and/or cytoskeletal defects [ 16 ], rather than a cell cycle G2 arrest. In addition to nuclear import and cytostatic activities, HIV-1 Vpr exhibits cytotoxic properties. Elevated intracellular expression or addition of extracellular Vpr or derived peptides results in proapoptotic effects in human cells including neurons [ 6 , 28 , 29 ] as well as cytotoxicity in budding and fission yeasts [ 30 , 31 ]. Jacotot et al . have provided evidence indicating that extracellular Vpr or peptides derived from Vpr C-terminus induce mitochondrial dysfunction in human cells by a mechanism involving a specific binding to the adenine nucleotide translocator (ANT), a component of the permeability transition pore complex (PTPC) in the mitochondrial membrane. The resulting mitochondrial membrane permeabilization (MMP) leads to a decreased membrane potential and the release of cytochrome c and apoptosis inducing factor (AIF) [ 32 , 33 ]. This Vpr-mediated MMP is thought to initiate cell death through both caspase-dependent and independent mechanisms in human cells as well as cytotoxicity in budding yeast [ 32 - 37 ]. In addition, it has also been shown that extracellular Vpr is capable of forming cation-selective ion channels in planar lipid bilayers, which can depolarize intact cultured neurons, thus leading to cell death [ 28 ]. In a previous report, we have shown that expression of genetically-selected glutathione-S-transferase (GST)-fused di-tryptophan (di-W)-containing peptides inhibited Vpr-mediated growth arrest in S. cerevisiae presumably by interacting with Vpr [ 38 ]. Interestingly, these, di-W-containing peptides were also able to inhibit Vpr biological activities, including nuclear import, cell cycle G2 arrest and apoptosis, in mammalian cells or HIV-1 infected T cells [ 38 ]. Even though the inhibitory effect of these di-W-containing peptides correlated with their ability to interact with Vpr in budding yeast, the detailed mechanism underlying their mode of action remains to be defined. In addition, it is still unclear whether the growth arrest activity of Vpr in budding yeast is related to specific biological activities of Vpr in human cells. In this study, we have performed a mutagenic analysis of Vpr to identify Vpr domains important for di-W peptide binding and cytostatic activity in S cerevisiae . Results reveal that the inhibitory di-W-containing peptides target specifically a functional domain of Vpr directly involved in growth arrest in budding yeast. Furthermore, several previously well-characterized Vpr mutants unable to induce cell-cycle dysregulation and/or apoptosis in mammalian cells still exhibit strong growth arrest activity in budding yeast, indeed suggesting that Vpr carries out distinct functions in S. cerevisiae . Results Analysis of Vpr sequence and/or structural determinants implicated in the interaction with di-W-containing peptides We have previously used a genetic selection system in S. cerevisiae budding yeast and selected a panel of di-W-containing GST-peptides that specifically inhibit Vpr-mediated yeast growth arrest function presumably through their ability to bind HIV-1 Vpr [ 38 ]. In this study, we further investigated the molecular mechanism of this inhibition using a newly selected GST-fused di-W peptide WWSFKSV (GST-B4), which displayed an enhanced ability to bind Vpr and inhibit its growth arrest activity in budding yeast (Fig. 1A and 1B ). Figure 1 GST-B4 peptide binds to HIV-1 Vpr in S. cerevisiae and rescues cell growth. (A) GST pull-down from yeast extracts. S. cerevisiae HP16 strain co-transformed with GST or GST-B4 plasmids and (R+) or (R-) Vpr expressor were metabolically-labeled with 150 μCi of 35 S-Translabel in Vpr-inducible medium. Half volume of the cell extract was used for GST pull-down, while the remaining lysates were subjected to immunoprecipitation with polyclonal anti-Vpr antiserum. Total and GST-bound radiolabeled Vpr proteins were detected by autoradiography after SDS-PAGE. (B) GST-B4 suppression of Vpr-induced cell growth arrest. Yeast co-transformants were grown in non-inducible selective medium for two days. Similar number of yeast cells were then serially diluted, spotted onto either Vpr non-inducible (Trp - /Ura - , 2% raf) or Vpr-inducible plates (Trp - /Ura - , 2% gal) and incubated for 3 to 5 days to evaluate their growth rates. This data is representative of results obtained in two independent experiments Structural studies performed with synthetic forms of Vpr indicate that Vpr is characterized by a well-defined gamma turn (14–16)-alpha helix (α-helix I: 17–33)-turn (34–36), followed by an alpha helix(α-helix II: 40–48)-loop (49–54)-alpha helix (α-helix III: 55–83) domain and ends with a very flexible C-terminal arginine-rich sequence [ 39 ]. The α-helical determinants where shown to be required for Vpr virion incorporation, nuclear localization and oligomerization [ 39 - 44 ] and are believed to be involved in heterologous protein binding [ 45 ]. The arginine rich C-terminal region of Vpr has not been shown to have a predicted structure, however this region harbors protein phosphorylation sites and plays an important role in cell cycle G2 arrest and the nuclear localization of the protein in mammalian cells [ 6 , 31 , 46 ]. To further investigate the sequence and/or structural requirement of Vpr for GST-B4 binding, mutations were introduced in p424Gal1-Vpr expressor to target different regions of Vpr (Fig. 2A ). The N-terminal Q3R mutant was shown to affect Vpr proapoptotic activity during HIV-1 replication [ 47 ]. Four amino acids Glu21, Leu23, Glu25 and Ala30 were separately changed to Lys or Phe (E21K, L23F, E25K and A30F) in order to disrupt the amphipaticity of the first α-helix [ 39 ] (Fig. 2A ). The F34I was introduced in a γ-turn region which is just after the α-helix I [ 39 ]. The R62P and I63K mutations introduced in the third helix were aimed at interfering with the integrity of the α-helix and are known to abolish Vpr nuclear localization [ 41 ]. Four mutants in the C-terminal region, including, R77Q, S79A, R80A and R87, 88, were generated to replace positively-charged arginine residues or to remove the critical phosphorylation site (Ser 79) of the protein. Vpr mutants S79A and R80A were reported to be defective for cell cycle G2 arrest activity in mammalian cells, while the proapoptotic activity of R77Q was severely attenuated [ 6 , 24 , 48 ]. In addition, a frameshift mutation (R77fs) [ 40 ] and a truncation mutation (R86stop), which prematurely terminate the protein at amino acid 77 and 86 respectively were also constructed. Figure 2 HIV-1 Vpr mutants exhibit differential GST-B4 binding abilities. Each Vpr mutant used in this study with the exact location of the introduced mutation is described (A). (B) GST pull-down using a panel of Vpr mutants. Assays were performed as described in Fig. 1A. Protein extracts were prepared from radiolabeled cells expressing GST (lanes 1–2) or GST-B4 proteins (lanes 3–17) alone (R-), or in presence of wild-type Vpr (R+) or different mutant proteins, as indicated. Vpr bound to GST-B4 (upper panel) and the total amount of Vpr as determined using immunoprecipitation with anti-Vpr antiserum (lower panel) were separated by SDS-PAGE and detected after autoradiography. (C) The percentage of GST-B4-bound Vpr relative to the total amounts of Vpr for each mutant was quantified by autoradiography scanning and the level of wild type Vpr bound to GST-B4 was arbitrarily set as 100%. These data are representative of at least two independent experiments. To determine the impact of the Vpr mutations on GST-B4 peptide binding, HP-16 yeast co-transformed with mutated-Vpr expressors and either GST or GST-B4 vectors were radio-labeled in Vpr-inducible medium and subjected to GST pull-down assays (Fig. 2B ), as described in Materials and Methods. Moreover, the amount of wild type Vpr or each mutant bound to GST-B4 peptide was evaluated by laser densitometric scanning of bands in autoradiograms and normalized to the total amounts of Vpr and GST proteins that were expressed in each transformants. The amounts of wild type Vpr bound to GST-B4 was arbitrarily set as 100% (Fig. 2C ). Results of figure 2B reveal that all Vpr mutants were expressed at comparable levels, as determined by Vpr immunoprecipitation of induced-cell lysates with the exception of Vpr (R77fs), which indeed was previously reported to be less stable than wild type Vpr [ 40 ] (Fig. 2B , lower panel). While no Vpr interacted with GST (Fig. 2B , upper panel, lane 2), similar levels of wild type Vpr, E25K, F34I, I63K, R77fs, R80A, R87, 88 and R86stop mutants were pulled-down with GST-B4 (Fig. 2B , upper panel and 2C ). Similar results were obtained for Vpr mutants Q3R, R77Q, S79A (data not shown). In contrast, E21K, L23F, A30F and R62P mutants, which are respectively located in α-helix I and α-helix III regions, were not co-pulled down with GST-B4 (Fig. 2B (upper panel, lanes 5, 7, 8 and 14) and 2C ). Taken together, these results suggest that the integrity of the N-terminal α-helix I and the α-helix III of Vpr are crucial for GST-B4 binding, whereas the C-terminal domain is dispensable for the interaction. Vpr mutants defective for GST-B4 binding are unable to arrest yeast cell growth We next tested the growth arrest activity of these Vpr mutants in HP-16 yeast. Growing yeast cells transformed with the empty vector (R-), wild-type (R+) or mutated Vpr expressors were serially diluted and spotted onto either a Vpr non-inducible plate (Trp - , 2% raf) or a Vpr-inducible plate (Trp - , 2% gal). Cell growth was evaluated following an incubation of 3–5 days at 30°C (Fig. 3 ). In Vpr non-inducible plate, all yeast transformants grew at similar rate (Fig. 3 , left panel). Upon galactose induction, while the empty vector (R-)-transformed yeast grew efficiently (Fig. 3 , lanes 1, 8 and 15), the wild-type Vpr (lanes 2, 9 and 16), the Q3R mutant and all proteins mutated in the C-terminal region, including R77Q, S79A, R80A exhibited a profound growth arrest activity (Fig. 3 , right panel (lanes 10,13, 14 and 18). Similar results were obtained for R-87,88 and R86stop mutants (data not shown), indicating that the C-terminal arginine-rich region of Vpr is not involved in budding yeast growth arrest activity. Of note, R77fs showed an impaired growth arrest activity (Fig. 3 , lane 17), which is most likely due to the shorter half-life of this truncated protein, as reported before [ 40 ]. In contrast, expression of helices I and III Vpr mutants, E21K, L23F, A30F and R62P, which displayed a strong attenuation of binding to GST-B4, did not lead to HP-16 budding yeast growth arrest (Fig. 3 , right panel, lanes 3, 4, 6 and 11). On the contrary, helix I or III mutants E25K, F34I, and I63K, which were able to interact with GST-B4, still exhibited growth arrest activity, even though at reduced levels as compared to wild-type Vpr (Fig. 3 , right panel, 5, 7 and 12). These results indicate that Vpr helices I and III represent an important functional domain involved in growth arrest in budding yeast. Figure 3 The growth arrest activity of different Vpr mutants. S. cerevisiae HP16 yeast was transformed with the p424Gal1 expressor alone (R-) or coding for wild-type (R+) or each mutant, as indicated at the left of photograph, and first grown in non-inducible medium for 2 days. Then, similar amounts of transformed yeast were serially 10× diluted and spotted onto either non-inducible (Trp - , 2% raf) or Vpr-inducible (Trp - , 2% gal) plates and incubated for 3 days to evaluate their growth rates. This data is representative of at least two independent experiments. The ability of each mutant to bind the B4 peptide is indicated on the right. (+) indicates efficient binding while (-) indicates absence of binding. GST-B4 inhibits the cytostatic activity of Vpr mutants and rescues cell growth To further investigate the correlation between the inhibitory effect of GST-B4 and its Vpr-binding ability, GST or GST-B4 were co-expressed with two GST-B4-binding defective Vpr mutants E21K and L23F or with two GST-B4-binding competent mutants E25K and F34I in HP-16 yeast and the resulting cell growth was monitored in Vpr-inducible plates as described above. In agreement with the data of figure 3 , in the presence of GST alone, mutants E25K and F34I induced significant yeast growth arrest, while such activity was severely impaired for B4-binding defective mutants E21K and L23F (Fig. 4 , left panel). In contrast, GST-B4 co-expression strongly inhibited the growth arrest activity of the wild type Vpr, E25K and F34I mutants and indeed restored their cell growth at a level comparable to that of yeast cells expressing E21K or L23F (Fig. 4 , right panel). A weak inhibitory activity of B4 was also observed with mutants E21K and L23F (lanes 3 and 4). It is possible that this may reflect a weak or instable interaction between B4 and Vpr mutants E21K and L23F, which could not be clearly detected in the binding experiments (Fig. 3 ). Overall, these results clearly indicate that GST-B4 specifically binds to structural determinants that are important for inducing cell growth arrest. Moreover, as described previously (38), the binding efficiency of B4 peptides correlates with the extent of the peptide inhibitory activity. Figure 4 Comparison of GST-B4-mediated inhibition of the growth arrest activity of different Vpr mutants. S. cerevisiae HP16 yeast co-expressing GST (left panel) or GST-B4 (right panel) and a panel of representative Vpr mutants, as indicated, were serially 10× diluted and plated on Vpr-inducible and selective (Trp - /Ura - , 2% gal) plates as described in Fig. 3. The respective cell growth was evaluated after a 3-day incubation. This data is representative for two independent experiments. Discussion We have previously shown that GST-fused di-W-containing peptides were able to interact with HIV-1 Vpr and as a result inhibit its multiple functions in budding yeast as well as in HIV-1 infected T cells [ 38 ]. In the present study we have further investigated the sequence and/or structural determinants required for Vpr/peptide interaction and determined their impact on Vpr cytostatic activity in budding yeast. Results clearly show that GST-fused B4 peptide interaction with Vpr involves the α-helical I and III structure of Vpr. Mutations affecting the integrity of these helical regions not only interfered with the interaction with GST-B4 peptide, but also failed to induce a cytostatic activity in budding yeast. Furthermore, Vpr mutants, including Q3R, R77Q, R80A and S79A, yet defective for cell-cycle arrest or apoptosis in mammalian cells, still induced a growth arrest in S. cerevisiae and displayed sensitivity to GST-B4 inhibition. Overall, these results indicate that GST-fused di-W-containing peptides directly target functional domains of HIV-1 Vpr responsible for inducing growth arrest in budding yeast and strongly suggest that the mechanism(s) underlying Vpr-induced cytostatic effect in budding yeast are distinct from those implicated in cell-cycle alteration and apoptosis in mammalian cells. Previous reports have indicated that the Vpr cytostatic activity in S. cerevisiae budding yeast was attributed to its last 63–96 amino acid (aa) and the critical domain was located in a conserved C-terminal HFRIGCRHSRIG sequence from aa 71 to 82 [ 15 ]. In contrast, our results showed that expression of a truncated Vpr encompassing aa 1 to 77 was sufficient to induce growth arrest (Fig. 3 ), suggesting that the sequence of HFRIG (aa 71 to 75), but not HSRIG (aa 78 to 82) and other C-terminal region of Vpr, may constitute one important determinant for this Vpr-induced phenotype. Consistently, a mutagenic analysis by Berglez et al ., revealed that substitution mutations of aa His71 or Gly75 in this HFRIG sequence abolished Vpr cytostatic effect in budding yeast [ 18 ]. Interestingly, our analysis clearly reveal that the N-terminal α-helix I and the α-helix III are both contributing to Vpr cytostatic effect, which is in agreement with a previous finding by Gu et al showing that the Vpr F34I mutant was unable to induce a growth arrest phenotype in budding yeast [ 16 ]. On the basis of the Vpr NMR structure reported by Wecker et al [ 39 ], mutations E21K, L23F, E25K and A30F located within the α-helix I (from aa 17 to 33) were designed to disrupt either the negatively-charged cluster or the hydrophobic interface. With the exception of E25K mutant, all other mutations in this N-terminal α-helical region lead to the loss of Vpr cytostatic function (Fig. 3 ). In addition, disruption of the third α-helix by introduction of a proline at position 62 (R62P) suppressed Vpr-induced growth arrest, suggesting that integrity of α-helices I and III was required for Vpr cytostatic activity in budding yeast. It was also noted that E25K and I63K still induced a low level of growth arrest compared to other helical region mutants (Fig. 3 ). It could be possible that E25K is somewhat external to the spatially-aligned acidic cluster D17-E21-E24 [ 39 ], and may be therefore less critical. Similarly, the I63K mutation may have a minor impact on the tridimensional structure of helix III as compared to the introduction of a proline residue as with the R62P Vpr mutant. One striking observation of this study is that the four mutants (E21K, L23F, A30F and R62P) located in the α-helical I and III regions of Vpr, which were defective for the cytostatic activity in budding yeast (Fig. 3 ) also lost the ability to interact with the inhibitory GST-B4 peptide (Fig. 2 ). It indicates that GST-B4 directly targets a critical functional domain, possibly a structural cluster comprising both of α-helical I and III, that is responsible for cytostatic activity. Interestingly, the sequence of GST-B4 (GST-WWSKKSV) reveals that, in addition to a conserved di-W motif [ 38 ], it also harbors an overlapping WxxF motif, which has been previously isolated by phage-display as a Vpr-binding motif and is present in Vpr-interacting protein uracil DNA glycosylase (UDG) [ 49 ]. Coincidentally, a bipartite domain encompassing Vpr amino acids 15–27 and 63–77 was also shown to be involved in UDG binding [ 50 ]. Based on these observations, it appears that similar regions of Vpr are involved in binding to UDG and GST-B4 through targeting of a WxxF element. However, E25K and F34I mutants, which were shown to be defective for UDG binding in two-hybrid assays [ 21 ], were still able to interact with GST-B4 in vivo . Such a difference may specifically rely on the hydrophobic di-W motif, which is not present in UDG [ 49 ]. Up to date, how HIV-1 Vpr induces a growth arrest in budding yeast remains an open question. During HIV-1 replication, the expression of Vpr has been shown to induce a cell cycle G2 arrest resulting from the inactivation of the mitotic p34cdc2/cyclinB complex [ 51 ]. In contrast, Vpr-mediated growth arrest in budding yeast is thought to occur through a distinct mechanism(s), since it occurs independently of any evident block at the G2/M transition [ 16 ]. In this study, we tested a panel of well-characterized Vpr mutants for their ability to growth arrest HP-16 budding yeast. Interestingly, Vpr mutants (S79A and R80A) which were previously shown to be as stable as wild type Vpr but defective for cell-cycle G2 arrest in human cells [ 6 , 19 , 24 ] still induced strong growth arrest in budding yeast. Conversely, L23F, and R62P mutants, which are competent for cytostatic effect in mammalian cells, [ 20 , 41 ] were unable to block yeast growth. Therefore, it can be concluded that Vpr structural determinants required for growth arrest in S. cerevisiae and human cells are clearly distinct, implying that different molecular mechanisms governs Vpr activities in these different cell species. Moreover, our study also demonstrates that Vpr cytostatic effect in budding yeast is not related to the cytotoxic activity of the viral protein. Vpr exhibits different cytotoxic properties that implicate distinct domains of the viral protein. First, wild-type Vpr and its first 40 N-terminal amino acids can form cation-selective ion channels in lipid bilayers [ 28 , 52 ]. Depolarization of the plasma membrane resulting from inward sodium current eventually induces killing of polarized cells such as neurons. On the other hand, apoptosis in T cells is thought to be triggered by transduction of full-length Vpr or its C-terminal 52–96 moiety into cells and involves mitochondrial membrane permeabilization [ 33 , 53 , 54 ]. Resulting loss of mitochondrial transmembrane potential then induces the release of apoptogenic proteins, leading to caspase-dependent (37,55,48) or caspase-independent [ 55 ] cell killing. The fact that both 17–33 and 55–83 alpha-helices are required for growth arrest in S. cerevisiae strongly suggests that the cytostatic effect observed in budding yeast is mechanistically distinct from effects resulting from ion channels formation or mitochondria permeabilization. Consistently, Q3R, R80A, R77Q Vpr mutants, which were previously shown to be as stable as wild type Vpr, but yet defective for apoptosis induction in human cells [ 6 , 47 , 48 ] were still able to block yeast growth in a B4-sensitive way. Conclusions Taken together, the results presented here provide evidence that Vpr triggers growth arrest in budding yeast by an undefined mechanism that is unrelated to Vpr-induced G2 arrest and apoptosis in mammalian cells. This Vpr-induced budding yeast growth arrest can be effectively inhibited by GST-fused di-W peptide through an interaction of di-W peptide with Vpr functional domain, which includes α helix I and III. These observations would support a model in which, Vpr interacts with a di-W-containing protein in S. cerevisiae to induce yeast growth arrest. The question that still remains unanswered at this point is whether this Vpr cytostatic activity in budding yeast can also play an important role during HIV-1 replication and viral pathogenesis and further investigations are currently underway to address this question. Materials and methods Yeast strain The S. cerevisiae yeast strain used in this study was the protease-deficient HP-16 strain ( MAT∝ ura3-52 his3Δ1 leu2 trp1Δ63 prb1-1122 pep4-3 prc1-407 ) [ 56 ]. Plasmid transformation was performed using the lithium acetate method [ 57 ]. Plasmids, antisera and chemicals The HIV-1 Vpr yeast expression plasmid (p424Gal1-Vpr) and the negative control plasmid p424Gal1-R - have been previously described [ 38 ]. To generate different p424Gal1-Vpr mutant expression plasmids, each of Vpr mutant cDNAs (Fig. 2A ) was generated by a two-steps polymerase chain reaction (PCR)-based method [ 40 ] by using a 5'-primer (5'-CTGCTAGCGGATAGATGGGA-3') harboring a BamH I site in front of the Vpr initiation codon, a 3'-primer (5'-GCATCGCTCGAGGATCTACTGGC-3') containing a Xho I site after the stop codon of Vpr and the complementary oligonucleotide primers containing the desired mutations. Amplified Vpr cDNA harboring specific mutations were then cloned into the p424Gal1 vector at BamH I/ Xho I sites. The Vpr mutants L23F, E25K, A30F, R62P, I63K, R77Q, R77fs and R80A were previously described [ 6 , 41 , 48 ]. The pPGK-GST plasmid was described previously [ 38 ] while the pPGK-GST-B4 expressor was isolated and purified from an S. cerevisiae HP-16 yeast colony that was resistant to HIV-1 Vpr-mediated growth arrest as previously described [ 38 ]. The rabbit anti-Vpr polyclonal serum was raised against bacterially expressed recombinant Vpr as described previously [ 58 ]. Galactose, raffinose and glucose were purchased from Sigma Inc. Evaluation of the growth arrest activity of Vpr mutants and the anti-Vpr activity of GST-peptide in budding yeast The experimental procedures to evaluate protein expression, Vpr growth arrest activity and the anti-Vpr activity of GST-fused di-W peptide were described previously [ 38 ]. Briefly, HP-16 yeast cells transformed with p424Gal1-wild-type/mutant Vpr plasmids or co-transformed with Vpr expressors and pPGK-GST-B4 plasmid were first grown in a Vpr non-inducible selective medium (Trp - or Trp - /Ura - , 2% raffinose (raf + )) for 2 days. Then, suspensions of transformed HP-16 yeast cells (adjusted at similar cell densities) were serially diluted and spotted onto either a Vpr non-inducible plate (Trp - or Trp - /Ura - , 2% raf) or a Vpr-inducible plate (Trp - or Trp - /Ura - , 2% gal) to evaluate the growth of each co-transformed HP16 population. GST pull-down assay and anti-Vpr immunoprecipitation HP16 co-transformants were radiolabeled with 150 μCi of 35 S-Translabel (ICN Inc.) in Vpr-inducible medium and lysed in CHAPS buffer as previously described [ 38 ]. Cell extracts were then subjected to GST pull-down assay [ 4 , 38 ]. Briefly, lysates were incubated with glutathione-sepharose 4B beads (Amerham Pharmacia Biotech Inc) for 2 hours at 4°C. Beads were washed 3 times and the radiolabeled protein complexes were eluted with an elution buffer (100mM reduced gluthathione, 120 mM NaCl, 100 mM Tris-HCl pH 8.5) by gentle shaking at 4°C for 1 hour. Eluted protein complexes were separated by SDS-PAGE and detected by autoradiography. For Vpr expression analysis, aliquots of labeled yeast lysates were immunoprecipitated with anti-Vpr antibodies as described previously [ 38 , 40 ]. Authors' contributions X-J Y designed the experiments, constructed most Vpr mutants and wrote the manuscript. NR carried out the binding assays and tested the effect of Vpr mutants on yeast cells growth. JL selected and characterized the B4 GST-di-W-containing peptide. GD participated in the design of the study and critically evaluated the manuscript. EAC participated in the design of the study and coordinated it. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC516023.xml |
546199 | Validating viral quasispecies with digital organisms: a re-examination of the critical mutation rate | Background In this report we re-examine some recent experiments with digital organisms to test some predictions of quasispecies theory. These experiments revealed that under high mutation rates populations of less fit organisms previously adapted to such high mutation rates were able to outcompete organisms with higher average fitness but adapted to low mutation rates. Results We have verified that these results do hold in the original conditions and, by extending the set of initial parameters, we have also detected that the critical mutation rate was independent of population size, a result that we have found to be dependent on a different, contingent factor, the initial fitness vector. Furthermore, in all but one case, the critical mutation rate is higher than the error threshold, a key parameter in quasispecies theory, which prevents its extrapolation to natural viral populations. Conclusion From these results we conclude that digital organisms are useful tools for investigating evolutionary patterns and processes including some predictions from the quasispecies theory. | Background RNA viruses are among the most infective pathogens affecting plants, animals and humans. Several of their features such as their reduced genomes, high genetic heterogeneity, large population sizes, short generation times and fast evolutionary rates place them among the best models for evolutionary and population genetic studies [ 1 , 2 ]. These same features explain why they are so difficult to eradicate. Many of them are able to establish chronic infections because their high mutation rates allow them to escape from the immune system pressure. As a consequence, selection, that translates in competition with the host and among viral variants, usually results in the persistence of the most infective, pathogenic or more persistent variants. The molecular bases for this genetic variability are three mechanisms differentially used by each kind of virus: mutation, homologous and non-homologous recombination and genome rearrangement [ 3 ]. Attempts to model the evolutionary dynamics of RNA viruses incorporate their most relevant features, such as large population sizes (due to their short replication times RNA viruses can reach population sizes of around 10 10 individuals in short times), high mutation rates (in the order of 0,1–1 mutations per genome and replication round (m/g/r) derived from lack of proof-read correction in the polymerase), and small genome sizes (ranging from 3 to 30 kilobases). For years RNA virus population dynamics has been studied under the classical population genetics framework [ 1 ], thus allowing the development of models that explained their evolution in terms of selection, mutation, genetic drift and, less importantly, migration within and among hosts. Under this framework theoretical predictions such as the Red Queen hypothesis [ 4 ], frequency-dependent selection [ 5 , 6 ] or clonal interference [ 7 ] have been demonstrated with experimental populations of viruses. Despite these achievements in the late 70's some results suggested that the evolution of RNA viruses might be better explained by a quasispecies model. The quasispecies concept was formulated by Eigen [ 8 ] in his studies on the evolution of the first replicons. The concept arises as an alternative to the neutral theory [ 9 ] which requires small population sizes and large genomes. A population of replicons with these characteristics cannot explore the whole neutral space of an adaptive landscape and, consequently, the stochastic differentiation of the molecules is possible. But in the case of molecules with small genomes and large population sizes (such as early replicons) the whole neutral space can be explored thus avoiding the effect of the genetic drift. This property along with high mutation rates allows quasispecies formation in viral populations. A quasispecies has been defined as a cloud of mutants organized around one or a few high fitness variants and with very low Hamming distances among them. The high mutation rates are the connective agent between the members of the quasispecies, with their frequencies depending on their replication fidelity and that of the rest of neighbor mutants. This mutational coupling implies that the object of natural selection is the quasispecies as a whole and not each individual variant. The quasispecies structure has three important implications [ 10 ]: - Selection acts upon the quasispecies as a whole and not upon individual variants. The result is that under appropriate conditions lower fitness variants can outcompete higher fitness ones (survival of the flattest vs. survival of the fittest). - Genetic drift has no relevant effects: their tiny genomes, large population sizes and high mutation rates allow the exploration of all the neutral space around the master sequence. - The average consensus sequence remains stable along quasispecies evolution. This model contrasts sharply with conventional population genetic models in which the existence of a large number of neutral mutations would lead to genetic drift of the population and the individual is the unit of selection rather than a cloud of related variants [ 11 ]. This last difference is most relevant when quasispecies theory is applied to real entities, such as RNA viruses, in two contexts. First, RNA viruses represent the vast majority of emerging pathogens and there is a growing interest in the application of evolutionary principles for the control, prevention and treatment of diseases caused by them [ 2 ]. Second, RNA viruses represent the best example of measurably evolving populations [ 12 ] and as such are widely used to experimentally test many postulates of evolutionary theory [ 2 , 13 ]. Hence, differences on the nature of the unit of evolution in RNA viruses may have important consequences in practical applications and experimental verification of evolutionary theory. The difficulty in experimentally testing some predictions of quasispecies theory has led to the search of alternative systems. In this work we have used digital organisms as an approximation to the population dynamics of RNA viruses. Digital organisms are self-replicating entities and compete for access to resources, in this case CPU cycles, as implemented in the AVIDA platform [ 14 , 15 ]. Avidians are programs (genomes) composed by arrays of logical instructions (genes) that allow them obtaining CPU cycles. There are 28 possible instructions. The number of instructions in a digital organism is equivalent to the genome size in a biological organism [ 16 , 17 ]. Many studies have been done using the AVIDA platform. The possibility that genome sizes change during the course of evolution and the rewards that they can obtain by the combination of functions have allowed investigations about the evolution of genomic complexity [ 18 ]. Furthermore the possibility of studying their evolution throughout long periods of adaptation and competition has allowed the reproduction of studies originally performed with other asexual organisms such as viruses and bacteria [ 19 ]. Other studies have attempted to determine the effect of each possible mutation on the fitness of a genome and the nature of their interactions [ 20 , 21 ]. In this work we have focused in a recent study by Wilke et al. [ 22 ] with digital organisms in which they concluded the validity of one of the principal tenets of quasispecies theory. The prediction is that less fit organisms can outcompete fitter organisms when mutation rates are high. The dynamics of their experiments consists of generating, from a common ancestor, pairs of organisms adapted to high (lower fitness organisms) and low (higher fitness organisms) mutation rates. Then, competition experiments between high and low fitness organisms are performed at different mutation rates. As the mutation rate increases, these experiments result in the winner being always the lower fitness variant. This indicates that previous adaptation to high mutation rates generates less fit but very robust variants. Therefore, under high mutation rates these variants generate a better adapted cloud of mutants. On the other hand, high fitness variants are in higher but steeper adaptation peaks and in consequence are more sensitive to mutation. Therefore at high mutation rates there is survival of the flattest and not survival of the fittest . Here we have extended the original experimental conditions in order to study the effect and interaction between three of the key factors in the quasispecies model: population and genome sizes and mutation rates. Our results indicate that chance events in the form of historical contingency play an important role in the evolution of these populations. Moreover we have established a new, corrected mutation rate necessary for quasispecies formation with a higher value than the original one. The implications of this correction are discussed. Results We considered three factors affecting the critical mutation rate in our digital organisms: genome size, population size and the influence of the initial fitness vector. This is a vector of randomly assigned priorities for the first time evaluation of each organism fitness, incorporated to prevent the system from collapsing if all organisms simultaneously try to enter the CPU, and can be interpreted as a historical, contingent factor in evolution. The range of genome sizes studied varied from 54 to 272 instructions (Table 1 ). Our exploratory experiments indicated that the initial fitness vector might have an important effect on the results of competition between pairs of organisms adapted to low and high mutation rates. This was most apparent when comparing results using the original, fixed initial vector used by Wilke et al. [ 22 ] for all the competitions involving the same pair of organisms and those obtained when the initial fitness vector was a random one, with different values for each experiment. Hence, in the original study for 3600 individuals the critical mutation rates varied between 0.88 and 3.66 with a mean value, normalized according to our criterion for estimating the critical mutation rate, of 1.386 (standard deviation, SD = 0.777). In our experiments for this same population size and random initial vectors, critical genomic mutation rates ranged between 0.5 and 3 but with a higher average value, 2.045 (SD = 0.757). Consequently, we decided to proceed with two series of experiments, one using the same initial fitness vector for all the competition experiments for each pair of organisms and the other with initial fitness randomly assigned in each competition. Table 1 The twelve digital organisms used in the experiments. Size reflects the number of instructions in the corresponding genomes (genome size). Organism Size C185 54 C212 62 C148 70 C119 86 C280 90 C238 92 C216 96 C149 108 C202 134 C295 207 C274 241 C222 272 Fixed initial fitness vector Table 2 shows the critical rates for organism and population size in the experiments with the same, fixed initial fitness vector for each pair of organisms. Only population sizes equal or lower than N = 3600 individuals were assayed, since the original experiments involved only 3600 individuals. Hence, it was impossible to assign the same initial fitness vector used by Wilke et al. [ 22 ] to larger population sizes. In three of the twelve competitions (Table 2 ) we encountered some differences with respect to the critical rate calculated by Wilke et al. [ 22 ]. In organisms C202 and C149 this rate was smaller (1.75 and 0.5 instead of 2.25 and 0.88, respectively) and larger for organism C238 (1.25 instead of 0.88). The remaining rates are equal to those obtained by Wilke et al. [ 22 ] and the differences are due to the better approximation obtained in the original paper through some extra experiments. In our case these additional experiments were not performed because we were more interested in comparing the rates between the two parts of our study. Table 2 Critical mutation rates using one fixed vector per organism. The corresponding values obtained by Wilke et al. [22] for 3600 individuals are shown in the last column ("Original"). Population size Organism 250 500 1250 2500 3600 Original C185 1.25 1.25 1.25 1.25 1.25 1.13 C212 1.25 1.25 1.25 1.25 1.25 1.13 C148 0.75 0.75 0.75 0.75 0.75 0.88 C119 1.75 1.75 1.75 1.75 1.75 1.75 C280 1.25 1.25 1.25 1.25 1.25 1.13 C238 1.25 1.25 1.25 1.25 1.25 0.88 C216 1.25 1.25 1.25 1.25 1.25 1.25 C149 0.5 0.5 0.5 0.5 0.5 0.88 C202 1.75 3 1.75 1.75 1.75 2.25 C295 1.75 1.75 1.75 1.75 1.75 1.88 C274 3 3 3 3 3 3.6 C222 3 3 3 3 3 3.6 Random initial fitness vector As expected from our preliminary results, the use of an initial random vector for each experiment and not for each organism resulted in clear differences with the results encountered by Wilke et al. [ 22 ]. These differences are shown in Table 3 , which presents a summary of the critical mutation rates obtained for each organism and population size using one random vector in each experiment and those originally with one fixed initial vector and N = 3600. In eight of the 11 cases studied the critical rate was higher than the original value. Only in two cases this value was equal to the one originally reported by Wilke et al. and in one case, for organism C222, it was lower. Table 3 summarizes the critical mutation rate encountered for each organism and population size (see also Fig. 1 ). The correlation between the critical mutation rate and population size for each organism allows the separation of the twelve organisms in three main groups (Table 4 ): (i) those with a significant, positive correlation rate (C212, C148, C119, and C202); (ii) organisms with no significant correlation (C185, C222, C280, C149, C216 and C295), and (iii) organism C238, which is the only one with a significant, negative correlation rate. Organism C274 was excluded from this analysis because it did not show a clear pattern of fixation. Table 3 Critical mutation rates using one random vector in each experiment. The last column ("Original") presents the results obtained by Wilke et al. [22] for 3600 individuals and one fixed vector in all the experiments with each organism. Population size Organism 250 500 1250 2500 3600 6400 10000 Original C185 2.25 2.25 2.25 2.25 2.25 2.25 2.25 1.13 C212 0.5 0.5 1.25 1.25 1.25 1.25 1.25 1.13 C148 0.5 0.5 1.25 1.25 1.75 1.75 1.75 0.88 C119 0.5 0.5 1.75 1.75 1.75 1.75 1.75 1.75 C280 1.75 2.25 1.75 2.25 2.25 2 2.25 1.13 C238 2 2 1.75 1.75 1.75 1.75 1.75 0.88 C216 3 3 3 3 3 3 3 1.25 C149 1.5 2 2.25 1.75 2.25 1.75 2 0.88 C202 0.5 0.5 0.5 0.5 2.75 2.75 2.75 2.25 C295 2.25 2.25 3 2.25 3 2.75 2.75 1.88 C274 No pattern 3.6 C222 0.5 0.5 0.5 0.5 0.5 0.5 0.5 3.5 Figure 1 Critical mutation rates using one fixed initial vector per organism. Critical mutation rate (U c ) versus population size (N) for each organism used in the experiments with one fixed initial fitness vector per experiment. In order to obtain exact replicates of the original simulation [22] we did not included population sizes larger than N = 3600. Table 4 Correlation between population size and critical mutation rate in digital organisms. Correlation coefficients (r) were calculated from the experiments with one random vector in each case (Table 3). An asterisk indicates a significant difference from r = 0 for α = 0.05. Two asterisks indicate a significant difference after Bonferroni's correction (α' = 0.0045). Organism r P-value C185 Constant C212 0.791 0.034 * C148 0.932 0.002 ** C119 0.791 0.034 * C280 0.487 0.268 C238 -0.791 0.034 * C216 Constant C149 0.277 0.547 C202 0.866 0.012* C295 0.552 0.199 C274 Not applicable C222 Constant Nevertheless, despite these differences between organisms we cannot conclude that there is a globally significant effect of population size on critical mutation rate. Using Bonferroni's correction for multiple comparisons we obtained a new significance level α' = 0.0045. Therefore only organism C148 has a significant, positive correlation (r = 0.932, P = 0.002). But the differences in the use of a random or fixed initial fitness vector are clear and can be observed by comparing Figures 1 and 2 where values of the critical mutation rate for each organism according to population size are represented. It seems that the use of one fixed initial vector per organism reduces variability in the results. Figure 2 Critical mutation rates using random initial vectors per organism. Critical mutation rate (U c ) versus population size (N) for each organism used in the experiments of one random initial fitness vector per organism. Table 5 shows the critical mutation values found for a population size of 10000 individuals. It can be observed that there is no correlation between critical mutation rate and genome size (r = -0.269, P = 0.424). For comparison, we also compiled similar data for RNA viruses (Table 6 ), including retroviruses, and we did not encounter a significant correlation (r = 0.636, P = 0.125) between genome size and mutation rate (Fig. 3 ). Table 5 Population size and critical mutation rate in digital organisms. Correlation (r = -0.267, P = 0.428) between critical mutation rates (U C ) calculated for a population size of N = 10000 individuals and genomic size of digital organisms. Organism Size U C C185 54 2.25 C212 62 1.25 C148 70 1.75 C119 86 1.75 C280 90 2.25 C238 92 1.75 C216 96 3 C149 108 2 C202 134 2.75 C295 207 2.75 C284 241 N.A. C222 272 0.5 Table 6 Population size and critical mutation rate in viruses. Correlation (r = 0.636, P = 0.125) between the experimentally calculated genomic mutation rate (μ g ) and genomic size of some RNA viruses (adapted from [35]). Virus Size (kb) μ g Lytic RNA viruses VSV [27] 11.2 1.07 Poliovirus [36] 7.4 0.81 Influenza A virus [36] 13.6 0.99 Retroviruses [37] Spleen necrosis virus 7.8 0.16 Molony murine leukemia virus 8.4 0.029 Rous sarcoma virus 9.3 0.43 HIV-1 [38] 9.2 0.22 Figure 3 Genomic mutation rates necessary for quasispecies formation for each of the eleven digital organisms with a population size N = 10000. Discussion The quasispecies model requires a series of conditions to be fulfilled. These requirements are related to four key factors: population size, mutation rate, genome size and neutrality [ 11 , 23 ]. In our study with digital organisms we have analyzed three of these factors and we have related them to known results in virus evolution. The three main conclusions derived from this study are: 1) The use of different initial fitness vectors for otherwise identical experiments results in unpredictable effects on critical mutation rates for different population sizes. These effects were not detected in the original experiments by Wilke et al. [ 22 ] and can alter their conclusions, as contingency, or historical factors, are introduced in the system through different initial conditions leading to different final outcomes. 2) There is no significant correlation between genome size and critical mutation rate. 3) The originally calculated critical mutation rates underestimate their real values. Despite the lack of a general correlation between critical mutation rate and population size, the comparison of Figures 1 and 2 reveals a clear difference with the initial study. By using one fixed initial fitness vector per organism, Wilke et al. [ 22 ] eliminated variability in the outcome of competition (Fig. 2 ). Our study results in different individual responses to changes in population size. In fact, only three of the organisms analyzed maintained a constant response to these changes when different random vectors were used in each of the experiments. Therefore it will be interesting to analyze why certain organisms are more strongly influenced by population size than others. Under the quasispecies model it is expected that increasing population size will favor the establishment of a quasispecies [ 11 ]. A large population size allows the exploration of the neutral space that surrounds the master sequence hence avoiding the effects of genetic drift. If a correlation between the two factors is to be expected, then it should be negative, as with larger population sizes a lower mutation rate is needed to maintain the equilibrium quasispecies structure. Nevertheless, theoretical and simulations results by Wilke et al. [ 22 ] indicate that this is not a true correlation but a phase transition as at low population sizes the critical mutation rate becomes more difficult to ascertain. Our results do not allow to discriminate between both alternatives but they show substantially more variability among organisms (Fig. 1 ) than the ones reported by Wilke et al. [ 22 ], hence pointing at a more complex scenario than that depicted in a simple phase transition. Another key factor in the quasispecies model is genome size. The establishment of the quasispecies is easier in populations with small genomes, as in these the number of neutral sites is reduced and therefore the neutral space is also smaller. However the relationship between critical mutation rate and genomic size to be expected is somewhat contradictory. On the one hand, larger genomes need higher mutation rates because the neutral and adaptive landscapes are larger. On the other hand, it is well known that large genomes require a stability not supplied by high mutation rates, hence the existence of an error threshold that will be discussed later. In fact Eigen [ 24 ] proposed that there should be a negative correlation between these two factors. However in our analyses we have found no such correlation neither in digital organisms nor in experimental data with RNA viruses (Tables 5 and 6 ), in agreement with [ 25 ]. However the absence of a significant correlation does not necessarily mean that there is no relationship between the two factors. Genomic mutation rates impose a limit on the maximum genome size but this does not imply that the best adaptive strategy is to reach the maximum variability attainable for the corresponding genome size [ 26 ]. Our correction to the critical mutation rates estimated in the original paper relates directly to the limits imposed by the mutation rate. In most cases Wilke et al. [ 22 ] obtained critical mutation rates larger than 1 (between 1.13 and 3.5). However, in our experiments we have found these rates to be even larger. This correction in the mutation rate needed for the establishment of a quasispecies is important because estimates of genomic mutation rates of RNA viruses are usually about or below 1 [ 27 , 28 ] (Table 6 ). This limit is known as the error threshold and is another key concept for quasispecies theory. It represents the mutation rate beyond which the information in the molecules would be lost due to degeneracy. The critical mutation rates obtained in the vast majority of cases here reported are larger than 2. If these values were similar in "real" virus populations then they would be beyond the error threshold and therefore the viral quasispecies would not be possible. Therefore, it is important to determine up to which point the comparison of mutation rates between viruses and digital organisms is valid. There are two extreme possibilities: either it is not valid, and therefore digital organisms cannot be invoked as a proof of the evolution of RNA viruses as quasispecies, or if the analogy is possible this means that, at least in the case of RNA viruses, the quasispecies is a theoretical possibility but the practical conditions needed are not met. The presence of an error threshold in viruses is a consequence of a trade-off between the maximization of variability (genomic mutation rate) and the maintenance of molecule integrity (genomic size). It is this trade-off, translated into an error threshold, which might prevent virus quasispecies formation. In this way, the error threshold would not be proof of their existence [ 29 ] but rather of their impossibility in RNA viruses. In conclusion, although some predictions from quasispecies theory are not fulfilled in our experiments, we do have observed the principal prediction that lower fitness competitors can win the competition to high fitness ones, but only under very high mutation rates. Recently, several papers [ 11 , 23 , 30 ] have pointed out the possibility that RNA viruses do not meet all the requirements for quasispecies persistence. The results from this study also suggest that the necessary mutation rates are not attainable either. One possible explanation is that viruses are necessarily more constrained in their evolution than digital organisms. Some experiments demonstrate that the variability found in natural isolates of RNA viruses is not correlated to their mutation rate because some form very conserved RNA secondary structures [ 31 ]. Similarly, it has been demonstrated the frequent selection of the same mutations in the HIV gag region in isolates from different patients, an indication of the limited adaptive solutions able to produce escape mutants to the immune response of cytotoxic T lymphocytes [ 32 ]. Further restrictions could be related with the mechanisms and routes of virus infection [ 33 ]. Analogies are very useful in science, but they have to be used cautiously. Similar features and dynamics between digital organisms and RNA viruses are tempting and usually lead to conclude that both kinds of entities are governed by the same laws. This is not necessarily the case, as practitioners of the comparative method know. In any case, digital organisms are an extraordinary system to experiment with controlled, repeatable evolution conditions and further work with them is necessary to ascertain which evolution features are of their own and which are of common application to other evolving entities. Methods Experimental design The project was started with the twelve pairs of organisms generated in a previous experiment [ 22 ] that had been adapted to two different mutational regimes. The 12 ancestral organisms originating each of the twelve pairs were adapted to low mutation rates (0.5 mutations/genome/replication round – m/g/r) and to high mutation rates (2 m/g/r) for 1000 generations. In all the cases the organisms adapted to a low mutation rate, denoted A, had a significantly larger fitness than their corresponding pair, adapted to a high mutation rate and denoted B. With these twelve pairs we followed the same experimental procedure designed by Wilke et al. [ 22 ]. Basically, we placed in competition equal numbers of A and B organisms during 50 generations. Unlike the original experiment, we did not restrict to a single population size (N = 3600) but we added four smaller (N = 250, 500, 1250, 2500) and, when possible, two larger (N = 6400 and 10000) sizes. The mutation rates under which the competitions were performed were 0.5, 1.0, 1.5, 2.0, 2.5 and 3.0 m/g/r. The A organisms carried a label such that we could follow their proportion in the population. Initial fitness vector In AVIDA organisms occupy the limiting environmental resource, the computer CPU, depending on their "fitness". In order to prevent the collapse of the system when all the competing organisms simultaneously try to use the CPU, there exists one feature designed to prevent the simultaneous replication of all organisms at the start of the competition, when all the organisms might be equally fit since they have not been tested yet in the environment. This is achieved by asynchronously introducing organisms in the competition system by assigning an initial fitness to each organism that introduces a small time lag in the accession to the CPU. For this, Wilke et al. [ 22 ] generated an initial fitness vector in the population for each pair of competing organisms. This vector was generated at random and assigned a different initial fitness for each of the 3600 individuals in the original competition. All the experiments for each pair of organisms were carried out with the same initial fitness vector. During exploratory experiments we noticed that this vector could play a decisive influence in the result of the competition. In consequence, we divided the study into two parts. In the first one we kept the vector assigned by Wilke et al. [ 22 ] to each pair of competing organisms, and we adapted it for other population sizes whenever possible (N = 250, 500, 1250, 2500 individuals). On the other hand, for all the population sizes (including N = 6400 and 10000 individuals) we generated a different random initial vector for each experimental replicate. Therefore, for this second part we generated 252 distinct vectors for pair of organisms in contrast to the five (one per population size) generated in the first part of our study or the single one generated by Wilke et al. [ 22 ] for N = 3600. Critical mutation rate determination The critical mutation rate is "the midpoint between the highest rate at where A prevailed and the lowest rate where B prevailed" [ 22 ]. It represents the rate at which the quasispecies effect is important. We measured this critical parameter as the average of the two rates at which a shift in the winner was observed. It is necessary to clarify the conceptual difference between the critical mutation rate and the error threshold. The first one is the rate at which the prediction of quasispecies theory that organisms with lower fitness can win the competition is fulfilled. However the error threshold is the genomic mutation rate beyond which the information in the molecules that compose the quasispecies loses sense due to mutational degeneracy [ 29 ]. In practical terms, this means that this is the maximum rate that the virus can support. The relationship between the two rates is clear: the critical mutation rate must be necessarily lower or equal than the error threshold because otherwise the quasispecies effects cannot be measured. AVIDA configuration We used versions 1.4 and 1.6 of the AVIDA program. Basically, digital organisms are chains of instructions that act over the CPU with the objective of reproducing as fast as possible. In this manner the CPU time becomes the limiting resource in their evolution. During replication their genomes can mutate and, as a result, a system with variation and therefore with selection and evolution is obtained. Genome sizes of the twelve pairs of digital organisms varied between 54 and 272 instructions (Table 1 ). AVIDA works with some input files that determine the characteristics of the world during the population's evolution. In this case we used the "COPY_MUT_PROB" in the "GENESIS" file, which is the mutation rate that results from dividing the genomic mutation rate by the genome size. In the "EVENT_LIST" file we specified the order of introduction of the individuals and marked each with a hereditary label (A = 1, B = 0). Generally 50 generations were enough for the fixation of the A or B organism in the population. Configuration files used in these experiments are available from the authors web site [ 39 ]. Statistical analysis Each mutation rate-population size combination was replicated six times. For the verification of the relation between the population size and the critical mutation rate we used Pearson's correlation coefficient and a significance level of 5% using Bonferroni's correction [ 34 ]. The same analysis was used for the correlation between genomic sizes and critical mutation rate. Both analyses were carried out with SPSS 11.0 (SPSS Inc.). Authors' contributions IC performed all the simulations, made the statistical analyses and wrote the first draft of the manuscript. AM contributed to the design of the experiments and the discussion of the results. FGC designed and supervised the experiments, discussed the results and their analyses and wrote the final version of the manuscript. All the authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC546199.xml |
512292 | The oncogenic fusion protein RUNX1-CBFA2T1 supports proliferation and inhibits senescence in t(8;21)-positive leukaemic cells | Background The fusion protein RUNX1-CBFA2T1 associated with t(8;21)-positive acute myeloid leukaemia is a potent inhibitor of haematopoetic differentiation. The role of RUNX1-CBFA2T1 in leukaemic cell proliferation is less clear. We examined the consequences of siRNA-mediated RUNX1-CBFA2T1 depletion regarding proliferation and clonogenicity of t(8;21)-positive cell lines. Methods The t(8;21)-positive cell line Kasumi-1 was electroporated with RUNX1-CBFA2T1 or control siRNAs followed by analysis of proliferation, colony formation, cell cycle distribution, apoptosis and senescence. Results Electroporation of Kasumi-1 cells with RUNX1-CBFA2T1 siRNAs, but not with control siRNAs, resulted in RUNX1-CBFA2T1 suppression which lasted for at least 5 days. A single electroporation with RUNX1-CBFA2T1 siRNA severely diminished the clonogenicity of Kasumi-1 cells. Prolonged RUNX1-CBFA2T1 depletion inhibited proliferation in suspension culture and G1-S transition during the cell cycle, diminished the number of apoptotic cells, but induced cellular senescence. The addition of haematopoetic growth factors could not rescue RUNX1-CBFA2T1-depleted cells from senescence, and could only partially restore their clonogenicity. Conclusions RUNX1-CBFA2T1 supports the proliferation and expansion of t(8;21)-positive leukaemic cells by preventing cellular senescence. These findings suggest a central role of RUNX1-CBFA2T1 in the maintenance of the leukaemia. Therefore, RUNX1-CBFA2T1 is a promising and leukaemia-specific target for molecularly defined therapeutic approaches. | Background The chromosomal translocation t(8;21) (q22;q22), which is associated with 10–15% of all cases of acute myeloid leukaemia, fuses the DNA binding domain of the transcription factor RUNX1 (also called AML1 or CBFα) to the almost complete open reading frame of CBFA2T1 (also named MTG8 or ETO) [ 1 , 2 ]. The resulting fusion protein RUNX1-CBFA2T1 (AML1/MTG8, AML1/ETO) interferes with haematopoetic gene expression by recruiting histone deacetylases via N-CoR and mSin3 to promoters, thereby inhibiting the transcription of the respective target gene [ 3 - 7 ]. Moreover, by directly binding to and sequestering transcription factors, such as SMAD3, C/EBPα or vitamin D receptor, RUNX1-CBFA2T1 interferes with signal transduction pathways controlling differentiation and proliferation [ 8 - 12 ]. Consequently, RUNX1-CBFA2T1 blocks myeloid differentiation and promotes self-renewal of haematopoetic progenitors [ 13 - 16 ]. The influence of RUNX1-CBFA2T1 on the control of proliferation and apoptosis is less clear. On the one hand, its ectopic expression in several cell types, including leukaemic cell lines such as U937, inhibits proliferation and enhances apoptosis [ 13 ]. On the other hand, RUNX1-CBFA2T1 may interfere with p53-dependent cell cycle arrest and apoptosis by suppressing the p53-stabilizing protein p14 ARF [ 17 ]. RUNX1-CBFA2T1 expression supports the expansion of haematopoetic progenitor cells, which has been mainly attributed to its anti-differentiation capabilities, but which may also depend on a proliferation supporting activity of RUNX1-CBFA2T1 [ 18 - 21 ]. RUNX1-CBFA2T1 alone is not sufficient to cause leukaemia [ 22 , 23 ]. Instead, secondary mutations have to be acquired in addition to RUNX1-CBFA2T1 to induce leukaemia [ 24 - 27 ]. Cellular senescence limits the proliferative capacity of cells and is characterized by an irreversible G1 arrest [ 28 ]. Senescent cells cannot be stimulated with mitogens to enter the S phase of the cell cycle. Nevertheless, senescent cells are still viable and metabolically active [ 29 ]. They can be distinguished from non-senescent cells by the expression of senescence-associated β-galactosidase activity, which can be detected at slightly acidic pH [ 30 ]. In the case of replicative senescence, cells enter G1 arrest after the telomeres have shortened below a critical length [ 29 ]. After exposure to stresses, cells may also undergo stress-induced senescence instead of apoptosis or transient cell cycle arrest [ 28 ]. The molecular mechanisms of senescence are still very incompletely understood. However, several regulators of cell cycle progression such as pRb, p53 or the cdk inhibitors p16 Ink4a or p27 Kip1 are involved in the establishment of senescence [ 31 ]. Furthermore, overexpression of oncogenic H-Ras in murine embryonic fibroblasts (MEFs) induce premature senescence in a PML-dependent fashion [ 32 , 33 ]. Similarly, overexpression of RUNX1 in MEFs induces senescence likely by upregulating p19 Arf [ 17 ]. However, control of senescence by an endogenously expressed oncogene such as RUNX1-CBFA2T1 in t(8;21)-positive leukaemic cells has not been shown yet. The specific inhibition of gene expression by small interfering RNAs (siRNAs) provides a new approach to investigate the functions of oncogenes in the development of cancer, thereby complementing other approaches such as ectopic (over-) expression [ 34 - 36 ]. We and others have used siRNAs to specifically down-modulate leukaemic fusion genes such as BCR-ABL or RUNX1-CBFA2T1 [ 14 , 37 - 39 ]. We have shown that the siRNA-mediated depletion of RUNX1-CBFA2T1 led to a sensitization towards myeloid differentiation inducing agents such as TGFβ and vitamin D 3 [ 14 ]. Here, we report the consequences of RUNX1-CBFA2T1 depletion on CD34 expression (a surface marker, indicator of differentiation), for the proliferation and clonogenicity of in t(8;21)-positive leukaemic cell lines. We demonstrate that RUNX1-CBFA2T1 supports the clonogenicity and proliferation of t(8;21) positive Kasumi-1 cells by interfering with the establishment of cellular senescence. Methods siRNAs The siRNAs siAGF1 and siAM targeting the fusion site of the RUNX1-CBFA2T1 mRNA, the mismatch control siAGF6, and the unrelated controls targeting luciferase (siGL2) (for sequences see ref. 13) or MLL-AF4 (siMLL14; sense 5'-AAA CCA AAA GAA AAG CAG ACC-3', antisense 5'-GGU CUG CUU UUC UUU UGG UUU UU-3') were synthesized by either Alnylam Europe AG (Kulmbach, Germany) or MWG Biotech (Ebersberg, Germany) and hybridized as described [ 14 ]. Cell culture and siRNA transfection The t(8;21)-positive acute myeloid leukaemia (AML) cell lines Kasumi-1 (Deutsche Sammlung für Mikroorganismen und Zellkulturen (DSMZ) No. ACC 220) [ 40 ] and SKNO-1 [ 41 ], and the t(8;21)-negative leukaemic lines MV4-11 (DSMZ No. ACC 102), NB4 (DSMZ No. ACC 207), HL60 (DSMZ No. ACC 3), U937 (DSMZ No. ACC 5) and K562 (DSMZ No. ACC 10) were cultivated and electroporated together with siRNAs as described [ 14 , 42 , 43 ]. Briefly, cells were electroporated in 100–500 μl culture medium at a density of 5 × 10 6 /ml in 4 mm electroporation cuvettes. SiRNAs were added immediately before electroporation. If not otherwise indicated, the siRNA concentration during electroporation was 100 nM. Electroporations were performed using a rectangle pulse EPI 2500 electroporator (Fischer, Heidelberg, Germany; ). Parameters were 330 V and 10 ms (milliseconds) for Kasumi-1 cells, and 350 V and 10 ms (milliseconds) for all other cell lines. Fifteen minutes after electroporation, cells were diluted twenty fold in culture medium and incubated at 37°C, 5% CO 2 and 92% humidity. Using this protocol, we routinely observe less than 10% of dead cells after electroporation with an siRNA transfection efficiency close to 100% [ 14 , 42 , 43 ]. Immunoblotting To obtain total cell lysates, cells were washed with phosphate-buffered saline, lysed in Urea lysis buffer (9 M urea, 4% (w/w) 3-[(3-Cholamidopropyl)-dimethylammonio]-propansulfonat (CHAPS), 1% (w/w) Dithiothreitol) and treated for 30 min on ice with 2.5 u/ml Benzonase (Merck, Darmstadt, Germany). Total (10–20 μg) and nuclear lysates (10 μg) were analyzed as described [ 14 ]. Band intensities were determined by scanning the exposed films followed by quantification using the Image Gauge 3.0 software (Fujifilm, Düsseldorf, Germany). The following antibodies were used for detection: AML1/RHD (2.5 mg/l; Oncogene Research PC285, Boston, USA); p27 C-19 (200 μg/l; Santa Cruz Biotechnology sc-9737; Heidelberg, Germany); Tubulin Ab-4 (1 mg/l, NeoMarkers MS-719-P0, Fremont, USA). Real-time RT-PCR Total RNA was isolated using the RNeasy Kit (Qiagen, Hilden, Germany). Reverse transcription was performed with random hexamers using MMLV-RT, RNase H - (Promega, Heidelberg, Germany) as suggested by the manufacturer. Real time PCRs were performed using Sybr-Green Mix (Applied Biosystems, Warrington, UK), 300 nM primers and 20% v/v diluted RT reaction mix using standard conditions on a 7000 Sequence detection system (Applied Biosystems, Foster City, CA, USA). The primer sequences for STAT1 (5'-CAT CAC ATT CAC ATG GGT GGA (forward primer) and 5'-GGT TCA ACC GCA TGG AAG TC (reverse primer)), for CD34 (5'-TCC AGA AAC GGC CAT TCA G (forward primer) and 5'-CCC ACC TAG CCG AGT CAC AA (reverse primer)), for G-CSF (5'-CCC ACC TTG GAC ACA CTG C (forward primer) and 5'-AGT TCT TCC ATC TGC TGC CAG (reverse primer)) and for GAPDH (5'-GAA GGT GAA GGT CGG AGT C (forward primer) and 5'-GAA GAT GGT GAT GGG ATT TC (reverse primer)) were designed with Primer-Express software (Applied Biosystems, Foster City, CA, USA). Proliferation, single cell expansion and colony formation assays To determine proliferation rates and doubling times, cells were counted using trypan blue. For single cell expansion assays, 24 h after electroporation cells were diluted to a density of 10 cells/ml and plated in 100 μl aliquots into 96 well tissue culture plates. Colony formation assays were performed 24 h after electroporation with a density of 5,000 cells/ml in semisolid medium (RPMI1640 containing 20% fetal calf serum and 0.5625% methyl cellulose). Colonies were counted 14 days after plating. The cytokine concentrations were 20 ng/ml for G-CSF and GM-CSF in the cell cycle and senescence assays, and 10 ng/ml for all growth factors in the colony formation assays. Cell cycle analysis Cell cycle analysis was performed as described [ 44 ]. Four days after electroporation, 10 6 cells were suspended in 200 μl citrate buffer (250 mM sucrose, 40 mM sodium citrate pH 7.6) followed by the addition of 800 μl staining solution (phosphate-buffered saline containing 20 mg/l propidium iodide, 0.5% (w/w) NonIdet P40, 500 μM EDTA) and 10 μl boiled RNase A (10 g/l). Cells were kept for 30 minutes on ice and analyzed by flow cytometry (FACSCalibur, Becton Dickinson, Heidelberg, Germany). Data analysis was performed with FCSPress 1.3 . Analysis of apoptosis and of CD34 expression For quantification of apoptotic cells, cells were stained with FITC-labeled annexin V as suggested by the supplier (Bender, Vienna, Austria). CD34 expression was monitored by staining with anti-human CD34-FITC (clone 581, BD Pharmingen #555388, Heidelberg, Germany). Analysis was performed by flow cytometry (FACSCalibur, Becton Dickinson, Heidelberg, Germany) using FCSPress 1.3 for data analysis. Results Time course of RUNX1-CBFA2T1 protein depletion by siRNAs Electroporation of t(8;21)-positive cells with the RUNX1-CBFA2T1 siRNA siAGF1, but not with the mismatch control siRNA siAGF6, resulted in a three- to fourfold reduction of RUNX1-CBFA2T1 fusion protein in both Kasumi-1 cells (fig. 1A , lanes 1–3) and SKNO-1 cells (fig. 1A , lanes 4–6). One day after electroporation with siAGF1, RUNX1-CBFA2T1 protein levels were already reduced by 75% (fig. 1B ). This finding suggests that RUNX1-CBFA2T1 protein has a half-life of less than 24 hours. The depletion of RUNX1-CBFA2T1 lasted for at least 7 days with a slow recovery of RUNX1-CBFA2T1 protein levels from day 5 on (fig. 1B ), in agreement with the previously reported time-course of siRNA-mediated RUNX1-CBFA2T1 mRNA reduction [ 14 ]. Transfection with siRNAs may induce an interferon response leading to non-specific effects on cellular processes such as proliferation or differentiation [ 45 ]. Since STAT1 is an interferon-stimulated gene, and since several siRNAs have been shown to induce STAT1 expression [ 45 ], we examined a possible induction of STAT1 gene expression by RUNX1-CBFA2T1 siRNAs or by mismatch control siRNAs in Kasumi-1 cells. Neither the electroporation with RUNX1-CBFA2T1 siRNA nor with the control siAGF6 caused a substantial change in STAT1 mRNA levels (data not shown), suggesting that these siRNAs do not induce an interferon response in Kasumi-1 cells. RUNX1-CBFA2T1 siRNAs decrease the clonogenic growth of Kasumi-1 cells In comparison to mock-transfected cells, electroporation with RUNX1-CBFA2T1 siRNAs resulted in a ten- to twenty fold decreased clonogenicity in both single cell expansion (data not shown) and colony formation assays (fig. 2 ). Electroporation with the mismatch control siRNA siAGF6 affected the clonogenicity in neither of these assays. We did not observe major differences in colony morphology. However, the colonies derived from cells treated with RUNX1-CBFA2T1siRNAs tended to be smaller in size than those derived from control cells (data not shown). We also examined the effects of RUNX1-CBFA2T1 siRNA on the clonogenicity of AML cell lines not expressing RUNX1-CBFA2T1. Colony formation of the t(8;21)-negative myeloid leukaemia cell lines MV4-11, NB4, HL60, U937 and K562 was not affected upon electroporation with RUNX1-CBFA2T1 siRNA (data not shown), which argues against a general, unspecific effect of the siRNAs used in this study on leukaemic clonogenicity. Therefore, the application of RUNX1-CBFA2T1 siRNAs is sufficient to inhibit the clonogenic growth of Kasumi-1 cells, indicating that RUNX1-CBFA2T1 is essential for the clonogenicity of these leukaemic cells. RUNX1-CBFA2T1 depletion reduces CD34 expression To examine the effects of RUNX1-CBFA2T1 on the progenitor status of t(8;21)-positive leukaemic cells, we analyzed the CD34 expression levels in siRNA-treated Kasumi-1 cells. When compared to mock-transfected cells, depletion of RUNX1-CBFA2T1 caused a twofold decrease in CD34 surface expression 7 days after electroporation (fig. 3 ). Control siRNA-treated cells contained only slightly lower amounts of CD34 surface marker. RUNX1-CBFA2T1 suppression inhibits proliferation in suspension cell culture In contrast to its effects on clonogenicity, a single electroporation with RUNX1-CBFA2T1 siRNA was not sufficient to substantially inhibit Kasumi-1 proliferation in suspension culture. However, repeating the electroporation every three to four days strongly inhibited cell proliferation (fig. 4A ). During a course of three to four electroporations with or without mismatch control siRNA, the cell numbers steadily increased with a doubling time of 3 days (fig. 4B ). RUNX1-CBFA2T1 siRNA treatment increased the doubling time twofold. Therefore, repetitive siRNA applications causing a long-lasting RUNX1-CBFA2T1 depletion inhibited cell proliferation of Kasumi-1 cells. These results suggest that RUNX1-CBFA2T1 is essential for the proliferation of t(8;21)-positive leukaemic cells in suspension culture. RUNX1-CBFA2T1 suppression interferes with cell cycle progression Next, we examined possible consequences of RUNX1-CBFA2T1 depletion for the cell cycle distribution of Kasumi-1 cells. Three days after an electroporation with RUNX1-CBFA2T1 siRNA we observed 25% less cells being in S phase (data not shown). Two or more consecutive electroporations with RUNX1-CBFA2T1 siRNA caused a reduction of the amount of S phase cells by 50% and a 30% reduction in G2/M phase cells, with a corresponding increase of the G1/G0 fraction by 20% (fig. 4C ). Treatment of Kasumi-1 cells with the mismatch control siRNA did not affect the cell cycle distribution. The changes in the cell cycle distribution are associated with changed expression levels of CDKN1B (p27 Kip1 ), a general inhibitor of Cyclin-CDK complexes and, thereby, of cell cycle progression. Two electroporations with RUNX1-CBFA2T1 siRNA caused a twofold increase in CDKN1B protein levels in RUNX1-CBFA2T1-depleted cells, but not in control cells (fig. 4D ). Therefore, siRNA-mediated inhibition of RUNX1-CBFA2T1 expression may interfere with the transition of Kasumi-1 cells from the G1 to the S phase of the cell cycle. RUNX1-CBFA2T1 controls apoptosis A possible reason for the observed inhibition of proliferation is the induction of apoptosis upon suppression of RUNX1-CBFA2T1. We addressed this possibility by examining the amount of apoptotic cells by annexin V staining and of hypodiploid (subG1) cells by flow cytometry. Electroporation of Kasumi-1 cells together with the RUNX1-CBFA2T1 siRNA siAGF1 neither increased the amount of hypodiploid cells nor the amount of annexin V positive cells when compared to control cells. Instead, after siAGF1 treatment we observed a slight but reproducible decrease in the amount of apoptotic cells in both assays. This decrease was more pronounced after two consecutive electroporations with siRNAs. Repeating the siAGF1 electroporation after 4 days resulted in a threefold decrease of the amount of annexin V stained cells (fig. 5 ) indicating that RUNX1-CBFA2T1 depletion has an antiapoptotic effect on Kasumi-1 cells, and that the reduced proliferation upon siRNA treatment is not related to increased apoptosis. RUNX1-CBFA2T1 siRNAs induce senescence in Kasumi-1 cells The observed G1 arrest upon RUNX1-CBFA2T1 depletion could be either reversible or irreversible. An irreversible G1 arrest is a hallmark of cellular senescence. To address the question of the nature of the G1 arrest, and to analyze a possible influence of RUNX1-CBFA2T1 on cellular senescence, we stained siRNA- and mock-treated cells for senescence-associated β-galactosidase activity. After two or more electroporations with RUNX1-CBFA2T1 siRNA, but not with control siRNA, a significant fraction of the Kasumi-1 cells stain positive for β-galactosidase (fig. 6A ). Cell counting revealed up to 50% of senescent cells, whereas mock or control siRNA treatment caused only a minor increase in senescent cells compared to untreated cells (fig. 6B ). Therefore, depletion of RUNX1-CBFA2T1 results in an increase in cellular senescence. This implies that the observed G1 arrest is irreversible for a major fraction of RUNX1-CBFA2T1 siRNA treated cells. Haematopoetic growth factors cannot rescue RUNX1-CBFA2T1-depleted cells in the long term A possible mechanism of how RUNX1-CBFA2T1 prevents cell cycle arrest and senescence is the autocrine production of growth factors. Kasumi-1 cells express, for instance, G-CSF (fig. 7A ), but not GM-CSF (data not shown). Depletion of RUNX1-CBFA2T1 caused a twofold decrease in G-CSF transcript levels (fig. 7A ). Therefore, we examined whether addition of haematopoetic growth factors rescued siRNA-treated cells from cell cycle arrest and senescence. Depletion of RUNX1-CBFA2T1 for 16 days reduced the fraction of S phase cells twofold compared to control cells. This decrease is independent of G-CSF or GM-CSF (fig. 7B ). When analyzing the cell cycle distribution siRNA or mock-treated cells in dependence on the duration of siRNA treatment, the fraction of S phase cells already decreases within 4 days of leukaemic fusion protein depletion and reaches its minimum after further 4 to 8 days (fig. 7C ). Addition of G-CSF, and particularly of GM-CSF delayed this decrease, but could not prevent it. Furthermore, in RUNX1-CBFA2T1 siRNA treated cells, neither G-CSF nor GM-CSF addition caused any substantial change in the generation of senescent cells compared to cells without growth factors (fig. 7D ). In each setting, the fraction of senescent cells increased from 2% at day 4 to 30–50% at days 12 and 16. Control groups contained between 1–10% of senescent cells. Finally, we examined the effects of several haematopoetic growth factors on colony formation of siRNA-treated Kasumi-1 cells. Supplementing the semisolid medium with the corresponding growth factors increased colony numbers for both control cells and RUNX1-CBFA2T1-depleted cells (fig. 8 ). G-CSF enhanced colony formation by control cells twofold, and by RUNX1-CBFA2T1-depleted cells threefold, and GM-CSF three- and fivefold, respectively. However, in comparison to control cells, neither G-CSF nor GM-CSF could completely restore colony formation of RUNX1-CBFA2T1-depleted cells. Taken together, the addition of G-CSF or GM-CSF delays, but does not prevent the G1 arrest in RUNX1-CBFA2T1-depleted cell. Furthermore, senescence is not inhibited by these growth factors. In line with these findings, neither of these growth factors completely restored the impaired clonogenicity of such cells. These results indicate that neither G-CSF nor GM-CSF can rescue leukaemic cells suffering of siRNA-mediated RUNX1-CBFA2T1 depletion. Discussion The leukaemic fusion protein RUNX1-CBFA2T1 inhibits haematopoetic differentiation by repressing gene expression via recruitment of histone deacetylases [ 4 - 7 ], and by sequestering transcription factors such as C/EBPα, SMADs or vitamin D 3 receptor [ 8 - 11 ]. In addition, RUNX1-CBFA2T1 inhibits cell proliferation by inducing apoptosis in a variety of different cell types [ 13 , 46 ], in spite of repressing the proapoptotic gene p14 Arf [ 17 ]. The notable exception are haematopoetic stem cells (HSCs); ectopic expression of RUNX1-CBFA2T1 causes an increased clonogenicity and results in their expansion both ex vivo and in vivo [ 20 , 21 ]. In addition to the block of differentiation, RUNX1-CBFA2T1 may also support proliferation, as, for instance, RUNX1-CBFA2T1-expressing HSCs maintain their telomere lengths and their CD34 expression [ 21 ]. We show that RUNX1-CBFA2T1 also supports proliferation, CD34 expression and colony formation in the t(8;21)-positive leukaemic cell line Kasumi-1. Thus, RUNX1-CBFA2T1 functions are similar in HSCs and in Kasumi-1. In the latter, support of proliferation is not caused by an inhibition of cell death. Instead, as observed with many other cell types [ 13 , 46 ], RUNX1-CBFA2T1 expression is associated with a certain extent of apoptosis. This observation disagrees with a report employing ribozymes to interfere with RUNX1-CBFA2T1 expression [ 47 ]. In contrast to our study, no clear reduction of either fusion transcript or protein was shown. It may, thus, be possible that the ribozyme effects observed in this study are not related to a suppression of RUNX1-CBFA2T1. The reduced proliferation of Kasumi-1 cells upon RUNX1-CBFA2T1 depletion is paralleled by a cell cycle arrest in G1. Particularly from a therapeutic point of view, it is important to know, whether this G1 block will be relieved by recovering RUNX1-CBFA2T1 expression, or by growth factors, or whether this arrest is irreversible. We observe that RUNX1-CBFA2T1 depleted cells show enhanced levels of senescence-associated β-galactosidase activity, a hallmark of cellular senescence. Therefore, the inhibition of G1 to S transition is associated by the induction of senescence indicating that this G1 arrest is irreversible. Interestingly, the elevated levels of p27 Kip1 (CDKN1B) during RUNX1-CBFA2T1 depletion are in line with the recently demonstrated central role of this CDK inhibitor in the establishment of senescence [ 49 , 50 ]. Furthermore, RUNX1 was shown to induce senescence in murine embryonic fibroblasts, probably by inducing p19 Arf [ 17 ]. Since RUNX1-CBFA2T1 inhibits the expression of p19 Arf and of its human homologue, p14 Arf [ 17 ], an increased expression of this regulator of p53 may also account for the induction of senescence in RUNX1-CBFA2T1-depleted Kasumi-1 cells. An important point to consider is a possible relation between the effects of RUNX1-CBFA2T1 on clonogenicity, proliferation and senescence, and its anti-differentiation activity. For instance, increased β-galactosidase activity may also be associated with late monocytic differentiation [ 48 ]. Moreover, later stages of myeloid differentiation are characterized by a stop of proliferation. Therefore, the observed inhibition of proliferation upon RUNX1-CBFA2T1 depletion might be directly related to myeloid differentiation. Kasumi-1 and SKNO-1 cells express the myeloid differentiation marker CD11b when treated with RUNX1-CBFA2T1 siRNA in combination with TGFβ and vitamin D 3 [ 14 ]. However, in the absence of differentiation inducers, even extended siRNA treatment only results in a small fraction of less than 5% of CD11b-positive cells [ 14 ], and data not shown) which cannot be accountable for the up to 50% β-galactosidase positive cells. Nevertheless, depletion of RUNX1-CBFA2T1 causes an onset of myeloid differentiation; therefore, a possible link between the antiproliferative effects of RUNX1-CBFA2T1 suppression and early differentiation events cannot be entirely excluded. The haematopoetic growth factors G-CSF and GM-CSF were shown to support clonogenicity and proliferation in Kasumi-1 cells [ 40 ]. However, neither of these factors prevented senescence in RUNX1-CBFA2T1-depleted cells. Furthermore, G1-S transition was only transiently restored; after extended terms of RUNX1-CBFA2T1 depletion, the fraction of S phase cells also diminished in the presence of G-CSF or GM-CSF. Finally, clonogenicity of RUNX1-CBFA2T1-depleted cells was not completely restored in the presence of G-CSF or GM-CSF. These findings show that at least some haematopoetic growth factors cannot rescue leukaemic cells treated with RUNX1-CBFA2T1 siRNAs, and suggest that targeting of leukaemic cells with such siRNAs in vivo might interfere with the development and maintenance of leukaemia. However, a remaining possibility might be that the stroma environment with its complex mixture of growth factors and cell-cell interactions may be able to rescue RUNX1-CBFA2T1-depleted cells. We are currently investigating this question both in cell culture and in vivo . A major point of concern for siRNA applications is the specificity of the corresponding siRNA [ 51 ]. Reasons for "off-target" effects may be (i) the siRNA-induced RNA cleavage of transcripts with imperfect homology to the siRNA [ 52 , 53 ], (ii) a possible competition with endogenously expressed micro RNAs (miRNAs), which may be involved in the control of, for instance, cell differentiation, for RISC protein components [ 54 ], and (iii) the induction of an interferon response [ 45 , 55 ]. However, neither RUNX1-CBFA2T1 siRNAs nor the mismatch control siRNA affected the clonogenicity of five different leukaemic cell lines not carrying the translocation t(8;21). Furthermore, expression of the interferon-inducible gene STAT1 is not changed upon electroporation with RUNX1-CBFA2T1 siRNA or mismatch control siRNA. These results do not argue for general "off-target" effects being responsible for the observed changes in cell proliferation and clonogenicity of RUNX1-CBFA2T1 siRNA-treated Kasumi-1 cells. Instead, the data suggest the siRNA-mediated depletion of this leukaemic fusion protein as the cause for the observed phenotypic changes. Conclusions In summary, we show that RUNX1-CBFA2T1 does not only inhibit the myeloid differentiation [ 14 ], but is also essential for the proliferation and clonogenicity of the t(8;21)-positive leukaemic cell line Kasumi-1 by interfering with the establishment of senescence. Since this cell line was derived from a patient suffering of leukaemia refractory to chemotherapy [ 40 ], these findings suggest a central role of RUNX1-CBFA2T1 not only in the expansion of preleukaemic progenitor cells, but also in the maintenance of the leukaemia. Moreover, they imply that this leukaemic fusion gene is a promising target for molecularly defined therapeutic approaches. Competing interests HPV is an employee of Alnylam Europe AG, which also provided some of the siRNAs used in this study. Authors' contributions NM performed the proliferation, senescence and cell cycle assays. BD analyzed the clonogenicity of growth-factor treated Kasumi-1 cells and of t(8;21) negative cell lines. HR performed the immunoblot analyses and was involved in the analysis of clonogenicity and apoptosis. CC examined the expression levels of CD34 and of STAT1. HPV, AG, GH and AN participated in the coordination of the study. JK and OH conceived and designed the experiments. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC512292.xml |
523846 | Different HIV Drugs Cause Different Lipid Profiles | null | Nevirapine and efavirenz are the most commonly prescribed of the class of antiretroviral drugs called non-nucleoside reverse transcriptase inhibitors (NNRTIs). Efavirenz has the advantage of once-daily dosing. In a recent study called the 2NN study (Lancet 363: 1253–1263), it appeared to be only marginally superior to nevirapine in terms of clinical success and virological suppression. Van Leth and colleagues have now shown that while nevirapine and efavirenz both raise high-density lipoprotein (HDL) cholesterol (the “good” type of cholesterol), the overall lipid profile is better with nevirapine than with efavirenz. “These data suggest that nevirapine may be preferable to efavirenz in HIV-infected adults with other cardiovascular risk factors,” says the study's academic editor, Andrew Carr of St. Vincent's Hospital in Darlinghurst, Australia. “However, perceived cardiovascular risk is only one factor that would affect the choice between these two drugs.” Van Leth and colleagues prospectively analyzed the lipids of patients enrolled in the 2NN study, a randomized, open-label efficacy study that included adults with HIV who had never been on antiretroviral drugs. All patients were given stavudine and lamivudine and were then randomized into three treatment groups: nevirapine, efavirenz, or both. For the lipid analysis, which was preplanned, the researchers included only the nevirapine and efavirenz groups (417 and 289 patients, respectively). This was because the 2NN study showed that simultaneous use of nevirapine and efavirenz should be avoided—the combination is associated with increased toxicity without increased efficacy. The increase in HDL cholesterol was significantly higher with nevirapine than with efavirenz. There was a decrease in the ratio of total cholesterol to HDL cholesterol with nevirapine and an increase with efavirenz. The study does not prove, however, that the rise in HDL cholesterol seen with NNRTIs (especially nevirapine) actually leads to a reduction in coronary heart disease. “There are no vascular functional data,” says Carr, “or clinical vascular endpoint data that confirm that the statistically significant lipid differences observed are clinically significant.” The study was funded by Boehringer Ingelheim, the manufacturer of nevirapine. The authors clearly state that the company had “a nonbinding input on issues of study design and analyses” but it had “no influence on reporting of the data or the decision to publish.” Despite its limitations, van Leth and colleagues' study “moves clinicians and patients away from ‘one-size-fits-all’ antiretroviral therapy,” says Carr. “It takes us further along the path of choice of antiretroviral therapy being individualized according to other patient comorbidities and risk factors, as well as therapy simplicity and side effects.” | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC523846.xml |
545490 | Mapping and mutation of the conserved DNA polymerase interaction motif (DPIM) located in the C-terminal domain of fission yeast DNA polymerase δ subunit Cdc27 | Background DNA polymerases α and δ play essential roles in the replication of chromosomal DNA in eukaryotic cells. DNA polymerase α (Pol α)-primase is required to prime synthesis of the leading strand and each Okazaki fragment on the lagging strand, whereas DNA polymerase δ (Pol δ) is required for the elongation stages of replication, a function it appears capable of performing on both leading and lagging strands, at least in the absence of DNA polymerase ε (Pol ε). Results Here it is shown that the catalytic subunit of Pol α, Pol1, interacts with Cdc27, one of three non-catalytic subunits of fission yeast Pol δ, both in vivo and in vitro . Pol1 interacts with the C-terminal domain of Cdc27, at a site distinct from the previously identified binding sites for Cdc1 and PCNA. Comparative protein sequence analysis identifies a protein sequence motif, called the DNA polymerase interaction motif (DPIM), in Cdc27 orthologues from a wide variety of eukaryotic species, including mammals. Mutational analysis shows that the DPIM in fission yeast Cdc27 is not required for effective DNA replication, repair or checkpoint function. Conclusions The absence of any detectable phenotypic consequences arising from mutation of the DPIM suggests that despite its evolutionary conservation, the interaction between the two polymerases mediated by this motif is a non-essential one. | Background Three conserved DNA polymerase enzymes whose activities are essential for complete chromosomal DNA replication have been identified through biochemical studies in mammalian systems [ 1 ] and combined genetic and biochemical studies in yeast [ 2 ]. During S-phase, the DNA polymerase α-primase complex synthesises the short RNA-DNA segment that is used to prime synthesis of the leading strand at the chromosomal replication origin and synthesis of each Okazaki fragment on the lagging strand. The short RNA segment is synthesised by the primase and then extended by 10–20 nucleotides by Pol α. The 3' end of the RNA-DNA primer is recognised by replication factor C (RFC), which displaces the Pol α-primase complex and catalyses the loading of the sliding clamp PCNA at the primer-template junction. PCNA then acts as a processivity factor for the Pol δ and/or Pol ε enzymes. The exact roles played by Pol δ and Pol ε remain unclear (for a recent perspective, see ref. [ 3 ] and references therein) but Pol δ is most likely responsible for lagging strand replication and may also play a role on the leading strand. Yeast lacking Pol ε catalytic activity are viable but are slow growing and somewhat impaired in chromosome replication [ 4 - 7 ]. In such cells, Pol δ is thought to perform the bulk of nascent DNA chain elongation, raising the possibility that this enzyme performs a similar function in wild-type cells. If this is the case, Pol ε could have a specialised role, at replication origins for example, or in the replication of sites of sister chromosome cohesion [ 3 ]. Each of the three essential polymerases is a multi-subunit entity, comprising a large catalytic subunit and a number of smaller subunits that are presumed to play either regulatory or structural roles [ 2 ]. Little is known of the biochemical functions of the smaller subunits but in yeast most are, like the three catalytic subunits, essential proteins. In the fission yeast Schizosaccharomyces pombe , the Pol α-primase and Pol δ complexes are both heterotetrameric in structure [ 2 ]. The catalytic subunits of the two complexes, Pol1 and Pol3 respectively, are members of the B family of DNA polymerases typified by bacteriophage T4 polymerase or Pfu [ 8 ]. Both complexes also contain related B-subunits. These proteins are members of a larger family of phosphotransferase and nuclease enzymes [ 9 , 10 ], although no enzymatic activity has been detected for either of the DNA polymerase subunits. In fission yeast, the B-subunit of Pol δ is the Cdc1 protein. Fission yeast Pol δ also contains two further subunits: the C-subunit Cdc27, which functions in part to link the polymerase to PCNA [ 11 , 12 ], and the D-subunit Cdm1. Of the four Pol δ subunits, only Cdm1 is non-essential [ 13 ]. Orthologues of all four of these proteins (Pol3, Cdc1, Cdc27 and Cdm1) make up mammalian Pol δ [ 14 , 15 ]. Perhaps surprisingly, there have been few reports of physical interactions between the various polymerase enzymes believed to be present at the eukaryotic replication fork. One exception to this comes from large-scale functional analysis of the budding yeast proteome where an interaction was uncovered between the catalytic subunit of the Pol α-primase complex, Pol1, and the C-subunit of the Pol δ complex in budding yeast, Pol32 [ 16 ]. Interaction between these proteins was detected using the two-hybrid system, raising the possibility that the interaction was not a direct one, but was instead mediated via a third protein factor or complex. Recently, the two-hybrid result was independently confirmed and a direct interaction between the budding yeast Pol α-primase and Pol δ demonstrated by mixing of the purified enzyme complexes followed by co-immunoprecipitation [ 17 ]. By this method, Pol α-primase and Pol δ were shown interact in a Pol32-dependent manner [ 17 ]. A form of Pol δ containing a mutant Pol32 protein lacking a 40 amino acid region of the C-terminal domain (Pol32-8, lacking amino acids 270–309) was also shown to be unable to co-immunoprecipitate with Pol α-primase [ 17 ], providing the first indication of the location of the Pol α binding site on Pol32. In this paper, it is shown that the orthologues of the budding yeast Pol1 and Pol32 proteins in fission yeast, Pol1 and Cdc27 respectively, also interact in the two-hybrid system. It is also shown that these proteins are capable of interacting directly with one another, in vitro , using purified recombinant proteins and peptides. Interaction is also seen between the human Cdc27 and Pol1 homologues, p66 and p180. The binding site for Pol1 maps to the extended C-terminal domain of the Cdc27 protein and requires the presence of a short protein sequence motif, which we have designated the DPIM, for DNA polymerase interaction motif. This short sequence is conserved in Cdc27 orthologues from a wide variety of eukaryotic species. Despite this evolutionary conservation, mutational inactivation of the Pol1 binding motif does not affect Cdc27 function in vivo . The implications of these results are discussed. Results Interaction between fission yeast Pol1 and Cdc27 To test for interaction between fission yeast Pol1 and Cdc27, the two-hybrid system was used. Full-length Cdc27 fused to the activation domain (AD) of the yeast Gal4 protein was tested for its ability to interact with amino acids 278–527 of fission yeast Pol1 fused to the DNA binding domain of the bacterial transcription factor LexA (see Materials and methods for details). Amino acids 278–527 correspond to the smallest region of budding yeast Pol1 shown to interact with Pol32 [ 16 ]. Reporter gene induction, measured by β-galactosidase activity assay, was observed in the presence of the Gal4 AD-Cdc27 and LexA-Pol1-278-527 (LexA-Pol1) proteins, but not when either protein alone was present (Figure 1 and Table 1A , upper part). Thus, as in budding yeast, the catalytic subunit of Pol α is able to interact with the C-subunit of Pol δ. Figure 1 Mapping of the Pol1 interaction site on Cdc27 using the two-hybrid system. Full-length and thirteen truncated Cdc27 proteins, expressed as Gal4 transcription activation domain fusions, were tested for their ability to interact with Cdc1-LexA, Pcn1-LexA and Pol1(278–512)-LexA baits in S. cerevisiae CTY10-5d. Interactions were monitored by β-galactosidase activity, as described in Material and methods, and are indicated as +++ (strong interaction, typically 80–100 Miller units of β-galactosidase activity), + (weak interaction, <20 Miller units), – (no interaction detectable above background, <1 Miller unit). The area between the broken vertical lines indicates the minimal Pol1 binding region, corresponding to amino acids 293–332 of Cdc27. The grey box indicates the minimal Cdc1 binding domain, amino acids 1–160; the black box represents the Pcn1 (PCNA) binding motif (amino acids 362–372). Table 1 Two-hybrid analysis. Prey and bait proteins were expressed as Gal4 AD and LexA BD fusions from pACT2 and pBTM116 respectively. Positive interactions corresponded to 50–100 Miller units (++) or 10–20 Miller units (+) of β-galactosidase activity. Negative interactions corresponded to < 2 units. See text for details. Prey Bait Bridge Interaction A. Fission yeast proteins Cdc27 - - - - Pol1(278–527) - - Cdc27 Pol1(278–527) - ++ Pcn1 Pol1(278–527) - - Pcn1 - Cdc27 - - Pol1(278–527) Cdc27 - Pcn1 Pol1(278–527) Cdc27 + B. Human proteins - Pol α (291–540) - - p66 (253–466) - - - p66 (356–466) - - - p66 (253–466) Pol α (291–540) - ++ p66 (356–466) Pol α (291–540) - ++ C. Cross-species interactions - Pol1(278–527) - - p66 (253–466) Pol1(278–527) - ++ p66 (356–466) Pol1(278–527) - ++ Cdc27 Pol α (291–540) - - D. Truncated Pol1 proteins Cdc27 Pol1(278–527) - ++ Cdc27 Pol1(328–527) - - Cdc27 Pol1(278–477) - - Cdc27 Pol1(278–487) - - Cdc27 Pol1(278–497) - - Cdc27 Pol1(278–507) - - Cdc27 Pol1(328–477) - - Cdc27 Pol1-TS13(278–527) - - Mapping the Pol1 binding site on Cdc27 Previously, minimal Cdc1 and PCNA binding regions on Cdc27 have been mapped [ 11 , 12 , 18 ]. Cdc1 binds within the globular 160 amino acid N-terminal domain of the 372 amino acid Cdc27 protein, whereas PCNA binds at the extreme C-terminus of Cdc27 at a conserved p21 Cip1 -like PCNA binding sequence, the PIP box. To map the Pol1 binding site, a series of thirteen truncated Cdc27 proteins fused to the Gal4 AD were tested against LexA-Pol1 in the two-hybrid system. The results of this analysis are shown in Figure 1 . Pol1 binds within the extended C-terminal domain of Cdc27, with the smallest construct capable of binding spanning a forty amino acid region, from amino acids 293–332. Therefore, the Pol1 binding region is distinct from both the globular N-terminal domain and C-terminal PIP box. In support of this, two-hybrid analysis showed that Cdc27 could bind to both Pol1 and Pcn1 (PCNA) proteins simultaneously: Gal4 AD-Pcn1 and LexA-Pol1 fusion proteins (which do not interact in the two-hybrid system) could be brought together by co-expression of Cdc27 (Table 1A , lower part). Cdc27-Pol1 interactions with recombinant proteins In order to test whether the interaction between Pol1 and Cdc27 was a direct one, purified recombinant Cdc27 and Pol1 proteins were assayed for their ability to interact in vitro . Purified GST-Cdc27-273-352 fusion protein [ 11 ] was tested for its ability to pull-down purified recombinant hexahistidine-tagged Pol1 (amino acids 278–527). The results (Figure 2 ) mirror those seen with the two-hybrid assays and provide the first evidence that the interaction between the Cdc27 and Pol1 proteins is a direct one, rather than being mediated via a third protein or set of proteins, such as one of the other subunits of the Pol α-primase or Pol δ complexes. Figure 2 Direct interaction between recombinant Pol1 (278–512) and Cdc27 proteins. A. Expression of H6-Pol1 (278–512) in E. coli detected by immunoblotting using anti-MRGS antibodies. Total protein extracts prepared from cells carrying the control plasmid pQE32-Pol1 (278–512)-REV (lanes 1, 2), in which the relevant pol1 + sequence is present in the wrong orientation, or pQE32-Pol1 (278–512) (lanes 3, 4), either before (lanes 1, 3) or after (lanes 2, 4) induction of protein expression by addition of IPTG. The position of the 25 kDa marker is shown. B. Purified GST and GST – Cdc27 (273–352) proteins detected by Coomassie staining following SDS-PAGE. C. Binding assays using GST – Cdc27 fusion proteins. Bound H6-Pol1 was detected by Western blotting as in part A. Lanes 1, 3, 5: binding of H6-Pol1 to GST. Lanes 2, 4, 6: binding of binding of H6-Pol1 to GST – Cdc27 (273–352). In each case the bound fraction corresponds to approximately 15–20% of the input. Lanes 1 and 2: assay performed in PBS containing 0.1% Triton X100; lanes 3 and 4: PBS containing 0.25% Triton X100; lanes 5 and 6: PBS containing 0.5% Triton X100. In the absence of detergent Pol1 binds equally well to both GST and GST-Cdc27 (273–352) proteins. Pol1 binding site sequence Previously, Cdc27 homologues from fission yeast [ 11 , 18 ], budding yeast [ 19 , 20 ], human, mouse [ 21 ], and Xenopus [ 15 ] have been identified and characterised. To extend this family, BLAST [ 22 ] and ψ-BLAST searches [ 23 , 24 ] were performed to identify putative Cdc27 family members from other eukaryotic species. In all, 24 protein sequences were identified, from organisms as diverse as vertebrates, worms, fungi and plants. Protein sequence alignments (Figure 3 ) of the C-terminal regions of these proteins identified a set of highly conserved amino acids that form a putative D NA p olymerase i nteraction m otif (or DPIM) with the consensus sequence D(D/E)-G -- (V/I)(T/S). The sequences flanking the DPIM are generally highly charged in character and some sequence conservation (particularly of charged amino acids) is apparent in these regions. Mutagenesis and peptide binding studies (described below) suggest that, in addition to the DPIM, some of these sequences may also play a role in binding Pol1 (see Discussion). Figure 3 Sequence alignment of Pol1 interacting region from fission yeast Cdc27 and homologues in other eukaryotic species. Conserved residues are boxed. Abbreviations and NCBI sequence accession numbers: Sp ( S. pombe , P30261), Sc ( S. cerevisiae , P47110), Nc ( Neurospora crassa , XP_328704), Ca ( Candida albicans , EAK99562), Gz ( Gibberella zeae , XP_384433), Um ( Ustilago maydis , XP_401381), Cn ( Cryptococcus neoformans , EAL21672), Dh ( Debaryomyces hansenii , CAG87841), Dm ( Drosophila melanogaster , AAD38629), Ss ( Sus scrofa , BF078337), Mm ( Mus musculus , Q9EQ28), Rn ( Rattus norvegicus , XP_215011), Cf ( Canis familiaris , CF411342), Hs ( Homo sapiens , Q15054), Ci ( Ciona intestinalis , AK114729), Fr ( Fugu rubripes , protein sequence derived by translation of clone M001240 at ), Tn ( Tetraodon nigroviridis , CAF97746), Dr ( Danio rerio , AAH76031), Xl ( Xenopus laevis , BAC82197), Att ( Ambystoma tigrinum tigrinum , CN059104), Gg ( Gallus gallus , BU121824 – note the additional glutamate residue within the DPIM), Ce ( C. elegans , Q21730), Os ( Oryza sativa , NP_913217), and At ( Arabadopsis thaliana , C96815). Note that in only a few cases (Sp, Sc, Hs, Mm, Xl) have the identities of these putative Cdc27 proteins been confirmed via purification and characterisation of Pol δ. However, all the sequences shown possess a canonical PCNA binding motif at or near their C-terminal ends (Q -- I -- FF), in common with the bone fide Cdc27 proteins. Interaction between human Pol1 and Cdc27 orthologues To examine whether the Pol1-Cdc27 (Pol1-Pol32) interaction observed in the yeasts was also conserved in higher eukaryotes, the catalytic subunit of human Pol α and the human Cdc27 orthologue p66/KIA00039 were assayed for interaction using the two-hybrid system. Amino acids 291–540 of the human Pol α catalytic subunit, corresponding to the minimum Cdc27 binding region (amino acids 278–527) in fission yeast Pol1, were expressed as a LexA fusion alongside Gal4 activation domain fusions of either the entire C-terminal domain of human p66 (amino acids 253–466) or the C-terminal 111 amino acids only (356–466). Interactions were tested by β-galactosidase assay. Both p66 constructs bound to Pol α (Table 1B ), indicating that the DNA polymerase interaction is a conserved feature. In addition, it was observed that human p66 was able to interact with fission yeast Pol1 (Table 1C ), suggesting that conserved sequences (or structural features) within the Cdc27/p66 C-terminal region, such as the DPIM, were important for the interaction. Mutational analysis of Pol1 binding site on Cdc27 Next, the importance for Pol1 binding of the conserved amino acids in the DPIM was examined. Eight mutant Cdc27 proteins (Cdc27-P1 through Cdc27-P7, and Cdc27-Q1, see Figure 4 ) were expressed as Gal4-Cdc27-273-352 fusion proteins and tested for their ability to bind to LexA-Pol1 (278–527) in the two-hybrid system. In each mutant protein one or more conserved amino acids is replaced with alanine. For example, the Cdc27-P1 mutant sees the conserved triplet DEE (residues 309–311) substituted with AAA. The results of this analysis are shown in Figure 4 . Mutating the conserved amino acids of the DPIM (mutants Cdc27-P1, P2 and Q1) completely abolished Pol1 binding in vivo by Cdc27. Similar reductions were seen with the Cdc27-P4 and Cdc27-P6 mutants, where the mutated residues are located N-terminal and C-terminal to the DPIM respectively. The conserved amino acids of the DPIM are therefore essential for Pol1 binding by Cdc27, although sequences flanking the conserved motif also play a role. Only three of the eight mutant proteins (Cdc27-P3, P5 and P7) were able to interact with LexA-Pol1 (278–527), though the strength of the interaction was reduced to 30–40% of the wild-type value for Cdc27-P3 and Cdc27-P5 (both of which affect residues N-terminal to the conserved motif), and to ~ 10% of wild-type for Cdc27-P7 (located C-terminal to the conserved motif). Immunoblotting showed that all the mutant proteins were present in yeast protein extracts at the same level as the wild-type Gal4-Cdc27 fusion protein (data not shown). Figure 4 Mutational analysis of Pol1 binding domain on Cdc27. A. Summary schematic of minimal Pol1 binding region on Cdc27, showing conserved amino acids (boxed) and the eight mutant alleles P1 through P7 and Q1. Each mutant sees two or three adjacent residues being replaced with the same number of alanines. In the case of P1 and P3 through P7, adjacent basic (P3, P4 and P5) or acidic (P1, P6 and P7) amino acids are mutated. B. Quantitation of β-galactosidase activity in liquid cultures. Key: C (pACT2 vector); W (pACT-Cdc27-273-352), mutants P1 – P7, plus Q1 (mutated forms of pACT-Cdc27-273-352); F (pACT-Cdc27, i.e. full-length Cdc27 fused to Gal4). The observation that sequences flanking the DPIM play a role in Pol1 binding was supported by the results of studies performed with a nested set of overlapping 20 mer peptides derived from the S. pombe Cdc27 sequence (see Figure 5A ). The peptides were tested for their ability to bind to an epitope-tagged form of the Pol1 protein (Pol1-13Myc, see Materials and methods) in fission yeast protein extracts. As can be seen in Figure 5B , Pol1-13myc is precipitated by peptides SpB, SpC and SpD, but not SpA and SpE. Peptide SpC spans the groups of basic amino acids N-terminal to the conserved motif, including those shown to be required for Pol1 binding as defined by the Cdc27-P3 and -P5 mutants above, but does not include the conserved sequence DEEGFLVT indicating that, in this in vitro situation, the conserved amino acids are not absolutely required for Pol1 binding, in contrast to what is observed in vivo . Similar results were observed with peptides derived for the human p66 protein sequence (Figure 5C , 5D ), again illustrating the potential for cross-species interaction between human p66 and fission yeast Pol1 (see Table 1C ). Figure 5 Peptide binding studies. A. Schematic of the structure of the Cdc27 protein (upper part) with the sequences of overlapping peptides SpA-SpE shown beneath. B. Immunoblot using anti-myc mAb showing pull-down of Pol1-13Myc protein from fission yeast protein extracts by Sp series peptides. C. Sequences of human peptide set HsA-HsE corresponding exactly to SpA-SpE above. D. Immunoblot using anti-myc mAb showing pull-down of Pol1-13Myc protein from fission yeast protein extracts by Hs series peptides. Mapping the Cdc27 binding site on Pol1 In an effort to map more precisely where within the Pol1 protein Cdc27 binds, several truncated forms of Pol1 (278–527) were constructed and tested as LexA fusions in the two-hybrid system (see Materials and methods for details). Removal of fifty amino acids from the N-terminus, or twenty amino acids from the C-terminus, of the Pol1 (278–527) protein was found to abolish the interaction with Cdc27 altogether (Table 1D ). Previously, the isolation of a temperature-sensitive mutant pol1 allele, pol1-ts13 , had been reported [ 25 ]. Sequence analysis revealed that this allele differed from the wild-type pol1 + by deletion of 9 bp from the ORF, resulting in loss of three amino acids (L470, S471, R472) from within the minimal Cdc27 binding domain defined above. In this study, the ability of a Pol1-TS13(278–527) bait to bind to Cdc27 was tested using the two-hybrid system. No interaction could be detected at a range of growth temperatures (Table 1D , and data not shown), indicating that amino acids 470–472 are required for Cdc27 binding by Pol1(278–527). To test whether overproduction of Cdc27 might rescue the temperature-sensitive phenotype of pol1-ts13 , these cells were transformed with plasmid pREP3X-Cdc27 [ 18 ] and transformants plated at restrictive and semi-restrictive temperatures on thiamine-free medium, to induce high-level expression of cdc27 + from the thiamine-repressible nmt promoter. No suppression of the pol1-ts13 phenotype was observed, however (data not shown). Indeed, no suppression was observed of any of three pol1 alleles that were analysed in this way, the others being pol1-1 [ 26 ] and pol1-H4 [ 27 ]. Expression of DPIM mutants in vivo To assay the in vivo role of the DPIM in Cdc27, four of the eight DPIM mutant alleles ( cdc27-P1 through cdc27-P4 ) were cloned into plasmid pREP3X, 3' to the repressible nmt1 promoter [ 28 , 29 ], and transformed into a cdc27 + / cdc27::his7 + diploid strain. Transformant colonies were then induced to sporulate and the spores plated on media containing thiamine, to repress the nmt1 promoter. Under these conditions, residual low level expression from the nmt1 promoter ensures that the level of Cdc27 protein present in the cell is comparable to that seen in wild-type cells [ 12 ]. Analysis of the meiotic products showed that the mutant Cdc27 proteins were able to support growth of cdc27Δ haploid cells; indeed, no phenotypic defects were apparent (data not shown). To confirm this, a fission yeast strain was constructed in which the endogenous cdc27 + gene was precisely replaced with the cdc27-Q1 mutant allele. In the Cdc27-Q1 mutant protein, the central five amino acids of the DPIM are replaced with alanine, resulting in loss of Cdc27-Pol1 interaction in the two-hybrid system (Figure 4 ). Construction of the cdc27-Q1 strain was achieved by first replacing one copy of cdc27 + with ura4 + in a diploid strain, before then replacing the cdc27::ura4 + allele with cdc27-Q1 via 5-FOA counterselection in the cdc27::ura4 + haploid carrying cdc27-Q1 on a plasmid (see Materials and methods for details). PCR analysis of genomic DNA using primers specific for wild-type cdc27 + and cdc27-Q1 sequences (see Figure 6 and legend) allowed the identification of haploid strains in which cdc27-Q1 was correctly integrated at the endogenous cdc27 + locus, and also allowed the cdc27-Q1 mutant to be conveniently followed through genetic crosses. Figure 6 Construction and analysis of cdc27-Q1 mutant yeast. A. Upper part: Schematic of the cdc27 + gene region (3.1 kb HindIII-BamHI region) showing location of oligonucleotides used for PCR amplification. (Key to oligonucleotides: A w = CDC27-Q1W-DIAG2, A m = CDC27-Q1M-DIAG2, B = CDC27-B, C = CDC27-SEQ2005, D = CDC27-H, X = CDC1-AB, and Y = CDC1-XY – see Material and methods for sequences). The boxes indicate approximate positions of cdc27 + exons. The NotI site shown is found in the cdc27-Q1 allele only. Lower part: Genomic DNA prepared from wild-type (WT) and cdc27-Q1 (Q1) strains was amplified using the primer pairs shown. The D+B PCR product from cdc27-Q1 alone can be digested with NotI (data not shown). The primer pair X and Y amplify an unrelated region of genome, and were included as a control. Molecular weight markers (kb) are shown to the right of the gel. B. Wild-type ( cdc27 + , left) and cdc27-Q1 (right) cells plated on YE medium and incubated for 3 days at 32°C. Mutation of the DPIM did not affect the efficiency of colony formation or growth rate. See text for details. Phenotypic analysis of cdc27-Q1 cells, in comparison to an otherwise isogenic cdc27 + control, revealed the following: cells carrying the cdc27-Q1 allele grew normally at a range of temperatures (18 – 36.5°C), with a generation time indistinguishable from wild-type (110 minutes at 32°C in YE medium). At 32°C, cdc27-Q1 cells underwent division at ~ 14.1 μm (compared to wild-type at ~ 14.4 μm). Further analysis showed that the cdc27-Q1 cells were indistinguishable from wild-type in all respects examined, including responses to the DNA replication inhibitor hydroxyurea (HU) and the DNA damaging agents methylmethane sulphonate (MMS), camptothecin (CPT), bleomycin sulphate (BMS) and UV light (see Materials and methods for details). The kinetics of cell division arrest in response to HU were also examined, but again no difference was detectable between cdc27-Q1 and wild-type strains (data not shown). That cell number increase is arrested in cdc27-Q1 cultures following treatment with HU, and that both wild-type and cdc27-Q1 cells become highly elongated under these conditions, is indicative of the presence of a functional DNA replication checkpoint in these cells. Taken together, these results strongly suggest that the Pol1-Cdc27 interaction does not play an essential role either during S-phase or in various DNA repair pathways in fission yeast. The consequences of introducing additional mutations into the cdc27-Q1 background were also investigated. cdc27-Q1 was crossed to strains carrying mutations in the other three subunits of Pol δ ( pol3-ts3 , cdc1-P13 , cdm1Δ ) and in the helicase-endonuclease Dna2 and its associated protein Cdc24 ( dna2-C2 and cdc24-M38 ). [ 13 , 18 , 30 - 33 ]. The double mutant strains were then analysed as described in Materials and methods. In every case examined, the properties of the double mutant with cdc27-Q1 were indistinguishable from the single mutant. The cdc27-Q1 mutation was also combined with rad3Δ [ 34 ] and cds1Δ [ 27 ] alleles, to create cdc27-Q1 rad3Δ and cdc27-Q1 cds1Δ double mutants, both of which were viable. The Rad3 and Cds1 proteins are key components of various DNA structure checkpoints in fission yeast [ 35 ]. That the cdc27-Q1 double mutants were viable indicates that cdc27-Q1 cells do not require the presence of a functional checkpoint for viability. When wild-type cells are treated with hydroxyurea, activation of Cds1 results in cell cycle arrest and replication fork stabilisation. In the absence of Cds1, however, replication forks are believed to collapse, resulting in loss of viability [ 35 ]. The rad3Δ and cds1Δ double mutants were tested for their sensitivity to HU and CPT, but as before, no differences were observed between single and double mutants with cdc27-Q1 (data not shown). Discussion In fission yeast, DNA polymerase δ is a multisubunit complex comprising a large catalytic subunit that is required for chromosomal replication as well as three smaller subunits, two of which are also essential for cell viability [ 13 , 18 , 36 ]. Understanding how the four subunits of the complex interact with one another, and how they interact with other components of the replication machinery, is an important goal. Once interactions are identified and the sites of interaction precisely mapped, reverse genetic analysis allows determination of the effects of disrupting individual protein-protein interactions on replisome function. In this paper we show that the C-subunit of fission yeast Pol δ, Cdc27, is able to interact both in vivo and in vitro with Pol1, the catalytic subunit of Pol α. The Pol1 binding site on Cdc27 has been mapped by in vivo and in vitro approaches and a region of 40 amino acids, from amino acids 293 – 332, has been shown to be sufficient for binding. Protein sequence alignments of Cdc27 homologues across this 40 amino acid region identify a short protein sequence motif (D -- G --VT) that is highly conserved and which is essential for Pol1 binding. This DNA polymerase interaction motif (DPIM) is flanked by relatively highly charged sequences. Ten basic amino acids are found flanking the DPIM in the fission yeast Cdc27 protein sequence, nine of which are located N-terminal to the central conserved motif. Similarly, there are ten acidic amino acids, all of which lie either within the conserved DPIM or C-terminal to it. Our mutagenesis data clearly implicates several of these charged groups in the binding to Pol1 (Figure 4B ). Secondary structure predictions suggest that the conserved DPIM is likely to form a turn or loop, raising the possibility that the sequences on either side of the conserved motif interact with one another. The distribution of positively and negatively charged amino acids in most, though not all, of the Cdc27 homologues in other species (Figure 3 ) is similar to that in fission yeast. The most obvious exception to this is in the C. albicans Cdc27 protein, where six acidic amino acids are found N-terminal, and seven basic amino acids C-terminal, to the DPIM. This organisation of charged residues appears again to be compatible with the notion the sequences on either side of the DPIM may interact with one another. The Cdc27 protein has an elongated shape, with a frictional ratio of 1.85 [ 12 ]. The same is true of its budding yeast orthologue Pol32 which has a frictional ratio of 2.22 [ 37 ]. The elongated shape of Cdc27 is due to the C-terminal region of the protein. The N-terminal Cdc1 binding domain, comprising amino acids 1–160 [ 11 ], behaves in solution as a globular protein [ 12 ]. The protein-protein interaction motif described in this study is the second to be mapped to the extended C-terminal domain. Previously we showed that Cdc27 interacts with PCNA via a conserved sequence at the extreme C-terminus of the protein [ 11 ]. Our results indicated that this interaction was essential for cell viability. Consistent with this, Cdc27-PCNA contact is vital for maximal polymerase processivity in vitro . However, recently we obtained evidence that the Cdc27-PCNA interaction is a non-essential one (H. Tanaka, G. Ryu, Y.-S. Seo and S.M., submitted). Recently, the results of a deletion analysis of Pol32, the budding yeast orthologue of Cdc27, were reported [ 17 ], including analysis of Pol32-Pol1 interaction. The results of these studies showed that amino acids 270–309 of Pol32 were required for Pol1 binding in the two-hybrid system. This region corresponds to amino acids 286–325 in fission yeast Cdc27 and includes the conserved DPIM sequence (Figure 3 ). Deletion of amino acids 250–289 (266–305 in Cdc27) greatly reduced Pol1 binding, consistent with the results reported here with the Cdc27-P3, -P4 and -P5 mutants (Figure 4 ), as did deleting amino acids 310–343 (326–363), consistent with the Cdc27-P7 mutant result. (As expected, the latter Pol32 deletion also disrupted binding to PCNA.) In addition, it was shown that that Pol α-primase and Pol δ could be co-immunoprecipitated following mixing of the purified complexes, but that this interaction was effectively abolished when Pol δ contained the 270–309 deletion of Pol32 rather than the full-length protein [ 17 ]. In this paper, cells expressing the DPIM mutant protein Cdc27-Q1 were shown to be no more sensitive then wild-type to HU, MMS, CPT, UV and BMS (see Results), while analysis of budding yeast cells expressing Pol32 lacking the DPIM (270–309 deletion) showed them to be no more sensitive than wild-type to HU and UV, and to show normal rates of mutagenesis following UV exposure [ 17 ]. Indeed, in all situations examined, S. pombe cdc27-Q1 cells were indistinguishable from wild-type. No genetic interactions were observed between cdc27-Q1 and various other DNA replication or checkpoint mutants, including the key checkpoint kinase Rad3. In conclusion, while the interaction between Pol α-primase and Pol δ mediated via Cdc27 could play an important role in coordinating the events of lagging strand synthesis, we have yet to obtain any evidence that this is the case. This raises two possibilities. First, that the observed interaction does not play an important role in chromosomal replication. We believe that this is unlikely to be the case, given the high degree of conservation of the DPIM sequence across evolution, in a region of the Cdc27 protein that is very poorly conserved at the primary sequence level. The second possibility is that there are multiple redundant interactions within the lagging strand machinery. In this case, an important role for the Cdc27-Pol1 interaction might only be uncovered when another protein-protein interaction is perturbed, in which case, the cdc27-Q1 mutation might be expected to be synthetically lethal or sick with a mutant that disrupted the overlapping redundant function. With the genetic tools available in yeast, this is a hypothesis that is readily testable. Conclusions In fission yeast, interaction between Pol α and Pol δ is mediated, at least in part, by direct binding of the Pol δ C-subunit Cdc27 to the Pol α catalytic subunit Pol1, and requires the presence of a short sequence motif (DPIM) in the C-terminal region of Cdc27. The DPIM is conserved in all known Cdc27 orthologues. Despite this, it has not been possible to identify any phenotypic consequences associated with deletion of the DPIM sequence, raising the possibility that the observed interaction does not play a crucial role in vivo . Methods Yeast strains, media and methods All fission yeast strains were as described previously, except for pol1-H4 [ 27 ] and pol1-1 [ 26 ], which were obtained from Dr J. Hayles (CR-UK, London, U.K.), and pol1-ts13 [ 25 ], which was obtained from Dr T.S.F. Wang (Stanford, U.S.A.). Note that this allele was originally designated polα-ts13 but is correctly renamed here to bring it in line with accepted nomenclature standards (see for further details). S. pombe media and techniques were essentially as described [ 38 ], with the following exceptions. Routine transformation of S. pombe was carried out by electroporation [ 39 ], whereas transformation for PCR-based gene targeting was accomplished using a modified lithium acetate method [ 40 ]. For two-hybrid analysis. S. cerevisiae CTY10-5d ( MATa ade2 met- trp1-901 leu2-3-112 his3-Δ200 gal4-gal80-URA3::lexA-LacZ ) was used [ 18 ]. S. cerevisiae was cultured in YPDA and SD medium. Plasmids for two-hybrid assay Two-hybrid interactions were monitored using the Gal4 transcription activation domain (prey) plasmid pACT2 (Clontech), the LexA DNA binding domain (bait) plasmid pBTM116, and S. cerevisiae lexA op-lacZ strain CTY10-5d, as described previously [ 18 ]. The Pol1 bait plasmid pBTM116-Pol1-(278–527) was constructed by amplifying sequences encoding amino acids 278–527 from plasmid pTZ19R-Pol1 (prepared by subcloning a 5925 bp SmaI-PstI fragment encompassing the entire pol1 + gene from S. pombe cosmid SPAC3H5, see , into plasmid pTZ19R) using oligonucleotides POL15 (5'-GTGTGGTTTG GGATCC CCCTATCACCAATGACACCTTTA-3') and POL13-2 (5'-GTGTGGTTTG GGATCC TACATCACCGTCATTGGAGGCGT-3'), restricting the PCR product with BamHI (sites underlined), cloning to pTZ19R (Fermentas) and sequencing to confirm the absence of errors. The 763 bp BamHI fragment was then transferred to pBTM116, to generate pBTM116-Pol1(278–527). Plasmid pBTM116-Pol1-TS13(278–527) was produced in a similar manner, except that the starting PCR template was genomic DNA prepared from pol1-ts13 cells [ 25 ], the final product thereby containing a 9 bp deletion in the pol1 + ORF, resulting in the loss of coding capacity for amino acids L470, S471 and R472. Plasmids expressing C-terminally truncated Pol1(278–527) proteins were constructed by PCR amplification using oligo POL15 in conjunction with POL25-3 (5'-GTGTGGTTTG GGATCC TAAGGACCCATAACTCTTCTACT-3'), POL13-487 (5'-GTGTGGTTTG GGATCC TAAAAATTTGGTTGTTGTATTTT-3'), POL1-497 (5'-GTGTGGTTTG GGATCC TACCGGCACCAACTAGCATTTTT-3') or POL1-507 (5'-GTGTGGTTTG GGATCCT AGTTCTGAGGTGACGAACATCC-3'), cloning directly to pBTM116 following BamHI cleavage of the PCR product, and sequencing. The resulting constructs encoded the following proteins as Lex A fusions: Pol1(278–477), Pol1(278–487), Pol1(278–497) and Pol1(278–507). A construct expressing an N-terminally truncated Pol1(278-527) protein, designated Pol1(328–527), was generated by amplification using POL15-2 (5'-GTGTGGTTTG GGATCC CCGGCTCATTGTGTCTATTTGGC-3') with POL13-2 (above). Amplification with POL15-2 and POL25-3 generated a construct with the potential to express a LexA-Pol1 protein truncated at both ends: LexA-Pol1(328–477). Sequences encoding the C-terminal region from human p66/KIAA0039 were amplified by PCR from plasmid pET19b-p66 (a gift of Dr P. Hughes, Villejuif, France) using either oligo P66-51 (5'-GTGTGGTTTG GGATCC CCTCAGAACAAGCAGTGAAAGAA-3') or P66-52 (5'-GTGTGGTTTG GGATCC CGTCTCCACCTCTTGAACCAGTG-3') with P66-3 (5'-GTGTGGTTTGGGATCCTTGGTCTTCACCCTTGACCACTC-3'). The PCR products were then restricted with BamHI, cloned to pACT2 and sequenced. The resulting plasmids encode, as Gal4 AD fusions, either the entire C-terminal domain of p66/KIAA0039 (amino acids 253–466, oligos P66-51 and P66-3) or a shorter region (amino acids 356–466, oligos P66-51 and P66-3). Sequences encoding the interacting domain of the catalytic domain of human Pol α were amplified from plasmid pBR322-Pol α (a gift of Dr T. Wang, Stanford, USA) using oligos HPOL1-5 (5'-GTGTGGTTTG GGATCC CCAAAGGGACCGTGTCCTACTTA-3') and HPOL1-3 (5'-GTGTGGTTTG GGATCC TACATCACGACAAGCGGTGGTGG-3'), restricted with BamHI, cloned to pBTM116 and sequenced. The resulting plasmid encodes amino acids 291–540 of the human protein as a LexA fusion. The Cdc27 prey plasmids were constructed in pACT2 (Clontech) and have been previously described, with the exception of mutants Cdc27-273-352-P1 through -P7, and Cdc27-273-352-Q1. The first four of these (P1-P4) were generated by PCR amplification from the pREP3X-Cdc27-P1 to -Cdc27-P4 plasmids described below. Oligonucleotides for PCR, with BamHI sites underlined: PMUT5 (5'-GTTTGTTGGT GGATCC CCACCGAAGCAAAATCTGCTGCA-3') and PMUT3 (5'-GTTGTGGGTG GGATCC TACTTCTTAGTTGCAATGTTTAC-3'). Mutant Q1 was generated by PCR amplification with same primers but using pHBLA-Cdc27-Q1 (below) as template. Plasmids expressing mutants P5-P7 were generated by overlap extension PCR using PMUT5 and PMUT3 together with the following mutagenic oligonucleotides (shown with mutated sequence underlined in top strand oligo): CDC27-P5F (5'-AAAGAAAAGTT GCAGCG TACGCGACAACGAAAG-3'), CDC27-P5R (5'-CTTTCGTTGTCGCGTACGCTGCAACTTTTCTTT-3'), CDC27-P6F (5'-TTGGTTACTAAG GCAGCAGCA GTCTGGGAATCA-3'), CDC27-P6R (5'-TGATTCCCAGACTGCTGCTGCCTTAGTAACCAA-3'), CDC27-P7F (5'-GAATCATTTTCT GCAGCTGCA AACATCTCAACT-3'), CDC27-P7R (5'-AGTTGAGATGTTTGCAGCTGCAGAAAATGATTC-3'). Two-hybrid assays Quantitative data for β-galactosidase activity was obtained as described previously [ 18 ] and is expressed in Miller units [ 41 ]. To analyse prey protein levels, yeast total protein extracts, prepared from the same cultures used for β-galactosidase assay, were subjected to SDS-PAGE and immunoblotted using antibodies against the HA epitope (12CA5, Roche Applied Science) present in proteins expressed from pACT2. Expression and purification of recombinant H6-Pol1 (278–527) The Pol1 (278–527) domain was expressed in E. coli with an N-terminal MRGSH 6 -tag to facilitate purification and detection. To achieve this, the 763 bp Pol1 (278–527)-encoding BamHI fragment described above was cloned into pQE32 (Qiagen). The resulting plasmid, pQE32-Pol1 (278–527), was transformed into E. coli M15 (pREP4) and recombinant protein expression induced in 500 ml cultures (OD 600 nm of 0.6) by addition of IPTG to a final concentration of 1 mM. Following incubation for 4 hours at 37°C, the cells were pelleted and the pellet resuspended in 40 ml of ice-cold buffer A (150 mM NaCl, 50 mM Tris-HCl pH 8.0, 20 mM imidazole) containing EDTA-free Complete™ inhibitors (Roche Applied Science) and 1 mM PMSF. The cells were lysed by sonication, then centrifuged at 25000 g for 15 minute at 4°C. The soluble supernatant was added to 4 ml of 50% (v/v) Ni-NTA agarose (Qiagen) in buffer A, mixed on a wheel at 4°C for one hour, then packed into a disposable chromatography column at 4°C. The column was drained of the flow-through and subsequently washed with 100 ml of buffer B (1 M NaCl, 50 mM Tris-HCl pH 8.0, 20 mM imidazole) containing Complete™ inhibitors and PMSF, then 100 ml of buffer A containing Complete™ inhibitors and PMSF, before the bound H6-Pol1 protein was eluted using 4 ml of buffer C (250 mM imidazole, 50 mM Tris-HCl pH 8.0). Following elution, samples were analysed by SDS-PAGE and the peak fractions pooled and dialysed overnight in PBS at 4°C. Protein concentration was determined by BCA assay (Pierce) versus BSA standards. Binding assays GST and GST-Cdc27-273-352 fusion proteins were prepared as described previously [ 11 ] from plasmid pGEX6P-1B, a modified version of pGEX6P-1 (Amersham Pharmacia) in which the reading frame of the polylinker is altered (S.M., unpublished). Sequences encoding wild-type and mutant forms of Cdc27-273-352 were subcloned into this vector from the equivalent pACT2 two-hybrid constructs described above. Binding assays were performed by mixing ~ 30 pmol of GST or GST-Cdc27-273-352 with 90 pmol of H6-Pol1 in PBS containing 1.0% Triton X100 for 1 hour at 4°C on a rotating wheel. The GST proteins were then precipitated using Glutathione Sepharose™ 4 Fast Flow resin (Amersham Biosciences) for 30 minutes at 4°C on a rotating wheel. Following extensive washing with PBS containing 0.1 – 1.0% Triton X100, SDS-PAGE sample buffer was added and the samples heated at 95°C for 5 minutes. Bound H6-Pol1 was visualised following electrophoresis of 15% SDS-PAGE gels by PAGE Blue G90 (Fluka) staining or by immunoblotting with anti-MRGS antibodies (Qiagen). Construction of Pol1-13Myc strain PCR-based gene targeting was used to tag the chromosomal pol1 + locus in an otherwise wild-type leu1-32 ura4-D18 h - strain with sequences encoding thirteen copies of the 9E10 (c-myc) mAb epitope, such that the encoded Pol1 protein (termed Pol1-13Myc) carries the 9E10 epitopes at its C-terminus. Oligonucleotides for amplification from pFA6a-13Myc-kanMX6 [ 40 ] were as follows: POL1-13MYC-5 (5'-TGCCATCAACAAAAATATCTCTCGAATAATGAACAAAAATGCGCGTGAATTTGTAGATATGGGACTGATATTTTCATCG CGGATCCCCGGGTTAATTAA -3', with plasmid-specific sequence underlined) and POL1-13MYC-3 (5'-GGCAATTCCCAAGTCTTTGAAACAGGTATTCCCATCAACATTTCTTGTACTGCATGAGCAAATATCTGTTCGAGGTGTC GAATTCGAGCTCGTTTAAAC -3'). Following transformation of ~ 10 μg of PCR product, correct G418-resistant integrants were identified by PCR amplification of genomic DNA using primer FCGPOL1-5 (5'-ATGTCGTGGAAGCGTTCATT-3') within the pol1 + ORF and KAN269R (5'-GATCGCAGTGGTGAGTAACCATGCATCATC-3') within the kanMX6 cassette, and by Western blotting using the mouse anti-Myc mAb 9E10 (Roche). Peptide binding assays Peptides were synthesized by Mimotopes (Australia). All peptides were synthesized to contain the sequence biotin-SGSG at the N-terminus. Peptide sequences: SpA AAPDEPQEIIKSVSGGKRRG; SpB PQEIIKSVSGGKRRGKRKVK; SpC KSVSGGKRRGKRKVKKYATT; SpD GKRRGKRKVKKYATTKDEEG: SpE KRKVKKYATTKDEEGFLVTK; HsA PKTEPEPPSVKSSSGENKRK; HsB EPPSVKSSSGENKRKRKRVL; HsC KSSSGENKRKRKRVLKSKTY; HsD ENKRKRKRVLKSKTYLDGEG; HsE RKRVLKSKTYLDGEGCIVTE. For binding assays, the peptides were initially bound to 10 μl Streptavidin-agarose beads in PBS in a final volume of 100 μl at room temperature for 1 hour with gentle agitation. After binding, the beads were washed 3 times with 100 μl PBS. Protein extracts (100 μg of fission yeast extract) were added to the beads, and incubated at 4°C for 1 hr with gentle agitation. The beads were pelleted by centrifugation at 2000 rpm for 3 minutes, and subsequently washed extensively with NP40 buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1% (v/v) NP-40), prior to final resuspension with 20 μl of SDS loading buffer. Following SDS-PAGE, Pol1-13Myc was visualised by immunoblotting with the mouse anti-Myc mAb 9E10 (Roche). Fission yeast protein extracts were prepared from 8 × 10 7 cells in mid-exponential growth in EMM medium. Cells were harvested, washed once in STOP buffer (150 mM NaCl, 50 mM NaF, 10 mM EDTA, 1 mM NaN 3 ) and disrupted in buffer A (10 mM sodium phosphate buffer pH 7.0, 1% Triton, 0.1% SDS, 1 mM EDTA, 150 mM NaCl, 1 mM PMSF) supplemented with Complete™ protease inhibitors (Roche), using a bead beater (Hybaid Ribolyser). Protein concentrations were determined at 595 nm using the BioRad protein assay reagent according to the manufacturer's instructions, with BSA as standard. Plasmids expressing DPIM mutants Plasmids pREP3X-Cdc27-P1 to -Cdc27-P4 were constructed by cloning mutagenised cdc27 + cDNAs (SalI-BamHI) from pTZ19R-Cdc27-cDNA [ 18 ] into pREP3X. Mutagenesis of pTZ19R-Cdc27-cDNA was accomplished using the MutaGene II mutagenesis kit (BioRad) according to the manufacturer's instructions. Oligonucleotides with mutated residues underlined: CDC27-P1 (5'-AAAAGTACGCGACAACGAAA GCCGCAGCA GGATTCTTGGTTACTAAGG-3'), CDC27-P2 (5'-CGACAACGAAAGATGAAGAA GCCGCC TTGGTTACTAAGGAAGAAG-3'), CDC27-P3 (5'-TCAAATCCGTATCCGGTGGA GCCGCAGCA GGGAAAAGAAAAGTTAAAAAG-3'), CDC27-P4 (5'-CCGGTGGAAAGAGACGTGGG GCCGCAGCA GTTAAAAAGTACGCGACAAC-3'). The resulting mutant alleles were tested for function by transforming a cdc27 + /cdc27::his7 + leu1-32/leu1-32 ura4-D18/ura4-D18 his7-366/his7-366 ade6-M210/ade6-M216 h - / h + diploid [ 11 ], transferring the transformants onto malt extract medium to induce sporulation, before finally plating the helicase-treated spores onto EMM plates supplemented with 5 μg/ml thiamine with/without histidine, at 32°C. Construction of DPIM mutant strain A 3.1 kb HindIII-BamHI genomic DNA fragment carrying the cdc27 + gene was first cloned into pTZ19R to make pHB-Cdc27. This vector was then modified by addition of the S. cerevisiae LEU2 gene and S. pombe ars1 , to make pHBLA-Cdc27. The cdc27 + gene was then subjected to oligonucleotide-directed in vitro mutagenesis using the QuikChange method (Stratagene) with oligonucleotides CDC27-QC1 (5'-GTACGCGACAACGAAA GCTGCAgcggccgcC TTGGTTACTAAGGAAGAAG-3', with mutated sequence underlined and NotI site in lower case) and CDC27-QC2 (5'-CTTCTTCCTTAGTAACCAA GgcggccgcTGCAGC TTTCGTTGTCGCGTAC-3'), to create plasmid pHBLA-Cdc27-Q1. This was then was transformed into a cdc27 + / cdc27::ura4 + diploid, and cdc27::ura4 + (pHBLA-Cdc27-Q1) haploids obtained following sporulation and regrowth. These were then plated on YE plates containing 1 mg/ml 5-FOA. 5-FOA resistant colonies were identified, purified and characterised by PCR amplification of genomic DNA (prepared using the method of Bähler and coworkers [ 40 ]), to ensure loss of the chromosomal ura4 + marker, its replacement with cdc27-Q1 , and loss of the pHBLA-Cdc27-Q1. Oligonucleotides for diagnostic PCR shown in Figure 6 : CDC27-Q1W-DIAG2 (5'-GCGACAACGAAA GATGAAGAAGGATTC -3'; note that the underlined region anneals only to the wild-type cdc27 + and not cdc27-Q1 ); CDC27-Q1M-DIAG2 (5'-GCGACAACGAAA GCTGCAGgcggccgcC -3'; underlined region anneals only to cdc27-Q1 and not to cdc27 + ; NotI site in lower case); CDC27-H (5'-ACTGGTAGAATTGCGTTCGCGCTC-3'); CDC27-B (5'-TCTAGGATCAGAGTGAACTGATTG-3'); CDC27-SEQ2005 (5'-AGGTTGTACTAACATTAACAG-3'). The resulting strain, cdc27-Q1 leu1-32 ura4-D18 his7-366 ade6-M216 h - was then analysed alongside the wild-type leu1-32 ura4-D18 his7-366 ade6-M216 h - , as described below. Phenotypic analysis Cells were grown to mid-exponential phase (~ 5 × 10 6 cells/ml) in YE medium and ~ 2000 cells plated on YE medium supplemented with varying concentrations of hydroxyurea (2.5, 5, 7.5, 10, 12.5 mM), methylmethanesulphonate (0.01, 0.005, 0.0025, 0.001, 0.0005, 0.00025, 0.0001%), bleomycin sulphate (2.5, 5, 7.5, 10, 12.5 mU/ml), or camptothecin (4, 4.5, 5, 5, 5.5, 6, 6.5, 7, 7.5, 8 μM). Cells were also plated on plates containing sub-lethal doses of both HU and CPT, specifically 7.5 mM HU with either 5.5 or 6 μM CPT, or 10 mM HU with either 5.5 or 6 μM CPT. To analyse the effects of HU treatment in liquid culture, HU was added to EMM medium to a final concentration of 12 mM. Cell number per ml of culture was monitored using a Coulter Z1 electronic particle counterTo test sensitivity to UV, 1000 cells were plated on EMM plates, allowed to dry for 20 minutes, then irradiated using either a Stratalinker UV source (Stratagene) over the range 0 – 250 J/m 2 . Following UV treatment, plates were placed immediately in the dark to avoid photoreversal. For all treatments, the efficiency of colony formation was determined after 4 days growth at 32°C. Growth rate (YE medium, 32°C) was determined by cell counting using a particle counter. Cell length at cell division was determined using a graduated eyepiece. DPIM double mutants Double mutants were constructed by standard methods, with the cdc27-Q1 allele being identified by PCR analysis of genomic DNA using CDC27-Q1W-DIAG2 and CDC27-Q1M-DIAG2 oligonucleotides described above. Double mutants were created with the following: pol3-ts3 [ 30 ], cdc1-P13 [ 42 ], cdm1::ura4 + [ 13 ], cdc24-M38 [ 43 ], dna2-C2 [ 31 ], rad3::ura4 + [ 34 ] and cds1::ura4 + [ 27 ]. List of abbreviations used DPIM (DNA polymerase interaction motif); PCNA (proliferating cell nuclear antigen). Authors' contributions In Edinburgh, SM conceived of the study, performed some of the experimental work and prepared and revised the final manuscript, while FG performed the remainder of the experimental work. In Dundee, EW and JRGP designed and carried out the peptide binding studies. All four authors read and approved the manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC545490.xml |
517507 | Comparison of IgG4 assays using whole parasite extract and BmR1 recombinant antigen in determining antibody prevalence in brugian filariasis | Background Brugia malayi is endemic in several Asian countries with the highest prevalence in Indonesia. Determination of prevalence of lymphatic filariasis by serology has been performed by various investigators using different kinds of antigen (either soluble worm antigen preparations or recombinant antigens). This investigation compared the data obtained from IgG4 assays using two different kinds of antigen in a study on prevalence of antibodies to B. malayi . Methods Serum samples from a transmigrant population and life long residents previously tested with IgG4 assay using soluble worm antigen (SWA-ELISA), were retested with an IgG4 assay that employs BmR1 recombinant antigen (BmR1 dipstick [Brugia Rapid™]). The results obtained with the two antigens were compared, using Pearson chi-square and McNemar test. Results There were similarities and differences in the results obtained using the two kinds of antigen (SWA and BmR1). Similarities included the observation that assays using both antigens demonstrated an increasing prevalence of IgG4 antibodies in the transmigrant population with increasing exposure to the infection, and by six years living in the area, antibody prevalence was similar to that of life-long residents. With regards to differences, of significance is the demonstration of similar antibody prevalence in adults and children by BmR1 dipstick whereas by SWA-ELISA the antibody prevalence in adults was higher than in children. Conclusions Results and conclusions made from investigations of prevalence of anti-filarial IgG4 antibody in a population would be affected by the assay employed in the study. | Background Lymphatic filariasis affects approximately 120 million people worldwide. Ten percent of these infections are attributed to Brugia malayi and Brugia timori [ 1 ]. Thick blood smear examination is the routine parasitological method used for diagnosis and prevalence studies in the Brugia endemic countries of Malaysia and Indonesia [ 2 , 3 ]. This diagnostic method depends on the detection of microfilariae in the peripheral blood, and due to the nocturnal periodicity of microfilaremia in these areas, requires nighttime collection and survey, which is often unpopular with the local population. Furthermore, this method is relatively insensitive [ 4 ] and difficult to perform accurately and with consistency in field situations. Conversely, serological diagnostic methods exhibit better sensitivity than detection of microfilaria by thick blood smear, allow the detection of amicrofilaraemic infections among "endemic normals" and afford daytime finger-prick blood sampling (thus overcoming the inconveniences associated with night blood sampling, thereby encouraging greater cooperation with the local population and facilitate field work) [ 5 ]. However, reports on antigen detection test for brugian filariasis have not demonstrated high levels of sensitivity [ 6 , 7 ]. Thus in the absence of a good antigen detection test for Brugia infection, anti-filarial IgG4 assay may be the next best alternative for detection of brugian filariasis [ 8 ]. Anti-filarial IgG4 levels have been demonstrated to be elevated in active filarial infection [ 9 - 13 ] and decline post-treatment [ 14 - 17 ]. Detection of anti-filarial IgG4 antibodies has also been used for epidemiological assessment of filariasis [ 13 , 18 , 19 ]. Studies assessing antibody prevalence of lymphatic filariasis have employed assays that use either soluble worm antigens or recombinant antigens [ 12 , 13 , 18 - 21 ]. These antigens may not bind to the same set of anti-filarial antibodies and probably display different cross-reactivities to antibodies against other infections. Thus differences in the antigens employed may be expected to affect the results of antibody prevalence studies. Therefore, the present study aimed to make direct comparison of two antigens i.e. soluble adult worm antigen (SWA) and a recombinant antigen (BmR1), in IgG4 assays on the same set of serum samples. Previously, an ELISA employing soluble adult worm antigen (SWA-ELISA) had been performed to determine prevalence of anti-filarial IgG4 antibodies on a set of serum samples from Indonesia. These samples were collected from: 1. A transmigrant population that migrated from a non-filarial endemic region to an area endemic for Brugian filariasis and; 2. Life-long residents of the Brugian endemic area [ 22 ]. In the present study a rapid test based on B. malayi recombinant antigen (BmR1 dipstick [Brugia Rapid™]) and detection of IgG4 antibodies were evaluated using the same set of serum samples. The BmR1 dipstick test has previously been shown to be highly specific and sensitive for the detection of brugian filariasis. In a study involving four international laboratories, the BmR1 dipstick was found to be 93% sensitive and 100% specific when tested with 535 serum samples from patients with various infections and healthy controls [ 8 ]. In another multicenter validation study, 97% sensitivity and 99% specificity were recorded when the BmR1 dipstick was tested with 753 serum samples [ 23 ]. The present study demonstrated that interpretations of some aspects of the seroepidemiology of filarial infection are affected by the kind of antigen employed in the assay. Thus this study, which utilized the same set of serum samples on assays using two kinds of antigens, highlights the role of the kind of antigen employed in the comparison of results of prevalence studies. Materials and methods Sera and study population The details on the sera and study population are as described previously [ 22 ]. Briefly, serum samples were collected cross-sectionally from a total of 247 transmigrants and 133 life-long residents (LLR) from Budong-budong, a district of Mamuju Regency in South-Sulawesi, Indonesia, which is endemic for nocturnal-periodic B. malayi [ 24 , 25 ]. The samples from the transmigrant population are valuable as they could help determine the pattern of acquisition of infection with increasing length of exposure to Brugian filariasis. The transmigrant population had traveled to their new homesteads in groups; they came from the same village or region in Bali or Lesser Sunda islands as part of the government-sponsored relocation programme. Each year a new settlement was founded close to the former one (between 10 and 20 kilometers) which accommodated groups of transmigrants from 2–3 different regions together with migrants from Polmas, an over-populated area in South Sulawesi, to promote integration of different tribes. Transmigrants were grouped together according to the year of arrival in the new settlement. A total of 6 transmigrant units, settled between several months and 6 years prior to the survey, were included in the study together with 2 villages of indigenous Sulawesians (LLR), which were situated closely to the transmigrant areas. Those aged ≤ 15 years were classified as children, while adults were classified as those aged 16 years and older. The mean age of children in transmigrant population and LLR population was 10.2 years and that of adults was 32.6 years. Soluble worm antigen (SWA) Adult B. malayi worms were purchased from TRS labs, Athens, Georgia, USA. Female worms were freeze dried, ground to powder, dissolved in phosphate buffered saline (PBS), homogenized and slowly stirred overnight at 4°C. The protein concentration was determined by 2,2'-biquinoline-4,4'-dicarboxylic acid disodium salt hydrate (BCA) method before storage at -20°C. BmR1 dipstick This BmR1 dipstick (Brugia Rapid™) was performed as described previously and according to the instructions of the manufacturer [16, Malaysian Bio Diagnostics Research Sdn. Bhd., Bangi, Selangor, Malaysia]. The BmR1 recombinant antigen was expressed from B m 17DIII DNA sequence, GenBank accession no. AF225296. Southern blot hybridization assays performed on cDNA libraries of L3, L4, mf, adult male and adult female B. malayi demonstrated that the DNA sequence is present in all of the five kinds of libraries (Rahmah et al ., unpublished data). Preliminary immunohistological studies suggest that the expressed antigen is found in the epithelial membranes of the adult female uterus (Rahmah et al ., unpublished data). Statistical Analysis The results obtained with BmR1 dipstick were compared to the results obtained previously with SWA-ELISA. The similarities and differences in the results were analyzed by comparing proportion of related samples using McNemar test; and comparing proportions of unrelated samples by using Pearson Chi-square (if indicated Fischer exact test was used instead). Results and Discussion Figure 1 shows the antibody prevalence to B. malayi , as determined by BmR1 dipstick assay and SWA-ELISA, in the transmigrant population who had resided for various lengths of time in the endemic area and the antibody prevalence in the LLR population. The first detection of IgG4 antibody by the BmR1 dipstick was recorded at 3 years post-exposure, the antibody prevalence increased from 0% in the new arrivals (≤ 1 month and 2–4 months) to 7.4%, 11.1%, 39.1% and 42 % in populations exposed to the infection for 3, 4, 5 and 6 years respectively. Comparison of IgG4 assays using the two kinds of antigen demonstrated similarities in three areas. First, using both SWA-ELISA and BmR1 dipstick, the total prevalence of specific IgG4 in the transmigrant population was found to increase with increasing length of residence in the endemic area. Thus the BmR1 dipstick test confirmed the previously reported finding which demonstrated that the development of anti-filarial IgG4 correlated with the duration of exposure in previously unexposed population [ 22 ]. Second, after a period of 5–6 years of being exposed to the filaria infection, anti-filarial IgG4 prevalence by BmR1 dipstick in the transmigrant population (42%) was comparable (p = 0.763) to the antibody prevalence in the LLR population (39.8%); this finding was also previously reported with SWA-ELISA [ 22 ]. Thus the results with BmR1 dipstick is in agreement with the previous finding that approximately 5–6 years of exposure is required for the anti-filarial IgG4 in the transmigrant population to reach levels comparable to life-long residents. This is also reported to be the period needed for the detection of microfilaria in the peripheral blood [ 22 ]. Third, the overall prevalence of anti-filarial IgG4 was found to be higher in males than in females, by both BmR1 dipstick (p = 0.019) and by SWA-ELISA (p = 0.001). Figure 1 IgG4 antibody prevalence in transmigrant population who resided in B. malayi endemic areas for various lengths of time and in life-long residents (LLR), as determined by BmR1 dipstick and SWA-ELISA. Differences between the two kinds of antigen were also observed (Figure 1 ). First, except for year 5 transmigrants (p = 0.146), the percent antibody prevalence recorded in the transmigrant population and in the LLR population by BmR1 dipstick were significantly lower than that detected by the SWA-ELISA i.e. year 3, p = 0.00; year 4, p = 0.00; year 6, p = 0.041; LLR, p = 0.00. Second, using SWA-ELISA, anti-filarial IgG4 was first detected at 2–4 months post-exposure in the population of transmigrants, while using BmR1 dipstick the first detection of IgG4 antibody was recorded after three years of residence in the endemic area. Figure 2a and 2b shows the IgG4 antibody prevalence among children and adult populations as determined by BmR1 dipstick and SWA-ELISA respectively. There was no significant difference detected in antibody prevalence between adults and children in the transmigrant population and in the LLR population when BmR1 dipstick was used (year 3, p = 0.594; year 4, p = 0.066; year 5, p = 0.907; year 6, p = 0.061; LLR, p = 0.074). Using SWA-ELISA, except for early transmigrant settlers (2–4 months residents), the IgG4 prevalence in the transmigrant population was reported to be significantly higher in adults than in children [ 22 ]; however in the LLR population, antibody prevalence in children was not significantly higher than in adults (p = 0.316). Figure 2 IgG4 antibody prevalence in transmigrant adult and children populations who resided in B. malayi endemic areas for various lengths of time and in life-long residents (LLR). a (top) IgG4 antibody prevalence as determined by BmR1 dipstick assay. b (bottom) IgG4 antibody prevalence as determined by SWA-ELISA. (Note: Previously published in Parasitology 2001; 122, pg. 636 [Reproduced with permission]). Thus the BmR1 dipstick test demonstrated that the establishment of anti-filarial IgG4 in the children and adults of the transmigrant population occur at comparable rates; this is not in agreement with the previous finding, using SWA-ELISA, that demonstrated IgG4 was established more rapidly in the adult population than in children [ 22 ]. The above observations may be due to a mixture of filarial antigens in SWA, some of which may recognize antibodies produced by exposed but not infected individuals, and/or by individuals who have cleared the infection either due to treatment or spontaneous death of worms. Although IgG4 detection significantly increases the specificity of antibody assays in brugian filariasis [ 12 ] the specificity of parasite extract-based assays and recombinant antigen-based assays may not be the same. In a recent study in a non-filaria endemic area in Brazil, the presence of Strongyloides antibody responses was found to be associated with higher antifilarial IgG4 responses in assay that uses crude filaria extract as compared to assay that uses B. malayi Bm14 recombinant antigen [ 26 ]. The discrepancy between the results of SWA-ELISA and BmR1 dipstick may also be partly due to the greater sensitivity of the former as compared to the latter. Out of 120 (of 381) discrepant results, 111 were positive by SWA-ELISA but negative by BmR1 dipstick. Out of these 111 samples, 38 had low ELISA titers (cut-off value > 4.02 but < 4.5). If the stringency of the cut-off value of the SWA-ELISA is increased from 4.02 to 4.5, then the discrepancy can be considerably reduced to 88 (of 381), with 73 positive by SWA-ELISA but negative by BmR1 dipstick. On the other hand, since BmR1 dipstick test has been reported to be highly specific [ 8 , 23 ], it is unclear why there were eight individuals who were positive by BmR1 dipstick but negative by SWA-ELISA. Due to the problem of maintaining the antigenicity of the BmR1 antigen when shipped from Malaysia to Netherlands, the dipstick (immunochromatography) assay, which can be transported at room temperature, was employed to test the BmR1 antigen. However the difference in the assay formats is unlikely to be the reason for the lower overall antibody prevalence levels seen with the latter. This is because we have previously shown that ELISA using BmR1 was less sensitive ( albeit equally specific) than BmR1 dipstick test [ 27 ] in detecting B. malayi infection. Thus in this study if the ELISA format had been used to determine prevalence of IgG4 antibodies to BmR1 instead of the dipstick format, the SWA-ELISA would still be detecting significantly more positives than the BmR1-ELISA. Differences between the two antigens were also observed in the pattern of IgG4 antibody prevalence in children (Figure 2a & 2b ). Using BmR1 dipstick, positivity of the test in children was first demonstrated at 3 years (2 of 35, 5.7%), followed by no positive child detected at year 4 (0 of 17), positive antibody prevalence at year 5 (3 of 8; 37.5%) which is significantly greater than at year 3 (p = 0.011). This is followed by a nonsignificant decrease (p = 0.589) in antibody prevalence by year 6 (6 of 22; 27.3%). IgG4 prevalence at year 6 and LLR (30 of 87; 34.5%) was also found to be not significantly different (p = 0.542). After year 4, the IgG4 prevalence in children seemed to achieve a stable level that was similar to the antibody prevalence in the LLR population children. In adults, infection was also initially detected at year 3 and there appeared to be a pattern of increasing antibody prevalence with increasing time of residence (9.1% at 3 years; 16.7% at year 4, 40% at year 5; 53.6 % at year 6). Furthermore, the detection rate at year 6 was not significantly different (p = 0.766) from that seen in the LLR population adults (50%). Thus using the BmR1 dipstick, the increasing total IgG4 prevalence with exposure to brugian filariasis was mostly due to the increasing positive cases in the adult population. It is tempting to speculate that the differences in the pattern observed in children and adults are due to the greater rate of aborted infections and/or spontaneous clearance of the infection in children than in adults. The overall lower worm burden in children, due to physiological differences, may enable higher rate of spontaneous clearance of infections in children than in adults. The difference in physiology between adults and children has been demonstrated by the greater number of natural killer cells, and T-and B-lymphocytes in children as compared to middle aged people [ 28 ]. In addition CD4/CD8 T-cells have also been reported to decline from a young age onwards [ 29 , 30 ]. However, since this is a cross-sectional study in which data for each period of residence were obtained from different groups of individuals, this hypothesis could not be confirmed. Conversely, results of the SWA-ELISA demonstrated initial antibody positivity in children at 2–4 months, followed by increasing antibody prevalence with time of residence. IgG4 prevalence at year 6 was found to be significantly lower than the antibody prevalence in children of the LLR population (p = 0.007). In the adult population, except for the earlier initial detection, a similar pattern was observed with the results of the BmR1 dipstick i.e. increasing rate of antibody prevalence in the transmigrants with exposure; and the IgG4 prevalence at year 6 is not significantly different (p = 0.430) than the antibody prevalence in the LLR adult population. Conclusions This study highlights that assays using both BmR1 and SWA antigens demonstrate an increasing prevalence of specific IgG4 antibodies in the transmigrant population with increasing length of residence in an area endemic for brugian filariasis, and, by six years residency that the antibody prevalence was similar to that observed in the LLR population. This study also documented three main differences in results derived from assays using two antigens i.e. 1. Earlier detection and higher rate of antibody prevalence by SWA-ELISA as compared to the results demonstrated by BmR1 dipstick; 2. Similar rate of acquisition of antibody prevalence in children and adults by BmR1 dipstick; whereas by SWA-ELISA adults were found to become antibody positive faster than in children; 3. By BmR1 dipstick the increasing total prevalence of IgG4 with exposure was primarily due to the adult population, whereas by SWA-ELISA this was attributed to both children and adult populations. This study demonstrates that some aspects of seroepidemiology of Brugia malayi infection may vary with the kind of antigen used in the assay. Thus comparison of results of different studies must take into account the kind of antigen employed, especially if one study uses native antigen and another uses recombinant antigen. It would be interesting to compare seroepidemiological data of IgG4 assays using two different recombinant antigens on the same population. Competing interests Rahmah Noordin is the inventor of the commercialized BmR1 dipstick test (Brugia Rapid™) Authors' contributions RN drafted the paper and supplied the BmR1 dipstick test; SW & AM supplied the sera and edited the paper; LBH performed most of the statistical analysis; ES and MY conceived the study and contributed significantly in editing of the paper: In addition ES performed the BmR1 dipstick test and some statistical analysis. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC517507.xml |
518973 | Emergence of new Salmonella Enteritidis phage types in Europe? Surveillance of infections in returning travellers | Background Among human Salmonella Enteritidis infections, phage type 4 has been the dominant phage type in most countries in Western Europe during the last years. This is reflected in Salmonella infections among Swedish travellers returning from abroad. However, there are differences in phage type distribution between the countries, and this has also changed over time. Methods We used data from the Swedish infectious disease register and the national reference laboratory to describe phage type distribution of Salmonella Enteritidis infections in Swedish travellers from 1997 to 2002, and have compared this with national studies conducted in the countries visited. Results Infections among Swedish travellers correlate well with national studies conducted in the countries visited. In 2001 a change in phage type distribution in S. Enteritidis infections among Swedish travellers returning from some countries in southern Europe was observed, and a previously rare phage type (PT 14b) became one of the most commonly diagnosed that year, continuing into 2002 and 2003. Conclusions Surveillance of infections among returning travellers can be helpful in detecting emerging infections and outbreaks in tourist destinations. The information needs to be communicated rapidly to all affected countries in order to expedite the implementation of appropriate investigations and preventive measures. | Background Salmonella Enteritidis is the most common serovar causing food-borne salmonellosis in humans, causing approximately 80% of salmonellosis cases reported in Europe [ 1 ]. During the 80s and early 90s, a steady increase in S. Enteritidis infections was reported in Europe and North America [ 2 - 4 ]. The most common phage types of S . Enteritidis varies between countries; while phage type (PT) 4 is reported to be dominant in most countries in Western Europe, PT 8 is common in North America and also a few European countries [ 5 - 7 ]. Epidemiological and environmental studies have implicated eggs and poultry products as primary risk factors for infection [ 3 , 8 ]. Approximately 70% of outbreaks caused by S. Enteritidis in Europe during the 90s, were related to eggs and egg products [ 1 ]. Based on these findings, prevention and control measures in the egg and poultry industry have been implemented in the European Union and in the US [ 9 , 10 ]. These measures seem to have been effective in reducing S . Enteritidis contamination of eggs [ 11 ] and are believed to have lead to a decrease in human incidence of S. Enteritidis in recent years [ 1 , 9 , 12 ]. In Sweden, about 85% of the reported salmonella infections are acquired during travel abroad and the levels in domestic animals and food products is low [ 13 ]. Therefore, trends in human salmonellosis in Sweden have mainly reflected trends in foreign travel and countries with popular package tour resorts account for the majority of infections. In this study, we investigate trends in travel related S. Enteritidis infections, describe the phage type distribution of S. Enteritidis isolated from Swedish travellers returning from abroad related to country of infection, and discuss possible reasons for the emergence of a new phage type of S. Enteritidis in 2001. A preliminary analysis of a subset of these data was published as a short letter in the Lancet in 2002 [ 14 ]. Methods In Sweden, salmonellosis is a mandatory notifiable disease. Both clinicians and laboratories are required to report a case to the infectious disease register at the Swedish Institute for Infectious Disease Control (SMI). Based on the patient's travel history, information on probable country of infection is collected on the notification forms. Diagnosis of salmonella infections is made at regional microbiology laboratories, and all isolates are submitted to the national reference laboratory at SMI for serotyping and phage typing. In this study we have included all cases of S. Enteritidis notified to SMI from January 1, 1997 through December 31, 2002. To investigate trends in travel-related infections, we collected information on air travel from the Swedish Civil Aviation Administration (CAA)[ 15 ]. The figures include all passengers carried on flights from any CAA airport to their first foreign destination, without indicating whether this destination is for transfer or the final destination. Using these figures as denominators, we calculated annual incidence rates for countries reported as place of infection for the three most popular countries for charter tourism among Swedes – Spain, Greece and Turkey. The incidence rates were calculated by dividing the number of cases reported as infected in the respective countries by number of flight passengers to that country. For the geographical description of dominant phage types in the different countries the analysis was restricted to infections acquired in countries in Europe during 1997 to 2002, and to countries from which more than ten cases were reported during this period. A phage type was considered dominant if it represented >30% of the isolates and was at least twice as common as the second most common phage type. If none of the phage types fulfilled these criteria, the two most common phage types that together represented >50% of the cases were defined as the dominant types. The pattern for 1997 to 2000 was compared with the pattern for 2001. Results Of 13,271 cases of S. Enteritidis infections notified during 1997–2002, 11,570 cases (87%) were reported as infected abroad, 1,032 (8%) were reported as infected in Sweden, and information of probable country of infection was not available for 669 (5 %). The total number of cases reported each year varied from 1,598 to 2,629 cases, with the highest being reported in 1999 and the lowest in 2002. Imported cases varied between 1,404 and 2,164. The most common countries of infection during the six-year period were Spain, Greece and Turkey, accounting for 34%, 8% and 4% of all cases, respectively. For six countries – 'Spain, Greece, Turkey, Poland, Thailand and Portugal – >50 cases of infection with S . Enteritidis among Swedish travellers were reported each year during 1997 to 2002. These are all popular destinations for leisure travel among Swedes. Figure 1 presents incidence rates for the three most popular countries for charter tourism, Spain, Greece and Turkey, using the number of flight passengers from Sweden to the first foreign destination as the denominator. The figure shows that the incidence rate among travellers to Spain and Turkey seemed to decrease during the six-year period, while for Greece the incidence peaked in 2001. Eighty-six percent (10,049) of the isolates from 1997 to 2001 were phage-typed, increasing from 75% in 1997 to 95% in 2001. The most common phage types over the period were PT 4 and PT 1 (Figure 2 ), accounting for 35% and 16% of all cases, respectively. In travellers returning from most countries in Western Europe, PT 4 was the dominant phage type. In Eastern Europe, PT 1 was dominant, and this phage type was also common among travellers returning from the Iberian Peninsula. PT 8 seemed to be more common among travellers returning from central European countries. In 2001 this pattern changed when PT 14b increased among travellers from several countries in Southern Europe (Figure 3b ). PT 14b was the third most common phage type among returning travellers in 2001, accounting for 13% of all isolates that were phage-typed that year (272/2,132), compared with 2 % of all that were phage-typed in the previous four years (154/7,917). This trend continued into 2002 when PT14b accounted for 9% (134/1,489) of all typed isolates. The majority of the PT 14b cases in 2001 and 2002 were reported among travellers returning from Greece (Figure 4 ), and this phage type accounted for 54%(157/293), 47%(83/176) and 42%(50/117) of all cases of S. Enteritidis from Greece in 2001, 2002 and 2003, respectively. Discussion We have described trends and phage type distribution of S. Enteritidis isolates among Swedish travellers infected abroad. Phage type 4 was the dominant phage type among returning travellers. There were, however, some differences in distribution between the countries and with time. Between 1997 and 2000, PT 1 dominated among travellers returning from Russia and the Baltic countries, PT 8 was commonly seen among travellers returning from some central European countries, and PT 4 dominated among travellers returning from most other European countries (Figure 3 ). In 2001 a change in phage type distribution was observed among Swedish travellers returning from some countries in South-Eastern Europe. PT 14b, a previously rare phage type, appeared to become predominant among travellers returning from Greece and also became more common among travellers from some other countries. During the six-year period, the S . Enteritidis incidence rate among travellers to Spain and Turkey appeared to decline. This trend is in contrast to surveillance data from Spain, which seem to show an increasing incidence over the same period [ 16 ]. The reason for this declining trend among tourists was not investigated, but may be related to an increased awareness among tourists concerning the prevention of traveller's diarrhoea or to improved food control efforts in some of the popular tourist resorts. There are some uncertainties when calculating incidence rates based on number of travellers. The number of travellers used as the denominator in the incidence rate calculations is based on the statistics of the airport of first landing after leaving Sweden. If the first airport is not the final destination, these travellers will not be included in the denominator for the destination country. Thus the calculated incidence rates for transit countries will be too low, while the incidence rates for the final destination countries will be too high. When comparing S. Enteritidis phage types isolated from Swedish travellers with studies conducted in the countries visited, isolates found in travellers were generally consistent with the dominant strains reported among inhabitants in the respective countries. Table 1 summarises the results of published studies on salmonellosis from a number of countries. The percentages of PT 4 and PT 6 among human S . Enteritidis infections in Austria and Denmark, respectively, have been reported to be decreasing in the last years [ 6 ]. These same trends were reflected among returning Swedish travellers. People returning from travel abroad may have a higher tendency to seek medical care and have a stool sample taken if an imported infection is suspected. In addition, visitors may be more susceptible to pathogens circulating in the community than the local inhabitants. The detection of new, emerging strains in travellers after returning to their home countries may therefore be helpful in detecting changes in the pathogen reservoir occurring in the countries visited, especially in tourist destinations. However, tourists have a tendency to aggregate in some smaller resorts that may have a different pathogen reservoir and rely on food supplies that are different from the rest of the country. This may lead to differences in risks and pathogens between the inhabitants and the tourists visiting the country that needs to be taken into consideration. The change in phage type distribution observed among Swedish travellers returning from some countries in Southern Europe in 2001 was not observed among inhabitants in the countries visited. In total, the two most common phage types among Swedish travellers were, as in previous years, PT 4 and PT 1. However, the third most common was PT 14b, a phage type hitherto uncommon in Sweden with only 20 to 40 cases reported annually prior to 2001. The majority of the cases were among travellers returning from Greece (90%). More cases were also reported among travellers returning from Spain and Bulgaria than in previous years. During the same time period, an increase in the same phage type had also been registered among Norwegian and Finnish travellers returning from Greece [ 17 ]. A request on Enter-Net (European network for the surveillance of enteric infections – Salmonella and VTEC O157) sent by the Norwegian Public Health Institute gave no response on increases of PT 14b in other countries. Spain reported an outbreak of the same phage in a school in January (unpublished data). But after this event, no further increase was noted. The UK later reported an increased number of the same phage type, both among travellers and among people who had been infected in the UK. However, there the 14b isolates of domestic origin were aerogenic, while isolates associated with travel to Greece were predominantly anaerogenic [ 18 ]. The isolates among Swedish travellers returning from Greece were also predominantly anaerogenic. No explanation for the sudden increase in this phage type among Nordic travellers to different countries in Southern Europe has been found. PT 14b is not a new phage type and outbreaks reported previously have been mainly related to eggs and egg-products (ice-cream, tiramisu [ 19 , 20 ]) or improper hygiene practices [ 21 , 22 ]. However, these outbreaks were localised, of limited duration, and the incriminated food products found and the outbreaks contained. The cause of the increase of this phage type among Nordic travellers in 2001 is still unclear. It may represent a geographically more widespread outbreak than previously described, possibly due to increased trade in food products, animals or animal feed across the borders. Another possible explanation for this increase may be that changes in PT 14b could have contributed to increased resistance or virulence factors, thereby facilitating the spread of this phage type in the environment. Alternatively, acquisition or loss of a plasmid or spontaneous mutations may have resulted in a conversion from another phage type to PT 14b. Such change has been described for other phage types [ 23 - 25 ] and a conversion from PT 8 to PT 14b has been described after inoculation into pathogen-free chicken [ 26 ]. Our data presented are limited by the small numbers of cases from each country investigated on an annual basis. It is therefore difficult to evaluate trends with any certainty. However, the possibility of using surveillance data of infections among returning travellers to detect emerging pathogens should be further investigated. In addition, data from countries that routinely collect information on travel could be pooled in order to increase the numbers of travel-related infections. In several countries outside Europe, laboratory capacity is limited and it may take a long time to detect the emergence of new pathogens or subtypes. Not all countries in Europe collect information on the probable place of infection and phage typing of Salmonella isolates is not routinely performed in some countries. However, if available, this information will be included in the data reported to Enter-Net. Data from this network has previously been useful in detecting travel-related outbreaks [ 27 , 28 ], and may also be useful in describing pathogen patterns in countries where laboratory capacities are limited or routine typing is not performed. Importantly, Enter-Net may expedite the dissemination of information concerning emerging pathogens. Conclusions This study demonstrates that surveillance of infections among returning travellers may be helpful in detecting emerging infections and outbreaks in tourist destinations, and provides some useful supplementary data about infectious diseases and trends in other geographical regions. Characterization of isolates from travellers can detect changes in the pathogen and antimicrobial resistance patterns in the destination country. This information may be an important supplement in countries where surveillance systems are deficient or lacking, or where the laboratories have limited capacity to do detailed sub-typing and resistance testing. In addition, infections and outbreaks among tourists may not always affect the local residents and therefore may not be detected by the local public health authorities. If proper investigations, and appropriate prevention and control measures are to be implemented in the countries visited, it is important that the surveillance information compiled from the traveller's home countries is rapidly communicated to the affected countries. Competing interests None declared. Authors' contributions KN performed the data analysis and drafted the manuscript. PJG, YA and BdJ participated in the design and coordination of the study. AO conducted typing and provided advice regarding laboratory issues. JG participated in the design and discussion, and provided advice on data analysis. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC518973.xml |
515310 | Intensive language training enhances brain plasticity in chronic aphasia | Background Focal clusters of slow wave activity in the delta frequency range (1–4 Hz), as measured by magnetencephalography (MEG), are usually located in the vicinity of structural damage in the brain. Such oscillations are usually considered pathological and indicative of areas incapable of normal functioning owing to deafferentation from relevant input sources. In the present study we investigated the change in Delta Dipole Density in 28 patients with chronic aphasia (>12 months post onset) following cerebrovascular stroke of the left hemisphere before and after intensive speech and language therapy (3 hours/day over 2 weeks). Results Neuropsychologically assessed language functions improved significantly after training. Perilesional delta activity decreased after therapy in 16 of the 28 patients, while an increase was evident in 12 patients. The magnitude of change of delta activity in these areas correlated with the amount of change in language functions as measured by standardized language tests. Conclusions These results emphasize the significance of perilesional areas in the rehabilitation of aphasia even years after the stroke, and might reflect reorganisation of the language network that provides the basis for improved language functions after intensive training. | Background Cerebrovascular stroke is a highly prevalent condition and the major cause of language impairment in adults. Immediately following a stroke about 38% of the affected population experience aphasia [ 1 ]. Spontaneous recovery is reported within the first six months after the event, while only minimal spontaneous improvements of language functions are expected after more than one year post-stroke [ 2 ]. Additional rehabilitation efforts have produced beneficial effects, as reported for speech and language therapy on the basis of different performance indices [ 3 , 4 ]. In accordance with recent progress in neurorehabilitation, which takes into account evidence of the brain's capacity for reorganization [ 5 - 7 ], intensive language training (several hours per week) seems to be the premise for substantial improvement of language functions in the chronic stage [ 8 ]. To date, the evaluation of impairment and recovery of function, including training-induced improvement in aphasia, has been based mainly on performance in neuropsychological tests. This is now being increasingly complemented by measures of brain function. Different mechanisms and time courses of recovery of language function after brain damage have been discussed. Hemodynamic imaging suggests the involvement of two mechanisms: (1) regression of diaschisis (reduced metabolism and function in areas connected with the damaged brain tissue, which have been cut off from essential input) and (2) functional reorganization of the neuronal networks involved in language processing. Regression of diachisis in perilesional and more distant regions have been shown to contribute to recovery of function particularly in early phases of the recovery process [ 9 ]. In contrast, "re"-recruitment (that is, reorganization) of perilesional areas of the left hemisphere [ 10 ] or reactiviation of left hemisphere network components [ 11 , 12 ] predict long-term recovery of language function. Moreover, recruitment of homotopic right-hemispheric areas may contribute to language recovery when the left-hemisphere language network components are permanently impaired [ 13 ]. However, it has been debated whether the recruitment of right hemispheric networks constitutes an additional potential for language processing or whether it is just a by-product of increased general activation. Others suggest this recruitment may even impair the recovery of left hemispheric areas, leading to a persistence of deficits [ 14 ]. Brain structures in the vicinity of structural lesions produce a larger amount of slow wave activity. This might be due to a loss of afferent input (e.g. from the lesion) or to a primary metabolic change within these perilesional areas [ 15 ]. These abnormal slow waves can be detected in the electroencephalogram (EEG) and, due to their focal generators, they can be localized using magnetic source imaging, a magnetencephalogram (MEG) based technique. In Abnormal Slow Wave Activity Mapping (ASWAM) [ 16 ], generators of abnormal slow waves are localized and mapped on to brain structures in order to identify areas that are active but incapable of normal function. A number of studies have demonstrated that focal slow waves indicate abnormality resulting from neurological damage such as contusions, tumors, or cerebrovascular stroke. In particular, abnormal slow wave activity in the delta-frequency range (1–4 Hz) has been found in areas adjacent to the structural lesion [ 17 - 19 ]. Since focal slow wave activity varies with changes in metabolism and blood flow due to the insult [ 19 , 20 ], it has been described as characteristic of a 'dysfunctional state' [ 21 ] of the neuronal tissue or a dysfunctional border zone with little ongoing information processing. In patients with brain tumors, this relationship between slow wave activity and metabolic changes was further elucidated by combining MEG and proton MR spectroscopic imaging [ 22 ]. A mild reduction of N-acetyl aspartate (NAA) and slight accumulation of lactate (Lac) was found in association with sources of focal slow wave activity in the border zones of the tumors, suggesting a border zone between seriously damaged and normal tissue with potential for re-recruitment in the course of the disease. The mapping of abnormal slow wave activity can be used not only to identify dysfunctional neuronal networks, but also to track changes in the course of recovery or treatment. For instance, de Jongh et al. [ 17 ] reported increased focal delta activity in the MEG before and a reduction after resection of brain tumors. The utility of 'abnormal slow wave mapping' (ASWAM) in diagnostics, recovery, or treatment evaluation may be validated by covariation with neuropsychological measures. Lewine et al. [ 23 ] found a correlation between symptom resolution and MEG-slow wave reduction in patients with minor traumatic brain injury (TBI) and Hensel et al. [ 24 ] reported a decrease of EEG-delta amplitude and dipole strength parallel to spontaneous recovery of language functions across the first year post stroke in aphasia patients. The present study employed ASWAM before and after intensive language training in aphasic patients. If ASWAM qualifies for the evaluation of treatment or training-supported rehabilitation in chronic aphasics, changes in the intensity and distribution of focally generated abnormal slow wave activity should vary with improvement of language function after a specific intervention. Aphasics were recruited from an ongoing project evaluating the effectiveness of an intensive language training program. This program combines the learning principles of shaping and the efficacy of concentrated training [ 25 , 6 ] while considering the principles of cortical reorganization [ 5 ]. In order to minimize any influence of spontaneous recovery on changes in the brain-function measure, only chronic aphasics were selected to participate either in 30 hours of Constrained-Induced Aphasia Therapy (CIAT) [ 25 ] or in 30 hours of massed model-based (MB) aphasia therapy [ 26 ]. All training sessions were scheduled within a two-week period. It was hypothesized that (a) aphasics would display an increased density of slow wave generators in the damaged (left) hemisphere before training, (b) this density would be reduced in the perilesional zone following language training and (c) there would be an improvement of language functions as evaluated by a standardized language test (Aachen Aphasia Test Battery, AAT) [ 27 ]. Results Language functions The average test performance of the entire patient group increased after language training, as indicated by the AAT profile (t(27) = 9.85, p < 0.0001, paired t-test, two tailed). Similar improvements were found for the Token Test (t(27) = 6.10, p < 0.0001). The average improvement of the profile score was 2.9 ± 1.3 points and 6.1 ± 5.3 points on the Token Test (T-scores). Twenty-five of the 28 patients improved on at least one subtest (N = 19) or subscale (N = 6) of the AAT. Maximum delta activity In 26 subjects the maximum activity of delta dipoles was found in the left hemisphere and in the vicinity of the structurally obvious lesion (as verified by structural MRT; see Figure 1 for three representative subjects). In one patient, the maximum delta activity was located in the right hemisphere anterior to the homologue of the lesion, a finding consistent across measurements. (The patient had a very mild amnesic aphasia, displayed the highest AAT profile score of the entire group [63.15] and showed the least amount of delta activity.) In another subject, the maximum delta activity was located at the posterior border of a large left fronto-temporal lesion due to an ischemic infarct of the middle cerebral artery in the first measurement. After training, the maximum delta activity was found in the right-hemispheric area anterior to the homologue of the lesion. Notably, both measurements showed that this patient had clusters of delta activity next to the lesion and its right hemispheric homologue. Left hemispheric Delta Dipole Density (DDD) decreased after training and increased in the right hemisphere, which might explain the shift of peak activity to the right. The location of this delta focus remained stable across the two measurements (Rho: x-axis: .69, p < 0.0001, y-axis: .85, p < 0.0001, z-axis: .73, p < 0.0001). The coordinates of maximum delta dipole density were exactly the same in eleven patients, while maximum delta activity shifted by one voxel in one of the three cardinal planes in eight patients, and by more than one voxel in nine patients. (As emphasized above, one patient displayed a reversal in hemispheric lateralization after training). Hemisphere-specific average delta activity Thresholds were significantly higher in the left hemisphere (F(1,54) = 49.03, p < 0.0001). Clusters of voxels with delta activity > 2 SD above the average DDD in a group of 25 healthy controls were found in 26 of the 28 patients in the left hemisphere before training. Such clusters were found in the right hemisphere in only 7 patients. In two patients only, delta activity in the right hemisphere exceeded left hemisphere activity. Average delta activity was significantly more pronounced in the left hemisphere before and after training (for the pre-measurement the main effect HEMISPHERE was F(1,54) = 55.35, p < 0.0001; for the post-measurement, F(1,54) = 46.55, p <0 .0001: Figure 2 ). Twelve patients showed an increase in left hemisphere delta activity after training, while a decrease occurred in sixteen patients (Figure 3 ). This diverging pattern became evident in the non-significant interaction TIME*HEMISPHERE (F(1,54) < 1). An increase in delta activity of the left hemisphere tended to covary with a longer amount of time since the lesion (F(1,26) = 3.69, p = 0.06). Changes in DDD relative to improvement of langauge functions "Magnitude of change" in the left hemisphere was more pronounced in those patients who displayed significant improvement in at least one subtest of the AAT (N = 19) compared to patients with minor improvements (in at least one subscale) or no improvements (N = 9; F(1,26) = 4.95, p < 0.05). Magnitude of change of left-hemispheric delta activity varied significantly with improvements in language functions (AAT profile: r = .60, p < 0.002; Token Test: r = .46, p <0 .02, Figure 4 ), while there was no correlation between right hemisphere magnitude of change and language measures (AAT profile: r = .-0.07; Token Test: r = 0.01). Discussion The present results provide further evidence that Abnormal Slow Wave Activity Mapping (ASWAM) discloses generators of abnormal slow waves. The mapping of abnormal slow wave activity on to brain structures allows for the identification of areas that are active, but not capable of normal function. This is, to the best of our knowledge, the first report of a re-test after controlled neuropsychological/-linguistic intervention within the same subjects. The comparison between the two measurements indicates a high reliability of peak locations in left hemispheric perilesional areas, even though successful training modified this activity in magnitude and spatial distribution. In almost all patients, the region surrounding the structural lesion continuously and reliably produced abnormal slow waves, whereas only very few patients presented slow wave activity distant from the structurally confirmed lesion. The amount of perilesional slow wave activity was markedly altered in patients who had improved after training, and the magnitude of this change was related to the changes in language functions. Substantial functional improvements were achieved even in chronic aphasic states by shaping procedures, constraint of non-verbal communication and massed practice. This replicates and extends the findings of Pulvermüller et al. [ 25 ]. The present results further suggest that similar improvements can be achieved irrespective of the particular training procedures (a comparable improvement occurred in the model-based group). Both strategies might produce their effects – at least in part – by reorganizing brain regions next to a lesion. Following Liepert et al. [ 28 ], who demonstrated with transcranial magnetic stimulation that constraint-induced (CI) movement training of the arm expanded the area of the brain involved in generating activity in the muscles of the hand, we might assume a similar mechanism for the presently observed language improvements after intensive speech and language training in aphasics, namely an increased number of increasingly functional areas. In contrast, some patients, though displaying language improvement after training, exhibited an increase of delta activity in the vicinity of the lesion. One explanation for this might be that the functional capabilities of the affected brain area remain disturbed, with no further potential to be restored or re-integrated in the language 'network', and this might inhibit or impair functionally intact regions. Further segregation of these continuously dysfunctional areas from the remaining network might then lead to improved language functions and consequently to increased delta activity. This hypothesis is supported by the correlation between language improvements and either decrease or increase of delta activity in perilesional areas. Moreover, the increase of delta activity was related to longer duration of disease. In most of the patients, re-integration of 'spared' brain areas into the language network should be completed in time (in patients exhibiting an increase of slow wave activity, the time-since-lesion averaged 55.4 ± 39.3 months, compared to 35 ± 13.9 months in patients exhibiting a decrease). Therefore, "dysfunctional" delta activity might be related, at least in a subgroup of chronic patients, to functionally more favorable outcomes. The increase of delta activity might be explained by reduced reciprocal exchange of information within functionally intact and permanently impaired network components. Conclusions Compared to the more conventional procedures for aphasia treatment in the chronic stage, the present training involved an intense use of language capabilities and a restraining of alternative, non-verbal methods of communication. In our opinion, any training that encourages speech production several hours a day over several days has the potential to be efficacious. Massed practice is likely to produce activity-dependent cortical reorganization, found to result from CI-movement therapy [ 28 - 31 ]. It is also presumed to be the basis for a long-term increase in the amount of use of the more-affected extremity and of improved language functions following short-term intensive training. Methods Subjects Twenty-eight patients suffering from chronic aphasia participated in the training (14 females, mean age 55 years, range 35–80 years; see Table 1 for clinical data). All patients were right-handed before brain injury, as assessed with the Edinburgh inventory [ 32 ]. In 20 patients, aphasia resulted from left-hemispheric ischemic stroke; in 8 patients it resulted from a hemorrhage affecting left-hemispheric areas. All patients were in a chronic state as defined by a time-since-lesion > 12 months. The average duration of the time-since-lesion was 43.78 months (range 12–156 months). Structural whole-head MRI was available in 26 patients and the scans were performed within the two week training period. For the other 2 patients, a left hemisphere lesion was verified by inspection of earlier MRI examination. Prior to training, aphasia was diagnosed according to guidelines of the Aachen Aphasia Test [ 27 ], and aphasic syndromes were classified as Wernicke (N = 4), Broca (N = 13), amnesic (N = 2) and global aphasia (N = 3). Six patients could not be classified according to the 4 syndromes given on the basis of the AAT. Aphasia was evaluated as mild (N = 11), moderate (N = 16), or severe (N = 1). Patients were recruited from the local rehabilitation centre (Kliniken Schmieder Allensbach & Konstanz) or from self-help groups, or were referred by neurologists and speech therapists. Design and procedure Since this report focuses on changes in slow wave activity, principles of speech and language training and results will only be summarized (a detailed description will be provided elsewhere). Training took place 3 hours/day for 10 consecutive days and included (for 18 patients) language exercises with increasing levels of difficulty [ 25 ] or (for 10 patients) model-based intervention (training based on the patients' functional deficit, with the main aim of gradually improving spoken word production). The patients received only language therapy during the two-week training period to ensure that changes in slow wave activity were not induced by improvement of potential comorbid neurological impairment (e.g. hemiplegia). Language function was evaluated by two sensitive measures of change of aphasia severity: the profile score and the Token Test of the Aachen Aphasia Test [ 27 ]. Tests were administered by trained psychologists or speech therapists one day before the onset of training and one day after the completion of training. Language function improved significantly after training in both groups (see results) regardless of the type of training (TREATMENT*TIME interaction for AAT profile: F(1,26) = .72, p > 0.3; for Token Test: F(1,26) = 3.59, p= 0.07), therefore data from the two groups were pooled for ASWAM. Training groups did not differ significantly with respect to age or time-since-lesion. Data acquisition and analysis Using a 148-channel whole-head neuromagnetometer (MAGNES™ 2500 WH, 4D Neuroimaging, San Diego, USA), MEG-measurements were collected twice: once on the day before training and once on the day after training. MEG was measured in a 5-minute resting period, during which subjects were asked to relax while staying awake, and to not engage in any specific mental activity. MEG recordings were obtained in a supine position. Subjects were asked to fixate a colored mark on the ceiling of the magnetically shielded room throughout the recording in order to avoid eye- and head-movement. A video camera installed inside the magnetically shielded room allowed for a monitoring of the subject's behavior and compliance throughout the experiment. Written informed consent was obtained from subjects prior to each MEG-session and the study was approved by the ethics committee of the University of Konstanz. The fiducial points, coils, and head shape were digitized with a Polhemus 3Space ® Fasttrack prior to each measurement. The subject's head position relative to the pickup coils of the sensor was estimated before and after each measurement. MEG was recorded with a sampling rate of 678.17 Hz, using a 0.1–200 Hz band-pass filter. For artifact control, eye movements (EOG) were recorded from four electrodes attached to the left and right outer canthus and above and below the right eye. The electrocardiogram (ECG) was monitored via electrodes attached to the right collarbone and the lowest left rib using a Synamps amplifier (NEUROSCAN ® ). Data reduction and analysis Data were reduced by a factor of 16 and digitally filtered for the delta (1.5–4.0 Hz) frequency band using a digital band pass filter (Butterworth filter of the order 6). Artifact-free time segments were determined by visual inspection. Single equivalent current dipoles were fitted for each time point in the selected artifact free segments (distance of time points 24 ms.). Five non-overlapping channel groups over left, right, center, anterior, posterior regions were chosen for dipole modeling. A homogeneous sphere, which gives the best least-squares fit to the digitized patient's headshape below the selected sensors, served as a model for the volume conductor. Dipole fit solutions at time points satisfying the following requirements were accepted: (1) a dipole moment (q) of 10 nAm < q < 100 nAm; (2) a goodness of fit (GOF) greater than 0.90. These restrictions should ensure that neither artifacts nor small amplitude biological noise would affect the results, and that only dipolar fields that were generated by focal sources were analyzed. Each data-set was divided into 1000 voxels, each of 20 mm 3 , using the AFNI-to3d-software (AFNI-Analysis of functional neuroimages [ 33 ]). For each patient, the percentage of dipoles in the delta frequency band per second in each voxel was z-transformed and statistically compared to the dipole density distribution of a group of 25 healthy controls, which were considered a 'norm' group [ 16 ]. Whole brain magnetic resonance images (TR = 19, TE = 5,6, Flip angle = 30°, FOV = 256 mm, 1 mm isotropic resolution) were acquired within the 2-week training period across a 256 mm slab from each subject using a Philips Gyroscan 1.5 Tesla scanner (Philips Medical Systems, Gyroscan ACS-T). MRIs were aligned to the coordinates of the MEG according to anatomical landmarks, coil positions and head shape information using the AFNI software. Focal abnormal slow wave activity and its changes after training were evaluated by the following measures Maximum delta activity A spatial clustering algorithm (FWHM, Filter Width Half Maximum, 60 mm) was applied to smooth the data. Maximum activity, i.e. the voxel with the highest percentage of delta dipoles, was determined using the AFNI subroutine 3dExtrema. The localization of this maximum was determined on the x- (medial-lateral), y- (anterior-posterior) and z- (inferior-superior) plane for the two measurements (pre- & post-training). Hemisphere-specific areas of high delta dipole density The localization of areas with high dipole densities (adjacent clusters of voxels) within each hemisphere was analyzed by applying a narrow filter of FWHM 20 mm to the original data. A narrow filter was used to minimize the influence of more distant voxels with low dipole density on the areas of higher density. First, the voxel with maximum activity was determined for each hemisphere in each patient. By setting thresholds according to the following criteria, voxels with high dipole density were extracted and averaged for each hemisphere: a. Only voxels with z-values within one standard deviation below the maximum of each patient were considered for averaging within each hemisphere. b. If the z-value was smaller than 2 standard deviations, voxels were not considered. c. If the peak density was below 2 SD, an iterative process was initiated. Thresholds were lowered until at least one voxel became apparent where the dipole densities were different before and after training. DDD changes relative to improvement of language functions A measure of the "Magnitude of change" was determined to evaluate changes in dipole density relative to changes in language functions pre- and post-treatment. The averaged (absolute) intensity of delta dipoles in voxels above threshold (in each hemisphere and each patient) was scaled by dividing the magnitude of change in each hemisphere (|T 2 -T 1 |) by the mean of both hemispheres before and after training ((T 1left +T 1right +T 2left +T 2right )/4). This number was then log transformed (to ensure a Gaussian distribution with respect to statistical tests) and submitted to statistical analyses. Two patients were excluded from this final analysis (12: predominantly right hemispheric delta, and 20: shift in lateralization). Statistics Changes in AAT test scores across the two assessments were verified by two-tailed t-tests. Stability of the location of the maximum delta activity was verified by correlation coefficients (Spearman's Rho) for the posterior-anterior, medial-lateral, and inferior-superior axes. Differences in thresholds and average DDD between hemispheres (for both measurements) were verified by means of analysis of variance (ANOVA), as were differences between patients that exhibited an increase or decrease of DDD concerning time-since-onset. Changes in language functions (Token Test, profile score AAT) relative to changes in DDD-magnitude were evaluated by Pearson correlation. Authors' contributions MM, TE and BR participated in the design of the study. MM was responsible for conducting the study, performed data analysis and drafted the manuscript. TE and BR participated in the discussion and general conclusions. CW provided knowledge of data analysis and wrote most of the scripts used for data analysis. DD and GB conducted therapy, assisted in collecting the data, and provided experience of therapeutic issues. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC515310.xml |
512286 | The six-minute walk test in community dwelling elderly: influence of health status. | Background The 6 minutes walk test (6MWT) is a useful assessment instrument for the exercise capacity of elderly persons. The impact of the health status on the 6MWT-distance in elderly, however, remains unclear, reducing its value in clinical settings. The objective of this study was to investigate to what extent the 6MWT-distance in community dwelling elderly is determined by health conditions. Methods One hundred and fifty-six community dwelling elderly people (53 male, 103 female) were assessed for health status and performed the 6MWT. After clinical evaluation, electrocardiography and laboratory examination participants were categorized into a stratified six-level classification system according to their health status, going from A (completely healthy) to D (signs of active disease at the moment of examination). Results The mean 6MWT-distance was 603 m (SD = 178). The 6MWT-distance decreased significantly with increasing age (ANOVA p = 0.0001) and with worsening health status (ANCOVA, corrected for age p < 0.001). A multiple linear regression model with health status, age and gender as independent variables explained 31% of the 6MWT-distance variability. Anthropometrical measures (stature, weight and BMI) did not significantly improve the prediction model. A significant relationship between 6MWT-distance and stature was only present in category A (completely healthy). Conclusions Significant differences in 6MWT-distance are observed according to health status in community-dwelling elderly persons. The proposed health categorizing system for elderly people is able to distinguish persons with lower physical exercise capacity and can be useful when advising physical trainers for seniors. | Background Aging results in an important decrease of muscle power and exercise capacity[ 1 ]. Therefore, elderly often function at the limit of their capacity in order to fulfill the activities of daily living [ 2 ]. Determination of the remaining physical capacity can be important in clinical decision-making. From previous studies [ 3 ] we observed that one in five elderly patients (70 years and over) is unable to execute the classical treadmill based exercise test, either for fear of falling or because of physical or cognitive limitations. The six-minute walking test (6MWT) is a valid alternative, evaluating the exercise capacity at levels corresponding more to efforts commonly performed by elderly during daily activities. The 6MWT has first been introduced as a functional exercise test by Lipkin in 1986 [ 4 ]. Its results are highly correlated with those of the 12 minutes walk test [ 5 ] from which it was derived [ 6 ] and with those of cycle ergometer or treadmill based exercise tests [ 3 ]. The 6MWT is also a valuable instrument to assess progression of functional exercise capacity in different clinical intervention studies [ 7 - 11 ]. The reliability of the test in healthy elderly persons is high (Intra Class Correlation = 0.93) [ 12 ] and it is considered as a valid and reliable test to assess the exercise capacity of elderly patients with chronic heart failure (CHF) and chronic obstructive pulmonary disease (COPD) [ 3 , 6 , 13 ]. Several authors studied the determining factors of the 6MWT-distance in healthy adults and propose either reference equations or normative data for the 6MWT-outcome [ 14 - 16 ]. Troosters et al. [ 15 ] found that age, gender, height and weight explained 66% of the 6MWT-distance variability in 51 healthy adults (free from injury and without history of hospitalization or chronic disease influencing exercise capacity) aged 50–85 years. In another study, Enright et al. [ 14 ] administered the 6MWT to 290 healthy persons between 40 and 80 years old (health based upon age<80 years; BMI<35 kg/m 2 ; ankle-arm blood pressure index <0.9; 1-second forced expiratory volume and absence of stroke history, use of diuretics and smoking). They report that gender-specific reference equations based upon age, height and weight explained 40% of the variance in 6MWT-distance, which was significantly less for men and women who were older and heavier, and for shorter men. Also, Rikli et al. [ 16 ], who measured the 6MWT-distance in respectively 1187 and 2721 community dwelling, functionally independent men and women (who were ambulatory without regular use of assistive device and without medical conditions or physical or cognitive limitations interfering with the test procedures) found a significant 5-year age-group decline in 6MWT-distance (after dividing the study population into gender-specific 5-year age categories) and significantly better scores for men compared to women. However, in a clinical setting one mainly has to deal with elderly having health problems and it can be presumed that their exercise capacity will generally be lower than that of healthy elderly. The available information concerning the impact of health status on the 6MWT-distance in elderly is limited to obesity [ 17 ], to the role of muscle strength in persons with mobility limitations [ 18 ] and to the influence of reduced aerobic capacity in patients with pulmonary or cardiovascular disease [ 13 , 19 , 20 ]. Since in current clinical practice geriatricians and physicians are mainly dealing with patients presenting a widespread and heterogeneous variety of co-morbidity, the results of the aforementioned studies offer only limited information concerning the prognostic value and the applicability of the 6MWT in clinical settings different from CHF and COPD. Therefore, we planned this study in order to investigate to what extent the 6MWT-distance in community dwelling elderly is influenced by a broader spectrum of health conditions. Methods Participants All members of a large Belgian Health Insurance Organization (BHIO) who registered for a health-conditioning week for seniors, organized by the BHIO, were invited by advertisement to participate in our study. The program of the health-conditioning week included general instructions for a healthy life-style and physical exercise classes. One hundred and fifty-six subjects (53 male, mean age 64.1 years, SD = 6.9; 103 female, mean age 65.5, SD = 7.7) volunteered to participate in the study and gave their informed consent. All participants were living in the community and belonged to the A-category according to Katz et al. [ 21 ] and thus were independent for basic activities of daily living. Health categories All participants underwent extensive health screening by medical doctors. First, blood & urine samples were collected after overnight fasting for determination of erythrocyte sedimentation rate, mean red blood cell corpuscular volume, leukocyte count (with differentiation), and concentration of haemoglobin, urea, alkaline phosphatase, glucose, Aspartate Aminotransferase (ASAT), Alanine Aminotransferase (ALAT), protein and electrophoresis for the blood samples and determination of protein, glucose and sediment for the urine samples (according to the SENIEUR protocol [ 22 ]). Second, by means of self-administered standardized questionnaires, which were completed by interview, information was obtained regarding medical history, actual diseases, medication use, tobacco and alcohol consumption. Next, all participants underwent physical examination and standard 12-lead electrocardiography (ECG). Based upon this information, participants were then classified into health categories by one of us (TM), before performing the 6MWT. All evaluations were performed on the same day. The classification system was originally developed in order to grade the participants according to the risk for dangerous complications during physical exercise and to allow physical therapists to adapt the scheduled program of the health-conditioning week (consisting in general instructions for a healthy life-style and physical exercise classes). Therefore, cardiovascular abnormalities were considered to present a higher risk than non-cardiovascular conditions. We distinguished four categories of decreasing health (see table 1 ). Subjects categorized as A were completely healthy and were considered as presenting no particular risk for any kind of physical exercise. An additional distinction can be made between those using no medication (A1) and those using preventive medication only (A2). This subdivision might be important in specific clinical contexts (e.g. assessment of Vitamin D levels in elderly); in the context of our study the distinction between health categories A1 and A2 is less relevant and therefore these participants will be considered together in all statistical analysis. Category B consisted of participants who were functioning normally, presented no major medical restrictions, but could be in need of special instructions for exercising due to their health status. Category B1 was accorded to participants having a disease that was non-cardiovascular and stable. Category B2 was given to participants using medication having cardiovascular effects. Subjects in category C had cardiovascular pathology or a history thereof; they were considered as having an increased risk of cardiovascular complications during exercise. Those belonging to category D were found to present signs of acute disease or exacerbation of chronic disease. If combinations of health conditions existed, subjects were classified in the worst health category. Table 1 Health categories Health Category Description * Clinical examples A A1 Completely healthy; no medication A2 Completely healthy; using only preventive medication Hormonal replacement therapy, aspirin, ... B B1 Functioning normally; presence of stabilised, non cardiovascular disease; absence of cardiovascular abnormalities treated hypothyroidism, stable diabetes, ... B2 Functioning normally; using medication with cardiovascular effect, no overt cardiovascular disease other than normalized arterial hypertension Arterial hypertension; β blocking agent, ... C (history of) cardio-vascular pathology or abnormal ECG. Bundle branch block; angina, CABG; ... D Presenting signs of acute or active disease at the moment of examination. bronchospasm, swollen joints, influenza, ... * Status after questioning, physical examination, ECG, and laboratory examination of blood, serum & urine according to the SENIEUR protocol [22]. CABG: coronary artery bypass graft Measurements Subjects were assessed for weight, height and body mass index [BMI = weight / (height) 2 ]. Before starting the health conditioning week, all participants performed the 6MWT following a protocol as previously described [ 3 ]. All participants were naive to the 6MWT. Each participant was tested individually and was constantly observed by a physical therapist, who was unaware of the health category attribution. The 6MWT was performed outdoors upon a hardened and flat surface following a circular circuit of 121 m. Participants were instructed to try to cover as much distance as possible within six minutes without running. They wore comfortable shoes and clothing and were allowed to rest or stop when necessary. Statistical analysis Statistical analysis was performed using the SPSS (for Windows, release 11.5.1) software package. All data subsets were assessed for the presence of a normal distribution (Kolmogorov-Smirnov Goodness of Fit Test p > 0.05) before using parametric analysis. Correlations between data subsets with a normal distribution were performed using Pearson's Correlation Coefficient; non-normally distributed datasets were analyzed for correlation using Kendall's Correlation Coefficient. Differences between data subsets were analyzed using Analysis of Variance (ANOVA), Analysis of Co-Variance (ANCOVA) and Students t-test. Bonferroni post-hoc test was performed to detect significant differences between subgroups. A multiple linear regression model was designed in order to explain the variability of the 6MWT-distance. Significance level was set at p < 0.05 [ 23 ]. Results The mean 6MWT-distance was 603 m (SD = 178) for the whole group (N = 156). Overall, male participants were significantly heavier (p < 0.001), taller (p < 0.001) and covered a significantly longer distance in six minutes (p = 0.022) than female participants. There was no significant difference in age or BMI between male and female participants (table 2 ). Table 2 Participants' characteristics. Male N = 53 mean (SD) Female N = 103 mean (SD) 6MWT (m) 647.9 (201.8) 579.3 * (159.8) Age (years) 64.1 (6.9) 65.5 (7.7) Weight (kg) 80.5 (11.0) 69.4 † (11.4) Height (cm) 171.3 (6.6) 158.2 † (6.3) BMI (kg/m 2 ) 27.4 (3.1) 27.8 (4.2) *p < 0.05, † p < 0.001 Since there was only one participant in health-category D , this subgroup was excluded from all statistic tests. Fifty-eight (37%) participants (44 in group A1 and 14 in group A2 ) could be considered as completely healthy. No significant difference was found regarding age or 6MWT-distance between participants of group A1 and A2 (mean age respectively 62.0 years SD = 6.6 and 62.7 years SD = 10.0, mean 6MWT-distance respectively 696.4 m SD = 150.7 and 686.9 m SD = 118.4 for A1 and A2 ). ANOVA testing (male and female considered together) revealed a significantly lower 6MWT-distance (p < 0.001) with worsening health category ( A → C ), even after correction for age (ANCOVA p < 0.001). Participants with poorer health status were significantly older (p = 0.0001) than those in better health (Figure 1 and Table 3 ). Figure 1 Impact of Health Status on 6-Minute Walk Distance in Community Dwelling Elderly. ■ Male Participants N = 53. Female Participants N = 102. Bars represent mean values ± SE. Significant decrease of 6 Minutes Walk Distance with worsening health category (ANCOVA corrected for age p < 0.01). * Significantly higher 6 MWT-distance than categories B 2 and C (Bonferroni post-hoc test p < 0.01) † Significantly higher 6 MWT-distance than category C (Bonferroni post-hoc test p < 0.05). Table 3 Characteristics of health groups. Category N Male/Female Age [years] mean (SD) A 58 20/38 62.2 *, † (6.7) B1 30 7/23 63.5 * (6.6) B2 35 11/24 69.0 (6.6) C 32 15/17 67.2 (6.6) D 1 0/1 73 All 156 53/103 65.1 (7.4) * Significantly younger than category. B 2 (Bonferroni post-hoc test p < 0.01) † Significantly younger than category. C (Bonferroni post-hoc test p < 0.01) Post-hoc tests allowed detecting significant differences in the 6MWT-distance between health-categories A & B 2 (p < 0.01), A & C (p < 0.01), and B1 & C (p < 0.05). Subjects in health-category A were significantly (p < 0.01) younger than those in category B2 and C ; those in health-category B1 were significantly (p < 0.01) younger than category B2 . There was a negative correlation between the 6MWT-distance and the participant's age for the whole group (r = -0.42, p < 0.001), and for both males (r = -0.32, p = 0.019) and females (r = -0.47, p < 0.001). A correlation between 6MWT-distance and age was present within health-categories A (r = -0.38; p = 0.004), and C (r = -0.40; p = 0.023), but not within category B1 (r = -0.21; p = 0.269) or B2 (r = -0.27; p = 0.121) (table 4 ). No statistically significant correlations were found between 6MWT-distance and weight or BMI for the whole group or within the different health status categories. Height was significantly correlated with 6MWT-distance (r = 0.33, p = 0.012, Table 4 ) only in the group of completely healthy participants (category A ). Table 4 Relationships between 6MWT-distance and participants' characteristics. Category Age Weight Height BMI All (N = 156) -0.42* -0.03 0.13 -0.14 A (N = 58) -0.38 † 0.10 0.33* -0.20 B1 (N = 30) -0.21 0.07 0.02 0.05 B2 (N = 35) -0.27 -0.10 -0.01 -0.10 C (N = 32) -0.40* -0.14 0.08 -0.29 Male -0.32* -0.19 -0.24 -0.07 Female -0.47* -0.09 0.15 -0.18 Values represent correlation coefficients. *p 0.05, † p < 0.01 When the 156 participants were divided in age categories, again a significant (p < 0.001) decrease in 6MWT-distance was observed with increasing age (figure 2 ). Subjects aged 75 years and more covered a significantly (p < 0.05) shorter distance in the 6MWT than those less than 65 years of age, regardless of their health status. Two-way ANOVA revealed no significant interaction between gender, health status and age concerning the 6MWT-distance (table 5 ). Figure 2 Impact of Age on 6-Minute Walk Distance in Community Dwelling Elderly. ■ Male Participants N = 53. Female Participants N = 102. Bars represent mean values ± SE. Significant decrease of 6 Minutes Walk (6 MWT) Distance with increasing age category (ANOVA p < 0.01). * Significantly higher 6 MWT-distance than categories 65–69 and 75+ (Bonferroni post-hoc test p < 0.05 and p < 0.01 respectively). Significantly higher 6 MWT-distance than category 75+ (Bonferroni post-hoc test p < 0.01). ‡ Significantly higher 6 MWT-distance than category 75+ (Bonferroni post-hoc test p < 0.05). Table 5 Two-way ANOVA for 6MWT-distance Source of Variation F-value P-value Health-category by Age-category 0.65 0.833 Health-category by Gender 0.20 0.937 Age-category by gender 0.40 0.807 In a first step a multiple linear regression model was computed with 6MWT-distance as dependent variable and with age, gender, weight, height and health status as independent variables. Analysis of this regression model (R 2 = 0.33) revealed that neither weight nor height was a significant coefficient, contributing to the model (t = -1.4, p = 0.16 and t = -0.98, p = 0.33 for weight and height respectively). Therefore, a new multiple linear regression model was computed without weight and height as independent variables explaining 31% of the variability of the 6MWT-distance (R 2 = 0.31). In figure 3 the measured individual 6MWT-distances are plotted against the predicted distances based upon the following regression equation: Figure 3 Predicted and Actual 6-Minute Walk Distance in Community Dwelling Elderly. The bullets represent individual values according to the attributed health category: Green bullets Health Category A, Yellow bullets Health Category B1, Red bullets Health Category B2 and Black bullets Health Category C. The lines represent the fit line and the 95% confidence interval. The predicted values are based upon the proposed multiple linear regression model 6MWT-distance predicted (m) = 1192 - (6 × age) - (57 × health category) -(69 × gender) with age expressed in years; 0 for male and 1 for female; 1 for health-category A , 2 for health-category B 1 , 3 for health-category B 2 , 4 for health-category C . The standard error of the estimate is 147 m. 6MWT-distance predicted (m) = 1192 - (6 × age) - (57 × health category) - (69 × gender) with age expressed in years; 0 for male and 1 for female; 1 for health-category A , 2 for health-category B1 , 3 for health-category B2 , 4 for health-category C . The standard error of the estimate was 147 m. In order to determine the impact of health status in the regression model, we have computed a model with 6MWT-distance as dependent variable and with age and gender as independent variables (R 2 = 0.18) in a first step, in which we introduced in a second step the entry of health status as supplementary independent variable. Using this method we obtained the partial correlation coefficient (partial r = -0.40, p < 0.001) between health status and 6MWT-distance. Discussion The 6MWT presents several interesting advantages for the evaluation of the exercise capacity in elderly people. Different authors have described reference equations and tables to predict the 6MWT-distance in healthy elderly subjects. Gender, age, weight and height of these subjects appear to explain a large proportion of the variability in the 6MWT-distance [ 14 - 16 ]. Advancing age, however, is predominantly accompanied by an increasing burden of pathology [ 24 ] and "apparently healthy" elderly persons, willing to participate in physical training sessions, actually present a large diversity in health status. This means that the exercise capacity and the risk for complications during exercise are not the same for each person who consider themselves able to perform physical exercise. Ideally, an exercise program must be established for each person individually considering all facets of his/her health condition. It is conceivable that healthier elderly will present a better physical exercise capacity than those with a worse health condition. Therefore it is to be expected that the performance on exercise tests will reflect the aforementioned difference in health status. However, the contribution of the health status of elderly participants to the variability of the distance walked in the 6MWT has not been extensively described. In order to distinguish completely healthy participants from those with a worse health status we used criteria derived from the SENIEUR protocol [ 22 ]. This protocol was originally established to select participants for studies concerning the age-related changes of the immune system and allows the distinction between age-related and disease-related changes. The literature provides no clear guidelines to categorize elderly persons attending exercise programs according to their health status. Therefore, we developed a classification system, which stratifies health categories corresponding to an increasing risk for complications during physical exercise. Since the SENIEUR protocol allows for the presence of certain diseases and the use of medication that has no influence on the immune system, in our classification these criteria were further adapted in order to identify the "completely healthy" participants (group A ) and to differentiate them from the "apparently healthy". The latter group was functioning normally, but could not be considered as completely healthy since participants either presented stabilized non-cardiovascular conditions (group B1 ) or were using cardiovascular medication, however, without any sign of active cardiovascular disease other than normalized hypertension and without significant abnormalities on ECG (group B2 ). When a history or signs of cardiovascular disease, other than hypertension, were noticed, participants were considered as belonging to group C . When evaluating large numbers of participants there is always the possibility of encountering individuals who are acutely ill; since they will not be able to fully participate in an exercise program, we choose to classify them separately (group D ). The classification system was primarily designed to allow the establishment of recommendations concerning the exercise schedule (type, duration and intensity) of elderly persons in the absence of direct medical supervision. Therefore, the classification system is rather conservative and it is easy for an individual to be considered at risk for complications. Roughly, participants in category A will have no particular limitations for exercising; for those in category B1 the instructions will vary with the nature of the health problem; those in category B2 will only exercise at higher intensity (e.g. up to 80% of maximal heart rate or higher) when guided by an instructor qualified for training elderly persons; those in category C will only be allowed to exercise under supervision of an instructor and with medical guidance of the training program; those in category D will not exercise unless cleared by a physician. Since any exercise schedule will depend upon its objective, we do not go into further details here for these recommendations. It is clear that elderly persons often termed as 'apparently healthy' do not correspond to criteria for being completely healthy. In our study, only 28% (group A1 = 44 of 156) or 37% (Groups A1 and A2 = 58 of 156) of the participants were completely healthy, depending on whether the use of preventive medication was taken into account. However, 79% (groups A , B1 and B2 = 123 of 156) were functioning normally in the community and had no overt history of cardiovascular disease. Rikli et al. (1999) [ 16 ] studied the 6MWT-distance in 7183 community-dwelling older adults (5048 male, 2135 female) aged 60–94 years. According to the inclusion criteria they describe, we can assume that the participants of that study meet our classification criteria of categories A , B1 and B2 . Our study, based upon a much smaller population sample, demonstrates that normative data based upon the population of the aforementioned authors is not representative for completely healthy elderly. Indeed, the mean 6MWT-distance of completely healthy participants (category A , mean age 62 years (SD = 7); mean 6MWT-distance 764 m (SD = 162) and 657 m (SD = 118) for male and female respectively) is much higher than the norm reported by Rikli et al. [ 16 ] for subjects in the same age category (60–64 years; mean 6MWT-distance 616 m (SD = 84) and 551 m (SD = 77) for male and female respectively). When age-matched categories with mixed health status are compared, our results (see figure 2 ) accord well with those of Rikli et al. [ 16 ]. Neither the studies of Enright et al. [ 14 ] or Troosters et al. [ 15 ] used the criteria of the SENIEUR protocol to select healthy individuals. This probably means that the health status of the populations they describe is heterogeneous and might also explain the considerable range in 6MWT-distance (383–820 m) encountered in the 'healthy' elderly (50–85 years old) participants of Troosters et al. [ 15 ] as well as the relatively low median 6MWT-distances (576 m and 494 m for male and female respectively) for 'healthy' adults (40–80 years old) in the study of Enright et al. [ 14 ]. Our study demonstrates that the proposed health-categorization is able to detect a significant reduction in physical capacity due to individual parameters such as medical history, medication use and current health status, other than age, gender (explaining only 18% of the variability) and anthropometrical parameters (no significant independent predictors when health status is included in the regression model). It is our opinion that the results of our study argue in favor of the validity of the proposed health categorization. At this moment, however, other data concerning the reliability of the classification system are not yet available. The results of the multiple linear regression analysis and the absence of a significant interaction (two-way ANOVA testing) between health status, age and gender confirm that each of these factors is an independent determinant of 6MWT-distance in the elderly. The proposed health categorization is able to discriminate the completely healthy elderly from the apparently healthy and to distinguish among the apparently healthy several categories. This results in a more diversified spectrum of the community dwelling elderly population than obtained with other categorizations like the New York Heart Association (NYHA)-classification for cardiovascular disease [ 25 ] and the American Heart Association (AHA) risk stratification criteria [ 26 ]. Currently, these systems are widely used for stratification purposes in physical training schedules. However, they were developed for the description of cardiac patients and do not consider other pathologies nor co-morbidities. A minor familiarization effect has been described for the 6MWT with better values obtained at a second testing [ 27 ]. Since repeated testing is less applicable in clinical settings, we chose to use the results from a single test administration when the participants were still naïve for the 6MWT. In our study the correlation of the 6MWT-distance with age, gender and stature as described by others [ 14 - 16 ] is confirmed in the group of the completely healthy participants (category A ). This relationship is attenuated or disappears in subjects presenting chronic pathology even without apparent functional limitations. Our study demonstrates that there exists a significant decrease in 6MWT-distance with increasing health problems. Since the main objective of our classification system was to stratify for the risk for cardiovascular or metabolic (e.g. hypoglycemia) complications during exercise, we did not expect that the performance on the 6MWT would vary with the health categories. It is probable that the relationship between health status and 6MWT-distance reflects the influence of the status of the cardiovascular system on the general fitness. Inversely, however, it can also be explained by the negative prognostic value of lack of fitness for the ulterior development of cardiovascular disease. It has, indeed, repeatedly been reported for healthy middle-aged adults that this risk is greater for the least fit individuals [ 28 , 29 ]. Since most of our participants were older than those in the risk studies, a certain amount of cardiovascular pathology did occur. In a multiple linear regression model age, gender and health status explained 31% of the variability of the 6MWT-distance in the population of our study. All three independent variables had a highly significant contribution to the model (age and gender together accounted for 18% of the variability, partial correlation between health status and 6MWT-distance = -0.40, p < 0.001). The addition of weight, height and BMI to the prediction model did not significantly improve the R 2 -value. This is different from other studies that describe regression equations for the 6MWT-distance in healthy elderly [ 14 , 15 ], and indicates that the health condition of community dwelling elderly is more influencing the 6MWT-distance than their anthropometrical predisposition (weight and height are not significant independent predictors for 6MWT-distance after considering health status, age and gender). The fact that these authors obtained a much higher R 2 -value (R 2 = 0.66 [ 15 ], R 2 = 0.42 for male and R 2 = 0.38 for female [ 14 ]) than ours (R 2 = 0.31) can be due to the more heterogeneous composition in health status and the repartition of the population in several health categories in our study. The remaining unexplained variability in 6MWT-distance might be found in the differences in skeletal muscle strength, the training levels and the physical activity levels of the participants. However, these parameters were not measured in our study. We propose to incorporate the present health classification for elderly people as outcome of the medical screenings preceding the start of a physical exercise program. As shown in our study, significant differences can be found in physical exercise capacity according to the health status. We used the 6MWT-distance, which is an established exercise test for elderly persons, to document the physical exercise capacity of our participants. Especially in the elderly, which show increased prevalence of co-morbidities (most frequently (cardio)vascular, metabolic or rheumatological), physical exercise programs should be established according to the individual exercise capacity. Following the health classification of the elderly participant, the training instructor can adapt the modalities of the training program in order to obtain the best results at the most appropriate training intensity. This should finally lead to concrete training guidelines for elderly people in relation to their health status, diminishing the risk for injuries or complications during physical exercise. Competing interests None declared. Authors' contributions TM conceived and coordinated the study, participated in the evaluation of the health condition of the participants, the analysis and the redaction. IB performed the statistical analysis and participated in the design and the redaction. ML participated in the evaluation of the health condition of the participants, the analysis and the redaction. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC512286.xml |
550670 | A case-control study to evaluate urinary tract complications in radical hysterectomy | Background This study has evaluated urinary tract injuries and dysfunction after Radical Hysterectomy (RH) performed in patients with cervical cancer and has compared the cystometric parameters and urinary complications occurring in these patients with those occurring in patients who had undergone Simple Hysterectomy (SH). Patients and methods A prospective case-control study was conducted to evaluate urinary tract injuries (intra-operative and post-operative) and dysfunction in 50 patients undergoing RH for cervical cancer and to compare them with the same parameters in 50 patients who underwent SH for benign disease. Results Mean age in the RH group was 46.3 years and in the SH group was 50.1 (p = 0.63). There were no bladder and urethral injuries in either group of patients. There was one intra-operative ureteral injury in the RH patients but none in those who underwent SH. (p < 0.05). In the two weeks after surgery, 15% of RH patients and 11% of SH patients had experienced a urinary tract infection urinary tract infection (p = 0.61). Two week after surgery 62% of RH patients had no urinary symptoms, compared to 84% in the SH group who did (p < 0.02). Urinary residual volume, first urinary sensation and maximal bladder capacity were higher in the RH group, but this was not statistically significant. The only case of a urinary fistula appeared in a patient who received 5000 cGy radiation therapy pre-operatively, but this spontaneously healed after 3 weeks of catheterization. Conclusions Intra-operative and post-operative urinary tract complications are comparable in patients undergoing RH and SH and an expert gynaecological oncologist might be able to further decrease complications. However, radiation therapy before surgery may increase the risk of complications. | Background Although the incidence of lower urinary tract complications after RH has been reported with variable rates, up to one half of patients undergoing RH experience at least one lower urinary tract symptom that develops after surgery and at a variable period of time [ 1 , 2 ]. Several retrospective studies have examined lower urinary tract dysfunction and traumatic injuries in patients who have undergone RH [ 3 , 4 ]. In this study, we prospectively evaluated intra-operative urinary tract injuries in addition to post-operative urinary tract dysfunction and infection at 2, 6, and 14 weeks following surgery. We also compared these findings with those at the same times in patients who underwent SH for benign disease. Patients and methods Between October 2000 and December 2002, 50 women who underwent RH and bilateral lymph node dissection (BLND) were considered eligible for inclusion in the study. All patients had squamous cell carcinoma of cervix (SCC Cx.) and were staged as being Stages I and II. The operations were performed by the same gynaecological surgeons, using the same standard technique (class III Piver & Rutledge). Pre-operative management was standardised for all patients. Preoperatively a detailed medical history, physical examination, routine laboratory tests, pelvic CT-scan (with intravenous and oral contrast), urine analysis, and urine culture were carried out. The exclusion criteria were; a history of voiding dysfunction, previous pelvic surgery, brain or spinal cord diseases, diabetes mellitus, and contraindications to urodynamic studies. The latter included a history of vesicoureteral reflux, hydronephrosis, frequent or recent urinary tract infection or urethral stricture. Patients received one pre-operative and three post-operative doses of a second generation cephalosporin (Cephazolin). The duration of surgery, amount of intra-operative haemorrhage and the occurrence of any organ injuries were recorded. A Foley's catheter was inserted at the time of surgery and was left in place for two weeks after surgery. The patient's urinary catheter was removed when their post-voiding residual volume was less than 75 ml. Water cystometry, urinalysis and urine culture were performed at 2, 6, 14 weeks after operation. The test for water cystometry was performed with the subject lying in a supine position. A 12F double-lumen catheter was introduced transurethrally into the bladder to withdraw residual urine. The pressure-volume relationship of the bladder was determined by filling the bladder with isotonic saline at a rate of 30–50 ml/min. The cystometry fill phase ceased when the patient experienced an urge to void urine, the first indication being leakage through the urethra, or a bladder volume of 600 ml (which ever occurred first). The volume at the termination of the fill-phase was designated as the maximum bladder capacity (MBC). We also assessed the bladder volume of each patient at their first desire to void (V desire, ml). Post-void residual urine volume (RU) was determined by transurethral catheterization after voiding had ceased. The presence or absence of any urinary symptoms was determined by both questionnaire and direct interview with the patient. Fifty patients with benign disease who had underwent SH, were evaluated at the same time periods in the same way for comparison with the RH group of patients. In the SH group of patients, the Foley's catheter was inserted for 24 hours after operation. Data were analysed by SPSS statistical software using the chi-square and Student's 't' test for data analysis. Results During this study, 50 patients with early stage cervical cancer and who underwent RH for cervical cancer and 50 patients who had undergone SH for benign disease were evaluated. Two patients in the RH group and 3 from the SH group were lost during the study. The mean ages and their BMIs (Body Mass Index) in two groups of patients were not statistically different. However, parity in the RH group was higher (p < 0.05) (Table 1 ). In the RH group, the stages for the cervical cancer were 65.1%, 23.2% and 11.6% for I, II A and II B stages, respectively. Patients who had stage I B2 or higher stages of cervical cancer received 4500–5000 cGy of irradiation pre-operatively. None of these patients received adjuvant radiation during the interval between surgery and performance of urodynamic studies. Table 1 Comparison of characteristics of Radical and Simple Hysterectomy groups of patients. Characteristics RH group SH group p Mean age 50.10 46.35 0.63 BMI 24.25 26.05 0.66 Parity 6 4 0.00 Blood loss (ml) 576 ml 416 ml 0.04 Mean operative time (min) 183. min 112. min 0.00 RH-Radical hysterectomy; SH-simple hyterectomy In the SH group, the most common pathological conditions requiring hysterectomy were as follows; dysfunctional uterine bleeding (47.83%), uterine myoma (21.7%), ovarian cyst (10.8%), chronic pelvic pain (4.3%), adenomyosis (4.3%), endometrial cancer (4.35%), CIN (4.35%) and molar pregnancy (2.17%). The average blood loss and mean operative time for both groups are shown in table 1 . There were no bladder and urethral injuries occurring during the primary operation in either of the two groups of patients. One patient (with stage Ib 1 cervical cancer) in the RH group had an intra-operative ureteral injury (which happened at the time of "unroofing" the distal part), and the ureteral anastomosis was carried out at that time. The urinary catheter and ureteral stent were removed four weeks after operation. Another patient (with stage Ib 2 cervical cancer) had received chemo-irradiation (5000 cGy) pre-operatively. She had a spontaneous urine leakage from the vagina approximately 2 months after surgery. A retrograde cystography revealed a minute vesico-vaginal fistula. After 3 weeks of continuous bladder drainage, the fistula resolved spontaneously and she had no urinary leakage at her follow-up visits. Post-operative positive urine culture and urinary symptoms (dysuria, frequency, nocturia and dribbling) are showed in table 2 . Urinary symptoms occurred more commonly in patients who had undergone pre-operative radiotherapy, but this difference was not statistically significant (table 3 ). The abnormal findings as regards water cystometry are shown in table 4 . Table 2 Postoperative urinary symptoms in RH and SH group of patients. RH group SH group p Positive U/C 1 st visit 15% 11% 0.06 Positive U/C 2 nd visit 31% 20% 0.00 Positive U/C 3 rd visit 11% 9% NS Urinary symptoms 1 st visit 31% 20% 0.00 Urinary symptoms 2 nd visit 40% 34% 0.07 Urinary symptoms 3 rd visit 30% 33% NS U/C-Urine Culture ; NS-Not significant; RH-Radical hysterectomy; SH-simple hyterectomy Table 3 Comparison of urinary symptoms between patients undergoing RH but with or without pre-operative radiotherapy. Urinary symptoms (Postoperatively) RH group XRT + RH p 2 weeks, 1 st visit 39% 44% 0.76 6 weeks, 2 nd visit 36% 44% 0.64 14 weeks, 3 rd visit 26% 44% 0.27 XRT-History of pre-operative radiotherapy; RH-Radical hysterectomy Table 4 Relative frequency of RV, MC, FS, SI and UTI in the patients. Abnormal RV (%) Abnormal FS (%) Abnormal MC (%) Abnormal SI (%) UTI (%) First visit* RH 4 66 68 22 31 SH 0 67 69 17 20 Second visit** RH 0 58 61 20 31 SH 0 66 70 16 20 Third visit*** RH 0 49 † 65 17 11 SH 4 72 † 65 18 9 *After discharging the drain **Four weeks after discharging the drain ***Eight weeks after discharging the drain †P-value = 0.02 Discussion Modern surgical techniques have resulted in a decrease in the incidence of lower urinary tract complications occurring as a result of radical hysterectomy. In particular, in recent times, various surgical strategies have been developed to avoid damaging the inferior segment of the cardinal ligament as well as the terminal bundle in the uterosacral and pubocervicovesical ligaments. This has made it possible for patients' lower urinary functions to return more rapidly to their pre-operative states [ 2 , 5 ]. However, transient post-operative urinary dysfunction involving urinary storage and evacuation function continues to be of concern [ 6 ]. In a study by Zaino and colleagues intra-operative complications were reported to occur as being 4.5% urinary tract and 8.7% other organs (nervous, haemorrhage, intestinal) [ 7 ]. Ralph et al , reported a 6.6% rate of intra-operative urinary tract injuries during radical surgery for cervical carcinoma [ 2 ]. In our study, we had no intestinal or bladder injuries occurring during radical hysterectomy. The only ureteral injury (2%) occurred during "unroofing" of the distal ureter and this was recognised and repaired immediately. Zaino et al , [ 7 ] reported a 20% risk of a post-operative urinary tract fistula after radical surgery [ 7 ] and this contrasts with a 4.4% risk of fistula in their series reported by Ralph [ 2 ]. In our study, the only fistula occurred in a patient who had received 5000 cGy radiotherapy before radical surgery, and with continuous bladder drainage for 3 weeks there was spontaneous healing of the fistula. The incidence of urinary tract infection (UTI) in our series was 11% by 14 weeks after surgery and this was comparable to that reported by Cardosi [ 8 ] but less than the figure of 20% documented by Abrao [ 9 ]. Also, Chen reported a 14% urinary tract infection rate following radical surgery [ 10 ]. In our study, urinary tract dysfunctions that followed radical surgery were that 4% had an abnormal post-voiding residual volume at the first post-operative visit. The first voiding sensation at the third visit was 49% and stress urinary incontinence was 17%. However, maximal capacity remained abnormal in 65% of cases by 14 weeks after surgery. Ralph et al reported that 67% of patients had impairment or absence of bladder sensation after a RH [ 2 ]. In the study from Chen et al , 84% of patients had an increased first desire to void and maximal capacity in the post-operative period [ 10 ]. Urinary symptoms in our study occurred in 20% 2 weeks after operation and which were higher in patients with pre-operative radiotherapy (although not statistically significant). Urinary symptoms remained high at the third post-operative visit, although they declined in patients who had undergone surgery alone. In our study, the patients mean age; BMI, parity, operative time, and blood loss were higher in those undergoing RH. The mean age of our patients was higher than patients in the study by Vervest et al , (mean was 45 years) [ 11 ]. Also in this study [ 11 ] the patients BMI of 23.2 was lower than that of the patients in our study. Cystometric parameters and intra-operative and post-operative complications showed little difference between patients having either RH or SH. The small number of patients in our study could have biased the results. However, in spite of the different gravidity and days of bladder drainage in the two groups of patients, the overall complication rate is relatively low in the RH patients. The data in this study requires confirmation from future multicentric studies with greater numbers of patients. In recent years, several studies support the role of a gynaecological oncologist who is specifically trained in such aspects of care and who can obtain optimal cytoreductive surgery in patients with ovarian carcinoma [ 12 , 13 ]. Therefore, it would seem that an experienced and appropriately trained gynaecological oncologist might achieve a complication rate for patients undergoing radical hysterectomy comparable with that reported by "general gynaecological surgeons.. Competing interests The author(s) declare that they have no competing interests. Authors' Contributions NB carried out the surgery and participated in drafting the manuscript. FG carried out follow-ups. HA participated in the design of the study and helped to draft the manuscript. HK helped in follow-ups and performed the statistical analyses. PH participated in the design of the study and helped to draft the manuscript. All authors read and approved the final manuscript. | /Users/keerthanasridhar/biomedlm/data/PMC000xxxxxx/PMC550670.xml |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.